Glucagon‐like peptide 1 (GLP‐1) receptor agonists for people with chronic kidney disease and diabetes

Abstract

Background

Approximately 40% of people with diabetes develop kidney failure and experience an accelerated risk of cardiovascular complications. Glucagon‐like peptide 1 (GLP‐1) receptor agonists are glucose‐lowering agents that manage glucose and weight control.

Objectives

We assessed the benefits and harms of GLP‐1 receptor agonists in people with chronic kidney disease (CKD) and diabetes.

Search methods

The Cochrane Kidney and Transplant Register of Studies was searched to 10 September 2024 through contact with the Information Specialist using search terms relevant to this review. Studies in the Register are identified through searches of CENTRAL, MEDLINE, and EMBASE, conference proceedings, the International Clinical Trials Registry Platform (ICTRP) Search Portal, and ClinicalTrials.gov.

Selection criteria

Randomised controlled studies were eligible if participants with diabetes and CKD were randomly allocated to a GLP‐1 receptor agonist, placebo, standard care or a second glucose‐lowering agent. CKD included all stages (from 1 to 5).

Data collection and analysis

Three authors independently extracted data and assessed the risk of bias using the risk of bias assessment tool 2. Pooled analyses using summary estimates of effects were obtained using a random‐effects model, and results were expressed as risk ratios (RR) and/or hazard ratio (HR) and their 95% confidence intervals (CI) for dichotomous outcomes and mean difference (MD) and 95% CI for continuous outcomes. The primary outcomes included death (all‐cause and cardiovascular), 3‐ and 4‐point major adverse cardiovascular events (MACE), kidney failure, composite kidney outcome, and severe hypoglycaemia. The secondary outcomes included non‐fatal or fatal myocardial infarction (MI) or stroke, non‐fatal peripheral arterial events, heart failure, hospitalisation due to heart failure, estimated glomerular filtration rate or creatinine clearance, doubling of serum creatinine, urine albumin‐to‐creatinine ratio, albuminuria progression, vascular access outcomes, body weight, body mass index, fatigue, life participation, peritoneal dialysis infection, peritoneal dialysis failure, adverse events, serious adverse events, withdrawal due to adverse events, HbA1c, sudden death, acute MI, ischaemic stroke, and coronary revascularisation. Confidence in the evidence was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach.

Main results

Forty‐two studies involving 48,148 participants were included. All studies were conducted on people with type 2 diabetes, and no studies were carried out on children. The median study age was 66 years. The median study follow‐up was 26 weeks. Six studies were conducted in people with CKD stages 1‐2, 11 studies in people with CKD stages 3‐5, one study in people on dialysis, and the remaining studies included people with both CKD stages 1‐2 and 3‐5. Risks of bias in the included studies for all the primary outcomes in studies that compared GLP‐1 receptor agonists to placebo were low in most methodological domains, except one study that was assessed at high risk of bias due to missing outcome data for death (all‐cause and cardiovascular). The overall risk of bias for all‐cause and cardiovascular death in studies that reported the treatment effects of GLP‐1 receptor agonists compared to standard care, dipeptidyl peptidase‐4 (DPP‐4) inhibitors or sodium‐glucose cotransporter 2 (SGLT2) inhibitors were assessed as unclear or at high risk of bias due to deviations from intended interventions or missing data. For GLP‐1 receptor agonists compared to insulin or another GLP‐1 receptor agonist, the risk of bias for all‐cause and cardiovascular death was low or unclear.

Compared to placebo, GLP‐1 receptor agonists probably reduced the risk of all‐cause death (RR 0.85, 95% CI 0.74 to 0.98; I² = 23%; 8 studies, 17,861 participants; moderate‐certainty evidence), but may have little or no effect on cardiovascular death (RR 0.84, 95% CI 0.68 to 1.05; I² = 42%; 7 studies, 17,801 participants; low‐certainty evidence).

Compared to placebo, GLP‐1 receptor agonists probably decreased 3‐point MACE (RR 0.84, 95% CI 0.73 to 0.98; I² = 65%; 4 studies, 19,825 participants; moderate‐certainty evidence), and 4‐point MACE compared to placebo (RR 0.77, 95% CI 0.67 to 0.89; 1 study, 2,158 participants; moderate‐certainty evidence). Based on absolute risks of clinical outcomes, it is likely that GLP‐1 receptor agonists prevent all‐cause death in 52 people with CKD stages 1‐2 and 116 in CKD stages 3‐5, cardiovascular death in 34 people with CKD stages 1‐2 and 71 in CKD stages 3‐5, while 95 CKD stages 1‐2 and 153 in CKD stages 3‐5 might experience a major cardiovascular event for every 1000 people treated over 1 year.

Compared to placebo, GLP‐1 receptor agonists probably had little or no effect on kidney failure, defined as starting dialysis or kidney transplant (RR 0.86, 95% CI 0.66 to 1.13; I² = 0%; 3 studies, 4,134 participants; moderate‐certainty evidence), or on composite kidney outcomes (RR 0.89, 95% CI 0.78 to 1.02; I² = 0%; 2 studies, 16,849 participants; moderate‐certainty evidence).

Compared to placebo, GLP‐1 receptor agonists may have little or no effect on the risk of severe hypoglycaemia (RR 0.82, 95% CI 0.54 to 1.25; I² = 44%; 4 studies, 6,292 participants; low‐certainty evidence).

The effects of GLP‐1 receptor agonists compared to standard care or other hypoglycaemic agents were uncertain.

No studies evaluated treatment on risks of fatigue, life participation, amputation or fracture.

Authors' conclusions

GLP‐1 receptor agonists probably reduced all‐cause death but may have little or no effect on cardiovascular death in people with CKD and diabetes. GLP‐1 receptor agonists probably lower major cardiovascular events, probably have little or no effect on kidney failure and composite kidney outcomes, and may have little or no effect on the risk of severe hypoglycaemia in people with CKD and diabetes.

Plain language summary

Does treatment with GLP‐1 receptor agonists prevent complications for people with chronic kidney disease and diabetes?

Key messages

• For people with chronic kidney disease (CKD) (a long‐term condition where the kidneys do not work effectively) also having diabetes (a lifelong condition that causes a person's blood sugar level to become too high) increases the chances of early death, heart attack, stroke and reduces a person's quality of life.

• Glucagon‐like peptide 1 receptor agonists (GLP‐1) (medicines used to lower blood glucose) probably reduce the risk of death due to any cause but may have little or no death due to a heart attack in people with CKD and diabetes.

• For combined major heart‐related events (stroke, nonfatal heart attack and death due to a heart attack), GLP‐1 probably lowers the risk of these events but probably has little or no effect on the risk of kidney failure or abnormally low blood sugar levels.

Why treat people with chronic kidney disease and diabetes with GLP‐1 receptor agonists?

Chronic kidney disease (CKD) (a long‐term condition where the kidneys do not work effectively) and diabetes (a lifelong condition that causes a person's blood sugar level to become too high) are chronic conditions that bring on many challenges for people, particularly when they have to manage both at the same time. Diabetes can accelerate the development of kidney disease and is the leading cause of kidney failure (a condition where the kidneys no longer function well enough to keep a person alive). GLP‐1 receptor agonists (GLP‐1) are medicines that lower blood glucose levels and may also have beneficial effects on high blood pressure and obesity, and decrease the chances of death, heart attack, stroke and kidney failure (a condition where the kidneys no longer function well enough to keep a person alive).

What did we want to find out?

We wanted to find out if GLP‐1 improves diabetes control and kidney function, decreases heart‐related complications such as heart attacks and stroke, and decreases the risk of kidney failure in people with both CKD and diabetes.

What did we do?

We searched for all trials that assessed the benefits and harms of GLP‐1 for treating people with both CKD and diabetes. We compared GLP‐1 with placebo (dummy medicine), usual care, and other glucose‐lowering medicines (for example, insulin).

What did we find?

We included 42 studies that randomised 48,148 adults (18 years or older) with CKD and diabetes. None of the studies included children. The number of people ranged from 7 to 14,691, and the age ranged from 51 to 71 years. On average, the duration of the studies was six months. Twenty‐one studies compared GLP‐1 to placebo, 16 compared GLP‐1 to a variety of glucose‐lowering medicines, and five compared GLP‐1 to standard care. Twenty‐eight studies were multinational.

Compared to placebo, GLP‐1 probably reduces the overall risk of dying due to any cause but may have little or no effect on the chances of dying from a heart‐related problem. When major heart‐related events were investigated together (death due to a heart attack, nonfatal heart attack and nonfatal stroke), GLP‐1 probably reduces the risk of these combined events occurring. GLP‐1 probably makes little or no difference to kidney failure and may make little or no difference to abnormally low blood glucose levels.

The effect of GLP‐1 compared to usual care and other glucose‐lowering medicines is unclear.

What are the limitations of the evidence?

We are moderately confident that GLP‐1 reduces the risk of dying due to any cause, reduces major heart‐related events, and makes little or no difference to kidney failure. We are less confident in its effect on death from a heart‐related problem or abnormally low blood sugar levels.

How up to date is the evidence?

The evidence is current to September 2024.

Summary of findings

Summary of findings 1. Summary of findings table ‐ GLP‐1 agonists compared to placebo for people with chronic kidney disease and diabetes.

GLP‐1 agonists compared to placebo for people with chronic kidney disease and diabetes
Patient or population: people with chronic kidney disease and diabetes Setting: Multinational Intervention: GLP‐1 agonists Comparison: placebo
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with GLP‐1 agonists
All‐cause death follow‐up: median 26 weeks	86 per 1000	73 per 1000 (64 to 85)	RR 0.85 (0.74 to 0.98)	17861 (8 RCTs)	⊕⊕⊕⊝ Moderatea	GLP‐1 receptor agonists probably reduced the risk of all‐cause death compared to placebo in people with all CKD stages and type 2 diabetes
Cardiovascular death follow‐up: median 26 weeks	56 per 1000	47 per 1000 (38 to 59)	RR 0.84 (0.68 to 1.05)	17801 (7 RCTs)	⊕⊕⊝⊝ Lowa,b	GLP‐1 receptor agonists may have little or no effect on the risk of cardiovascular death compared to placebo in people with all CKD stages and type 2 diabetes
3‐point MACE follow‐up: median 137 weeks	131 per 1000	110 per 1000 (96 to 129)	RR 0.84 (0.73 to 0.98)	19825 (5 RCTs)	⊕⊕⊕⊝ Moderatec	MACE 3‐point was defined in all included studies as the composite cardiovascular outcome, including cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke. GLP‐1 receptor agonists probably decreased MACE 3‐point compared to placebo
4‐point MACE follow‐up: median 200 weeks	304 per 1000	234 per 1000 (204 to 271)	RR 0.77 (0.67 to 0.89)	2158 (1 RCT)	⊕⊕⊕⊝ Moderated	MACE 4‐point was defined as the composite cardiovascular outcome including cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, plus coronary revascularisation or hospitalisation for unstable angina pectoris or heart failure. GLP‐1 receptor agonists probably decreased MACE 4‐point compared to placebo in people with CKD stages 3‐5 and type 2 diabetes
Kidney failure follow‐up: mean 610 weeks	53 per 1000	45 per 1000 (35 to 60)	RR 0.86 (0.66 to 1.13)	4134 (3 RCTs)	⊕⊕⊕⊝ Moderatee	GLP‐1 receptor agonists probably had little or no effect on kidney failure compared to placebo in people with all CKD stages and type 2 diabetes
Composite kidney outcome follow‐up: median 185 weeks	51 per 1000	45 per 1000 (40 to 52)	RR 0.89 (0.78 to 1.02)	16849 (2 RCTs)	⊕⊕⊕⊝ Moderateb	GLP‐1 receptor agonists probably had little or no effect on kidney composite outcomes compared to placebo in people with all CKD stages and type 2 diabetes
Severe hypoglycaemia follow‐up: median 116 weeks	37 per 1000	31 per 1000 (20 to 47)	RR 0.82 (0.54 to 1.25)	6292 (4 RCTs)	⊕⊕⊝⊝ Lowb,c	GLP‐1 receptor agonists may have little or no effect on the risk of severe hypoglycaemia compared to placebo in people with all CKD stages and type 2 diabetes
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_447355969649398767.

^a Downgraded one level for risk of bias (as risk of bias was high for some domains)
^b Downgraded one level for imprecision (as 95% CI contains potential benefit and harm and crossed the null)
^c Downgraded one level for moderate heterogeniety among studies
^d Downgraded one level since a single study reported this outcome
^e Downgraded two levels for imprecision (as few events and participant reported, and 95% CI contains potential benefit and harm that crossed the null)

Background

Description of the condition

Diabetes affects over 422 million people worldwide (WHO 2022), and it is estimated that its prevalence could reach up to 700 million people by 2045 (Saeedi 2019). Type 2 diabetes accounts for approximately 90% of cases of diagnosed diabetes and occurs because of insulin deficiency caused by pancreatic β‐cell dysfunction and tissue insulin resistance (Chatterjee 2017). Diabetes is a risk factor for cardiovascular diseases and chronic kidney disease (CKD) and increases 3‐ to 4‐fold the risk of all‐cause and cardiovascular death (Raghavan 2019). Diabetes impairs life participation and quality of life (Aschalew 2020) and contributes to substantial costs for the healthcare system, especially in low‐ and middle‐income countries (Moucheraud 2019).

CKD is characterised by decreased estimated glomerular filtration rate (eGFR) below 60 mL/min/1.73 m² and/or structural kidney or urinary abnormalities. The number of people diagnosed with CKD has dramatically increased globally over the past two decades (GBD 2017). Approximately 30% to 40% of people with diabetes develop kidney failure (previously called end‐stage kidney disease) (ANZDATA 2018; JSDT 2018; USRDS 2018), requiring kidney replacement therapy (KRT), including haemodialysis (HD), peritoneal dialysis (PD) and kidney transplantation (Collins 2008). People with CKD and diabetes experience a very high risk of death and cardiovascular events, including heart failure (Radbill 2008).

Description of the intervention

Glucagon‐like peptide 1 (GLP‐1) receptor agonists are glucose‐lowering agents that achieve glucose and weight control, and are associated with a lower risk of death and major cardiovascular outcomes in people with diabetes (Mirabelli 2021; Palmer 2021; Zheng 2018). GLP‐1 receptor agonists are incretin hormone mimetics that activate the GLP‐1 receptor, located at the surface of the pancreatic β‐cell, increasing insulin synthesis and reducing glucagon secretion, food intake and gastric emptying (Drucker 2018a).

The American Diabetes Association (ADA) recommends GLP‐1 receptor agonists as a first‐line treatment for type 2 diabetes, especially in people with obesity or cardiovascular disease (ADA 2022). Other guidelines suggest GLP‐1 receptor agonists in patients who are unable to use sodium‐glucose cotransporter 2 (SGLT2) inhibitors or metformin (Chen 2022). Previous studies have shown GLP‐1 receptor agonists are also effective in people with diabetes in all stages of CKD in improving blood glucose management, decreasing body weight, slowing the decline of GFR, and preventing the onset of albuminuria or severe hypoglycaemia (Górriz 2020). However, previous guidelines showed discordant recommendations for using GLP‐1 receptor agonists because the benefits and harms were still uncertain in this population. To date, Kidney Disease: Improving Global Outcomes (KDIGO) guidelines recommend GLP‐1 receptor agonists in people with type 2 diabetes and eGFR < 60 mL/min/1.73 m² to meet individualised glucose targets despite first‐line treatment with metformin and SGLT2 inhibitors. Although gastrointestinal effects might limit their use (KDIGO 2022), GLP‐1 receptor agonists may also be effective for weight management among people with diabetes and CKD, but their benefits in people with type 1 diabetes have not been investigated (KDIGO 2022). The ADA and KDIGO guidelines agree that GLP‐1 receptor agonists may reduce albuminuria and major cardiovascular events in people with type 2 diabetes and CKD, and recommend GLP‐1 receptor agonists as second‐line therapy, although large and powered randomised controlled trials (RCTs) are currently lacking (de Boer 2022).

How the intervention might work

GLP‐1 receptor agonists reduce kidney and cardiovascular endpoints in people with type 2 diabetes (Marx 2022; Yin 2020), and have beneficial effects on hypertension, obesity, and lipid profile (Drucker 2018b). Recent evidence suggests that GLP‐1 receptor agonists (including exenatide, liraglutide, semaglutide, and dulaglutide) may reduce death, major adverse cardiovascular events (MACE), progression of albuminuria, or GFR reduction in people with CKD and type 2 diabetes (Guja 2020; Mann 2020a; Mann 2020b; Tuttle 2018). Although the mechanism for the renoprotective effect has not been clearly understood, it is likely that GLP‐1 receptor agonists ameliorate oxidative stress and inflammation (Kawanami 2020).

However, GLP‐1 receptor agonists have potential adverse events, mainly gastrointestinal symptoms (such as nausea, vomiting and diarrhoea), that are dose‐dependent and varied based on the route of administration (Sun 2015). Other common adverse events include headache, nasopharyngitis, or injection site reactions, but usually, they do not lead to treatment discontinuation (Filippatos 2014).

To date, there is limited evidence on the efficacy and safety of GLP‐1 receptor agonists in people with concomitant diabetes and CKD in primary kidney and cardiovascular endpoints. The safety and efficacy of GLP‐1 receptor agonists might be different in people with CKD compared to other high‐risk populations because of the higher prevalence of inflammation, oxidative stress, mineral‐bone disease, proteinuria, or multidrug interactions (Lovre 2017).

Why it is important to do this review

In a previous review by Lo 2018, a broad focus on the efficacy and safety of all glucose‐lowering agents (including SGLT2 inhibitors, dipeptidyl peptidase‐4 (DPP‐4) inhibitors, GLP‐1 receptor agonists, glitazones, glinide, and insulin) in people with CKD (stages 3 to 5) and diabetes was taken. We have separated the reviews due to the difference in pharmacokinetics, pharmacodynamics and drug‐specific characteristics. The benefits and harms of each class of glucose‐lowering agents will be evaluated in new, separate reviews. This review focused on GLP‐1 receptor agonists. This review included people with type 1 and 2 diabetes and CKD and those with early stages of CKD (eGFR stages 1 and 2). People with diabetes were included either if they reported an eGFR < 90 mL/min/1.73 m² regardless of albuminuria or with microalbuminuria/macroalbuminuria regardless of eGFR, according to the KDIGO guidelines for CKD classification (KDIGO 2012).

Recently, new studies have been done in this setting, necessitating an updated evidence summary.

Objectives

This review aims to assess the benefits and harms of GLP‐1 receptor agonists in people with CKD and diabetes.

Methods

Criteria for considering studies for this review

Types of studies

All RCTs and quasi‐RCTs (RCTs in which allocation to treatment was obtained by alternation, use of alternate medical records, date of birth or other predictable methods) evaluating the effects of GLP‐1 receptor agonists versus placebo, standard care or a different glucose‐lowering agent in people with both CKD and diabetes were eligible, regardless of follow‐up duration. Cross‐over and cluster RCT were also eligible.

Types of participants

Inclusion criteria

We included adults (≥ 18 years) and children (< 18 years) with type 1 and 2 diabetes and all stages of CKD (from 1 to 5), including people treated with dialysis. People with diabetes and albuminuria or structural kidney abnormalities, regardless of eGFR, were also included.

RCTs enrolling people with diabetes and CKD were included regardless of the severity of albuminuria. We included people with diabetes who have either moderately increased albuminuria (formerly known as microalbuminuria ‐ A2) (equivalence of an albumin excretion of 30 to 300 mg/day) or severely increased albuminuria (formerly known as macroalbuminuria ‐ A3) (albumin excretion > 300 mg/day). People with diabetes were included either if they reported moderately or severely increased albuminuria regardless of eGFR or with eGFR < 90 mL/min/1.73 m² regardless of albuminuria, according to the KDIGO guidelines for CKD classification (KDIGO 2012). Studies reporting a mixed population, including people with CKD with or without diabetes, were eligible, but only data from people with any stage of CKD and diabetes were included in this review. Specifically, we considered exploring in more detail studies evaluating the effects of GLP‐1 receptor agonists in people with either diabetes or any stage of CKD only to be sure that a subset of patients of our interest could be included. Inclusion and exclusion criteria were assessed carefully for these studies to evaluate if any information could be collected to confirm that at least one participant could be included in a subset of people with CKD and diabetes. If yes, we included the study and extracted data only if it clearly related to the subset of participants of our interest; if no data were explicitly available to this population, the study was included, but no data were extracted. Authors were contacted to get more information on the population of our interest in a study addressing a mixed population, but no data was extracted if no clear information was provided.

Exclusion criteria

Recipients of a kidney transplant were ineligible. We also did not include evaluations of GLP‐1 receptor agonists in people with pre‐diabetes or polycystic kidney disease. We did not include studies evaluating GLP‐1 receptor agonists in people with gestational diabetes.

Types of interventions

We planned to evaluate the following treatment comparisons.

1. GLP‐1 receptor agonists versus placebo

2. GLP‐1 receptor agonists versus standard care (including usual other glucose‐lowering agents or no treatment)

3. GLP‐1 receptor agonists versus different glucose‐lowering agents (including metformin, insulin, sulphonylurea, glinides, glitazones, ɑ‐glucosidase inhibitors, DPP‐4 inhibitors, SGLT2 inhibitors, other GLP‐1 agonists, amylin analogues, and bromocriptine)

All GLP‐1 receptor agonists were eligible, including but not limited to liraglutide, semaglutide, exenatide, dulaglutide, and lixistenatide. We evaluated GLP‐1 receptor agonists given via any route (e.g. oral or injectable) at any frequency.

Types of outcome measures

We did not exclude studies based on non‐reporting of outcomes of interest.

The outcomes selected included the relevant SONG core outcome sets as specified by the Standardised Outcomes in Nephrology initiative (SONG 2017) and the core outcome sets for diabetes according to the KDIGO guidelines (KDIGO 2022). As per standard practice, we have included only the core outcome set identified by the SONG initiative, which is the minimum outcome set that should be reported in all clinical studies in nephrology as prioritised by patients, caregivers and health professionals (SONG 2017). We have also added all the relevant outcomes according to those reported in the KDIGO 2022 guideline to make our findings and reporting consistent with the current guideline. To avoid misclassification of outcomes that could be reported using different definitions, we provided definitions for outcomes (e.g. 3‐ and 4‐point MACE, composite kidney outcomes) as reported by the study authors.

Primary outcomes

1. All‐cause death

2. Cardiovascular death

3. 3‐point and 4‐point MACE, or a combination of MACE, as reported by the authors

4. Kidney failure requiring KRT (dialysis or kidney transplant) or eGFR < 15 mL/min/1.73 m²

5. Composite kidney outcomes, as reported by the authors

6. Severe hypoglycaemia was defined as hypoglycaemia requiring third‐party assistance.

Secondary outcomes

1. Non‐fatal or fatal myocardial infarction (MI), as defined by the study protocol

2. Non‐fatal or fatal stroke, as defined by the study protocol

3. Non‐fatal peripheral arterial events, as defined by the study protocol

4. Heart failure and hospitalisation due to heart failure

5. eGFR or creatinine clearance (CrCl) (any measure)

6. Doubling of serum creatinine(SCr)

7. Urine albumin‐to‐creatinine ratio (UACR)

8. Albuminuria progression: onset of albuminuria, moderately increased to severely increased albuminuria, including UACR and albumin excretion

9. Vascular access outcomes (e.g. failure, thrombosis, loss of unassisted patency such as stenosis/occlusions, and need for access intervention)

10. Body weight

11. Body mass index (BMI)

12. Fatigue

13. Life participation

14. PD infection

15. PD failure

16. Serious adverse events

17. Withdrawal due to adverse events

18. Adverse events (e.g. hypoglycaemia, lactic acidosis, amputation, bone fractures, acute kidney injury (AKI), diabetic ketoacidosis)

19. Glycated haemoglobin (HbA1c)

20. Sudden death

21. Acute MI

22. Ischaemic stroke

23. Coronary revascularisation.

Search methods for identification of studies

Electronic searches

The Cochrane Kidney and Transplant (CKT) Register of Studies was searched by the CKT Information Specialist using search terms relevant to this review (up to 10 September 2024). The Register contains studies identified from the following sources.

1. Cochrane Central Register of Controlled Trials (CENTRAL)

   • CENTRAL includes records from EMBASE, Clinicaltrials.gov and the International Clinical Trials Registry Platform (ICTRP) Search Portal.

2. Weekly searches of MEDLINE OVID SP using Cochrane Kidney and Transplant's scope.

3. Handsearched records from kidney‐related journals up to 2021.

Studies contained in the Register are identified through search strategies for CENTRAL, MEDLINE, and EMBASE based on the scope of CKT. Details of these strategies, as well as a list of handsearched journals, conference proceedings and current awareness alerts, are available in the "Specialised Register" section of information about Cochrane Kidney and Transplant.

Additional search strategies have been designed by the Information Specialist for MEDLINE, EMBASE and CENTRAL. Please see Appendix 1 for search terms used.

Searching other resources

1. Reference lists of review articles, relevant studies and clinical practice guidelines.

2. Contacting relevant individuals/organisations seeking information about unpublished or incomplete studies.

3. Grey literature sources (e.g. abstracts, dissertations and theses), in addition to those already included in the Cochrane Kidney and Transplant Register of Studies, were searched.

Data collection and analysis

Selection of studies

The search strategy described was used to obtain titles and abstracts of studies that may be relevant to the review. The titles and abstracts were screened independently by three authors (PN, DT, GP), who discarded studies that were not applicable. However, studies and reviews that might include relevant data or information on trials were retained initially. Three authors (PN, DT, GP) independently assess retrieved abstracts and, if necessary, the full text of these studies to determine which studies satisfy the inclusion criteria. Disagreements were resolved in consultation with another author (SG). Authors were contacted to confirm the eligibility of the studies, as appropriate.

Data extraction and management

Data extraction was carried out independently by three authors (PN, DT, GP) using standard data extraction forms. Disagreements were resolved in consultation with another author (SG). Studies reported in non‐English language journals were translated before assessment. Where more than one publication of one study existed, reports were grouped together, and the publication with the most complete data was used in the analyses. Where relevant outcomes were only published in earlier versions, these data were used. If data was not clearly reported by authors (e.g. unspecified MI or stroke), data was reported/assessed as it was, without combining in other outcomes categories (e.g. fatal or nonfatal MI or stroke) to avoid overestimation of the findings. Any discrepancies between the published versions were highlighted.

The following characteristics of each eligible study were extracted.

Methods

• Study design

• Study time frame

• Duration of follow‐up

• Country

• Setting

Participants

• Inclusion criteria

• Exclusion criteria

• Number (randomised/analysed)

• Mean age ± SD (years)

• Sex (M/F)

Interventions

• Intervention group

• Control group

• Co‐interventions

Outcomes

• Primary outcomes

• Secondary outcomes

Notes

• Funding source

• Conflicts of interest

• Trial registration number

Assessment of risk of bias in included studies

The following items were independently assessed by three authors (PN, DT, GP) using the risk of bias assessment tool version 2 (RoB2) (Higgins 2022; Sterne 2019).

We assessed the effect of the assignment on the intervention. We analysed participants in the intervention groups to which they were randomised, regardless of the intervention they actually received. We included all randomised participants in the outcome analyses, based on the intention‐to‐treat principle.

We assessed the risk of bias of all primary and secondary outcomes using the following five domains (Higgins 2023a; Higgins 2023b; Sterne 2019).

• Bias arising from the randomisation process

• Bias due to deviations from intended interventions

• Bias due to missing outcome data

• Bias in the measurement of the outcome

• Bias in the selection of the reported results.

For cross‐over studies, we used the standard version of RoB2 (Sterne 2019). For signalling questions within each domain for each outcome, we provided one of the five possible answers in the RoB2 tool ('Yes', 'Probably yes', 'No', 'Probably no' and 'No information'). After this step, we made a judgement for each domain according to the algorithm result as 'Low risk of bias', 'Some concerns' or 'High risk of bias' for each outcome assessed in this review. The overall risk of bias judgement for each outcome was also included. The three levels of judgement for an overall rating are as follows.

• Low risk of bias: the trial was judged to be at low risk of bias for all domains for this result

• Some concerns: the trial was judged to raise some concerns in at least one domain for this result, but is not at high risk of bias for any of the remaining domains

• High risk of bias: the trial was judged to be at high risk of bias in at least one domain for this result, or the study was judged to have some concerns for multiple domains in a way that substantially lowers confidence in the result.

For cluster RCTs, we planned to assess the risk of bias associated with one additional domain, specific to this study design (Eldridge 2024).

• Bias arising from the timing of identification and recruitment of individual participants in relation to the timing of randomisation.

We used the RoB2 Microsoft Excel tool to store the data.

Measures of treatment effect

For dichotomous outcomes (all‐cause death, cardiovascular death, non‐fatal MI, non‐fatal stroke, non‐fatal peripheral arterial events, kidney failure, serious adverse events, severe hypoglycaemia, 3‐ or 4‐point MACE, heart failure, hospitalisation due to heart failure, doubling of SCr, albuminuria progression, composite kidney outcomes, vascular access outcomes, PD infection, PD failure, withdrawn due to adverse events, and adverse events) results were reported as risk ratios (RR) or hazard ratios (HR) with 95% confidence intervals (CI). Where continuous scales of measurement were used to assess the effects of treatment (eGFR or CrCl, body weight, BMI, fatigue, life participation, and HbA1c), the mean difference (MD) was reported.

Treatment estimates reported as HR or RR with 95% CI were both reported separately and summarised to describe the treatment effect of GLP‐1 receptor agonists on cardiovascular, kidney, and safety outcomes. Both measures of association were used to capture all data pertaining to the therapeutic effect of GLP‐1 receptor agonists. However, data on HR was reported in a separate table to improve the readability of the results.

Absolute effects were calculated considering the assumed risk with allocation to placebo.

Studies analysing change scores or end‐of‐treatment values were included in meta‐analyses together with studies including endpoint outcome data. Missing standard deviations were calculated from standard errors, as available.

Unit of analysis issues

For cross‐over studies, if no data was reported on washout periods for both time periods, we extracted data from the end of the first period of treatment, if available. For cluster RCTs the unit of analysis was the cluster. In order to avoid a unit of analysis error, we reanalysed the data by taking into account the intra‐cluster correlation (ICC). If these data were not provided, then we excluded the study (Higgins 2023b).

Dealing with missing data

Any further information required from the original author team or study sponsor was requested by written correspondence (e.g. emailing the corresponding author/s) and any relevant information obtained in this manner was included in the review. Evaluation of important numerical data such as screened, randomised patients as well as intention‐to‐treat, as‐treated and per‐protocol population was carefully performed. Attrition rates, for example, drop‐outs, losses to follow‐up and withdrawals were investigated. Issues of missing data and imputation methods (e.g. last‐observation‐carried‐forward) were critically appraised (Higgins 2022).

Assessment of heterogeneity

We analysed clinical heterogeneity, reporting any possible variability in participants, interventions, and outcomes in the included studies. We assessed methodological heterogeneity by comparing the distribution of important study factors (allocation concealment, blinding of outcome assessment, loss to follow‐up, and treatment type).

We also assessed the heterogeneity by visual inspection of the forest plot. We quantified statistical heterogeneity using I² statistic, which describes the percentage of total variation across studies that is due to heterogeneity rather than sampling error (Higgins 2003). A guide to the interpretation of I² values was reported as follows.

• 0% to 40%: might not be important

• 30% to 60%: may represent moderate heterogeneity

• 50% to 90%: may represent substantial heterogeneity

• 75% to 100%: considerable heterogeneity.

The importance of the observed value of I² depends on the magnitude and direction of treatment effects and the strength of evidence for heterogeneity (e.g. P value from the Chi² test or a CI for I²) (Higgins 2022).

Assessment of reporting biases

If possible, funnel plots were used to assess the potential existence of small study effects in the absence of meta‐analysis heterogeneity (Higgins 2022).

Data synthesis

Data was pooled using the random‐effects model that provided a result that may be viewed as an ‘average intervention effect’, where this average was explicitly defined according to an assumed distribution of effects across studies. We assumed that they followed a normal distribution. The assumption implies that the observed differences among study results are due to a combination of the play of chance.

However, the fixed‐effect model was also used to ensure the robustness of the model chosen and susceptibility to outliers. The random‐effect and the fixed‐effect model give identical results when there is no heterogeneity among the studies.

Subgroup analysis and investigation of heterogeneity

Subgroup analysis was used to explore possible sources of heterogeneity. Heterogeneity among participants could be related to age (adults versus children), stage of kidney disease (stages 1‐2, stages 3‐5 not requiring dialysis, HD, PD), and presence of comorbidities (hypertension and cardiovascular disease). Heterogeneity in treatments could be related to prior agent(s) used, type, dose and duration of therapy, and the target HbA1c during therapy (≤ 7% or > 7%). Subgroup analysis was performed for different stages of CKD. Adverse effects were tabulated and assessed with descriptive techniques, as they were likely to be different for the various agents used. Where possible, the risk difference with 95% CI was calculated for each adverse effect, either compared to a placebo, standard care or another glucose‐lowering agent.

Sensitivity analysis

We performed sensitivity analyses in order to explore the influence of the following factors on effect size.

• Repeating the analysis excluding unpublished studies

• Repeating the analysis, taking into account the high risk of bias to explore the effect of the methodological quality of studies on the overestimation of the treatment effect, using the overall risk of bias

• Repeating the analysis, excluding any very long (> 3 years, that is 156 weeks) or large studies (> 2000 participants), to establish how much they dominate the results

• Repeating the analysis excluding studies using the following filters: diagnostic criteria, the language of publication, source of funding (industry versus other), and country.

Summary of findings and assessment of the certainty of the evidence

We present the main results of the review in the summary of findings (SOF) tables. These tables present key information concerning the certainty of the evidence, the magnitude of the effects of the interventions examined, and the sum of the available data for the main outcomes (Schünemann 2023a). The SOF tables also included an overall grading of the evidence related to each of the main outcomes using the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) approach (GRADE 2008; GRADE 2011). We used the GRADE approach for the preparation of the SOF tables, using the GRADEpro GDT software (GRADEpro GDT). We reported outcomes considering the longest median or mean follow‐up and the range of follow‐up for each of the outcomes assessed.

The GRADE approach defines the certainty of a body of evidence as the extent to which one can be confident that an estimate of effect or association is close to the true quantity of specific interest. This was assessed by three authors (PN, DT, GP). The certainty of a body of evidence involves consideration of within‐trial risk of bias (methodological quality), directness of evidence, heterogeneity, precision of effect estimates, and risk of publication bias (Schünemann 2023b). We have presented the following outcomes in the SOF tables comparing GLP‐1 receptor agonists to placebo in people with CKD and diabetes.

• All‐cause death

• Cardiovascular death

• 3‐point MACE

• 4‐point MACE

• Kidney failure

• Kidney composite outcomes

• Severe hypoglycaemia

Results

Description of studies

The following section contains broad/brief descriptions of the studies considered in this review. For further details on each individual study. See Characteristics of included studies; Characteristics of excluded studies; Characteristics of ongoing studies. Twenty (46.5%) studies reported insufficient data, and we have contacted the authors to get further data. However, only the authors of four studies replied and/or were able to provide additional data, while authors from the remaining 16 studies did not reply.

Results of the search

After searching the Specialised Register (up to 10 September 2024), a total of 631 records were identified. After screening titles and abstracts, 456 records were excluded (not RCTs (3); wrong intervention (450)). Full‐text review of the remaining 175 records resulted in 42 studies (163 reports) included, three studies (3 reports) were excluded (Ahren 2013; Bolli 2014; Rosenstock 2014), and nine ongoing studies were identified (ChiCTR2200059717; ChiCTR2400080751; NCT05536804; NCT06182891; NCT06555146; NL‐OMON46820; OTID 2024; REMODEL 2022; RESET1 2024). These nine studies will be assessed in a future update of this review.

See Figure 1.

1.

Flow diagram of the studies identified by the search strategy. *The ongoing studies will be reassessed in the next version of this review.

Included studies

Forty‐two studies (48,148 participants) were included. Selvarajah 2024a and Selvarajah 2024b are related to the same study, but they reported separately to extract outcome data on two different GLP‐1 receptor agonists.

See Characteristics of included studies.

Study design, setting and characteristics

• Three studies (van der Aart‐van 2023; von Scholten 2017; Wajdlich 2024) had a cross‐over design, and the remaining studies were parallel‐group RCTs.

• One study (GetGoal studies 2017 ‐ pooled) included five RCTs, but analyses on subgroup patients with CKD and diabetes were pooled.

• One study has been split in two (Selvarajah 2024a; Selvarajah 2024b) to extract outcome data from two different GLP‐1 receptor agonists.

• Studies were conducted from 2013 to 2024 in China (Liu 2022; Ma 2024; Wang 2020a; Zhang 2012a; Zhou 2019a), Denmark (Idorn 2013; SEMPA 2023; Sivalingam 2024; von Scholten 2017), Ireland (Neff 2016), Italy (Tuttolomondo 2021), Poland (Wajdlich 2024), the Netherlands (van der Aart‐van 2023), and the USA (GRADE 2022). The remaining 28 studies were performed in multinational settings.

• Study follow‐up ranged from one day to five years, with a median of 26 weeks.

• Thirty‐two studies received funding from pharmaceutical companies (AMPLITUDE‐O 2021; AWARD‐7 2017; AWARD‐PEDS 2022; Dekkers 2021; DUAL I 2014; DUAL II 2014; DUAL III 2017; DUAL IV 2014; DUAL V 2016; ELIXA 2018; EXSCEL 2017; FLOW 2023; GetGoal studies 2017 ‐ pooled; GetGoal‐O 2015; Harmony Outcomes 2018; Idorn 2013; LEADER 2017; Leiter 2014; LIRA‐PRIME 2019; LIRA‐RENAL 2016; Muskiet 2019; PIONEER 5 2019; PIONEER 6 2019; REWIND 2019; Selvarajah 2024a; SOUL 2023; SUSTAIN‐6 2016; SUSTAIN 8 2019; Tuttolomondo 2021; van der Aart‐van 2023; von Scholten 2017; Wang 2020a).

• One study (Neff 2016) was only available as a conference proceedings abstract.

Study participants

• Six studies (Dekkers 2021; Ma 2024; Muskiet 2019; SUSTAIN 8 2019; Zhang 2012a; Zhou 2019a) were conducted in people with CKD stages 1‐2, 11 studies (AWARD‐7 2017; DUAL I 2014; DUAL II 2014; DUAL III 2017; DUAL IV 2014; DUAL V 2016; ELIXA 2018; GetGoal‐O 2015; LIRA‐RENAL 2016; Parker 2022; PIONEER 5 2019) in people with CKD stages 3‐5, one study (Idorn 2013) in people on dialysis (both HD and PD), and the remaining studies included people with both CKD stages 1‐2 and stages 3‐5.

• All studies were conducted on people with type 2 diabetes, and no studies were carried out on children.

• The study sample size varied from 7 to 14,691 participants, with a median of 92 participants.

• The mean study age ranged from 51 years to 71 years, with a median of 66 years.

• Comorbidities were rarely reported. Four studies (AWARD‐PEDS 2022; Idorn 2013; Leiter 2014; Wang 2020a) included people with and without cardiovascular comorbidities at baseline, while five studies (Idorn 2013; Leiter 2014; LIRA‐RENAL 2016; Ma 2024; Wang 2020a) were performed on people with and without hypertension. Five studies reported pre‐specified HbA1c targets of < 6.5% (Leiter 2014), < 7% (GRADE 2022; PIONEER 5 2019), between 6.5% and 7% (AWARD‐PEDS 2022), and both < 7% and < 7.5% (LIRA‐RENAL 2016), respectively, but no information was provided for the remaining studies.

Interventions

GLP‐1 receptor agonists versus placebo

Twenty‐one studies compared GLP‐1 receptor agonists to placebo.

• Albiglutide: one study (Harmony Outcomes 2018)

• Cotadutide: one study (Parker 2022)

• Dulaglutide: three studies (AWARD‐PEDS 2022; DUAL IV 2014; REWIND 2019)

• Efpeglenatide administered at different doses: one study (AMPLITUDE‐O 2021)

• Exenatide: one study (EXSCEL 2017)

• Liraglutide: seven studies (Dekkers 2021; Idorn 2013; LEADER 2017; LIRA‐RENAL 2016; Ma 2024; von Scholten 2017; Wajdlich 2024)

• Lixisenatide: three studies (GetGoal studies 2017 ‐ pooled^#; GetGoal‐O 2015; ELIXA 2018)

• Semaglutide: seven studies (FLOW 2023; PIONEER 5 2019; PIONEER 6 2019; Selvarajah 2024a*; SEMPA 2023; SOUL 2023; SUSTAIN‐6 2016)

• Cotadutide administered at different doses: one study (Selvarajah 2024b*).

^#GetGoal studies 2017 ‐ pooled included the pooled analyses of five RCTs.

*Selvarajah 2024a and Selvarajah 2024b are related to the same study but reported separately to extract outcome data on two different GLP‐1 receptor agonists.

GLP‐1 receptor agonists versus standard care

Five studies compared GLP‐1 receptor agonists to standard care.

• Liraglutide versus oral antidiabetic drugs: one study (LIRA‐PRIME 2019)

• Liraglutide versus standard diabetes care including renin‐angiotensin‐aldosterone system inhibitors or antagonists: one study (Neff 2016)

• Liraglutide versus no treatment: one study (Liu 2022)

• Exenatide to metformin plus olmesartan (Zhou 2019a)

• Dulaglutide versus no treatment: one study (Tuttolomondo 2021).

GLP‐1 receptor agonists versus DPP‐4 inhibitors

Two studies compared GLP‐1 receptor agonists to DPP‐4 inhibitors.

• Albiglutide versus sitagliptin: one study (Leiter 2014)

• Liraglutide versus sitagliptin: one study (GRADE 2022).

GLP‐1 receptor agonists versus insulin

Eight studies compared GLP‐1 receptor agonists to insulin.

• Dulaglutide administered at different doses versus insulin glargine: one study (AWARD‐7 2017)

• Exenatide versus insulin glargine: one study (Muskiet 2019)

• Exenatide versus insulin lispro: one study (Wang 2020a)

• Liraglutide plus insulin degludec versus insulin degludec: three studies (DUAL I 2014; DUAL II 2014; DUAL V 2016)

• Liraglutide versus insulin degludec: one study (DUAL I 2014)

• Liraglutide versus insulin degludec: one study (DUAL IV 2014)

• Liraglutide versus insulin glargine: one study (GRADE 2022).

GLP‐1 receptor agonists versus SGLT‐2 inhibitors

Four studies compared GLP‐1 receptor agonists to SGLT‐2 inhibitors.

• Semaglutide versus canagliflozin: one study (SUSTAIN 8 2019)

• Semaglutide versus empagliflozin: one study (SEMPA 2023)

• Exenatide versus dapagliflozin: one study (van der Aart‐van 2023)

• Liraglutide versus dapaglifozin: one study (Ma 2024).

GLP‐1 receptor agonist versus another GLP‐1 receptor agonist

• Semaglutide versus different doses of cotadutide: one study (Selvarajah 2024a; Selvarajah 2024b).

Selvarajah 2024a and Selvarajah 2024b are related to the same study but reported separately to extract outcome data on two different GLP‐1 receptor agonists.

GLP‐1 receptor agonists versus sulfonylurea

Two studies compared GLP‐1 receptor agonists to sulfonylurea.

• Exenatide versus glimepiride: one study (Zhang 2012a)

• Liraglutide versus glimepiride: one study (GRADE 2022).

GLP‐1 receptor agonists plus insulin versus GLP‐1 receptor agonists alone

Two studies compared GLP‐1 receptor agonists plus insulin to GLP‐1 receptor agonists alone.

• Liraglutide plus insulin degludec to liraglutide alone: two studies (DUAL I 2014; DUAL III 2017).

GLP‐1 receptor agonists versus GLP‐1 receptor agonists plus SGLT‐2 inhibitors

Two studies compared GLP‐1 receptor agonists to GLP‐1 receptor agonists plus SGLT‐2 inhibitors.

• Exenatide versus exenatide plus dapagliflozin: one study (van der Aart‐van 2023)

• Semaglutide versus semaglutide plus empagliflozin: one study (SEMPA 2023).

GLP‐1 receptor agonists plus SGLT‐2 inhibitors versus SGLT‐2 inhibitors

One study compared GLP‐1 receptor agonists plus SGLT‐2 inhibitors to SGLT‐2 inhibitors.

• Exenatide plus dapagliflozin versus dapagliflozin: one study (van der Aart‐van 2023).

Summary of outcomes

• All‐cause death: 14 studies (AWARD‐7 2017; EXSCEL 2017; FLOW 2023; Idorn 2013; LEADER 2017; Leiter 2014; LIRA‐RENAL 2016; Liu 2022; Ma 2024; Neff 2016; PIONEER 5 2019; Selvarajah 2024a*; Selvarajah 2024b*; Sivalingam 2024; Zhou 2019a).

• Cardiovascular death: 13 studies (AWARD‐7 2017; EXSCEL 2017; FLOW 2023; Idorn 2013; LEADER 2017; Leiter 2014; LIRA‐RENAL 2016; Liu 2022; Ma 2024; Neff 2016; PIONEER 5 2019; Selvarajah 2024a*; Selvarajah 2024b*; Zhou 2019a).

• 3‐point MACE: seven studies (AMPLITUDE‐O 2021; FLOW 2023; EXSCEL 2017; Harmony Outcomes 2018; LEADER 2017; PIONEER 6 2019; SUSTAIN‐6 2016)

• 4‐point MACE: two studies (ELIXA 2018; LEADER 2017)

• Kidney failure: four studies (AWARD‐7 2017; FLOW 2023; LIRA‐RENAL 2016; PIONEER 5 2019)

• Composite kidney outcomes: three studies (EXSCEL 2017; LEADER 2017; REWIND 2019)

• severe hypoglycaemia (defined as hypoglycaemia requiring third‐party assistance): seven studies (AWARD‐7 2017; FLOW 2023; LEADER 2017; Leiter 2014; LIRA‐RENAL 2016; PIONEER 5 2019; Wang 2020a).

• Adverse events were inconsistently reported, and no studies reported outcome data on fatigue, life participation, amputation or fracture.

Note: *Selvarajah 2024a and Selvarajah 2024b were related to the same study.

Data were reported either as RR or HR.

Excluded studies

Three studies were excluded because they enrolled the wrong population (Ahren 2013; Bolli 2014; Rosenstock 2014).

Ongoing studies

Nine studies are ongoing and will be assessed in a future update of this review.

• Semaglutide versus insulin glargine (ChiCTR2200059717)

• Semaglutide versus insulin detemir (ChiCTR2400080751)

• Tirepatide versus placebo (NCT05536804)

• Dulaglutide versus other hypoglycaemic agents (not GLP‐1) (NCT06182891

• Semaglutide versus placebo (NCT06555146; REMODEL 2022; RESET1 2024)

• Exenatide versus dapagliflozin versus exenatide plus dapagliflozin (NL‐OMON46820)

• Liraglutide versus dapagliflozin versus liraglutide plus dapagliflozin versus liraglutide plus dapagliflozin plus intensive lifestyle approach (OTID 2024)

Risk of bias in included studies

In the majority of domains, the risks of bias in the included studies were low or unclear.

• Bias arising from the randomisation process was at low risk in 22 studies.

• Bias due to deviations from intended interventions was at low risk in 19 studies.

• Bias due to missing outcome data was at low risk in seven studies.

• Bias in the measurement of the outcome was at low risk in 16 studies.

• Bias in selection of the reported result was at low risk in 18 studies.

GLP‐1 receptor agonists versus placebo

Overall, the risk of bias for all domains was considered to be at low risk of bias except for Idorn 2013, which was assessed at high risk of bias due to missing outcome data for death (all‐cause and cardiovascular).

GLP‐1 receptor agonists versus standard care

The overall risks of bias for all‐cause and cardiovascular death were assessed as unclear or at high risk due to concerns about deviations from intended interventions or missing outcome data, probably due to low number of events, participants and short follow‐up duration to assess critical outcomes.

GLP‐1 receptor agonists versus DDP‐4 inhibitors

Leiter 2014 was judged to be at high risk of bias for death (all‐cause and cardiovascular) and severe hypoglycaemia due to missing outcome data, whilst other domains seemed to be free from other sources of bias.

GLP‐1 receptor agonists versus insulin

AWARD‐7 2017 was judged to be at low or unclear risk of bias for all for death (all‐cause and cardiovascular), kidney failure and severe hypoglycaemia, although the primary outcomes have been assessed at high risk of bias due to deviations from intended interventions.

GLP‐1 receptor agonists versus SGLT2 inhibitors

Ma 2024 was judged to be at low risk of bias for all‐cause and cardiovascular death, although blinding was not clearly reported.

GLP‐1 receptor agonist versus another GLP‐1 receptor agonist

For Selvarajah 2024a, the risk of bias for all‐cause and cardiovascular death was judged to be unclear due to uncertainty in allocation concealment.

Effects of interventions

See: Table 1

GLP‐1 receptor agonists versus placebo

Eighteen studies (AMPLITUDE‐O 2021; ELIXA 2018; EXSCEL 2017; FLOW 2023; GetGoal studies 2017 ‐ pooled; GetGoal‐O 2015; Harmony Outcomes 2018; Idorn 2013; LEADER 2017; LIRA‐PRIME 2019; LIRA‐RENAL 2016; Ma 2024; PIONEER 5 2019; PIONEER 6 2019; REWIND 2019; Selvarajah 2024a; Selvarajah 2024b; Sivalingam 2024; SUSTAIN‐6 2016) compared GLP‐1 receptor agonists to placebo, during a median follow‐up of 26 weeks.

See Table 1.

Death

• GLP‐1 receptor agonists probably reduced the risk of all‐cause death compared to placebo (RR 0.85, 95% CI 0.74 to 0.98; I² = 23%; 8 studies, 17,861 participants; moderate‐certainty evidence; Analysis 1.1) in people with all CKD stages and type 2 diabetes. Similar results were seen for the subgroups CKD stages 1‐2 (Analysis 1.1.1) and CKD stages 3‐5 (Analysis 1.1.2). The results of the meta‐analysis of HR in trials are shown in Table 2 (HR 0.84, 95% CI 0.71 to 0.98; I² = 31%; 4 studies; moderate‐certainty evidence; Analysis 1.2).

• GLP‐1 receptor agonists may have little or no effect on the risk of cardiovascular death compared to placebo (RR 0.84, 95% CI 0.68 to 1.05; I² = 42%; 7 studies, 17,801 participants; low‐certainty evidence; Analysis 1.3) in people with all CKD stages and type 2 diabetes. Similar results were seen for the subgroups CKD stages 1‐2 (Analysis 1.3.1) and CKD stages 3‐5 (Analysis 1.3.2). The results of the meta‐analysis of HR in trials are shown in Table 2 (HR 0.90, 95% CI 0.73 to 1.12; I² = 54%; 3 studies; low‐certainty evidence; Analysis 1.4).

1.1. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 1: All‐cause death

1. Hazard ratio (HR) for studies comparing GLP‐1 receptor agonists with placebo at time‐to‐event.

Outcome	Number of studies	Participants	HR (95% CI)	I2	GRADE
All‐cause death (Analysis 1.2)	4 RCTs	Number not clearly reported (people with all CKD stages and type 2 diabetes)	0.84 (0.71 to 0.98)	31%	Moderate certainty
Cardiovascular death (Analysis 1.4)	3 RCTs	Number not clearly reported (people with all CKD stages and type 2 diabetes)	0.90 (0.73 to 1.12)	54%	Low certainty
MACE 3‐point (Analysis 1.6)	7 RCTs	Number not clearly reported (people with all CKD stages and type 2 diabetes)	0.80 (0.73 to 0.89)	53%	Moderate certainty
MACE 4‐point (Analysis 1.8)	2 RCTs	Number not clearly reported (people with all CKD stages and type 2 diabetes)	0.86 (0.74 to 1.01)	58%	Moderate certainty
Kidney failure (Analysis 1.10)	1 RCT	Number not clearly reported (people with all CKD stages and type 2 diabetes)	0.84 (0.63 to 1.12)	‐	Moderate certainty
Composite kidney outcome (Analysis 1.12)	2 RCTs	Number not clearly reported (people with all CKD stages and type 2 diabetes)	0.78 (0.62 to 0.97)	60%	Moderate certainty
Severe hypoglycaemia (Analysis 1.14)	1 RCT	Number not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	0.63 (0.43 to 0.91)	‐	High certainty
Nonfatal myocardial infarction (Analysis 1.16)	2 RCTs	Number not clearly reported (people with all CKD stages and type 2 diabetes)	0.82 (0.63 to 1.06)	45%	Moderate certainty
All myocardial infarction (Analysis 1.20)	1 RCT	Number not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	0.73 (0.55 to 0.98)	‐	High certainty
Nonfatal stroke (Analysis 1.22)	2 RCTs	Number not clearly reported (people with all CKD stages and type 2 diabetes)	0.93 (0.59 to 1.47)	74%	Moderate certainty
All strokes (Analysis 1.24)	1 RCT	Number not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	0.53 (0.36 to 0.79)	‐	Moderate certainty
Hospitalisation due to heart failure (Analysis 1.29)	1 RCT	Number not clearly reported (people with all CKD stages and type 2 diabetes)	0.84 (0.62 to 1.14)	54%	Low certainty
Withdrawal due to adverse events (Analysis 1.40)	2 RCTs	Number not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	1.62 (0.24 to 11.11)	92%	Low certainty
Acute kidney injury (Analysis 1.44)	2 RCTs	2,482 participants (people with CKD stages 3‐5 and type 2 diabetes)	0.83 (0.62 to 1.11)	0%	Moderate certainty
Coronary revascularisation (Analysis 1.50)	1 RCT	Number not clearly reported (people with all CKD stages and type 2 diabetes)	0.82 (0.69 to 0.98)	0%	High certainty

CI: confidence interval; CKD: chronic kidney disease; GRADE: Grading of Recommendations Assessment, Development and Evaluation; HR: hazard ratio; MACE: major cardiovascular events; RCT: randomised controlled trial

1.2. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 2: All‐cause death (Hazard ratio)

1.3. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 3: Cardiovascular death

1.4. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 4: Cardiovascular death (Hazard ratio)

Major adverse cardiovascular events

3‐point MACE was defined in all included studies as the composite cardiovascular outcome, including cardiovascular death, nonfatal MI, or nonfatal stroke.

• GLP‐1 receptor agonists probably decreased 3‐point MACE compared to placebo (RR 0.84, 95% CI 0.73 to 0.98; I² = 65%; 4 studies, 19,825 participants; moderate‐certainty evidence; Analysis 1.5) in people with all CKD stages and type 2 diabetes. Similar results were seen for the subgroups CKD stages 1‐2 (Analysis 1.5.1) and CKD stages 3‐5 (Analysis 1.5.2). The meta‐analysis based on HR values has been reported in Table 2 (HR 0.80, 95% CI 0.73 to 0.89; I² = 53%; 7 studies; moderate‐certainty evidence; Analysis 1.6).

1.5. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 5: 3‐point MACE

1.6. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 6: 3‐point MACE (Hazard ratio)

4‐point MACE was defined as the composite cardiovascular outcome including cardiovascular death, nonfatal MI, nonfatal stroke, plus coronary revascularisation or hospitalisation for unstable angina pectoris or heart failure.

• GLP‐1 receptor agonists probably decreased 4‐point MACE compared to placebo (RR 0.77, 95% CI 0.67 to 0.89; 1 study, 2,158 participants; moderate‐certainty evidence; Analysis 1.7) in people with CKD stages 3‐5 and type 2 diabetes. The meta‐analysis based on HR values has been reported in Table 2 (HR 0.86, 95% CI 0.74 to 1.01; I² = 58%; 2 studies; moderate‐certainty evidence; Analysis 1.8).

1.7. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 7: 4‐point MACE

1.8. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 8: 4‐point MACE (Hazard ratio)

Kidney failure

GLP‐1 receptor agonists probably had little or no effect on kidney failure, defined as the need for KRT, compared to placebo (RR 0.86, 95% CI 0.66 to 1.13; I² = 0%; 3 studies, 4,134 participants; moderate‐certainty evidence; Analysis 1.9) in people with all CKD stages and type 2 diabetes. Similar results were seen for the subgroup CKD stages 3‐5 (Analysis 1.9.2).

1.9. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 9: Kidney failure

The meta‐analysis based on HR values has been reported in Table 2 (HR 0.84, 95% CI 0.63 to 1.12; 1 study; moderate‐certainty evidence; Analysis 1.10).

1.10. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 10: Kidney failure (Hazard ratio)

Composite kidney outcomes

LEADER 2017 defined the composite kidney outcome as doubling of SCr or KRT. REWIND 2019 defined the composite kidney outcome as the development of severely increased albuminuria (development of UACR >33.9 mg/mmol in people with a lower baseline value). EXSCEL 2017 defined the composite kidney outcome as 40% eGFR decline, KRT, or kidney death.

GLP‐1 receptor agonists probably had little or no effect on kidney composite outcomes compared to placebo (RR 0.89, 95% CI 0.78 to 1.02; I² = 0%; 2 studies, 16,849 participants; moderate‐certainty evidence; Analysis 1.11) in people with all CKD stages and type 2 diabetes. Similar results were seen for the subgroups CKD stages 1‐2 (Analysis 1.11.1) and CKD stages 3‐5 (Analysis 1.11.2).

1.11. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 11: Composite kidney outcome

The meta‐analysis based on HR values has been reported in Table 2 (HR 0.78, 95% CI 0.62 to 0.97; I² = 60%; 2 studies; moderate‐certainty evidence; Analysis 1.12)

1.12. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 12: Composite kidney outcome (Hazard ratio)

Severe hypoglycaemia

GLP‐1 receptor agonists may have little or no effect on the risk of severe hypoglycaemia compared to placebo (RR 0.82, 95% CI 0.54 to 1.25; I² = 44%; 4 studies, 6,292 participants; low‐certainty evidence; Analysis 1.13) in people with all CKD stages and type 2 diabetes. Similar results were seen for the subgroup CKD stages 3‐5 (Analysis 1.13.1).

1.13. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 13: Severe hypoglycaemia

The meta‐analysis based on HR values has been reported in Table 2 (HR 0.63, 95% CI 0.43 to 0.91; 1 study; high‐certainty evidence; Analysis 1.14).

1.14. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 14: Severe hypoglycaemia (Hazard ratio)

Myocardial infarction

• GLP‐1 receptor agonists probably had little to no effect on nonfatal MI compared to placebo (RR 0.76, 95% CI 0.57 to 1.01; 1 study, 2,158 participants; moderate‐certainty evidence; Analysis 1.15) in people with CKD stages 3‐5 and type 2 diabetes. The results of the meta‐analysis of HR in trials are shown in Table 2 (HR 0.82, 95% CI 0.63 to 1.06; I² = 45%; 2 studies; moderate‐certainty evidence; Analysis 1.16). Similar results were seen for the subgroups CKD stages 1‐2 (Analysis 1.16.1) and CKD stages 3‐5 ( Analysis 1.16.2).

• GLP‐1 receptor agonists had uncertain effects on fatal MI compared to placebo (RR 2.96, 95% CI 0.12 to 72.21; 1 study, 324 participants; very‐low‐certainty evidence; Analysis 1.17) in people with CKD stages 3‐5 and type 2 diabetes.

• GLP‐1 receptor agonists probably have little or no effect on MI (undefined) (RR 0.76, 95% CI 0.58 to 1.00; I² = 0%; 2 studies, 2,435 participants; moderate‐certainty evidence; Analysis 1.18) in people with CKD stages 3‐5 and type 2 diabetes.

• GLP‐1 receptor agonists may have little or no effect on fatal or nonfatal MI compared to placebo (RR 0.93, 95% CI 0.81 to 1.07; I² = 34%; 2 studies, 16,849 participants; low‐certainty evidence; Analysis 1.19) in people with all CKD stages and type 2 diabetes. Similar results were seen for the subgroups CKD stages 1‐2 (Analysis 1.19.1) and CKD stages 3‐5 (Analysis 1.19.2). HR values have been reported in Table 2 (HR 0.73, 95% CI 0.55 to 0.98; 1 study; high‐certainty evidence; Analysis 1.20).

1.15. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 15: Nonfatal myocardial infarction

1.16. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 16: Nonfatal myocardial infarction (Hazard ratio)

1.17. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 17: Fatal myocardial infarction

1.18. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 18: Myocardial infarction (undefined)

1.19. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 19: Fatal or nonfatal myocardial infarction

1.20. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 20: All myocardial infarction (Hazard ratio)

Stroke

• GLP‐1 receptor agonists probably decreased nonfatal stroke compared to placebo (RR 0.53, 95% CI 0.34 to 0.81; 1 study, 2,158 participants; moderate‐certainty evidence; Analysis 1.21) in people with CKD stages 3‐5 and type 2 diabetes. The meta‐analysis based on HR values has been reported in Table 2 (HR 0.93, 95% CI 0.59 to 1.47; I² = 74%; 2 studies; moderate‐certainty evidence; Analysis 1.22). Similar results are seen for the subgroups CKD stages 1‐2 (Analysis 1.22.1) and CKD stages 3‐5 (Analysis 1.22.2).

• GLP‐1 receptor agonists may have little or no effect on combined fatal or nonfatal stroke compared to placebo (RR 0.79, 95% CI 0.55 to 1.14; I² = 66%; 3 studies, 17,173 participants; low‐certainty evidence; Analysis 1.23) in people with all CKD stages and type 2 diabetes. Similar results are seen for the subgroups CKD stages 1‐2 (Analysis 1.23.1) and CKD stages 3‐5 (Analysis 1.23.2). The meta‐analysis based on HR values has been reported in Table 2 (HR 0.53, 95% CI 0.36 to 0.79; 1 study; moderate‐certainty evidence; Analysis 1.24).

• GLP‐1 receptor agonists had uncertain effects on ischaemic stroke compared to placebo (RR 0.34, 95% CI 0.01 to 8.27; 1 study, 329 participants; very‐low‐certainty evidence; Analysis 1.25) in people with CKD stages 3‐5 and type 2 diabetes.

• GLP‐1 receptor agonists had uncertain effects on haemorrhagic stroke compared to placebo (RR 2.96, 95% CI 0.12 to 72.21; 1 study, 324 participants; very‐low‐certainty evidence; Analysis 1.26) in people with CKD stages 3‐5 and type 2 diabetes.

1.21. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 21: Nonfatal stroke

1.22. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 22: Nonfatal stroke (Hazard ratio)

1.23. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 23: Fatal or nonfatal stroke

1.24. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 24: Fatal or nonfatal stroke (Hazard ratio)

1.25. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 25: Ischaemic stroke

1.26. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 26: Haemorrhagic stroke

Heart failure

GLP‐1 receptor agonists had uncertain effects on heart failure compared to placebo (RR 2.94, 95% CI 0.12 to 71.46; 1 study, 277 participants; very‐low‐certainty evidence; Analysis 1.27) in people with CKD stages 3‐5 and type 2 diabetes.

1.27. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 27: Heart failure

Hospitalisation due to heart failure

GLP‐1 receptor agonists had little or no effect on hospitalisation due to heart failure compared to placebo (RR 0.88, 95% CI 0.72 to 1.07; I² = 32%; 3 studies, 17,173 participants; high‐certainty evidence; Analysis 1.28) in people with all CKD stages and type 2 diabetes. Similar results were seen in the subgroups for CKD stages 1‐2 (Analysis 1.28.1) and CKD stages 3‐5 (Analysis 1.28.2).

1.28. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 28: Hospitalisation due to heart failure

The meta‐analysis based on HR values has been reported in Table 2 (HR 0.84, 95% CI 0.62 to 1.14; I² = 54%; 1 study; low‐certainty evidence; Analysis 1.29).

1.29. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 29: Hospitalisation due to heart failure (Hazard ratio)

Change in glomerular filtration rate

GLP‐1 receptor agonists had uncertain effects on eGFR compared to placebo (MD 0.99%, 95% CI ‐1.07 to 3.05; 1 study, 584 participants; very‐low‐certainty evidence; Analysis 1.30) in people with CKD stages 3‐5 and type 2 diabetes.

1.30. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 30: Change in eGFR [%]

Urine albumin‐creatinine ratio

GLP‐1 receptor agonists have uncertain effects on UACR compared to placebo (RR 2.94, 95% CI 0.12 to 71.49; 1 study, 277 participants; very‐low‐certainty evidence; Analysis 1.31) in people with CKD stages 3‐5 and type 2 diabetes.

1.31. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 31: Increased urinary albumin‐creatinine ratio

Body weight

• GLP‐1 receptor agonists may decrease body weight compared to placebo (MD ‐5.34 kg, 95% CI ‐8.72 to ‐1.96; 1 study, 80 participants; low‐certainty evidence; Analysis 1.32) in people with CKD stages 1‐2.

• GLP‐1 receptor agonists may decrease the change in body weight compared to placebo (MD ‐2.50 kg, 95% CI ‐3.33 to ‐1.67; 1 study, 324 participants; low‐certainty evidence; Analysis 1.33) (MD ‐0.86 kg, 95% CI ‐1.46 to ‐0.26; I² = 51%; 3 studies; low‐certainty evidence; Analysis 1.34) in people with all CKD stages and type 2 diabetes. Similar results were seen in the subgroups for CKD stages 1‐2 (Analysis 1.34.1), CKD stages 3‐5 (Analysis 1.34.2), and people on dialysis (Analysis 1.34.3).

1.32. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 32: Body weight [kg]

1.33. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 33: Change in body weight [kg]

1.34. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 34: Change in body weight [kg]

Body mass index

• GLP‐1 receptor agonists may decrease BMI compared to placebo (MD ‐2.47 kg/m², 95% CI ‐3.35 to ‐1.59; 1 study, 80 participants; low‐certainty evidence; Analysis 1.35) in people with CKD stages 1‐2.

• GLP‐1 receptor agonists may decrease the change in BMI compared to placebo (MD ‐0.71 kg/m², 95% CI ‐1.10 to ‐0.32; I² = 74%; 2 studies, 572 participants; low‐certainty evidence; Analysis 1.36) (MD ‐0.51 kg/m², 95% CI ‐0.83 to ‐0.19; 1 study; low‐certainty evidence; Analysis 1.37) in people with CKD stages 3‐5 and type 2 diabetes.

1.35. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 35: BMI [kg/m²]

1.36. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 36: Change in BMI [kg/m²]

1.37. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 37: Change in BMI [kg/m²]

Serious adverse events

GLP‐1 receptor agonists probably have little or no effect on serious adverse events compared to placebo (RR 1.01, 95% CI 0.65 to 1.59; I² = 0%; 2 studies, 601 participants; moderate‐certainty evidence; Analysis 1.38) in people with CKD stages 3‐5 and type 2 diabetes.

1.38. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 38: Serious adverse events

Withdrawal due to adverse events

GLP‐1 receptor agonists have uncertain effects on withdrawal due to adverse events compared to placebo (RR 1.34, 95% CI 0.83 to 2.16; I² = 77%; 7 studies, 6,678 participants; very‐low‐certainty evidence; Analysis 1.39) in all CKD stages and type 2 diabetes. Similar results were seen in the subgroup for CKD stages 3‐5 (Analysis 1.39.2).

1.39. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 39: Withdrawal due to adverse events

The meta‐analysis based on HR values has been reported in Table 2 (HR 1.62, 95% CI 0.24 to 11.11; I² = 92%; 2 studies, low‐certainty evidence; Analysis 1.40).

1.40. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 40: Withdrawal due to adverse events (Hazard ratio)

Hypoglycaemia

GLP‐1 receptor agonists have uncertain effects on hypoglycaemia compared to placebo (RR 1.10, 95% CI 0.42 to 2.84; I² = 49%; 3 studies, 661 participants; very‐low‐certainty evidence; Analysis 1.41) in people with all CKD stages and type 2 diabetes.

1.41. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 41: Hypoglycaemia

Lactic acidosis

The effects of GLP‐1 receptor agonists on lactic acidosis compared to placebo is uncertain (PIONEER 5 2019, 324 participants; Analysis 1.42), as no events were reported.

1.42. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 42: Lactic acidosis

Acute kidney injury

Compared to placebo, the risk of AKI was uncertain with GLP‐1 receptor agonists (RR 0.88, 95% CI 0.11 to 6.75; I² = 11%; 2 studies, 571 participants; very‐low‐certainty evidence; Analysis 1.43) in people with all CKD stages and type 2 diabetes.

1.43. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 43: Acute kidney injury

The meta‐analysis based on HR values has been reported in Table 2 (HR 0.83, 95% CI 0.62 to 1.11; I² = 0%; 2 studies, 2,482 participants; low‐certainty evidence; Analysis 1.44).

1.44. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 44: Acute kidney injury (Hazard ratio)

Diabetic ketoacidosis

GLP‐1 receptor agonists had uncertain effects on the risk of diabetic ketoacidosis compared to placebo (RR 2.94, 95% CI 0.12 to 71.46; 1 study, 277 participants; very‐low‐certainty evidence; Analysis 1.45) in people with CKD stages 3‐5 and type 2 diabetes.

1.45. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 45: Diabetic ketoacidosis

HbA1c (at the end of treatment or change from baseline)

• GLP‐1 receptor agonists had uncertain effects on HbA1c compared to placebo (MD ‐0.64%, 95% CI ‐1.91 to 0.64; I² = 95%; 3 studies, 373 participants; very‐low‐certainty evidence; Analysis 1.46) in people with all CKD stages and type 2 diabetes. In a subgroup analysis, Ma 2024 reported HbA1c was lower with GLP‐1 receptor agonists compared to placebo in people with CKD stages 1‐2 (MD ‐2.01%, 95% CI ‐2.58 to ‐1.44; 40 participants; Analysis 1.46.1).

• GLP‐1 receptor agonists probably increased change in HbA1c compared to placebo (MD ‐0.72%, 95% CI ‐0.89 to ‐0.55; I² = 0%; 2 studies, 587 participants; moderate‐certainty evidence; Analysis 1.47) in people with CKD stages 3‐5 and type 2 diabetes.

• GLP‐1 receptor agonists may lower HbA1c compared to placebo (MD ‐0.64%, 95% CI ‐0.81 to ‐0.48; I² = 55%; 5 studies; low‐certainty evidence; Analysis 1.48) in people with all CKD stages and type 2 diabetes. Similar results were seen in the subgroups for CKD stages 1‐2 (Analysis 1.48.1) and CKD stages 3‐5 (Analysis 1.48.2), but not in people on dialysis (Analysis 1.48.3).

1.46. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 46: HbA1c [%]

1.47. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 47: Change in HbA1c [%]

1.48. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 48: Change in HbA1c [%]

Coronary revascularisation

GLP‐1 receptor agonists probably have little or no effect on coronary revascularisation compared to placebo (RR 0.84, 95% CI 0.63 to 1.11; 1 study, 2,158 participants; moderate‐certainty evidence; Analysis 1.49) in people with CKD stages 3‐5 and type 2 diabetes.

1.49. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 49: Coronary revascularisation

The meta‐analysis based on HR values has been reported in Table 2 (HR 0.82, 95% CI 0.69 to 0.98; I² = 0%; 1 study; high‐certainty evidence; Analysis 1.50).

1.50. Analysis.

Comparison 1: GLP‐1 receptor agonists versus placebo, Outcome 50: Coronary revascularisation (Hazard ratio)

No other secondary outcomes were reported for this comparison.

GLP‐1 receptor agonists versus standard care

Three studies (Liu 2022; Neff 2016; Zhou 2019a) compared GLP‐1 receptor agonists to standard care in people with all CKD stages and type 2 diabetes during a median follow‐up of 24 weeks. The certainty of the evidence was low to very low.

Death

• GLP‐1 receptor agonists had uncertain effects on all‐cause death compared to standard care alone (RR 0.33, 95% CI 0.02 to 7.32; 3 studies, 184 participants; very‐low‐certainty evidence; Analysis 2.1) in people with all CKD stages and type 2 diabetes.

• GLP‐1 receptor agonists had uncertain effects on cardiovascular death compared to standard care alone (RR 0.33, 95% CI 0.02 to 7.32; 3 studies, 184 participants; very‐low‐certainty evidence; Analysis 2.2) in people with all CKD stages and type 2 diabetes.

2.1. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 1: All‐cause death

2.2. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 2: Cardiovascular death

Urine albumin‐creatinine ratio

Zhou 2019a reported GLP‐1 receptor agonists may reduce UACR compared to standard care alone (MD ‐17.27 mg/g, 95% CI ‐25.21 to ‐9.33; 1 study, 80 participants; low‐certainty evidence; Analysis 2.3) in people with CKD stages 1‐2 and type 2 diabetes.

2.3. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 3: Urinary albumin‐creatinine ratio [mg/g]

Microalbuminuria

Zhou 2019a reported GLP‐1 receptor agonists may decrease moderately increased albuminuria compared to standard care alone (MD ‐9.25 mg/L, 95% CI ‐14.57 to ‐3.93; 1 study, 80 participants; low‐certainty evidence; Analysis 2.4) in people with CKD stages 1‐2 and type 2 diabetes.

2.4. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 4: Microalbuminuria [mg/L]

Urinary albumin excretion rate

GLP‐1 receptor agonists may decrease urinary albumin excretion rate compared to standard care alone (MD ‐40.79 µg/min, 95% CI ‐52.77 to ‐28.80; 2 studies, 100 participants; low‐certainty evidence; Analysis 2.5) in people with all CKD stages and type 2 diabetes.

2.5. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 5: Urinary albumin excretion rate [µg/min]

Body mass index

Liu 2022 reported GLP‐1 receptor agonists may decrease BMI compared to standard care alone (MD ‐1.61 kg/m², 95% CI ‐3.00 to ‐0.22; 1 study, 84 participants; low‐certainty evidence; Analysis 2.6) in people with all CKD stages and type 2 diabetes.

2.6. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 6: BMI [kg/m²]

Serious adverse events

Neff 2016 reported GLP‐1 receptor agonists had uncertain effects on serious adverse events compared to standard care alone (RR 1.00, 95% CI 0.07 to 13.87; 1 study, 20 participants; very‐low‐certainty evidence; Analysis 2.7) in people with all CKD stages and type 2 diabetes.

2.7. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 7: Serious adverse events

Withdrawal due to adverse events

The effects of GLP‐1 receptor agonists compared to standard care alone on withdrawal due to adverse events were not estimable (Neff 2016; 20 participants; Analysis 2.8) since zero events were reported in people with all CKD stages and type 2 diabetes.

2.8. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 8: Withdrawal due to adverse events

Hypoglycaemia

Neff 2016 reported GLP‐1 receptor agonists had uncertain effects on hypoglycaemia compared to standard care alone (RR 2.00, 95% CI 0.21 to 18.69; 1 study, 20 participants; very‐low‐certainty evidence; Analysis 2.9) in people with all CKD stages and type 2 diabetes.

2.9. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 9: Hypoglycaemia

Acute kidney injury

Neff 2016 reported GLP‐1 receptor agonists had uncertain effects on AKI compared to standard care alone (RR 1.00, 95% CI 0.07 to 13.87; 1 study, 20 participants; very‐low‐certainty evidence; Analysis 2.10) in people with all CKD stages and type 2 diabetes.

2.10. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 10: Acute kidney injury

HbA1c (at the end of treatment or change from baseline)

• GLP‐1 receptor agonists may decrease HbA1c compared to standard care alone (MD ‐0.69%, 95% CI ‐0.88 to ‐0.50; 2 studies, 164 participants; low‐certainty evidence; Analysis 2.11) in people with all CKD stages and type 2 diabetes.

• Zhou 2019a reported GLP‐1 receptor agonists may decrease change in HbA1c compared to standard care alone (MD 0.22%, 95% CI 0.12 to 0.32; 1 study; low‐certainty evidence; Analysis 2.12) in people with CKD stages 1‐2 and type 2 diabetes.

2.11. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 11: HbA1c [%]

2.12. Analysis.

Comparison 2: GLP‐1 receptor agonists versus standard care, Outcome 12: Change in HbA1c [%]

No other primary or secondary outcomes were reported for this comparison.

GLP‐1 receptor agonists versus DPP‐4 inhibitors

Leiter 2014 compared GLP‐1 receptor agonists to DPP‐4 inhibitors in people with all CKD stages and type 2 diabetes at 52 weeks. The certainty of the evidence was low.

Death

• GLP‐1 receptor agonists may make little or no difference to all‐cause death compared to DPP‐4 inhibitors (RR 0.99, 95% CI 0.25 to 3.91; 1 study, 495 participants; low‐certainty evidence; Analysis 3.1) in people with all CKD stages and type 2 diabetes.

• GLP‐1 receptor agonists may make little or no difference to cardiovascular death compared to DPP‐4 inhibitors (RR 1.98, 95% CI 0.18 to 21.65; 1 study, 495 participants; low‐certainty evidence; Analysis 3.2) in people with all CKD stages and type 2 diabetes.

3.1. Analysis.

Comparison 3: GLP‐1 receptor agonists versus DPP‐4 inhibitors, Outcome 1: All‐cause death

3.2. Analysis.

Comparison 3: GLP‐1 receptor agonists versus DPP‐4 inhibitors, Outcome 2: Cardiovascular death

Severe hypoglycaemia

GLP‐1 receptor agonists may make little or no difference to severe hypoglycaemia compared to DPP‐4 inhibitors (RR 0.25, 95% CI 0.03 to 2.19; 1 study, 495 participants; low‐certainty evidence; Analysis 3.3) in people with all CKD stages and type 2 diabetes.

3.3. Analysis.

Comparison 3: GLP‐1 receptor agonists versus DPP‐4 inhibitors, Outcome 3: Severe hypoglycaemia

Fatal stroke

GLP‐1 receptor agonists may make little or no difference to fatal stroke compared to DPP‐4 inhibitors (RR 0.33, 95% CI 0.01 to 8.05; 1 study, 495 participants; low‐certainty evidence; Analysis 3.4) in people with all CKD stages and type 2 diabetes.

3.4. Analysis.

Comparison 3: GLP‐1 receptor agonists versus DPP‐4 inhibitors, Outcome 4: Fatal stroke

Serious adverse events

GLP‐1 receptor agonists may make little or no difference to serious adverse events compared to DPP‐4 inhibitors (RR 0.88, 95% CI 0.56 to 1.37; 1 study, 495 participants; low‐certainty evidence; Analysis 3.5) in people with all CKD stages and type 2 diabetes.

3.5. Analysis.

Comparison 3: GLP‐1 receptor agonists versus DPP‐4 inhibitors, Outcome 5: Serious adverse events

Withdrawal due to adverse events

GLP‐1 receptor agonists may make little or no difference to withdrawal due to adverse events compared to DPP‐4 inhibitors (RR 0.99, 95% CI 0.59 to 1.65; 1 study, 495 participants; low‐certainty evidence; Analysis 3.6) in people with all CKD stages and type 2 diabetes.

3.6. Analysis.

Comparison 3: GLP‐1 receptor agonists versus DPP‐4 inhibitors, Outcome 6: Withdrawal due to adverse events

Hypoglycaemia

Leiter 2014 reported GLP‐1 receptor agonists may increase hypoglycaemia compared to DPP‐4 inhibitors (RR 1.52, 95% CI 1.06 to 2.18; 1 study, 495 participants; low‐certainty evidence; Analysis 3.7) in people with all CKD stages and type 2 diabetes.

3.7. Analysis.

Comparison 3: GLP‐1 receptor agonists versus DPP‐4 inhibitors, Outcome 7: Hypoglycaemia

HbA1c (end of treatment or change from baseline)

• Leiter 2014 reported GLP‐1 receptor agonists may slightly lower HbA1c compared to DPP‐4 inhibitors (MD ‐0.41%, 95% CI ‐0.70 to ‐0.12; 1 study, 478 participants; low‐certainty evidence; Analysis 3.8) in people with all CKD stages and type 2 diabetes.

• Leiter 2014 reported GLP‐1 receptor agonists may slightly decrease change in HbA1c compared to DPP‐4 inhibitors (MD ‐0.27%, 95% CI ‐0.70 to 0.16; 1 study, 231 participants; low‐certainty evidence; Analysis 3.9.2) in people with CKD stages 3‐5 and type 2 diabetes, but not CKD stages 1‐2 (MD ‐0.06%, 95% CI ‐0.27 to 0.15; Analysis 3.9.1)

3.8. Analysis.

Comparison 3: GLP‐1 receptor agonists versus DPP‐4 inhibitors, Outcome 8: HbA1c [%]

3.9. Analysis.

Comparison 3: GLP‐1 receptor agonists versus DPP‐4 inhibitors, Outcome 9: Change in HbA1c [%]

No other primary or secondary outcomes were reported for this comparison.

GLP‐1 receptor agonists versus insulin

Six studies (AWARD‐7 2017; DUAL I 2014; DUAL II 2014; DUAL V 2016; Muskiet 2019; Wang 2020a) compared GLP‐1 receptor agonists to insulin in people with either CKD stages 1‐2 or CKD stages 3‐5 and type 2 diabetes for 26 weeks. The certainty of the evidence was moderate to very low.

Death

• GLP‐1 receptor agonists had uncertain effects on all‐cause death compared to insulin (RR 0.76, 95% CI 0.28 to 2.11; 1 study, 576 participants; very‐low‐certainty evidence; Analysis 4.1) in people with CKD stages 3‐5 and type 2 diabetes. HR values have been reported in Table 3 (HR 1.02, 95% CI 0.39 to 2.67; 1 study; very‐low‐certainty evidence; Analysis 4.2).

• GLP‐1 receptor agonists had uncertain effects on cardiovascular death compared to insulin (RR 1.02, 95% CI 0.31 to 3.33; 1 study, 576 participants; very‐low‐certainty evidence; Analysis 4.3) in people with CKD stages 3‐5 and type 2 diabetes. HR values have been reported in Table 3 (HR 1.02, 95% CI 0.31 to 3.36; 1 study; very‐low‐certainty evidence; Analysis 4.4).

4.1. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 1: All‐cause death

2. Hazard ratio (HR) for studies comparing GLP‐1 receptor agonists with insulin at time‐to‐event.

Outcome	Number of studies	Participants	HR (95% CI)	I2	GRADE
All‐cause death (Analysis 4.2)	1 RCT	Number of participants was not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	1.02 (0.39 to 2.67)	‐	Low certainty
Cardiovascular death (Analysis 4.4)	1 RCT	Number of participants was not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	1.02 (0.31 to 3.36)	‐	Low certainty

CI: confidence interval; CKD: chronic kidney disease; GRADE: Grading of Recommendations Assessment, Development and Evaluation; HR: hazard ratio; RCT: randomised controlled trial.

4.2. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 2: All‐cause death (Hazard ratio)

4.3. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 3: Cardiovascular death

4.4. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 4: Cardiovascular death (Hazard ratio)

Kidney failure

GLP‐1 receptor agonists had uncertain effects on kidney failure compared to insulin (RR 1.02, 95% CI 0.19 to 5.50; 1 study, 576 participants; very‐low‐certainty evidence; Analysis 4.5) in people with CKD stages 3‐5 and type 2 diabetes.

4.5. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 5: Kidney failure

Severe hypoglycaemia

GLP‐1 receptor agonists had uncertain effects on severe hypoglycaemia compared to insulin (RR 0.66, 95% CI 0.36 to 1.22; I² = 0%; 2 studies, 668 participants; very‐low‐certainty evidence; Analysis 4.6) in people with all CKD stages and type 2 diabetes.

4.6. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 6: Severe hypoglycaemia

Myocardial infarction

GLP‐1 receptor agonists had uncertain effects on MI compared to insulin (RR 1.02, 95% CI 0.31 to 3.33; 1 study, 576 participants; very‐low‐certainty evidence; Analysis 4.7) in people with CKD stages 3‐5 and type 2 diabetes.

4.7. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 7: Myocardial infarction

Heart failure

GLP‐1 receptor agonists had uncertain effects on heart failure compared to inhibitors (RR 1.53, 95% CI 0.06 to 37.32; 1 study, 576 participants; very‐low‐certainty evidence; Analysis 4.8) in people with CKD stages 3‐5 and type 2 diabetes.

4.8. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 8: Heart failure

Change in glomerular filtration rate

GLP‐1 receptor agonists may reduce the change in eGFR compared to insulin (MD 3.18 mL/min, 95% CI 0.34 to 6.02; I² = 87%; 2 studies, 657 participants; low‐certainty evidence; Analysis 4.9) (MD 5.83%, 95% CI 4.38 to 7.28; 1 study, 81 participants; very‐low‐certainty evidence; Analysis 4.10) in people with all CKD stages and type 2 diabetes.

4.9. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 9: Change in eGFR [mL/min]

4.10. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 10: Change in eGFR [%]

Glomerular filtration rate loss

GLP‐1 receptor agonists had uncertain effects on eGFR loss compared to insulin (RR 0.72, 95% CI 0.45 to 1.16; 1 study, 576 participants; very‐low‐certainty evidence; Analysis 4.11) in people with CKD stages 3‐5 and type 2 diabetes.

4.11. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 11: eGFR loss

Change in creatinine clearance

GLP‐1 receptor agonists had uncertain effects on change in CrCl compared to insulin (MD 1.20 mL/min, 95% CI 0.07 to 2.33; 1 study, 576 participants; very‐low‐certainty evidence; Analysis 4.12) in people with CKD stages 3‐5 and type 2 diabetes.

4.12. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 12: Change in creatinine clearance [mL/min]

Urinary albumin‐creatinine ratio

Wang 2020a reported GLP‐1 receptor agonists may decrease UACR compared to insulin (MD ‐28.40%, 95% CI ‐37.55 to ‐19.25; 1 study, 81 participants; low‐certainty evidence; Analysis 4.13) (MD ‐607.81 mg/g, 95% CI ‐727.04 to ‐488.58; 1 study, 81 participants; low‐certainty evidence; Analysis 4.14) in people with all CKD stages and type 2 diabetes.

4.13. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 13: Urinary albumin‐creatinine ratio [%]

4.14. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 14: Urinary albumin‐creatinine ratio [mg/g]

Urinary albumin excretion rate

Wang 2020a reported GLP‐1 receptor agonists may decrease urine albumin excretion rate compared to insulin (MD ‐29.71%, 95% CI ‐37.20 to ‐22.22; 1 study, 81 participants; low‐certainty evidence; Analysis 4.15) (MD ‐159.67 mg/24 hours, 95% CI ‐293.45 to ‐25.89; 1 study, 81 participants; low‐certainty evidence; Analysis 4.16) in people with all CKD stages and type 2 diabetes.

4.15. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 15: Urinary albumin excretion rate [%]

4.16. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 16: Urinary albumin excretion rate [mg/24 h]

Change in body weight

It was uncertain whether GLP‐1 receptor agonists had any effects on change in body weight compared to insulin (MD ‐1.68 kg, 95% CI ‐3.72 to 0.36; I² = 95%; 2 studies, 630 participants; very‐low‐certainty evidence; Analysis 4.17) (MD ‐3.80 kg, 95% CI ‐4.80 to ‐2.80; 1 study; very‐low‐certainty evidence; Analysis 4.18) in people with all CKD stages and type 2 diabetes.

4.17. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 17: Change in body weight [kg]

4.18. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 18: Change in body weight [kg]

Body weight increase

GLP‐1 receptor agonists had uncertain effects on body weight increase compared to insulin (RR 0.57, 95% CI 0.30 to 1.10; 1 study, 576 participants; very‐low‐certainty evidence; Analysis 4.19) in people with CKD stages 3‐5 and type 2 diabetes.

4.19. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 19: Body weight increase

Serious adverse events

GLP‐1 receptor agonists had uncertain effects on serious adverse events compared to insulin (RR 0.74, 95% CI 0.40 to 1.37; I² = 12%; 2 studies, 668 participants; very‐low‐certainty evidence; Analysis 4.20) in people with all CKD stages and type 2 diabetes.

4.20. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 20: Serious adverse events

Withdrawal due to adverse events

It was uncertain whether GLP‐1 receptor agonists had any effect on withdrawal due to adverse events compared to insulin (RR 1.20, 95% CI 0.39 to 3.66; 1 study, 92 participants; very‐low‐certainty evidence; Analysis 4.21) in people with all CKD stages and type 2 diabetes.

4.21. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 21: Withdrawal due to adverse events

Hypoglycaemia

GLP‐1 receptor agonists probably resulted in less hypoglycaemia compared to insulin (RR 0.69, 95% CI 0.52 to 0.90; I² = 26%; 2 studies, 668 participants; moderate‐certainty evidence; Analysis 4.22) in people with all CKD stages and type 2 diabetes.

4.22. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 22: Hypoglycaemia

Acute kidney injury

GLP‐1 receptor agonists had uncertain effects on AKI compared to insulin (RR 2.03, 95% CI 0.44 to 9.47; 2 studies, 583 participants; very‐low‐certainty evidence; Analysis 4.23) in people with all CKD stages and type 2 diabetes.

4.23. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 23: Acute kidney injury

Change in HbA1c

• It was uncertain whether GLP‐1 receptor agonists had any effects on change in HbA1c compared to insulin (MD 0.08%, 95% CI ‐0.23 to 0.39; I² = 83%; 2 studies, 630 participants; very‐low‐certainty evidence; Analysis 4.24) (MD ‐0.86%, 95% CI ‐1.59 to ‐0.13; I² = 90%; 4 studies; very‐low‐certainty evidence; Analysis 4.25) in people with all CKD stages and type 2 diabetes.

• GLP‐1 receptor agonists had uncertain effects on change in HbA1c compared to insulin (MD ‐1.10 mmol/mol, 95% CI ‐3.68 to 1.48; 1 study, 549 participants; very‐low‐certainty evidence; Analysis 4.26) in people with CKD stages 3‐5 and type 2 diabetes.

4.24. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 24: Change in HbA1c [%]

4.25. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 25: Change in HbA1c [%]

4.26. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 26: Change in HbA1c [mmol/mol]

Ischaemic stroke

GLP‐1 receptor agonists had uncertain effects on ischaemic stroke compared to insulin (RR 3.56, 95% CI 0.19 to 68.65; 1 study, 576 participants; very‐low‐certainty evidence; Analysis 4.27) in people with CKD stages 3‐5 and type 2 diabetes.

4.27. Analysis.

Comparison 4: GLP‐1 receptor agonists versus insulin, Outcome 27: Ischaemic stroke

No other primary or secondary outcomes were reported for this comparison.

GLP‐1 receptor agonists versus SGLT2 inhibitors

One study (Ma 2024) compared GLP‐1 receptor agonists to SGLT2 inhibitors in people with CKD stages 1‐2 and type 2 diabetes during a follow‐up of 12 weeks. The certainty of the evidence was low to very low.

Death

• The effects of GLP‐1 receptor agonists on all‐cause death compared to SGLT2 inhibitors were not estimable (no events; 1 study, 80 participants; very low‐certainty evidence; Analysis 5.1).

• The effects of GLP‐1 receptor agonists on cardiovascular death compared to SGLT2 inhibitors were not estimable (no events; 1 study, 80 participants; very low‐certainty evidence; Analysis 5.2).

5.1. Analysis.

Comparison 5: GLP‐1 receptor agonists versus SGLT2 inhibitors, Outcome 1: All‐cause death

5.2. Analysis.

Comparison 5: GLP‐1 receptor agonists versus SGLT2 inhibitors, Outcome 2: Cardiovascular death

Withdrawal due to adverse events

The effects of GLP‐1 receptor agonists on withdrawal due to adverse events compared to SGLT2 inhibitors were not estimable (no events; 1 study, 80 participants; very low‐certainty evidence; Analysis 5.3).

5.3. Analysis.

Comparison 5: GLP‐1 receptor agonists versus SGLT2 inhibitors, Outcome 3: Withdrawal due to adverse events

HbA1c

Ma 2024 reported GLP‐1 receptor agonists may decrease HbA1c compared to SGLT2 inhibitors (MD ‐0.90%, 95% CI ‐1.65 to ‐0.15; 1 study, 80 participants; low‐certainty evidence; Analysis 5.4).

5.4. Analysis.

Comparison 5: GLP‐1 receptor agonists versus SGLT2 inhibitors, Outcome 4: HbA1c [%]

Body weight

Ma 2024 reported GLP‐1 receptor agonists may increase body weight compared to SGLT2 inhibitors (MD 7.43 kg, 95% CI 4.41 to 10.45; 1 study, 80 participants; low‐certainty evidence; Analysis 5.5).

5.5. Analysis.

Comparison 5: GLP‐1 receptor agonists versus SGLT2 inhibitors, Outcome 5: Body weight [kg]

Body mass index

Ma 2024 reported GLP‐1 receptor agonists may increase BMI compared to SGLT2 inhibitors (MD 1.53 kg/m², 95% CI 0.63 to 2.43; 1 study, 80 participants; low‐certainty evidence; Analysis 5.6).

5.6. Analysis.

Comparison 5: GLP‐1 receptor agonists versus SGLT2 inhibitors, Outcome 6: BMI [kg/m²]

No other primary or secondary outcomes were reported for this comparison.

GLP‐1 receptor agonist versus another GLP‐1 receptor agonist

One study (Selvarajah 2024a) compared a GLP‐1 receptor agonist (semaglutide) to another GLP‐1 receptor agonist (cotadutide) in people with all CKD stages and type 2 diabetes during a follow‐up of 26 weeks. The certainty of the evidence was very low.

Death

• A GLP‐1 receptor agonist (semaglutide) had uncertain effects on all‐cause death compared to another GLP‐1 receptor agonist (cotadutide) (RR 0.66, 95% CI 0.03 to 13.52; 1 study, 196 participants; very‐low‐certainty evidence; Analysis 6.1).

• A GLP‐1 receptor agonist (semaglutide) had uncertain effects on cardiovascular death compared to another GLP‐1 receptor agonist (cotadutide) (RR 1.10, 95% CI 0.05 to 26.58; 1 study, 196 participants; very‐low‐certainty evidence; Analysis 6.2).

6.1. Analysis.

Comparison 6: GLP‐1 receptor agonist versus another GLP‐1 receptor agonist, Outcome 1: All‐cause death

6.2. Analysis.

Comparison 6: GLP‐1 receptor agonist versus another GLP‐1 receptor agonist, Outcome 2: Cardiovascular death

Withdrawal due to adverse events

A GLP‐1 receptor agonist (semaglutide) had uncertain effects on withdrawal due to adverse events compared to another GLP‐1 receptor agonist (cotadutide) (RR 1.38, 95% CI 0.61 to 3.12; 1 study, 196 participants; very‐low‐certainty evidence; Analysis 6.3).

6.3. Analysis.

Comparison 6: GLP‐1 receptor agonist versus another GLP‐1 receptor agonist, Outcome 3: Withdrawal due to adverse events

Acute kidney injury

The effects of a GLP‐1 receptor agonist (semaglutide) on AKI compared to another GLP‐1 receptor agonist (cotadutide) were not estimable (no events; 1 study, 196 participants; very low‐certainty evidence; Analysis 6.4).

6.4. Analysis.

Comparison 6: GLP‐1 receptor agonist versus another GLP‐1 receptor agonist, Outcome 4: Acute kidney injury

Sensitivity analyses for GLP‐1 receptor agonists versus placebo

Sensitivity analyses, taking into account the high overall risk bias (Table 4), very long studies (Table 5), very large studies (Table 6), and funding (Table 7), did not show substantial differences compared with our data.

3. Sensitivity analysis removing studies with overall high risk of bias: GLP‐1 recepeto agonists versus placebo.

Outcome	Number of studies	Participants	Metrics (95% CI)	I2	GRADE
All‐cause death (Analysis 8.1)	7 RCTs	17,837 participants	RR 0.85 (0.74 to 0.98)	23%	High certainty
Cardiovascular death (Analysis 8.2)	6 RCTs	17,777 participants	RR 0.84 (0.68 to 1.05)	42%	Low certainty
Change in body weight [kg] (Analysis 8.3)	2 RCTs	Number not clearly reported (people with all CKD stages and type 2 diabetes)	MD ‐0.65 kg (‐1.10 to ‐0.21)	25%	Moderate certainty
HbA1c [%] (Analysis 8.4)	2 RCTs	357 participants	MD ‐0.99% (‐2.95 to 0.97)	98%	Very low certainty
Change in HbA1c [%] (Analysis 8.5)	3 RCTs	Number not clearly reported (people with all CKD stages and type 2 diabetes)	MD ‐0.69% (‐0.85 to ‐0.52)	55%	Low certainty

CI: confidence interval; CKD: chronic kidney disease; GRADE: Grading of Recommendations Assessment, Development and Evaluation; HbA1c: glycated haemoglobin; MD: mean difference; RR: risk ratio; RCT: randomised controlled trial.

4. Sensitivity analysis removing any very long study: GLP‐1 receptor agonist versus placebo.

Outcome	Number of studies	Participants	Metrics (95% CI)	I2	GRADE
All‐cause death (Analysis 9.1)	6 RCTs	1,012 participants	RR 1.52 (0.41 to 5.59)	0%	Low certainty
All‐cause death (HR) (Analysis 9.2)	2 RCTs	Number not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	HR 1.42 (0.19 to 10.88)	40%	Low certainty
Cardiovascular death (Analysis 9.3)	5 RCTs	952 participants	RR 1.05 (0.18 to 6.01)	0%	Low certainty
Cardiovascular death (HR) (Analysis 9.4)	1 RCT	Number not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	HR 0.99 (0.06 to 16.33)	‐	Low certainty
MACE 3‐point (Analysis 9.5)	2 RCTs	2,976 participants	RR 0.75 (0.60 to 0.94)	0%	High certainty
MACE 3‐point (HR) (Analysis 9.6)	4 RCTs	Number not clearly reported (people with all CKD stages and type 2 diabetes)	HR 0.76 (0.67 to 0.87)	7%	Moderate certainty
MACE 4‐point (HR) (Analysis 9.7)	1 RCT	Number not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	HR 1.01 (0.80 to 1.28)	‐	Moderate certainty
Kidney failure (Analysis 9.8)	2 RCTs	601 participants	RR 0.79 (0.31 to 2.02)	0%	Moderate certainty
Composite kidney outcome (HR) (Analysis 9.9)	1 RCT	Number not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	HR 0.88 (0.73 to 1.06)	‐	Moderate certainty
Severe hypoglycaemia (Analysis 9.10)	2 RCTs	601 participants	RR 2.94 (0.12 to 71.46)	‐	Moderate certainty
Myocardial infarction (Analysis 9.11)	1 RCT	277 participants	HR 2.94 (0.12 to 71.46)	‐	Moderate certainty
Fatal or nonfatal stroke (Analysis 9.12)	1 RCT	324 participants	RR 0.99 (0.06 to 15.66)	‐	Moderate certainty
Hospitalisation due to heart failure (Analysis 9.13)	1 RCT	324 participants	HR 0.33 (0.01 to 8.02)	‐	Moderate certainty
Withdrawal due to adverse events (Analysis 9.14)	5 RCTs	988 participants	RR 2.09 (1.06 to 4.14)	38%	Low certainty
Withdrawal due to adverse events (HR) (Analysis 9.15)	1 RCT	Number not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	HR 4.65 (1.62 to 13.31)	‐	Moderate certainty
AKI (HR) (Analysis 9.16)	1 RCT	324 participants	HR 1.98 (0.18 to 21.78)	‐	Moderate certainty

CI: confidence interval; CKD: chronic kidney disease; GRADE: Grading of Recommendations Assessment, Development and Evaluation; AKI: acute kidney injury; MACE: major cardiovascular events; RR: risk ratio; HR: hazard ratio; RCT: randomised controlled trial.

5. Sensitivity analysis removing any very large study: GLP‐1 recepto agonist versus placebo.

Outcome	Number of studies	Participants	Metrics (95% CI)	I2	GRADE
All‐cause death (Analysis 10.1)	6 RCTs	1,012 participants	RR 1.52 (0.41 to 5.59)	0%	Moderate certainty
All‐cause death (HR) (Analysis 10.2)	2 RCTs	Number not clearly reported (people with CKD 3‐5 stages and type 2 diabetes)	HR 1.42 (0.19 to 10.88)	40%	Low certainty
Cardiovascular death (Analysis 10.3)	5 RCTs	952 participants	RR 1.05 (0.18 to 6.01)	0%	Low certainty
Cardiovascular death (HR) (Analysis 10.4)	1 RCT	Number not clearly reported (people with CKD 3‐5 stages and type 2 diabetes)	HR 0.99 (0.06 to 16.33)	‐	Moderate certainty
MACE 3‐point (Analysis 10.5)	1 RCT	939 participants	RR 0.85 (0.58 to 1.24)	‐	Moderate certainty
MACE 3‐point (HR) (Analysis 10.6)	1 RCT	Number not clearly reported (people with CKD 3‐5 stages and type 2 diabetes)	HR 0.84 (0.57 to 1.24)	‐	Moderate certainty
Kidney failure (Analysis 10.7)	2 RCTs	601 participants	RR 0.79 (0.31 to 2.02)	0%	Moderate certainty
Severe hypoglycaemia (Analysis 10.8)	2 RCTs	601 participants	RR 2.94 (0.12 to 71.46)	‐	Moderate certainty
Myocardial infarction (Analysis 10.9)	1 RCT	277 participants	RR 2.94 (0.12 to 71.46)	‐	Moderate certainty
Fatal or nonfatal stroke (Analysis 10.10)	1 RCT	324 participants	RR 0.99 (0.06 to 15.66)	‐	Moderate certainty
Hospitalisation due to heart failure (Analysis 10.11)	1 RCT	324 participants	RR 0.33 (0.01 to 8.02)	‐	Moderate certainty
Withdrawal due to adverse events (Analysis 10.12)	5 RCTs	988 participants	RR 2.09 (1.06 to 4.14)	38%	Low certainty
Withdrawal due to adverse events (HR) (Analysis 10.13)	1 RCT	Number not clearly reported (people with CKD stages 3‐5 and type 2 diabetes)	HR 4.65 (1.62 to 13.31)	‐	Moderate certainty
AKI (HR) (Analysis 10.14)	1 RCT	324 participants	HR 1.98 (0.18 to 21.78)	‐	Moderate certainty

CI: confidence interval; CKD: chronic kidney disease; GRADE: Grading of Recommendations Assessment, Development and Evaluation; AKI: acute kidney injury; MACE: major cardiovascular events; RR: risk ratio; HR: hazard ratio; RCT: randomised controlled trial.

6. Sensitivity analysis removing studies with funding: GLP‐1 receptor agonist versus placebo.

Outcome	Number of studies	Participants	Metrics (95% CI)	I2	GRADE
HbA1c [%] (Analysis 11.1)	1 RCT	80 participants	MD ‐2.01% (‐2.58 to ‐1.44)	‐	Low certainty

CI: confidence interval; GRADE: Grading of Recommendations Assessment, Development and Evaluation; HbA1c: glycated haemoglobin; MD: mean difference; RCT: randomised controlled trial

Sensitivity analyses were not possible due to little data and studies for unpublished studies, diagnostic criteria, language of publication, and country.

Discussion

Summary of main results

We identified 42 studies evaluating GLP‐1 receptor agonists in 48,148 people with CKD and type 2 diabetes. Most studies compared GLP‐1 receptor agonists to placebo in people with GFR ≥ 60 mL/min/1.73 m². Risks of bias in the included studies were low or unclear, leading to generally moderate to low‐certainty evidence.

GLP‐1 receptor agonists probably reduced all‐cause death but may have little or no effect on cardiovascular death in people with CKD and diabetes. GLP‐1 receptor agonists probably lower major cardiovascular events (3‐ or 4‐point MACE), and probably have little or no effect on kidney failure. GLP‐1 receptor agonists probably have little or no effect on composite kidney outcome, and may have little or no effect on the risk of severe hypoglycaemia in people with CKD and diabetes. GLP‐1 receptor agonists may decrease body weight and BMI compared to placebo. Based on absolute risks of clinical outcomes, it is likely that GLP‐1 receptor agonists prevent all‐cause death in 52 people with CKD stages 1‐2 and 116 in CKD stages 3‐5, cardiovascular death in 34 people with CKD stages 1‐2 and 71 in CKD stages 3‐5, while 95 CKD stages 1‐2 and 153 in CKD stages 3‐5 might experience a major cardiovascular event for every 1000 people treated over 1 year (Table 1).

The effects of GLP‐1 receptor agonists compared to placebo on heart failure, hospitalisation due to heart failure, GFR, UACR, withdrawal from treatment, hypoglycaemia, lactic acidosis, or diabetic ketoacidosis were uncertain. Adverse events were inconsistently reported.

The effects of GLP‐1 receptor agonists compared to standard care or other hypoglycaemic agents (DPP‐4 inhibitors, insulin, SGLT2 inhibitors) were uncertain.

No studies evaluated treatment on risks of fatigue, life participation, amputation or fracture.

Overall completeness and applicability of evidence

Evidence from existing studies was frequently of moderate or low certainty. About a third of the included studies were in people with CKD stages 3‐5. Many studies compared different treatment doses. No studies were conducted on children or on people with type 1 diabetes.

The definitions used for the composite kidney outcome were quite different between studies, and this could prevent a firm conclusion from being reached in this setting.

The 2022 KDIGO Guideline suggested GLP‐1 receptor agonists may be preferentially used in patients with obesity, type 2 diabetes, and CKD to promote weight loss (KDIGO 2022). GLP‐1 receptor agonists are effective glucose‐lowering agents in people with type 2 diabetes and CKD, reducing the risk of cardiovascular events and albuminuria, and slowing eGFR decline (KDIGO 2022). In this review, most studies compared GLP‐1 receptor agonists with placebo, and the majority of the included studies were conducted in a small sample size, were of short‐term duration, had methodological limitations, or were primarily designed to evaluate surrogate outcomes. No studies reported patient‐centred outcomes (e.g. fatigue or life participation). Adverse events related to treatment were not systematically reported in the included studies (Table 8). No studies compared different types and doses of GLP‐1 receptor agonists and data on head‐to‐head comparisons were limited.

7. Adverse events reported in the included studies.

Study ID	Intervention group	Control group
AMPLITUDE‐O 2021	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
AWARD‐7 2017	Diarrhoea (63/382), nausea (65/382), GFR decreased (35/382), hypertension (27/382), UTI (26/382), peripheral oedema (23/382), hyperkalaemia (20/382), dizziness (19/382), vomiting (42/382), constipation (22/382), cough (14/382)	Diarrhoea (14/194), nausea (9/194), GFR decreased (25/194), hypertension (21/194), UTI (21/194), peripheral oedema (15/194), hyperkalaemia (13/194), dizziness (10/194), vomiting (6/194), constipation (14/194), cough (14/194)
AWARD‐PEDS 2022	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
Dekkers 2021	AE were not clearly reported	AE were not clearly reported
DUAL I 2014	AE were not reported only for participants with CKD and diabete	AE were not reported only for participants with CKD and diabetes
DUAL II 2014	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
DUAL III 2017	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
DUAL IV 2014	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
DUAL V 2016	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
ELIXA 2018	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
EXSCEL 2017	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
FLOW 2023	Serious AEs (805/1767), AEs leading to discontinuation of treatment (232/1767), AKI (153/1767), malignant tumors (103/1767), pancreatic cancer (5/1767)	Serious AEs (893/1766), AEs leading to discontinuation of treatment (209/1766), AKI (175/1766), malignant tumors (89/1766), pancreatic cancer (6/1766)
GetGoal studies 2017 ‐ pooled	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
GetGoal‐O 2015	AE were not clearly reported	AE were not clearly reported
GRADE 2022	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
Harmony Outcomes 2018	Severe hypoglycaemia (31/4717), pancreatitis (10/4717), thyroid cancer (0/4717), haematological neoplasia (9/4717), pancreatic cancer (6/4717), serious GI events (92/4717), renal impairment (279/4717)	Severe hypoglycaemia (55/4715), pancreatitis (7/4715), thyroid cancer (0/4715), haematological neoplasia (5/4715), pancreatic cancer (5/4715), serious GI events (87/4715), renal impairment (319/4715)
Idorn 2013	Hypoglycaemic events were not clearly reported	Hypoglycaemic events were not clearly reported
LEADER 2017	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
Leiter 2014	Albiglutide (249 participants analysed): any AE (208), UTI (23), diarrhoea (25), upper–respiratory tract infection (14), nasopharyngitis (14), hypertension (14), anaemia (16), peripheral oedema (14), constipation (15), nausea (12), vomiting (4), pancreatitis (1), thyroid cancer (0)	Sitagliptin (246 participants analysed): any AE (205), UTI (23), diarrhoea (20), upper–respiratory tract infection (16), nasopharyngitis (23), hypertension (19), anaemia (10), peripheral oedema (8), constipation (6), nausea (8), vomiting (3), pancreatitis (0), thyroid cancer (0)
LIRA‐PRIME 2019	AE were not clearly reported	AE were not clearly reported
LIRA‐RENAL 2016	Data was referred from 140 participants: AE (107), GI (50), nausea (29), vomiting (17), diarrhoea (10), GFR decrease (9), cardiac disorders (5)	Data was referred from 137 participants: AE (94), GI (24), nausea (6), vomiting (3), diarrhoea (4), GFR decrease (7), cardiac disorders (4)
Liu 2022	Not reported	Not reported
Ma 2024	No AE occurred	No AE occurred
Muskiet 2019	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
Neff 2016	Data referred out of 10 participants: diarrhoea (2), dizziness (0)	Data referred out of 10 participants: diarrhoea (0), dizziness (1)
Parker 2022	AE were not clearly reported	AE were not clearly reported
PIONEER 5 2019	Data was referred to 163 participants: AE (122), SAE (20), nausea (31), constipation (19), vomiting (19), diarrhoea (17), headache (10), GI (19)	Data was referred to 161 participants: AEs (109), SAE (18), nausea (12), constipation (6), vomiting (192), diarrhoea (6), headache (8), GI (3)
PIONEER 6 2019	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
REWIND 2019	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
Selvarajah 2024a	Cotadutide 100 ug/d: any AE (45/55), SAE (5/55), nausea (6/55), vomiting (2/55), diarrhoea (4/55), hypoglycaemia (8/55) Cotadutide 300 ug/d: any AE (38/48), SAE (5/48), nausea (6/48), vomiting (2/48), diarrhoea (4/48), hypoglycaemia (9/48) Cotadutide 600 ug/d: any AE (39/48), SAE (5/48), nausea (13/48), vomiting (7/48), diarrhoea (3/48), hypoglycaemia (13/48) Semaglutide 1.0 mg/wk: any AE (39/45), SAE (4/45), nausea (11/45), vomiting (6/45), diarrhoea (6/45), hypoglycaemia (3/45)	Any AE (39/51), SAE (5/51), nausea (6/51), vomiting (2/51), diarrhoea (1/51), hypoglycaemia (10/51)
SEMPA 2023	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
Sivalingam 2024	Any AE (21/30), SAE (11/30), nausea (8/30), vomiting (1/30), diarrhoea (1/30), hypoglycaemia (1/30), hyperglycaemia (4/30), LUTS (2/30), neoplasms (0/30), fatigue (6/30)	Any AE (18/30), SAE (5/30), nausea (4/30), vomiting (0/30), diarrhoea (0/30), hypoglycaemia (2/30), hyperglycaemia (0/30), LUTS (0/30), neoplasms (0/30), fatigue (1/30)
SOUL 2023	Not reported	Not reported
SUSTAIN‐6 2016	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
SUSTAIN 8 2019	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
Tuttolomondo 2021	AE were not clearly reported	AE were not clearly reported
van der Aart‐van 2023	AE were not reported only for participants with CKD and diabetes	AE were not reported only for participants with CKD and diabetes
von Scholten 2017	SAE (3/27)	SAE (1/27)
Wajdlich 2024	AE were not clearly reported	AE were not clearly reported
Wang 2020a	Data was reported out of 46 participants: AE (16 in the intervention group) (P = 0.015), GI events (15)	Data was reported out of 46 participants: AE (6), GI events (0)
Zhang 2012a	AE were not clearly reported	AE were not clearly reported
Zhou 2019a	AE were not clearly reported	AE were not clearly reported

AE: adverse events; AKI: acute kidney injury; CKD: chronic kidney disease; GFR: glomerular filtration rate; GI: gastrointestinal; LUTS: lower urinary tract symptoms; SAE: severe adverse events; UTI: urinary tract infection

LIRA‐RENAL 2016 appeared to be an outlier for some analyses, probably because the study reported a high withdrawal rate (25%) in both treatment groups and included an older and frailer population compared to other studies, which may lead to more adverse events. In addition, the short study duration prevented predicting long‐term glycaemic responses.

Quality of the evidence

A standard Cochrane RoB2 assessment tool and GRADE (GRADE 2008) were used to evaluate the certainty of the evidence. The majority of studies had a low or unclear risk of bias for most of the domains assessed.

• The risk of bias for the primary outcomes in studies comparing GLP‐1 receptor agonists to placebo was assessed as low risk of bias, although missing outcome data for death (all‐cause and cardiovascular) was assessed to be at high risk in one study.

• The overall risk of bias for all‐cause and cardiovascular death in studies that reported the treatment effects of GLP‐1 receptor agonists compared to standard care were assessed as unclear or at high risk of bias due to deviations from intended interventions or missing outcome data that might come from the low number of events, participants and short follow‐up to assess key outcomes.

• GLP‐1 receptor agonists compared to DPP‐4 inhibitors were judged to be at high risk of bias for death (all‐cause and cardiovascular) and severe hypoglycaemia due to missing outcome data.

• GLP‐1 receptor agonists compared to insulin were judged to be at low or unclear risk of bias for death (all‐cause and cardiovascular), kidney failure and severe hypoglycaemia, although the primary outcomes were assessed to be at high risk of bias due to deviations from intended interventions.

• GLP‐1 receptor agonists compared to SGLT2 inhibitors were judged to be at low risk of bias for all‐cause and cardiovascular death, although blinding was not clearly reported.

• GLP‐1 receptor agonists compared to another GLP‐1 receptor agonist were judged to have an uncertain risk of bias for all‐cause and cardiovascular death due to allocation concealment.

Overall, the evidence of the effects of GLP‐1 receptor agonists in people with CKD and diabetes is limited, and a low number of studies and events prevented exploring the presence of moderate or substantial heterogeneity in the analyses. Generally, subgroup and sensitivity analyses could not be conducted due to insufficient data. Adverse events were rarely and inconsistently reported in the included studies.

Our review has some limitations.

1. It was possible that some potential studies have been missed due to publication bias (e.g. conference abstracts, dissertations).

2. Some assumptions on classification for author‐self‐reported outcomes (3‐ or 4‐point MACE and a composite kidney outcome) or CKD subgroups extracted from studies reporting mixed populations or from those where only GFR values have been reported may underestimate our confidence in reporting.

3. Missing data were inevitable, despite our efforts to minimise missing results in specific syntheses, and this may prevent the generalisability of our results.

Potential biases in the review process

Our review was carried out using standard Cochrane methods using a sensitive search without language restriction provided by an expert Information Specialist. Each step was completed independently by three investigators to reduce the chance of misclassification and mistakes in the adjudication of the evidence, and to increase our confidence in the data collection process. Study authors were contacted to obtain further data. Some studies did not report key outcomes in a format suitable for meta‐analysis.

Agreements and disagreements with other studies or reviews

Previous studies have shown partly different results compared to our systematic review. In fact, a systematic review showed that treatment with GLP‐1 receptor agonists was not associated with evidence of a reduction in the 3‐point MACE (odds ratio (OR) 0.80, 95% CI 0.59 to 1.07; P = 0.13) among people with CKD (eGFR < 60 mL/min) and type 2 diabetes (Kelly 2022). However, a network meta‐analysis on phase III or IV RCTs, reported that GLP‐1 receptor agonists reduced the risks of MACE and the kidney‐specific composite outcome but did not show benefits for cardiovascular death and hospitalisation due to heart failure compared with placebo (Cao 2022).

Findings from a network meta‐analysis (13 studies on over 30,000 participants) confirmed that GLP‐1 receptor agonists were not associated with a decreased risk of cardiovascular and kidney events in people with CKD and type 2 diabetes (Yamada 2021). This study focused only on GLP‐1 receptor agonists and SGLT2 inhibitors in people with CKD stages 3‐5.

A network meta‐analysis showed that placebo‐controlled trials evaluating GLP‐1 receptor agonists, SGLT2 or DPP‐4 inhibitors in people with diabetic kidney disease showed that cardiovascular and kidney events were lower in people treated with SGLT2 inhibitors compared to those treated with GLP‐1 receptor agonists or DPP‐4 inhibitors (Cao 2022). Another large network meta‐analysis that addressed the benefits and harms of both GLP‐1 receptor agonists and SGLT2 inhibitors demonstrated that those pharmacological classes reduced the risk of all‐cause (OR 0.88, 95% CI 0.83 to 0.94) and cardiovascular death (OR 0.88, 95% CI 0.80 to 0.96), nonfatal MI, and kidney failure compared to placebo (Palmer 2021). However, the P value was not reported in the analyses, and the study included people with and without CKD with type 2 diabetes. Therefore, we had insufficient data to evaluate these aspects.

Authors' conclusions

Implications for practice.

GLP‐1 receptor agonists probably reduced the risk of all‐cause death, but the evidence is still poor to evaluate the effects on cardiovascular death in people with CKD and diabetes. The use of GLP‐1 receptor agonists probably showed benefits in decreasing 3‐ and 4‐point MACE, and probably had little or no effect on kidney failure and on composite kidney outcome. GLP‐1 receptor agonists may have little or no effect on the risk of severe hypoglycaemia, but may decrease body weight or BMI in this population. Adverse events were inconsistently and rarely reported in the included studies, thus preventing us from assessing the risk of GLP‐1 receptor agonist treatment in people with CKD and diabetes. The efficacy and safety of GLP‐1 receptor agonists in association with other glucose‐lowering treatments compared to other classes were poor. There was limited evidence to inform decision‐making in children, people with type 1 diabetes or those receiving dialysis, including both haemodialysis and peritoneal dialysis.

Implications for research.

Future studies should address the benefits and harms of using GLP‐1 receptor agonists in people with CKD and diabetes, focusing on core outcomes as prioritised by patients, caregivers and health professionals (SONG 2017) to better inform decision‐making. Future well‐designed and powered RCTs need to be designed and conducted to assess the effects of GLP‐1 receptor agonists on treating people with CKD and type 1 and 2 diabetes. Newer studies should provide information on key patient‐reported outcomes, including fatigue and life participation, and provide data on body weight and BMI in this setting. Future studies of GLP‐1 receptor agonists comparing them with standard care or providing a head‐to‐head comparison with other hypoglycaemic medications will increase our certainty of the current evidence. Economic evaluation of GLP‐1 receptor agonists in the treatment of diabetes would inform decision‐making by policy‐makers and stakeholders in people with CKD and diabetes.

History

Protocol first published: Issue 4, 2023

Risk of bias

Risk of bias for analysis 1.1 All‐cause death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.1.1 CKD stages 1‐2

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356 (3.6%) and 303/7396 (4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: not reported. However, participants and/or investigators could be aware of treatment assigned

Low risk of bias

Comment: all participants completed the study and were included into the analysis

Some concerns

Comment: not reported. However, objective and subjective outcomes were reported

Some concerns

Comment: prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain

Subgroup 1.1.2 CKD stages 3‐5

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356 (3.6%) and 303/7396 (4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Quote: "Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets."

Low risk of bias

Comment: ITT was performed, however some participants discontinued

Low risk of bias

Comment: outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: all prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: all outcomes that were planned in the methods were reported in the results

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups

Low risk of bias

Comment: all outcomes that were planned in the methods were reported in the results

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 1.1.3 Dialysis (HD/PD)

Idorn 2013

Low risk of bias

Quote: “Patients and control subjects were assigned to receive either liraglutide or pIacebo according to a computer‐generated randomisation list provided by Novo Nordisk”.

Low risk of bias

Quote: “Participants, Investigators, and healthcare staff were blinded for the allocated treatment and remained so until the last patient’s last visit”.

High risk of bias

Comment: > 5% lost to follow‐up with discrepancies between groups

Some concerns

Comment: objective and subjective outcomes were reported

Low risk of bias

Comment: all outcomes that were planned in the methods were reported in the results

High risk of bias

The study is judged to be high risk of bias in at least one domain for this result.

Subgroup 1.1.4 All CKD stages

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: all participants completed the study and were included into the analysis

Some concerns

Comment: not reported. However, objective and subjective outcomes were reported

Low risk of bias

Comment: all prespecified and key outcomes were reported

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: all participants completed the study and were included into the analysis

Some concerns

Comment: not reported. However, objective and subjective outcomes were reported

Low risk of bias

Comment: all prespecified and key outcomes were reported

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain

Sivalingam 2024

Low risk of bias

Quote: "One person not involved in the study had access to the computer generated random semaglutide/placebo allocation sequence."

Low risk of bias

Quote: "Investigators, participants and treating physicians were blinded to treatment allocation."

Low risk of bias

Quote: "All participants completed the study and were included into the analysis." Comment: 54/60 participants were included in the analysis. 28/30 participants were included in the intervention group and 26/30 In the control group. However, data seemed to be reported on the ITT

Some concerns

Comment: not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: all prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.2 All‐cause death (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.2.1 CKD stages 1‐2

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.2.2 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 1.2.3 All CKD stages

FLOW 2023

Low risk of bias

Quote: "Central interactive Web‐based response system."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: ITT was performed

Low risk of bias

Quote: "Analyses were conducted by the sponsor and were independently verified with the use of the original data by Statogen Consulting." "An independent data and safety monitoring committee reviewed the data and recommended early completion of the trial for efficacy."

Low risk of bias

Comment: all prespecified and key outcomes were reported

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.3 Cardiovascular death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.3.1 CKD stages 1‐2

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Low risk of bias

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Low risk of bias

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Low risk of bias

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 1.3.2 CKD stages 3‐5

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 1.3.3 Dialysis (HD/PD)

Idorn 2013

Low risk of bias

Quote: “Patients and control subjects were assigned to receive either liraglutide or pIacebo according to a computer‐generated randomisation list provided by Novo Nordisk”.

Low risk of bias

Quote: “Participants, Investigators, and healthcare staff were blinded for the allocated treatment and remained so until the last patient’s last visit”.

High risk of bias

Comment: >5% lost to follow up with discrepancies between groups.

Some concerns

Comment: Objective and subjective outcomes were reported.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 1.3.4 All CKD stages

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.4 Cardiovascular death (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.4.1 CKD stages 1‐2

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.4.2 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 1.4.3 All CKD stages

FLOW 2023

Low risk of bias

Quote: "Central interactive Web‐based response system."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: ITT was performed.

Low risk of bias

Quote: "Analyses were conducted by the sponsor and were independently verified with the use of the original data by Statogen Consulting." "An independent data and safety monitoring committee reviewed the data and recommended early completion of the trial for efficacy."

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.5 3‐point MACE.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.5.1 CKD stages 1‐2

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.5.2 CKD stages 3‐5

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

SUSTAIN‐6 2016

Low risk of bias

Quote: "A randomisation session will be carried out by using the IV/WRS (interactive voice/web response system)."

Low risk of bias

Quote: "Double‐blind study."

Low risk of bias

Comment: Overall, 15/1648 (1.5%) of the semaglutide group and 40/1649 (2.4%) of the control group did not complete the trial. The rate of treatment discontinuation was higher in the semaglutide group (214/1648, 13.0%) than the placebo group (110/1649, 6.6%), which was mainly from gastrointestinal disorders. ITT was performed.

Low risk of bias

Quote: "The components of the primary endpoint and other key safety outcomes will be subject to blinded external adjudication, and an independent Data Monitoring Committee (DMC) will oversee the safety of trial participants."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.5.3 All CKD stages

AMPLITUDE‐O 2021

Low risk of bias

Quote: "Randomization was performed through an interactive Web‐response system and was stratified according to current or potential future use of SGLT2 inhibitors (i.e., current use at the time of randomisation, SGLT2 inhibitor likely to be added to therapy, or SGLT2 inhibitor unlikely to be added to therapy) to minimize between group differences in therapy with various cardioprotective medications."

Low risk of bias

Quote: "Double‐blind". "Only the data monitoring committee had access to unmasked data until the database lock." "Efpeglenatide or placebo was added to each participant’s current therapy in a blinded manner." Comment: the study was reported as double‐blind and it was likely that all people involved (excluding data monitoring) were blinded

Low risk of bias

Comment: all participants were included in the ITT analysis. Discontinuation was reported in sufficient detail to perform adjudication in people with CKD and diabetes.

Low risk of bias

Quote: "Only the data monitoring committee had access to unmasked data until the database lock." Comment: objective and subjective outcomes were reported. It was not stated if the data monitoring was blinded to the treatment assigned

Some concerns

Comment: prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported in people with CKD and diabetes

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.6 3‐point MACE (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.6.1 CKD stages 1‐2

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Harmony Outcomes 2018

Low risk of bias

Quote: "Patients were assigned in a 1:1 ratio to receive subcutaneous injections of albiglutide or placebo (with matched administration device, diluent, and volume injected) once a week, according to a sequestered, fixed, computer‐generated randomisation code that used balanced permuted blocks of treatment group allocations, without stratification. Investigators used an interactive voice or web response system to obtain treatment assignment."

Low risk of bias

Quote: "All investigators and patients involved in the trial were masked to treatment group."

Low risk of bias

Comment: overall, 1161/4731 (24.5%) of albiglutide group and 1318/4732 (27.9%) of placebo group discontinued treatment. The proportion of withdrawal were 2.3% in albiglutide group and 3.2% in placebo group. Regarding the reason for treatment discontinuation, the ratios of AEs were almost the same between the arms; albiglutide group (8.6%), placebo (6.4%). ITT was performed

Low risk of bias

Quote: "An IDMC (Independent Data Monitoring Committee) will have study oversight to ensure participant safety and scientific integrity of the data (Section 9.8), an independent Cardiovascular Endpoint Committee (CEC) blinded to treatment allocation will adjudicate cardiovascular outcome events and a Pancreatitis Adjudication Committee will adjudicate potential events of pancreatitis."

Low risk of bias

Comment: all outcomes that were planned in the methods were reported in the results

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results..

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.6.2 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Harmony Outcomes 2018

Low risk of bias

Quote: "Patients were assigned in a 1:1 ratio to receive subcutaneous injections of albiglutide or placebo (with matched administration device, diluent, and volume injected) once a week, according to a sequestered, fixed, computer‐generated randomisation code that used balanced permuted blocks of treatment group allocations, without stratification. Investigators used an interactive voice or web response system to obtain treatment assignment."

Low risk of bias

Quote: "All investigators and patients involved in the trial were masked to treatment group."

Low risk of bias

Comment: overall, 1161/4731 (24.5%) of albiglutide group and 1318/4732 (27.9%) of placebo group discontinued treatment. The proportion of withdrawal were 2.3% in albiglutide group and 3.2% in placebo group. Regarding the reason for treatment discontinuation, the ratios of AEs were almost the same between the arms; albiglutide group (8.6%), placebo (6.4%). ITT was performed

Low risk of bias

Quote: "An IDMC (Independent Data Monitoring Committee) will have study oversight to ensure participant safety and scientific integrity of the data (Section 9.8), an independent Cardiovascular Endpoint Committee (CEC) blinded to treatment allocation will adjudicate cardiovascular outcome events and a Pancreatitis Adjudication Committee will adjudicate potential events of pancreatitis."

Low risk of bias

Comment: all outcomes that were planned in the methods were reported in the results

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results..

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

PIONEER 6 2019

Low risk of bias

Quote: "Randomization was performed using an interactive voice/web response system (IV/WRS)".

Low risk of bias

Comment: Double‐blind study.

Low risk of bias

Comment: 5/1591 (0.3%) of canagliflozin group and 6/1592 (0.4%) of placebo group did not complete the study. The rates (15.3% and 9.9% for semaglutide and placebo, respectively) and causes of discontinuation of study drugs were almost the same between the groups.

Low risk of bias

Comment: An independent data monitoring committee evaluated unblinded trial data.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

SUSTAIN‐6 2016

Low risk of bias

Quote: "A randomisation session will be carried out by using the IV/WRS (interactive voice/web response system)."

Low risk of bias

Quote: "Double‐blind study."

Low risk of bias

Comment: overall, 15/1648 (1.5%) of the semaglutide group and 40/1649 (2.4%) of the control group did not complete the trial. The rate of treatment discontinuation was higher in the semaglutide group (214/1648, 13.0%) than the placebo group (110/1649, 6.6%), which was mainly from gastrointestinal disorders. ITT was performed

Low risk of bias

Quote: "The components of the primary endpoint and other key safety outcomes will be subject to blinded external adjudication, and an independent Data Monitoring Committee (DMC) will oversee the safety of trial participants."

Low risk of bias

Comment: all outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result..

Subgroup 1.6.3 All CKD stages

FLOW 2023

Low risk of bias

Quote: "Central interactive Web‐based response system."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: ITT was performed.

Low risk of bias

Quote: "Analyses were conducted by the sponsor and were independently verified with the use of the original data by Statogen Consulting." "An independent data and safety monitoring committee reviewed the data and recommended early completion of the trial for efficacy."

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

AMPLITUDE‐O 2021

Low risk of bias

Quote: "Randomization was performed through an interactive Web‐response system and was stratified according to current or potential future use of SGLT2 inhibitors (i.e., current use at the time of randomisation, SGLT2 inhibitor likely to be added to therapy, or SGLT2 inhibitor unlikely to be added to therapy) to minimize between group differences in therapy with various cardioprotective medications."

Low risk of bias

Quote: "Double‐blind". "Only the data monitoring committee had access to unmasked data until the database lock." "Efpeglenatide or placebo was added to each participant’s current therapy in a blinded manner." Comment: The study was reported as double blind and it was likely that all people involved (excluding data monitoring) were blinded.

Low risk of bias

Comment: All participants were included in the ITT analysis. However, discontinuation was reported in sufficient detail to perform adjudication in people with CKD and diabetes.

Low risk of bias

Quote: "Only the data monitoring committee had access to unmasked data until the database lock." Comment: Objective and subjective outcomes were reported. It was not stated if the data monitoring was blinded to the treatment assigned.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported in people with CKD and diabetes.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.7 4‐point MACE.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.7.1 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.8 4‐point MACE (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.8.1 CKD stages 1‐2

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.8.2 CKD stages 3‐5

ELIXA 2018

Low risk of bias

Quote: "The randomised treatment kit number list will be generated centrally by Sanofi. The Investigational Product (IP) (lixisenatide or placebo) will be packaged in accordance with this list." […] "The Supply Chain Trial Manager (SCTM) will provide the randomized treatment kit number list and the Study Biostatistician will provide the randomisation scheme to the centralized treatment allocation system. Then, this centralised treatment allocation system will generate the patient randomisation list according to which it will allocate the treatments to the patients." "Two types of centralized treatment allocation system will be used, the Interactive Voice Response System (IVRS) and the Interactive Web Response System (IWRS) depending on the choice of the site."

Low risk of bias

Quote: "Lixisenatide and placebo were indistinguishable. Patients, study staff, and individuals involved in analysis of trial data were masked to treatment allocation.”

Low risk of bias

Comment: the rates of discontinuation were 105/3034 (3.5%) and 110/3034 (3.6%) for lixisenatide group and placebo group, respectively. The rate of adverse events leading to discontinuation of study drugs were 20.6% and 22.1% for lixisenatide group and placebo group, respectively. ITT was performed

Low risk of bias

Quote: "Separate independent committees whose members were unaware of the study group assignments adjudicated potential cardiovascular, pancreatic, and allergic events."

Low risk of bias

Quote: "Separate independent committees whose members were unaware of the study group assignments adjudicated potential cardiovascular, pancreatic, and allergic events."

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.9 Kidney failure.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.9.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 1.9.2 All CKD stages

FLOW 2023

Low risk of bias

Quote: "Central interactive Web‐based response system."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: ITT was performed.

Low risk of bias

Quote: "Analyses were conducted by the sponsor and were independently verified with the use of the original data by Statogen Consulting." "An independent data and safety monitoring committee reviewed the data and recommended early completion of the trial for efficacy."

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.10 Kidney failure (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.10.1 All CKD stages

FLOW 2023

Low risk of bias

Quote: "Central interactive Web‐based response system."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: ITT was performed.

Low risk of bias

Quote: "Analyses were conducted by the sponsor and were independently verified with the use of the original data by Statogen Consulting." "An independent data and safety monitoring committee reviewed the data and recommended early completion of the trial for efficacy."

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.11 Composite kidney outcome.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.11.1 CKD stages 1‐2

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.11.2 CKD stages 3‐5

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.12 Composite kidney outcome (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.12.1 CKD stages 1‐2

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.12.2 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

REWIND 2019

Low risk of bias

Quote: "Randomisation was done by a computer‐generated random code with an interactive web response system with stratification by site."

Low risk of bias

Quote: "All investigators and participants were masked to treatment allocation."

Low risk of bias

Comment: overall, 17/4949 of dulaglutide group and 17/4952 of placebo group did not complete the study. The rates (9.1% [451/4949] and 6.3% [310/4952] for dulaglutide and placebo, respectively). Incidences of adverse events were almost the same between the groups. ITT was performed

Low risk of bias

Quote: "An independent clinical endpoint committee (CEC) will adjudicate CV events, pancreatitis events, thyroid evaluations that result in a biopsy or thyroidectomy, and all deaths."

Low risk of bias

Comment: all prespecified and key outcomes were reported

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.13 Severe hypoglycaemia.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.13.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 1.13.2 All CKD stages

FLOW 2023

Low risk of bias

Quote: "Central interactive Web‐based response system."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: ITT was performed.

Low risk of bias

Quote: "Analyses were conducted by the sponsor and were independently verified with the use of the original data by Statogen Consulting." "An independent data and safety monitoring committee reviewed the data and recommended early completion of the trial for efficacy."

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.14 Severe hypoglycaemia (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.14.1 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.15 Nonfatal myocardial infarction.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.15.1 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.16 Nonfatal myocardial infarction (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.16.1 CKD stages 1‐2

FLOW 2023

Low risk of bias

Quote: "Central interactive Web‐based response system."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: ITT was performed.

Low risk of bias

Quote: "Analyses were conducted by the sponsor and were independently verified with the use of the original data by Statogen Consulting." "An independent data and safety monitoring committee reviewed the data and recommended early completion of the trial for efficacy."

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.16.2 CKD stages 3‐5

FLOW 2023

Low risk of bias

Quote: "Central interactive Web‐based response system."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: ITT was performed.

Low risk of bias

Quote: "Analyses were conducted by the sponsor and were independently verified with the use of the original data by Statogen Consulting." "An independent data and safety monitoring committee reviewed the data and recommended early completion of the trial for efficacy."

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.17 Fatal myocardial infarction.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.17.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.18 Myocardial infarction (undefined).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.18.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.19 Fatal or nonfatal myocardial infarction.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.19.1 CKD stages 1‐2

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.19.2 CKD stages 3‐5

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.20 All myocardial infarction (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.20.1 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.21 Nonfatal stroke.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.21.1 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.22 Nonfatal stroke (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.22.1 CKD stages 1‐2

FLOW 2023

Low risk of bias

Quote: "Central interactive Web‐based response system."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: ITT was performed.

Low risk of bias

Quote: "Analyses were conducted by the sponsor and were independently verified with the use of the original data by Statogen Consulting." "An independent data and safety monitoring committee reviewed the data and recommended early completion of the trial for efficacy."

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.22.2 CKD stages 3‐5

FLOW 2023

Low risk of bias

Quote: "Central interactive Web‐based response system."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: ITT was performed.

Low risk of bias

Quote: "Analyses were conducted by the sponsor and were independently verified with the use of the original data by Statogen Consulting." "An independent data and safety monitoring committee reviewed the data and recommended early completion of the trial for efficacy."

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.23 Fatal or nonfatal stroke.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.23.1 CKD stages 1‐2

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.23.2 CKD stages 3‐5

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.24 Fatal or nonfatal stroke (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.24.1 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: all outcomes that were planned in the methods were reported in the results

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.25 Ischaemic stroke.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.25.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result

Risk of bias for analysis 1.27 Heart failure.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.27.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.28 Hospitalisation due to heart failure.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.28.1 CKD stages 1‐2

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.28.2 CKD stages 3‐5

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.29 Hospitalisation due to heart failure (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.29.1 CKD stages 1‐2

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.29.2 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.30 Change in eGFR [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.30.1 CKD stages 3‐5

ELIXA 2018

Low risk of bias

Quote: "The randomised treatment kit number list will be generated centrally by Sanofi. The Investigational Product (IP) (lixisenatide or placebo) will be packaged in accordance with this list." […] "The Supply Chain Trial Manager (SCTM) will provide the randomized treatment kit number list and the Study Biostatistician will provide the randomisation scheme to the centralized treatment allocation system. Then, this centralised treatment allocation system will generate the patient randomisation list according to which it will allocate the treatments to the patients." "Two types of centralized treatment allocation system will be used, the Interactive Voice Response System (IVRS) and the Interactive Web Response System (IWRS) depending on the choice of the site."

Low risk of bias

Quote: "Lixisenatide and placebo were indistinguishable. Patients, study staff, and individuals involved in analysis of trial data were masked to treatment allocation.”

Low risk of bias

Comment: The rates of discontinuation were 105/3034 (3.5%) and 110/3034 (3.6%) for lixisenatide group and placebo group, respectively. The rate of adverse events leading to discontinuation of study drugs were 20.6% and 22.1% for lixisenatide group and placebo group, respectively. ITT was performed.

Low risk of bias

Quote: "Separate independent committees whose members were unaware of the study group assignments adjudicated potential cardiovascular, pancreatic, and allergic events."

Low risk of bias

Quote: "Separate independent committees whose members were unaware of the study group assignments adjudicated potential cardiovascular, pancreatic, and allergic events."

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.31 Increased urinary albumin‐creatinine ratio.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.31.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.32 Body weight [kg].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.32.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.33 Change in body weight [kg].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.33.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.34 Change in body weight [kg].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.34.1 CKD stages 1‐2

GetGoal studies 2017 ‐ pooled

Some concerns

Comment: No information were reported on sequence generation and allocation concealment.

Low risk of bias

Comment: All studies included in the pooled analyses were double‐blind studies.

Low risk of bias

Quote: "Analyses of efficacy endpoints were performed using the pooled modified intent‐to‐treat (mITT) populations." Comment: Slight difference in rate of discontinuation between the intervention and control group (6% and 4%, respectively) but the reasons were balanced. ITT was performed.

Low risk of bias

Comment: "A data monitoring committee supervised the conduct of the study by an ongoing review of unblinded safety and main efficacy parameters. An allergic reaction adjudication committee (ARAC) performed blinded assessment of potentially allergic or allergic like reactions."

Some concerns

Comment: Only post‐hoc analysis was available, and all prespecified analyses were not reported.

Some concerns

The study is judged to be at some risk of bias in at least one domain for this result.

Subgroup 1.34.2 CKD stages 3‐5

GetGoal studies 2017 ‐ pooled

Some concerns

Comment: No information were reported on sequence generation and allocation concealment.

Low risk of bias

Comment: All studies included in the pooled analyses were double‐blind studies.

Low risk of bias

Quote: "Analyses of efficacy endpoints were performed using the pooled modified intent‐to‐treat (mITT) populations." Comment: Slight difference in rate of discontinuation between the intervention and control group (6% and 4%, respectively) but the reasons were balanced. ITT was performed.

Low risk of bias

Comment: "A data monitoring committee supervised the conduct of the study by an ongoing review of unblinded safety and main efficacy parameters. An allergic reaction adjudication committee (ARAC) performed blinded assessment of potentially allergic or allergic like reactions."

Some concerns

Comment: Only post‐hoc analysis was available, and all prespecified analyses were not reported.

Some concerns

The study is judged to be at some risk of bias in at least one domain for this result.

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 1.34.3 Dialysis (HD/PD)

Idorn 2013

Low risk of bias

Quote: “Patients and control subjects were assigned to receive either liraglutide or pIacebo according to a computer‐generated randomisation list provided by Novo Nordisk”.

Low risk of bias

Quote: “Participants, Investigators, and healthcare staff were blinded for the allocated treatment and remained so until the last patient’s last visit”.

High risk of bias

H Comment: >5% lost to follow up with discrepancies between groups.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 1.35 BMI [kg/m²].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.35.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.36 Change in BMI [kg/m²].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.36.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.37 Change in BMI [kg/m²].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.37.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.38 Serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.38.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.39 Withdrawal due to adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.39.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 1.39.2 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 1.39.3 All CKD stages

FLOW 2023

Low risk of bias

Quote: "Central interactive Web‐based response system."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: ITT was performed.

Low risk of bias

Quote: "Analyses were conducted by the sponsor and were independently verified with the use of the original data by Statogen Consulting." "An independent data and safety monitoring committee reviewed the data and recommended early completion of the trial for efficacy."

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Sivalingam 2024

Low risk of bias

Quote: "One person not involved in the study had access to the computer generated random semaglutide/placebo allocation sequence."

Low risk of bias

Quote: "Investigators, participants and treating physicians were blinded to treatment allocation."

Low risk of bias

Quote: "All participants completed the study and were included into the analysis." Comment: 54/60 participants were included in the analysis. 28/30 participsnts were included in the interevention group and 26/30 In the control group. However, data were seemed to be reported on the ITT

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.40 Withdrawal due to adverse events (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.40.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.41 Hypoglycaemia.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.41.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 1.41.2 All CKD stages

Sivalingam 2024

Low risk of bias

Quote: "One person not involved in the study had access to the computer generated random semaglutide/placebo allocation sequence."

Low risk of bias

Quote: "Investigators, participants and treating physicians were blinded to treatment allocation."

Low risk of bias

Quote: "All participants completed the study and were included into the analysis." Comment: 54/60 participants were included in the analysis. 28/30 participsnts were included in the interevention group and 26/30 In the control group. However, data were seemed to be reported on the ITT

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.42 Lactic acidosis.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.42.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.43 Acute kidney injury.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.43.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 1.43.2 All CKD stages

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.44 Acute kidney injury (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.44.1 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 1.45 Diabetic ketoacidosis.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.45.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.46 HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.46.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 1.46.2 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 1.46.3 Dialysis (HD/PD)

Idorn 2013

Low risk of bias

Quote: “Patients and control subjects were assigned to receive either liraglutide or pIacebo according to a computer‐generated randomisation list provided by Novo Nordisk”.

Low risk of bias

Quote: “Participants, Investigators, and healthcare staff were blinded for the allocated treatment and remained so until the last patient’s last visit”.

High risk of bias

Comment: >5% lost to follow up with discrepancies between groups.

Some concerns

Comment: Objective and subjective outcomes were reported.

Low risk of bias

Comment: Prespecified outcomes were reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 1.47 Change in HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.47.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 1.48 Change in HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.48.1 CKD stages 1‐2

GetGoal studies 2017 ‐ pooled

Some concerns

Comment: No information were reported on sequence generation and allocation concealment.

Low risk of bias

Comment: All studies included in the pooled analyses were double‐blind studies.

Low risk of bias

Quote: "Analyses of efficacy endpoints were performed using the pooled modified intent‐to‐treat (mITT) populations." Comment: Slight difference in rate of discontinuation between the intervention and control group (6% and 4%, respectively) but the reasons were balanced. ITT was performed.

Low risk of bias

Comment: "A data monitoring committee supervised the conduct of the study by an ongoing review of unblinded safety and main efficacy parameters. An allergic reaction adjudication committee (ARAC) performed blinded assessment of potentially allergic or allergic like reactions."

Some concerns

Comment: Only post‐hoc analysis was available, and all prespecified analyses were not reported.

Some concerns

The study is judged to be at some risk of bias in at least one domain for this result.

Subgroup 1.48.2 CKD stages 3‐5

GetGoal‐O 2015

Some concerns

Sequence generation and allocation concealmet methods were not reported in sufficient detail to permit judgement.

Low risk of bias

Quote: "Double‐blinded". Comment: The study was reported as double blind but no further details were reported.

High risk of bias

Comment: Not reported in sufficient detail for people with CKD and diabetes. In the overall population a high rate of particpants discontinued. ITT was not reported.

Some concerns

Comment: Objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported although not all data were extractable. Only HbA1c was extractable. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

High risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

GetGoal studies 2017 ‐ pooled

Some concerns

Comment: No information were reported on sequence generation and allocation concealment.

Low risk of bias

Comment: All studies included in the pooled analyses were double‐blind studies.

Low risk of bias

Quote: "Analyses of efficacy endpoints were performed using the pooled modified intent‐to‐treat (mITT) populations." Comment: Slight difference in rate of discontinuation between the intervention and control group (6% and 4%, respectively) but the reasons were balanced. ITT was performed.

Low risk of bias

Comment: "A data monitoring committee supervised the conduct of the study by an ongoing review of unblinded safety and main efficacy parameters. An allergic reaction adjudication committee (ARAC) performed blinded assessment of potentially allergic or allergic like reactions."

Some concerns

Comment: Only post‐hoc analysis was available, and all prespecified analyses were not reported.

Some concerns

The study is judged to be at some risk of bias in at least one domain for this result.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 1.48.3 Dialysis (HD/PD)

Idorn 2013

Low risk of bias

Quote: “Patients and control subjects were assigned to receive either liraglutide or pIacebo according to a computer‐generated randomisation list provided by Novo Nordisk”.

Low risk of bias

Quote: “Participants, Investigators, and healthcare staff were blinded for the allocated treatment and remained so until the last patient’s last visit”.

High risk of bias

Comment: >5% lost to follow up with discrepancies between groups.

Some concerns

Comment: Objective and subjective outcomes were reported.

Low risk of bias

Comment: Prespecified outcomes were reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 1.49 Coronary revascularisation.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.49.1 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 1.50 Coronary revascularisation (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.50.1 CKD stages 1‐2

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 1.50.2 CKD stages 3‐5

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 2.1 All‐cause death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.1.1 CKD stages 1‐2

Zhou 2019a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported.

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 2.1.2 All CKD stages

Neff 2016

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. No information on imbalance between groups were reported.

High risk of bias

Quote: "Open‐label".

High risk of bias

Comment: 10/10 participants in the treatment group 1 and 8/10 participants in the treatment group 2 completed the study.

Some concerns

Comment: Objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Liu 2022

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. There was no imbalance between

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 2.2 Cardiovascular death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.2.1 CKD stages 1‐2

Zhou 2019a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported.

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 2.2.2 All CKD stages

Neff 2016

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. No information on imbalance between groups were reported.

High risk of bias

Quote: "Open‐label".

High risk of bias

Comment: 10/10 participants in the treatment group 1 and 8/10 participants in the treatment group 2 completed the study.

Some concerns

Comment: Objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Liu 2022

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. There was no imbalance between

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 2.3 Urinary albumin‐creatinine ratio [mg/g].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.3.1 CKD stages 1‐2

Zhou 2019a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported.

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 2.4 Microalbuminuria [mg/L].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.4.1 CKD stages 1‐2

Zhou 2019a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported.

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 2.5 Urinary albumin excretion rate [µg/min].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.5.1 All CKD stages

Neff 2016

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. No information on imbalance between groups were reported.

High risk of bias

Quote: "Open‐label".

High risk of bias

Comment: 10/10 participants in the treatment group 1 and 8/10 participants in the treatment group 2 completed the study.

Some concerns

Comment: Objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Liu 2022

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. There was no imbalance between

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 2.6 BMI [kg/m²].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.6.1 All CKD stages

Liu 2022

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. There was no imbalance between

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 2.7 Serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.7.1 All CKD stages

Neff 2016

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. No information on imbalance between groups were reported.

High risk of bias

Quote: "Open‐label".

High risk of bias

Comment: 10/10 participants in the treatment group 1 and 8/10 participants in the treatment group 2 completed the study.

Some concerns

Comment: Objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 2.8 Withdrawal due to adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.8.1 All CKD stages

Neff 2016

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. No information on imbalance between groups were reported.

High risk of bias

Quote: "Open‐label".

High risk of bias

Comment: 10/10 participants in the treatment group 1 and 8/10 participants in the treatment group 2 completed the study.

Some concerns

Comment: Objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 2.9 Hypoglycaemia.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.9.1 All CKD stages

Neff 2016

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. No information on imbalance between groups were reported.

High risk of bias

Quote: "Open‐label".

High risk of bias

Comment: 10/10 participants in the treatment group 1 and 8/10 participants in the treatment group 2 completed the study.

Some concerns

Comment: Objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 2.10 Acute kidney injury.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.10.1 All CKD stages

Neff 2016

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. No information on imbalance between groups were reported.

High risk of bias

Quote: "Open‐label".

High risk of bias

Comment: 10/10 participants in the treatment group 1 and 8/10 participants in the treatment group 2 completed the study.

Some concerns

Comment: Objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 2.11 HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.11.1 CKD stages 1‐2

Zhou 2019a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported.

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 2.11.2 All CKD stages

Liu 2022

Some concerns

Comment: Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. There was no imbalance between

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups

Some concerns

Comment: Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 2.12 Change in HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.12.1 CKD stages 1‐2

Zhou 2019a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported.

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 3.1 All‐cause death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 3.1.1 All CKD stages

Leiter 2014

Low risk of bias

Quote: “An interactive voice response system was used for the blinded randomisation, which was based on a sequestered fixed randomisation schedule.”

Low risk of bias

Quote: “Double‐blind.”

High risk of bias

Comment: The ratios of discontinuation were almost balanced between the groups (68/253 [26.9%] in the sitagliptin group and 51/254 [20.1%] in the albiglutide group). The rate of withdrawal was nearly twice in the sitagliptin group (10.3%) compared to the albiglutide group (4.7%).

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results. However, the majority of data were reported at 26 weeks only, according to the protocol.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 3.2 Cardiovascular death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 3.2.1 All CKD stages

Leiter 2014

Low risk of bias

Quote: “An interactive voice response system was used for the blinded randomisation, which was based on a sequestered fixed randomisation schedule.”

Low risk of bias

Quote: “Double‐blind.”

High risk of bias

Comment: The ratios of discontinuation were almost balanced between the groups (68/253 [26.9%] in the sitagliptin group and 51/254 [20.1%] in the albiglutide group). The rate of withdrawal was nearly twice in the sitagliptin group (10.3%) compared to the albiglutide group (4.7%).

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results. However, the majority of data were reported at 26 weeks only, according to the protocol.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 3.3 Severe hypoglycaemia.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 3.3.1 All CKD stages

Leiter 2014

Low risk of bias

Quote: “An interactive voice response system was used for the blinded randomisation, which was based on a sequestered fixed randomisation schedule.”

Low risk of bias

Quote: “Double‐blind.”

High risk of bias

Comment: The ratios of discontinuation were almost balanced between the groups (68/253 [26.9%] in the sitagliptin group and 51/254 [20.1%] in the albiglutide group). The rate of withdrawal was nearly twice in the sitagliptin group (10.3%) compared to the albiglutide group (4.7%).

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results. However, the majority of data were reported at 26 weeks only, according to the protocol.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 3.4 Fatal stroke.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 3.4.1 All CKD stages

Leiter 2014

Low risk of bias

Quote: “An interactive voice response system was used for the blinded randomisation, which was based on a sequestered fixed randomisation schedule.”

Low risk of bias

Quote: “Double‐blind.”

High risk of bias

Comment: The ratios of discontinuation were almost balanced between the groups (68/253 [26.9%] in the sitagliptin group and 51/254 [20.1%] in the albiglutide group). The rate of withdrawal was nearly twice in the sitagliptin group (10.3%) compared to the albiglutide group (4.7%).

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results. However, the majority of data were reported at 26 weeks only, according to the protocol.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 3.5 Serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 3.5.1 All CKD stages

Leiter 2014

Low risk of bias

Quote: “An interactive voice response system was used for the blinded randomisation, which was based on a sequestered fixed randomisation schedule.”

Low risk of bias

Quote: “Double‐blind.”

High risk of bias

Comment: The ratios of discontinuation were almost balanced between the groups (68/253 [26.9%] in the sitagliptin group and 51/254 [20.1%] in the albiglutide group). The rate of withdrawal was nearly twice in the sitagliptin group (10.3%) compared to the albiglutide group (4.7%).

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results. However, the majority of data were reported at 26 weeks only, according to the protocol.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 3.6 Withdrawal due to adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 3.6.1 All CKD stages

Leiter 2014

Low risk of bias

Quote: “An interactive voice response system was used for the blinded randomisation, which was based on a sequestered fixed randomisation schedule.”

Low risk of bias

Quote: “Double‐blind.”

High risk of bias

Comment: The ratios of discontinuation were almost balanced between the groups (68/253 [26.9%] in the sitagliptin group and 51/254 [20.1%] in the albiglutide group). The rate of withdrawal was nearly twice in the sitagliptin group (10.3%) compared to the albiglutide group (4.7%).

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results. However, the majority of data were reported at 26 weeks only, according to the protocol.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 3.7 Hypoglycaemia.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 3.7.1 All CKD stages

Leiter 2014

Low risk of bias

Quote: “An interactive voice response system was used for the blinded randomisation, which was based on a sequestered fixed randomisation schedule.”

Low risk of bias

Quote: “Double‐blind.”

High risk of bias

Comment: The ratios of discontinuation were almost balanced between the groups (68/253 [26.9%] in the sitagliptin group and 51/254 [20.1%] in the albiglutide group). The rate of withdrawal was nearly twice in the sitagliptin group (10.3%) compared to the albiglutide group (4.7%).

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results. However, the majority of data were reported at 26 weeks only, according to the protocol.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 3.8 HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 3.8.1 CKD stages 1‐2

Leiter 2014

Low risk of bias

Quote: “An interactive voice response system was used for the blinded randomisation, which was based on a sequestered fixed randomisation schedule.”

Low risk of bias

Quote: “Double‐blind.”

High risk of bias

Comment: The ratios of discontinuation were almost balanced between the groups (68/253 [26.9%] in the sitagliptin group and 51/254 [20.1%] in the albiglutide group). The rate of withdrawal was nearly twice in the sitagliptin group (10.3%) compared to the albiglutide group (4.7%).

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results. However, the majority of data were reported at 26 weeks only, according to the protocol.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 3.8.2 CKD stages 3‐5

Leiter 2014

Low risk of bias

Quote: “An interactive voice response system was used for the blinded randomisation, which was based on a sequestered fixed randomisation schedule.”

Low risk of bias

Quote: “Double‐blind.”

High risk of bias

Comment: The ratios of discontinuation were almost balanced between the groups (68/253 [26.9%] in the sitagliptin group and 51/254 [20.1%] in the albiglutide group). The rate of withdrawal was nearly twice in the sitagliptin group (10.3%) compared to the albiglutide group (4.7%).

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results. However, the majority of data were reported at 26 weeks only, according to the protocol.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 3.9 Change in HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 3.9.1 CKD stages 1‐2

Leiter 2014

Low risk of bias

Quote: “An interactive voice response system was used for the blinded randomisation, which was based on a sequestered fixed randomisation schedule.”

Low risk of bias

Quote: “Double‐blind.”

High risk of bias

Comment: The ratios of discontinuation were almost balanced between the groups (68/253 [26.9%] in the sitagliptin group and 51/254 [20.1%] in the albiglutide group). The rate of withdrawal was nearly twice in the sitagliptin group (10.3%) compared to the albiglutide group (4.7%).

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results. However, the majority of data were reported at 26 weeks only, according to the protocol.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 3.9.2 CKD stages 3‐5

Leiter 2014

Low risk of bias

Quote: “An interactive voice response system was used for the blinded randomisation, which was based on a sequestered fixed randomisation schedule.”

Low risk of bias

Quote: “Double‐blind.”

High risk of bias

Comment: The ratios of discontinuation were almost balanced between the groups (68/253 [26.9%] in the sitagliptin group and 51/254 [20.1%] in the albiglutide group). The rate of withdrawal was nearly twice in the sitagliptin group (10.3%) compared to the albiglutide group (4.7%).

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results. However, the majority of data were reported at 26 weeks only, according to the protocol.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.1 All‐cause death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.1.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.2 All‐cause death (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.2.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.3 Cardiovascular death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.3.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open‐­label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.4 Cardiovascular death (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.4.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.5 Kidney failure.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.5.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.6 Severe hypoglycaemia.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.6.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 4.6.2 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.7 Myocardial infarction.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.7.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.8 Heart failure.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.8.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.9 Change in eGFR [mL/min].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.9.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 4.9.2 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.10 Change in eGFR [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.10.1 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.11 eGFR loss.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.11.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.12 Change in creatinine clearance [mL/min].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.12.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.13 Urinary albumin‐creatinine ratio [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.13.1 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.14 Urinary albumin‐creatinine ratio [mg/g].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.14.1 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.15 Urinary albumin excretion rate [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.15.1 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.16 Urinary albumin excretion rate [mg/24 h].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.16.1 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.17 Change in body weight [kg].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.17.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 4.17.2 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.18 Change in body weight [kg].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.18.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.19 Body weight increase.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.19.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.20 Serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.20.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 4.20.2 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.21 Withdrawal due to adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.21.1 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.22 Hypoglycaemia.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.22.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 4.22.2 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.23 Acute kidney injury.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.23.1 CKD stages 1‐2

Muskiet 2019

Some concerns

Sequence generation and allocation concealment methods were not reported in sufficient detail to permit judgement. No information on imbalance between groups were reported.

High risk of bias

Quote: "Open‐label".

Some concerns

Not reported in sufficient detail to perform adjudication were reported on attrition (post hoc analysis).

Some concerns

Objective and subjective outcomes were reported.

Some concerns

Prespecified outcomes were reported. Clinically‐relevant outcomes that would be expected for this type of intervention were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 4.23.2 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.24 Change in HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.24.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 4.24.2 All CKD stages

Wang 2020a

Low risk of bias

Quote: "Based on randomized (1:1), open‐label, parallel‐group, controlled study design, a block randomisation method was applied in this study." "To prevent selection bias, we developed an allocation concealment mechanism, whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work." No imbalance between treatment groups were apparent.

High risk of bias

Quote: "Open‐label study".

High risk of bias

Quote: "Individuals were screened and 92 individuals were randomized and allocated to 2 groups (46 individuals per group). Eleven patients were excluded from the effective analysis due to missing follow‐up data after randomisation. Eight patients discontinued IP: premature intervention discontinued due to AE (N = 6), withdrawal of consent (N = 1), and lost to follow‐up (N = 1). These 8 patients were included in the FAS analysis but not in the PPS analysis.Eighty‐one (88.0%) patients (43 in the intervention group, 38 in the control group) in total were included for the FAS analysis and 73 (79.3%) (37 patients in the intervention group, 36 patients in the control group) patients were included in the PPS." Comment: Fig 1 showed that 43/46 participants in the treatment group and 38/46 participants in the control group were included into the FAS analysis. 37/46 participants in the treatment group and 36/46 participants in the control group were included into the PPS analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.25 Change in HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.25.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

DUAL II 2014

Low risk of bias

Quote: "A central interactive voice/web system, participants were randomly allocated 1:1 to receive (treatment)."

Low risk of bias

Quote: "Treatment was blinded for investigators and participants via use of visually identical trial drugs. Blinding was maintained for all involved in the trial (including titration, event adjudication, and calcitonin monitoring) until the database was released for statistical analyses."

Low risk of bias

Comment: Some participants discontinued but ITT performed in the majority of participants. The rates of discontinuation were 32/207 (15.5%) and 35/206 (17.0%) for IDegLira group and insulin degludec group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almos the same between the groups.

Low risk of bias

Quote: "An external independent event adjudication committee (EAC) performed ongoing adjudication of cardiovascular events, pancreatitis, neoplasms, and thyroid disease requiring thyroidectomy."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

DUAL I 2014

Low risk of bias

Quote: "Randomisation was done with an interactive voice or web system, with stratified cation by concomitant oral antidiabetic treatment, baseline HbA1c (≤8·3% and >8·3%), and whether or not the patient was to participate in a meal test sub study"

High risk of bias

Quote: "Patients and all other investigators were not masked to treatment assignment."

Low risk of bias

Comment: The rates of discontinuation were high. However, all participants were included into the analyses.

Low risk of bias

Quote: "Treatment assignment was masked for a safety committee (responsible for safety surveillance), an independent external committee that adjudicated selected adverse events, and personnel involved in defining the analysis sets until the database was released for statistical analysis."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

DUAL V 2016

Low risk of bias

Quote: "Patients were randomised 1:1 via an interactive voice/web response system to receive (treatment)."

High risk of bias

Comment: An open‐label trial.

Low risk of bias

Comment: Withdrawal rate > 10% but all participants were included in the analyses.

Low risk of bias

Quote: "The event adjudication committee, who adjudicated cardiovascular, neoplasm, thyroid disease, or pancreatitis events were blinded to randomised treatment."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.26 Change in HbA1c [mmol/mol].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.26.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 4.27 Ischaemic stroke.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 4.27.1 CKD stages 3‐5

AWARD‐7 2017

Low risk of bias

Quote: "Participants were randomly assigned (1:1:1) by use of a computer­‐generated random sequence with an interactive response system."

High risk of bias

Quote: "An open­‐label study with respect to assignment to dulaglutide or insulin, but with the investigators and patients masked to the dose of dulaglutide."

Low risk of bias

Comment: ITT was performed. All but one participant in dulaglutide 1.5 mg group were included in the safety analyses. Less than 5% of participants were excluded from the ITT. Reason for discontinuation were provided and were quite high in all groups.

Low risk of bias

Quote: "Death and cardiovascular, pancreatitis, and kidney events were prospectively adjudicated by an independent clinical endpoint committee on the basis of a prespecified adjudication process (adjudication committee members were masked to treatment assignments)."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Risk of bias for analysis 5.1 All‐cause death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 5.1.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 5.2 Cardiovascular death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 5.2.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 5.3 Withdrawal due to adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 5.3.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 5.4 HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 5.4.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 5.5 Body weight [kg].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 5.5.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 5.6 BMI [kg/m²].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 5.6.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 6.1 All‐cause death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 6.1.1 All CKD stages

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 6.2 Cardiovascular death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 6.2.1 All CKD stages

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 6.3 Withdrawal due to adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 6.3.1 All CKD stages

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 6.4 Acute kidney injury.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 6.4.1 All CKD stages

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 8.1 All‐cause death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 8.1.1 CKD stages 1‐2

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 8.1.2 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 8.1.3 All CKD stages

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Sivalingam 2024

Low risk of bias

Quote: "One person not involved in the study had access to the computer generated random semaglutide/placebo allocation sequence."

Low risk of bias

Quote: "Investigators, participants and treating physicians were blinded to treatment allocation."

Low risk of bias

Quote: "All participants completed the study and were included into the analysis." Comment: 54/60 participants were included in the analysis. 28/30 participsnts were included in the interevention group and 26/30 In the control group. However, data were seemed to be reported on the ITT

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 8.2 Cardiovascular death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 8.2.1 CKD stages 1‐2

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Low risk of bias

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Low risk of bias

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Low risk of bias

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 8.2.2 CKD stages 3‐5

EXSCEL 2017

Low risk of bias

Quote: "An interactive voice‐response system assigned patients on the basis of computer generated block randomization."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: The rates of non‐completers were 262/7356(3.6%) and 303/7396(4.1%) for exenatide group and placebo group, respectively. The rate and cause of adverse events leading to discontinuation of study drugs were almost the same between the groups (43.0% and 45.2% for exenatide group and placebo group, respectively). ITT was performed.

Low risk of bias

Quote: "The events which constitute the principal endpoints of this trial will be adjudicated by the Clinical Events Classification Committee (CEC), coordinated through the Duke Clinical Research Institute (DCRI), which will be comprised of approximately 5‐7 physicians and a coordinator."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

LEADER 2017

Low risk of bias

Quote: "A randomisation session will be carried out for all subjects by using the Interactive Voice/Web Response System."

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: Overall, 139/4668 (3.0%) of the liraglutide group and 159/4672 (3.4%) of the control group did not complete the study. ITT was performed.

Low risk of bias

Quote: "External Event Adjudication Committees (EAC) will be constituted for the trial to perform ongoing adjudication of the below events in an independent and blinded manner."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 8.2.3 All CKD stages

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 8.3 Change in body weight [kg].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 8.3.1 CKD stages 1‐2

GetGoal studies 2017 ‐ pooled

Some concerns

Comment: No information were reported on sequence generation and allocation concealment.

Low risk of bias

Comment: All studies included in the pooled analyses were double‐blind studies.

Low risk of bias

Quote: "Analyses of efficacy endpoints were performed using the pooled modified intent‐to‐treat (mITT) populations." Comment: Slight difference in rate of discontinuation between the intervention and control group (6% and 4%, respectively) but the reasons were balanced. ITT was performed.

Low risk of bias

Comment: "A data monitoring committee supervised the conduct of the study by an ongoing review of unblinded safety and main efficacy parameters. An allergic reaction adjudication committee (ARAC) performed blinded assessment of potentially allergic or allergic like reactions."

Some concerns

Comment: Only post‐hoc analysis was available, and all prespecified analyses were not reported.

Some concerns

The study is judged to be at some risk of bias in at least one domain for this result.

Subgroup 8.3.2 CKD stages 3‐5

GetGoal studies 2017 ‐ pooled

Some concerns

Comment: No information were reported on sequence generation and allocation concealment.

Low risk of bias

Comment: All studies included in the pooled analyses were double‐blind studies.

Low risk of bias

Quote: "Analyses of efficacy endpoints were performed using the pooled modified intent‐to‐treat (mITT) populations." Comment: Slight difference in rate of discontinuation between the intervention and control group (6% and 4%, respectively) but the reasons were balanced. ITT was performed.

Low risk of bias

Comment: "A data monitoring committee supervised the conduct of the study by an ongoing review of unblinded safety and main efficacy parameters. An allergic reaction adjudication committee (ARAC) performed blinded assessment of potentially allergic or allergic like reactions."

Some concerns

Comment: Only post‐hoc analysis was available, and all prespecified analyses were not reported.

Some concerns

The study is judged to be at some risk of bias in at least one domain for this result.

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 8.4 HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 8.4.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 8.4.2 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 8.5 Change in HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 8.5.1 CKD stages 1‐2

GetGoal studies 2017 ‐ pooled

Some concerns

Comment: No information were reported on sequence generation and allocation concealment.

Low risk of bias

Comment: All studies included in the pooled analyses were double‐blind studies.

Low risk of bias

Quote: "Analyses of efficacy endpoints were performed using the pooled modified intent‐to‐treat (mITT) populations." Comment: Slight difference in rate of discontinuation between the intervention and control group (6% and 4%, respectively) but the reasons were balanced. ITT was performed.

Low risk of bias

Comment: "A data monitoring committee supervised the conduct of the study by an ongoing review of unblinded safety and main efficacy parameters. An allergic reaction adjudication committee (ARAC) performed blinded assessment of potentially allergic or allergic like reactions."

Some concerns

Comment: Only post‐hoc analysis was available, and all prespecified analyses were not reported.

Some concerns

The study is judged to be at some risk of bias in at least one domain for this result.

Subgroup 8.5.2 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

GetGoal studies 2017 ‐ pooled

Some concerns

Comment: No information were reported on sequence generation and allocation concealment.

Low risk of bias

Comment: All studies included in the pooled analyses were double‐blind studies.

Low risk of bias

Quote: "Analyses of efficacy endpoints were performed using the pooled modified intent‐to‐treat (mITT) populations." Comment: Slight difference in rate of discontinuation between the intervention and control group (6% and 4%, respectively) but the reasons were balanced. ITT was performed.

Low risk of bias

Comment: "A data monitoring committee supervised the conduct of the study by an ongoing review of unblinded safety and main efficacy parameters. An allergic reaction adjudication committee (ARAC) performed blinded assessment of potentially allergic or allergic like reactions."

Some concerns

Comment: Only post‐hoc analysis was available, and all prespecified analyses were not reported.

Some concerns

The study is judged to be at some risk of bias in at least one domain for this result.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 9.1 All‐cause death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.1.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 9.1.2 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 9.1.3 Dialysis (HD/PD)

Idorn 2013

Low risk of bias

Quote: “Patients and control subjects were assigned to receive either liraglutide or pIacebo according to a computer‐generated randomisation list provided by Novo Nordisk”.

Low risk of bias

Quote: “Participants, Investigators, and healthcare staff were blinded for the allocated treatment and remained so until the last patient’s last visit”.

High risk of bias

Comment: >5% lost to follow up with discrepancies between groups.

Some concerns

Comment: Objective and subjective outcomes were reported.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 9.1.4 All CKD stages

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Sivalingam 2024

Low risk of bias

Quote: "One person not involved in the study had access to the computer generated random semaglutide/placebo allocation sequence."

Low risk of bias

Quote: "Investigators, participants and treating physicians were blinded to treatment allocation."

Low risk of bias

Quote: "All participants completed the study and were included into the analysis." Comment: 54/60 participants were included in the analysis. 28/30 participsnts were included in the interevention group and 26/30 In the control group. However, data were seemed to be reported on the ITT

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 9.2 All‐cause death (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.2.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 9.3 Cardiovascular death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.3.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Low risk of bias

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Low risk of bias

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Low risk of bias

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 9.3.2 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 9.3.3 Dialysis (HD/PD)

Idorn 2013

Low risk of bias

Quote: “Patients and control subjects were assigned to receive either liraglutide or pIacebo according to a computer‐generated randomisation list provided by Novo Nordisk”.

Low risk of bias

Quote: “Participants, Investigators, and healthcare staff were blinded for the allocated treatment and remained so until the last patient’s last visit”.

High risk of bias

Comment: >5% lost to follow up with discrepancies between groups.

Some concerns

Comment: Objective and subjective outcomes were reported.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 9.3.4 All CKD stages

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 9.4 Cardiovascular death (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.4.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 9.5 3‐point MACE.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.5.1 CKD stages 3‐5

SUSTAIN‐6 2016

Low risk of bias

Quote: "A randomisation session will be carried out by using the IV/WRS (interactive voice/web response system)."

Low risk of bias

Quote: "Double‐blind study."

Low risk of bias

Comment: Overall, 15/1648 (1.5%) of the semaglutide group and 40/1649 (2.4%) of the control group did not complete the trial. The rate of treatment discontinuation was higher in the semaglutide group (214/1648, 13.0%) than the placebo group (110/1649, 6.6%), which was mainly from gastrointestinal disorders. ITT was performed.

Low risk of bias

Quote: "The components of the primary endpoint and other key safety outcomes will be subject to blinded external adjudication, and an independent Data Monitoring Committee (DMC) will oversee the safety of trial participants."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 9.5.2 All CKD stages

AMPLITUDE‐O 2021

Low risk of bias

Quote: "Randomization was performed through an interactive Web‐response system and was stratified according to current or potential future use of SGLT2 inhibitors (i.e., current use at the time of randomisation, SGLT2 inhibitor likely to be added to therapy, or SGLT2 inhibitor unlikely to be added to therapy) to minimize between group differences in therapy with various cardioprotective medications."

Low risk of bias

Quote: "Double‐blind". "Only the data monitoring committee had access to unmasked data until the database lock." "Efpeglenatide or placebo was added to each participant’s current therapy in a blinded manner." Comment: The study was reported as double blind and it was likely that all people involved (excluding data monitoring) were blinded.

Low risk of bias

Comment: All participants were included in the ITT analysis. Discontinuation was reported in sufficient detail to perform adjudication in people with CKD and diabetes.

Low risk of bias

Quote: "Only the data monitoring committee had access to unmasked data until the database lock." Comment: Objective and subjective outcomes were reported. It was not stated if the data monitoring was blinded to the treatment assigned.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported in people with CKD and diabetes.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 9.6 3‐point MACE (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.6.1 CKD stages 1‐2

Harmony Outcomes 2018

Low risk of bias

Quote: "Patients were assigned in a 1:1 ratio to receive subcutaneous injections of albiglutide or placebo (with matched administration device, diluent, and volume injected) once a week, according to a sequestered, fixed, computer‐generated randomisation code that used balanced permuted blocks of treatment group allocations, without stratification. Investigators used an interactive voice or web response system to obtain treatment assignment."

Low risk of bias

Quote: "All investigators and patients involved in the trial were masked to treatment group."

Low risk of bias

Comment: Overall, 1161/4731 (24.5%) of albiglutide group and 1318/4732 (27.9%) of placebo group discontinued treatment. The proportion of withdrawal were 2.3% in albiglutide group and 3.2% in placebo group. Regarding the reason for treatment discontinuation, the ratios of AEs were almost the same between the arms; albiglutide group (8.6%), placebo (6.4%). ITT was performed.

Low risk of bias

Quote: "An IDMC (Independent Data Monitoring Committee) will have study oversight to ensure participant safety and scientific integrity of the data (Section 9.8), an independent Cardiovascular Endpoint Committee (CEC) blinded to treatment allocation will adjudicate cardiovascular outcome events and a Pancreatitis Adjudication Committee will adjudicate potential events of pancreatitis."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Subgroup 9.6.2 CKD stages 3‐5

Harmony Outcomes 2018

Low risk of bias

Quote: "Patients were assigned in a 1:1 ratio to receive subcutaneous injections of albiglutide or placebo (with matched administration device, diluent, and volume injected) once a week, according to a sequestered, fixed, computer‐generated randomisation code that used balanced permuted blocks of treatment group allocations, without stratification. Investigators used an interactive voice or web response system to obtain treatment assignment."

Low risk of bias

Quote: "All investigators and patients involved in the trial were masked to treatment group."

Low risk of bias

Comment: Overall, 1161/4731 (24.5%) of albiglutide group and 1318/4732 (27.9%) of placebo group discontinued treatment. The proportion of withdrawal were 2.3% in albiglutide group and 3.2% in placebo group. Regarding the reason for treatment discontinuation, the ratios of AEs were almost the same between the arms; albiglutide group (8.6%), placebo (6.4%). ITT was performed.

Low risk of bias

Quote: "An IDMC (Independent Data Monitoring Committee) will have study oversight to ensure participant safety and scientific integrity of the data (Section 9.8), an independent Cardiovascular Endpoint Committee (CEC) blinded to treatment allocation will adjudicate cardiovascular outcome events and a Pancreatitis Adjudication Committee will adjudicate potential events of pancreatitis."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

PIONEER 6 2019

Low risk of bias

Quote: "Randomization was performed using an interactive voice/web response system (IV/WRS)".

Low risk of bias

Comment: Double‐blind study.

Low risk of bias

Comment: 5/1591 (0.3%) of canagliflozin group and 6/1592 (0.4%) of placebo group did not complete the study. The rates (15.3% and 9.9% for semaglutide and placebo, respectively) and causes of discontinuation of study drugs were almost the same between the groups.

Low risk of bias

Comment: An independent data monitoring committee evaluated unblinded trial data.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

SUSTAIN‐6 2016

Low risk of bias

Quote: "A randomisation session will be carried out by using the IV/WRS (interactive voice/web response system)."

Low risk of bias

Quote: "Double‐blind study."

Low risk of bias

Comment: Overall, 15/1648 (1.5%) of the semaglutide group and 40/1649 (2.4%) of the control group did not complete the trial. The rate of treatment discontinuation was higher in the semaglutide group (214/1648, 13.0%) than the placebo group (110/1649, 6.6%), which was mainly from gastrointestinal disorders. ITT was performed.

Low risk of bias

Quote: "The components of the primary endpoint and other key safety outcomes will be subject to blinded external adjudication, and an independent Data Monitoring Committee (DMC) will oversee the safety of trial participants."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result..

Subgroup 9.6.3 All CKD stages

AMPLITUDE‐O 2021

Low risk of bias

Quote: "Randomization was performed through an interactive Web‐response system and was stratified according to current or potential future use of SGLT2 inhibitors (i.e., current use at the time of randomisation, SGLT2 inhibitor likely to be added to therapy, or SGLT2 inhibitor unlikely to be added to therapy) to minimize between group differences in therapy with various cardioprotective medications."

Low risk of bias

Quote: "Double‐blind". "Only the data monitoring committee had access to unmasked data until the database lock." "Efpeglenatide or placebo was added to each participant’s current therapy in a blinded manner." Comment: The study was reported as double blind and it was likely that all people involved (excluding data monitoring) were blinded.

Low risk of bias

Comment: All participants were included in the ITT analysis. However, discontinuation was reported in sufficient detail to perform adjudication in people with CKD and diabetes.

Low risk of bias

Quote: "Only the data monitoring committee had access to unmasked data until the database lock." Comment: Objective and subjective outcomes were reported. It was not stated if the data monitoring was blinded to the treatment assigned.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported in people with CKD and diabetes.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 9.7 4‐point MACE (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.7.1 CKD stages 3‐5

ELIXA 2018

Low risk of bias

Quote: "The randomised treatment kit number list will be generated centrally by Sanofi. The Investigational Product (IP) (lixisenatide or placebo) will be packaged in accordance with this list." […] "The Supply Chain Trial Manager (SCTM) will provide the randomized treatment kit number list and the Study Biostatistician will provide the randomisation scheme to the centralized treatment allocation system. Then, this centralised treatment allocation system will generate the patient randomisation list according to which it will allocate the treatments to the patients." "Two types of centralized treatment allocation system will be used, the Interactive Voice Response System (IVRS) and the Interactive Web Response System (IWRS) depending on the choice of the site."

Low risk of bias

Quote: "Lixisenatide and placebo were indistinguishable. Patients, study staff, and individuals involved in analysis of trial data were masked to treatment allocation.”

Low risk of bias

Comment: The rates of discontinuation were 105/3034 (3.5%) and 110/3034 (3.6%) for lixisenatide group and placebo group, respectively. The rate of adverse events leading to discontinuation of study drugs were 20.6% and 22.1% for lixisenatide group and placebo group, respectively. ITT was performed.

Low risk of bias

Quote: "Separate independent committees whose members were unaware of the study group assignments adjudicated potential cardiovascular, pancreatic, and allergic events."

Low risk of bias

Quote: "Separate independent committees whose members were unaware of the study group assignments adjudicated potential cardiovascular, pancreatic, and allergic events."

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 9.8 Kidney failure.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.8.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 9.9 Composite kidney outcome (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.9.1 CKD stages 3‐5

REWIND 2019

Low risk of bias

Quote: "Randomisation was done by a computer‐generated random code with an interactive web response system with stratification by site."

Low risk of bias

Quote: "All investigators and participants were masked to treatment allocation."

Low risk of bias

Comment: Overall, 17/4949 of dulaglutide group and 17/4952 of placebo group did not complete the study. The rates (9.1% [451/4949] and 6.3% [310/4952] for dulaglutide and placebo, respectively). Incidences of adverse events were almost the same between the groups. ITT was performed.

Low risk of bias

Quote: "An independent clinical endpoint committee (CEC) will adjudicate CV events, pancreatitis events, thyroid evaluations that result in a biopsy or thyroidectomy, and all deaths."

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 9.10 Severe hypoglycaemia.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.10.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 9.11 Myocardial infarction.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.11.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 9.12 Fatal or nonfatal stroke.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.12.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 9.13 Hospitalisation due to heart failure.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.13.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 9.14 Withdrawal due to adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.14.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 9.14.2 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 9.14.3 All CKD stages

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Sivalingam 2024

Low risk of bias

Quote: "One person not involved in the study had access to the computer generated random semaglutide/placebo allocation sequence."

Low risk of bias

Quote: "Investigators, participants and treating physicians were blinded to treatment allocation."

Low risk of bias

Quote: "All participants completed the study and were included into the analysis." Comment: 54/60 participants were included in the analysis. 28/30 participsnts were included in the interevention group and 26/30 In the control group. However, data were seemed to be reported on the ITT

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 9.15 Withdrawal due to adverse events (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.15.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 9.16 Acute kidney injury (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 9.16.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 10.1 All‐cause death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.1.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 10.1.2 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 10.1.3 Dialysis (HD/PD)

Idorn 2013

Low risk of bias

Quote: “Patients and control subjects were assigned to receive either liraglutide or pIacebo according to a computer‐generated randomisation list provided by Novo Nordisk”.

Low risk of bias

Quote: “Participants, Investigators, and healthcare staff were blinded for the allocated treatment and remained so until the last patient’s last visit”.

High risk of bias

Comment: >5% lost to follow up with discrepancies between groups.

Some concerns

Comment: Objective and subjective outcomes were reported.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 10.1.4 All CKD stages

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Sivalingam 2024

Low risk of bias

Quote: "One person not involved in the study had access to the computer generated random semaglutide/placebo allocation sequence."

Low risk of bias

Quote: "Investigators, participants and treating physicians were blinded to treatment allocation."

Low risk of bias

Quote: "All participants completed the study and were included into the analysis." Comment: 54/60 participants were included in the analysis. 28/30 participsnts were included in the interevention group and 26/30 In the control group. However, data were seemed to be reported on the ITT

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 10.2 All‐cause death (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.2.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 10.3 Cardiovascular death.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.3.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Low risk of bias

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Low risk of bias

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Low risk of bias

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 10.3.2 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 10.3.3 Dialysis (HD/PD)

Idorn 2013

Low risk of bias

Quote: “Patients and control subjects were assigned to receive either liraglutide or pIacebo according to a computer‐generated randomisation list provided by Novo Nordisk”.

Low risk of bias

Quote: “Participants, Investigators, and healthcare staff were blinded for the allocated treatment and remained so until the last patient’s last visit”.

High risk of bias

Comment: >5% lost to follow up with discrepancies between groups.

Some concerns

Comment: Objective and subjective outcomes were reported.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

High risk of bias

The study is judged to be at high risk of bias in at least one domain for this result.

Subgroup 10.3.4 All CKD stages

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 10.4 Cardiovascular death (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.4.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 10.5 3‐point MACE.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.5.1 CKD stages 3‐5

SUSTAIN‐6 2016

Low risk of bias

Quote: "A randomisation session will be carried out by using the IV/WRS (interactive voice/web response system)."

Low risk of bias

Quote: "Double‐blind study."

Low risk of bias

Comment: Overall, 15/1648 (1.5%) of the semaglutide group and 40/1649 (2.4%) of the control group did not complete the trial. The rate of treatment discontinuation was higher in the semaglutide group (214/1648, 13.0%) than the placebo group (110/1649, 6.6%), which was mainly from gastrointestinal disorders. ITT was performed.

Low risk of bias

Quote: "The components of the primary endpoint and other key safety outcomes will be subject to blinded external adjudication, and an independent Data Monitoring Committee (DMC) will oversee the safety of trial participants."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result.

Risk of bias for analysis 10.6 3‐point MACE (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.6.1 CKD stages 3‐5

SUSTAIN‐6 2016

Low risk of bias

Quote: "A randomisation session will be carried out by using the IV/WRS (interactive voice/web response system)."

Low risk of bias

Quote: "Double‐blind study."

Low risk of bias

Comment: Overall, 15/1648 (1.5%) of the semaglutide group and 40/1649 (2.4%) of the control group did not complete the trial. The rate of treatment discontinuation was higher in the semaglutide group (214/1648, 13.0%) than the placebo group (110/1649, 6.6%), which was mainly from gastrointestinal disorders. ITT was performed.

Low risk of bias

Quote: "The components of the primary endpoint and other key safety outcomes will be subject to blinded external adjudication, and an independent Data Monitoring Committee (DMC) will oversee the safety of trial participants."

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Low risk of bias

The study is judged to be at low risk of bias for all domains for this result..

Risk of bias for analysis 10.7 Kidney failure.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.7.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 10.8 Severe hypoglycaemia.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.8.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 10.9 Myocardial infarction.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.9.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 10.10 Fatal or nonfatal stroke.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.10.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 10.11 Hospitalisation due to heart failure.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.11.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 10.12 Withdrawal due to adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.12.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Subgroup 10.12.2 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Subgroup 10.12.3 All CKD stages

Selvarajah 2024a

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Selvarajah 2024b

Some concerns

Comment: Study was described as randomised, but a method of randomisation was not reported

Low risk of bias

Quote: "Double blind."

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Sivalingam 2024

Low risk of bias

Quote: "One person not involved in the study had access to the computer generated random semaglutide/placebo allocation sequence."

Low risk of bias

Quote: "Investigators, participants and treating physicians were blinded to treatment allocation."

Low risk of bias

Quote: "All participants completed the study and were included into the analysis." Comment: 54/60 participants were included in the analysis. 28/30 participsnts were included in the interevention group and 26/30 In the control group. However, data were seemed to be reported on the ITT

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 10.13 Withdrawal due to adverse events (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.13.1 CKD stages 3‐5

LIRA‐RENAL 2016

Low risk of bias

Quote: "Trial patients who met the eligibility criteria at screening were randomised (1:1), using a sponsor‐provided telephone‐ or Web‐based randomisation system, to receive (treatment)"

Low risk of bias

Quote: "Double blind study."

Low risk of bias

Comment: Discontinuation rates were almost the same between two groups: 35/140 (25%) from the liraglutide group versus 34/137 (24.8%) in the control group. However, the high rate of adverse events in the liraglutide group compared to the control group (13.6% versus 2.9%) might contribute to the attrition bias. ITT was performed.

Some concerns

Comment: No information provided detailing specifically how outcome assessors were kept blind to treatment groups.

Low risk of bias

Comment: All outcomes that were planned in the methods were reported in the results.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Risk of bias for analysis 10.14 Acute kidney injury (Hazard ratio).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 10.14.1 CKD stages 3‐5

PIONEER 5 2019

Low risk of bias

Quote: "Randomisation was done by use of an interactive web‐response system, which allocated dispensing unit numbers for each patient. Randomisation was stratified by background glucose‐lowering medication (metformin alone, sulphonylurea with or without metformin, or basal insulin with or without metformin) and renal function (eGFR 45–59 mL/min per 1·73 m² [CKD‐EPI stage 3A] or 30–44 mL/min per 1·73 m² [stage 3B]; at least 40% of patients had to be at stage 3B at screening)."

Low risk of bias

Comment: Double‐blind study. Patients and site staff were masked to assignment, which was maintained by use of visually identical oral semaglutide and placebo tablets.

Low risk of bias

Comment: ITT was performed, however some participants discontinued.

Low risk of bias

Comment: Outcome assessment was undertaken by an independent event adjudication committee and a central laboratory blinded to the participants allocation assessed blood glucose.

Low risk of bias

Comment: All prespecified and key outcomes were reported.

Low risk of bias

The study is judged to be at low risk of bias in at least one domain for this result.

Risk of bias for analysis 11.1 HbA1c [%].

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 11.1.1 CKD stages 1‐2

Ma 2024

Low risk of bias

Quote: "Randomization of patients was performed using a computer‐generated random number table. […] This randomization process ensures that the allocation of patients to each group is completely random, reducing potential bias and thereby improving the scientific nature and reliability of the study."

Some concerns

Comment: Not reported. However, participants and/or investigators could be aware of treatment assigned.

Low risk of bias

Comment: All participants completed the study and were included into the analysis.

Some concerns

Comment: Not reported. However, objective and subjective outcomes were reported.

Some concerns

Comment: Prespecified outcomes were reported. However, major clinical outcomes (death, kidney failure and cardiovascular events) were not reported.

Some concerns

The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

Appendices

Appendix 1. Electronic search strategies

Database	Search terms
CENTRAL	"renal replacement therapy":ti,ab,kw "renal replacement therapy‐dependent renal disease":ti,ab,kw "renal insufficiency":ti,ab,kw "renal impairment":ti,ab,kw (kidney next disease* or renal next disease* or kidney next failure or renal next failure):ti,ab,kw dialysis:ti,ab,kw hemodialysis:ti,ab,kw hemofiltration:ti,ab,kw hemodiafiltration:ti,ab,kw (ESRF or ESKF or ESRD or ESKD):ti,ab,kw (CKF or CKD or CRF or CRD):ti,ab,kw (CAPD or CCPD or APD):ti,ab,kw (predialysis or pre‐dialysis):ti,ab,kw uremi*:ti,ab,kw albuminuri*:ti,ab,kw (microalbuminuri* or macroalbuminuri*):ti,ab,kw proteinuri*:ti,ab,kw {or #1‐#17} diabetes:ti,ab,kw #18 and #19 diabetic next nephropath*:ti,ab,kw "diabetic kidney disease":ti,ab,kw "diabetic renal disease":ti,ab,kw {or #21‐#23} #20 or #24 "glucagon‐like peptide‐1 receptor agonist":ti,ab,kw "glucagon‐like peptide 1":ti,ab,kw GLP‐1 next receptor next agonist*:ti,ab,kw "GLP1‐RA":ti,ab,kw incretin next mimetic:ti,ab,kw liraglutide:ti,ab,kw exenatide:ti,ab,kw lixisenatide:ti,ab,kw albiglutide:ti,ab,kw dulaglutide:ti,ab,kw semaglutide:ti,ab,kw cotadutide:ti,ab,kw tirzepatide:ti,ab,kw pramlintide:ti,ab,kw taspoglutide:ti,ab,kw efpeglenatideti,ab,kw {or #26‐#41} #25 and #42 in Trials
MEDLINE	Renal Replacement Therapy/ Renal Insufficiency/ exp Renal Insufficiency, Chronic/ dialysis.tw,kf. (hemodialysis or haemodialysis).tw,kf. (hemofiltration or haemofiltration).tw,kf. (hemodiafiltration or haemodiafiltration).tw,kf. (kidney disease* or renal disease* or kidney failure or renal failure).tw,kf. (ESRF or ESKF or ESRD or ESKD).tw,kf. (CKF or CKD or CRF or CRD).tw,kf. (CAPD or CCPD or APD).tw,kf. (predialysis or pre‐dialysis).tw,kf. Uremia/ (uremic or ur?emia).tw,kf. (microalbuminuri* or macroalbuminuri*).tw,kf. (proteinuri* or albuminuri*).tw,kf. or/1‐16 diabetes mellitus/ exp Diabetes Mellitus, Type 1/ exp Diabetes Mellitus, Type 2/ diabetes.tw,kf. or/18‐21 and/17,22 Diabetic Nephropathies/ diabetic nephropath*.tw,kf. diabetic kidney disease*.tw,kf. diabetic renal disease*.tw,kf. or/24‐27 or/23,28 exp Glucagon‐Like Peptide 1/ glucagon‐like peptide‐1 receptor agonist*.tw,kf. GLP‐1 receptor agonist*.tw,kf. GLP1‐RA*.tw,kf. incretin mimetic*.tw,kf. liraglutide.tw,kf. Exenatide/ exenatide.tw,kf. lixisenatide.tw,kf. albiglutide.tw,kf. dulaglutide.tw,kf. semaglutide.tw,kf. cotadutide.tw,kf. tirzepatide.tw,kf. pramlintide.tw,kf. taspoglutide.tw,kf. efpeglenatide.tw,kf. or/30‐46 and/29,47 MEDLINE search terms have been combined with the Cochrane highly sensitive search terms for RCTS (Cochrane Handbook Chapter 4).
EMBASE	exp renal replacement therapy/ chronic kidney disease/ kidney failure/ chronic kidney failure/ mild renal impairment/ stage 1 kidney disease/ moderate renal impairment/ severe renal impairment/ end stage renal disease/ renal replacement therapy‐dependent renal disease/ (hemodialysis or haemodialysis).tw. (hemofiltration or haemofiltration).tw. (hemodiafiltration or haemodiafiltration).tw. dialysis.tw. (CAPD or CCPD or APD).tw. (kidney disease* or renal disease* or kidney failure or renal failure).tw. (CKF or CKD or CRF or CRD).tw. (ESRF or ESKF or ESRD or ESKD).tw. (predialysis or pre‐dialysis).tw. Uremia/ ur?emi*.tw. exp albuminuria/ (microalbuminuri* or macroalbuminuri*).tw. (proteinuri* or albuminuri*).tw. or/1‐24 diabetes mellitus/ insulin dependent diabetes mellitus/ non‐insulin dependent diabetes mellitus/ diabetes.tw. or/26‐29 and/25,30 diabetic nephropathy/ diabetic nephropath*.tw. diabetic kidney disease*.tw. diabetic renal disease*.tw. or/32‐35 or/31,36 exp glucagon like peptide 1 receptor agonist/ glucagon like peptide 1 receptor agonist.tw. GLP‐1 receptor agonist*.tw. GLP1‐RA*.tw. incretin mimetic*.tw. liraglutide.tw. exenatide.tw. lixisenatide.tw. albiglutide.tw. dulaglutide.tw. semaglutide.tw. cotadutide.tw. tirzepatide.tw. pramlintide.tw. taspoglutide.tw. efpeglenatide.tw. or/38‐53 and/37,54 EMBASE search terms have been combined with the Cochrane highly sensitive search terms for RCTS (Cochrane Handbook Chapter 4)

Data and analyses

Comparison 1. GLP‐1 receptor agonists versus placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 All‐cause death	9	17861	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.74, 0.98]
1.1.1 CKD stages 1‐2	2	11594	Risk Ratio (M‐H, Random, 95% CI)	0.79 [0.68, 0.91]
1.1.2 CKD stages 3‐5	4	5936	Risk Ratio (M‐H, Random, 95% CI)	0.90 [0.68, 1.19]
1.1.3 Dialysis (HD/PD)	1	24	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
1.1.4 All CKD stages	3	307	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.11, 6.36]
1.2 All‐cause death (Hazard ratio)	4		Hazard Ratio (IV, Random, 95% CI)	0.84 [0.71, 0.98]
1.2.1 CKD stages 1‐2	1		Hazard Ratio (IV, Random, 95% CI)	0.99 [0.79, 1.24]
1.2.2 CKD stages 3‐5	3		Hazard Ratio (IV, Random, 95% CI)	0.82 [0.43, 1.55]
1.2.3 All CKD stages	1		Hazard Ratio (IV, Random, 95% CI)	0.80 [0.67, 0.95]
1.3 Cardiovascular death	8	17801	Risk Ratio (M‐H, Random, 95% CI)	0.84 [0.68, 1.05]
1.3.1 CKD stages 1‐2	2	11594	Risk Ratio (M‐H, Random, 95% CI)	0.78 [0.65, 0.94]
1.3.2 CKD stages 3‐5	4	5936	Risk Ratio (M‐H, Random, 95% CI)	0.89 [0.60, 1.33]
1.3.3 Dialysis (HD/PD)	1	24	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
1.3.4 All CKD stages	2	247	Risk Ratio (M‐H, Random, 95% CI)	0.41 [0.02, 9.85]
1.4 Cardiovascular death (Hazard ratio)	3		Hazard Ratio (IV, Random, 95% CI)	0.90 [0.73, 1.12]
1.4.1 CKD stages 1‐2	1		Hazard Ratio (IV, Random, 95% CI)	0.97 [0.80, 1.18]
1.4.2 CKD stages 3‐5	2		Hazard Ratio (IV, Random, 95% CI)	1.07 [0.84, 1.36]
1.4.3 All CKD stages	1		Hazard Ratio (IV, Random, 95% CI)	0.71 [0.56, 0.90]
1.5 3‐point MACE	4	19825	Risk Ratio (M‐H, Random, 95% CI)	0.84 [0.73, 0.98]
1.5.1 CKD stages 1‐2	1	11514	Risk Ratio (M‐H, Random, 95% CI)	0.88 [0.79, 0.98]
1.5.2 CKD stages 3‐5	3	6274	Risk Ratio (M‐H, Random, 95% CI)	0.86 [0.66, 1.12]
1.5.3 All CKD stages	1	2037	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.53, 0.93]
1.6 3‐point MACE (Hazard ratio)	7		Hazard Ratio (IV, Random, 95% CI)	0.80 [0.73, 0.89]
1.6.1 CKD stages 1‐2	3		Hazard Ratio (IV, Random, 95% CI)	0.76 [0.63, 0.91]
1.6.2 CKD stages 3‐5	5		Hazard Ratio (IV, Random, 95% CI)	0.86 [0.72, 1.02]
1.6.3 All CKD stages	2		Hazard Ratio (IV, Random, 95% CI)	0.77 [0.64, 0.92]
1.7 4‐point MACE	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.7.1 CKD stages 3‐5	1	2158	Risk Ratio (M‐H, Random, 95% CI)	0.77 [0.67, 0.89]
1.8 4‐point MACE (Hazard ratio)	2		Hazard Ratio (IV, Random, 95% CI)	0.86 [0.74, 1.01]
1.8.1 CKD stages 1‐2	1		Hazard Ratio (IV, Random, 95% CI)	0.89 [0.77, 1.02]
1.8.2 CKD stages 3‐5	2		Hazard Ratio (IV, Random, 95% CI)	0.86 [0.64, 1.15]
1.9 Kidney failure	3	4134	Risk Ratio (M‐H, Random, 95% CI)	0.86 [0.66, 1.13]
1.9.1 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	0.79 [0.31, 2.02]
1.9.2 All CKD stages	1	3533	Risk Ratio (M‐H, Random, 95% CI)	0.87 [0.66, 1.15]
1.10 Kidney failure (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
1.10.1 All CKD stages	1		Hazard Ratio (IV, Random, 95% CI)	0.84 [0.63, 1.12]
1.11 Composite kidney outcome	2	16849	Risk Ratio (M‐H, Random, 95% CI)	0.89 [0.78, 1.02]
1.11.1 CKD stages 1‐2	1	11514	Risk Ratio (M‐H, Random, 95% CI)	0.88 [0.73, 1.08]
1.11.2 CKD stages 3‐5	2	5335	Risk Ratio (M‐H, Random, 95% CI)	0.90 [0.75, 1.08]
1.12 Composite kidney outcome (Hazard ratio)	2		Hazard Ratio (IV, Random, 95% CI)	0.78 [0.62, 0.97]
1.12.1 CKD stages 1‐2	1		Hazard Ratio (IV, Random, 95% CI)	0.59 [0.44, 0.80]
1.12.2 CKD stages 3‐5	2		Hazard Ratio (IV, Random, 95% CI)	0.87 [0.75, 1.01]
1.13 Severe hypoglycaemia	4	6292	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.54, 1.25]
1.13.1 CKD stages 3‐5	3	2759	Risk Ratio (M‐H, Random, 95% CI)	0.65 [0.46, 0.93]
1.13.2 All CKD stages	1	3533	Risk Ratio (M‐H, Random, 95% CI)	1.02 [0.68, 1.53]
1.14 Severe hypoglycaemia (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
1.14.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	0.63 [0.43, 0.91]
1.15 Nonfatal myocardial infarction	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.15.1 CKD stages 3‐5	1	2158	Risk Ratio (M‐H, Random, 95% CI)	0.76 [0.57, 1.01]
1.16 Nonfatal myocardial infarction (Hazard ratio)	2		Hazard Ratio (IV, Random, 95% CI)	0.82 [0.63, 1.06]
1.16.1 CKD stages 1‐2	2		Hazard Ratio (IV, Random, 95% CI)	0.59 [0.18, 1.89]
1.16.2 CKD stages 3‐5	2		Hazard Ratio (IV, Random, 95% CI)	0.80 [0.64, 1.02]
1.17 Fatal myocardial infarction	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.17.1 CKD stages 3‐5	1	324	Risk Ratio (M‐H, Random, 95% CI)	2.96 [0.12, 72.21]
1.18 Myocardial infarction (undefined)	2	2435	Risk Ratio (M‐H, Random, 95% CI)	0.76 [0.58, 1.00]
1.18.1 CKD stages 3‐5	2	2435	Risk Ratio (M‐H, Random, 95% CI)	0.76 [0.58, 1.00]
1.19 Fatal or nonfatal myocardial infarction	2	16849	Risk Ratio (M‐H, Random, 95% CI)	0.93 [0.81, 1.07]
1.19.1 CKD stages 1‐2	1	11514	Risk Ratio (M‐H, Random, 95% CI)	0.99 [0.85, 1.15]
1.19.2 CKD stages 3‐5	2	5335	Risk Ratio (M‐H, Random, 95% CI)	0.87 [0.68, 1.13]
1.20 All myocardial infarction (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
1.20.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	0.73 [0.55, 0.98]
1.21 Nonfatal stroke	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.21.1 CKD stages 3‐5	1	2158	Risk Ratio (M‐H, Random, 95% CI)	0.53 [0.34, 0.81]
1.22 Nonfatal stroke (Hazard ratio)	2		Hazard Ratio (IV, Random, 95% CI)	0.93 [0.59, 1.47]
1.22.1 CKD stages 1‐2	2		Hazard Ratio (IV, Random, 95% CI)	1.15 [0.84, 1.56]
1.22.2 CKD stages 3‐5	2		Hazard Ratio (IV, Random, 95% CI)	0.82 [0.33, 2.06]
1.23 Fatal or nonfatal stroke	3	17173	Risk Ratio (M‐H, Random, 95% CI)	0.79 [0.55, 1.14]
1.23.1 CKD stages 1‐2	1	11514	Risk Ratio (M‐H, Random, 95% CI)	0.75 [0.59, 0.95]
1.23.2 CKD stages 3‐5	3	5659	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.41, 1.63]
1.24 Fatal or nonfatal stroke (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
1.24.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	0.53 [0.36, 0.79]
1.25 Ischaemic stroke	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.25.1 CKD stages 3‐5	1	329	Risk Ratio (M‐H, Random, 95% CI)	0.34 [0.01, 8.27]
1.26 Haemorrhagic stroke	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.26.1 CKD stages 3‐5	1	324	Risk Ratio (M‐H, Random, 95% CI)	2.96 [0.12, 72.21]
1.27 Heart failure	1		Risk Ratio (IV, Random, 95% CI)	Subtotals only
1.27.1 CKD stages 3‐5	1	277	Risk Ratio (IV, Random, 95% CI)	2.94 [0.12, 71.46]
1.28 Hospitalisation due to heart failure	3	17173	Risk Ratio (M‐H, Random, 95% CI)	0.88 [0.72, 1.07]
1.28.1 CKD stages 1‐2	1	11514	Risk Ratio (M‐H, Random, 95% CI)	0.86 [0.67, 1.09]
1.28.2 CKD stages 3‐5	3	5659	Risk Ratio (M‐H, Random, 95% CI)	0.89 [0.62, 1.27]
1.29 Hospitalisation due to heart failure (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	0.84 [0.62, 1.14]
1.29.1 CKD stages 1‐2	1		Hazard Ratio (IV, Random, 95% CI)	0.98 [0.73, 1.32]
1.29.2 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	0.72 [0.54, 0.96]
1.30 Change in eGFR [%]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.30.1 CKD stages 3‐5	1	584	Mean Difference (IV, Random, 95% CI)	0.99 [‐1.07, 3.05]
1.31 Increased urinary albumin‐creatinine ratio	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.31.1 CKD stages 3‐5	1	277	Risk Ratio (M‐H, Random, 95% CI)	2.94 [0.12, 71.46]
1.32 Body weight [kg]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.32.1 CKD stages 1‐2	1	80	Mean Difference (IV, Random, 95% CI)	‐5.34 [‐8.72, ‐1.96]
1.33 Change in body weight [kg]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.33.1 CKD stages 3‐5	1	324	Mean Difference (IV, Random, 95% CI)	‐2.50 [‐3.33, ‐1.67]
1.34 Change in body weight [kg]	3		Mean Difference (IV, Random, 95% CI)	‐0.86 [‐1.46, ‐0.26]
1.34.1 CKD stages 1‐2	1		Mean Difference (IV, Random, 95% CI)	‐0.50 [‐0.89, ‐0.11]
1.34.2 CKD stages 3‐5	2		Mean Difference (IV, Random, 95% CI)	‐0.89 [‐1.72, ‐0.06]
1.34.3 Dialysis (HD/PD)	1		Mean Difference (IV, Random, 95% CI)	‐2.20 [‐3.87, ‐0.53]
1.35 BMI [kg/m2]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.35.1 CKD stages 1‐2	1	80	Mean Difference (IV, Random, 95% CI)	‐2.47 [‐3.35, ‐1.59]
1.36 Change in BMI [kg/m2]	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.36.1 CKD stages 3‐5	2	572	Mean Difference (IV, Random, 95% CI)	‐0.71 [‐1.10, ‐0.32]
1.37 Change in BMI [kg/m2]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.37.1 CKD stages 3‐5	1		Mean Difference (IV, Random, 95% CI)	‐0.51 [‐0.83, ‐0.19]
1.38 Serious adverse events	2	601	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.65, 1.59]
1.38.1 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.65, 1.59]
1.39 Withdrawal due to adverse events	8	6678	Risk Ratio (M‐H, Random, 95% CI)	1.34 [0.83, 2.16]
1.39.1 CKD stages 1‐2	1	80	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
1.39.2 CKD stages 3‐5	3	2759	Risk Ratio (M‐H, Random, 95% CI)	1.96 [0.52, 7.41]
1.39.3 All CKD stages	4	3839	Risk Ratio (M‐H, Random, 95% CI)	1.11 [0.93, 1.32]
1.40 Withdrawal due to adverse events (Hazard ratio)	2		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
1.40.1 CKD stages 3‐5	2		Hazard Ratio (IV, Random, 95% CI)	1.62 [0.24, 11.11]
1.41 Hypoglycaemia	3	661	Risk Ratio (M‐H, Random, 95% CI)	1.10 [0.42, 2.84]
1.41.1 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	1.33 [0.37, 4.78]
1.41.2 All CKD stages	1	60	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.05, 5.22]
1.42 Lactic acidosis	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.42.1 CKD stages 3‐5	1	324	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
1.43 Acute kidney injury	3	571	Risk Ratio (M‐H, Random, 95% CI)	0.88 [0.11, 6.75]
1.43.1 CKD stages 3‐5	1	324	Risk Ratio (M‐H, Random, 95% CI)	1.98 [0.18, 21.57]
1.43.2 All CKD stages	2	247	Risk Ratio (M‐H, Random, 95% CI)	0.23 [0.01, 5.51]
1.44 Acute kidney injury (Hazard ratio)	2		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
1.44.1 CKD stages 3‐5	2	2482	Hazard Ratio (IV, Random, 95% CI)	0.83 [0.62, 1.11]
1.45 Diabetic ketoacidosis	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.45.1 CKD stages 3‐5	1	277	Risk Ratio (M‐H, Random, 95% CI)	2.94 [0.12, 71.46]
1.46 HbA1c [%]	3	373	Mean Difference (IV, Random, 95% CI)	‐0.64 [‐1.91, 0.64]
1.46.1 CKD stages 1‐2	1	80	Mean Difference (IV, Random, 95% CI)	‐2.01 [‐2.58, ‐1.44]
1.46.2 CKD stages 3‐5	1	277	Mean Difference (IV, Random, 95% CI)	‐0.01 [‐0.20, 0.18]
1.46.3 Dialysis (HD/PD)	1	16	Mean Difference (IV, Random, 95% CI)	0.10 [‐0.61, 0.81]
1.47 Change in HbA1c [%]	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
1.47.1 CKD stages 3‐5	2	587	Mean Difference (IV, Random, 95% CI)	‐0.72 [‐0.89, ‐0.55]
1.48 Change in HbA1c [%]	5		Mean Difference (IV, Random, 95% CI)	‐0.64 [‐0.81, ‐0.48]
1.48.1 CKD stages 1‐2	1		Mean Difference (IV, Random, 95% CI)	‐0.50 [‐0.68, ‐0.32]
1.48.2 CKD stages 3‐5	4		Mean Difference (IV, Random, 95% CI)	‐0.74 [‐0.87, ‐0.61]
1.48.3 Dialysis (HD/PD)	1		Mean Difference (IV, Random, 95% CI)	0.10 [‐0.61, 0.81]
1.49 Coronary revascularisation	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.49.1 CKD stages 3‐5	1	2158	Risk Ratio (M‐H, Random, 95% CI)	0.84 [0.63, 1.11]
1.50 Coronary revascularisation (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	0.82 [0.69, 0.98]
1.50.1 CKD stages 1‐2	1		Hazard Ratio (IV, Random, 95% CI)	0.83 [0.67, 1.02]
1.50.2 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	0.82 [0.61, 1.10]

Comparison 2. GLP‐1 receptor agonists versus standard care.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 All‐cause death	3	184	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.02, 7.32]
2.1.1 CKD stages 1‐2	1	80	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
2.1.2 All CKD stages	2	104	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.02, 7.32]
2.2 Cardiovascular death	3	184	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.02, 7.32]
2.2.1 CKD stages 1‐2	1	80	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
2.2.2 All CKD stages	2	104	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.02, 7.32]
2.3 Urinary albumin‐creatinine ratio [mg/g]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
2.3.1 CKD stages 1‐2	1	80	Mean Difference (IV, Random, 95% CI)	‐17.27 [‐25.21, ‐9.33]
2.4 Microalbuminuria [mg/L]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
2.4.1 CKD stages 1‐2	1	80	Mean Difference (IV, Random, 95% CI)	‐9.25 [‐14.57, ‐3.93]
2.5 Urinary albumin excretion rate [µg/min]	2	100	Mean Difference (IV, Random, 95% CI)	‐40.79 [‐52.77, ‐28.80]
2.5.1 All CKD stages	2	100	Mean Difference (IV, Random, 95% CI)	‐40.79 [‐52.77, ‐28.80]
2.6 BMI [kg/m2]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
2.6.1 All CKD stages	1	84	Mean Difference (IV, Random, 95% CI)	‐1.61 [‐3.00, ‐0.22]
2.7 Serious adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.7.1 All CKD stages	1	20	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.07, 13.87]
2.8 Withdrawal due to adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2.8.1 All CKD stages	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2.9 Hypoglycaemia	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.9.1 All CKD stages	1	20	Risk Ratio (M‐H, Random, 95% CI)	2.00 [0.21, 18.69]
2.10 Acute kidney injury	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.10.1 All CKD stages	1	20	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.07, 13.87]
2.11 HbA1c [%]	2	164	Mean Difference (IV, Random, 95% CI)	‐0.69 [‐0.88, ‐0.50]
2.11.1 CKD stages 1‐2	1	80	Mean Difference (IV, Random, 95% CI)	‐0.67 [‐0.97, ‐0.37]
2.11.2 All CKD stages	1	84	Mean Difference (IV, Random, 95% CI)	‐0.71 [‐0.95, ‐0.47]
2.12 Change in HbA1c [%]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
2.12.1 CKD stages 1‐2	1		Mean Difference (IV, Random, 95% CI)	0.22 [0.12, 0.32]

Comparison 3. GLP‐1 receptor agonists versus DPP‐4 inhibitors.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 All‐cause death	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.1.1 All CKD stages	1	495	Risk Ratio (M‐H, Random, 95% CI)	0.99 [0.25, 3.91]
3.2 Cardiovascular death	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.2.1 All CKD stages	1	495	Risk Ratio (M‐H, Random, 95% CI)	1.98 [0.18, 21.65]
3.3 Severe hypoglycaemia	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.3.1 All CKD stages	1	495	Risk Ratio (M‐H, Random, 95% CI)	0.25 [0.03, 2.19]
3.4 Fatal stroke	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.4.1 All CKD stages	1	495	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.01, 8.05]
3.5 Serious adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.5.1 All CKD stages	1	495	Risk Ratio (M‐H, Random, 95% CI)	0.88 [0.56, 1.37]
3.6 Withdrawal due to adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.6.1 All CKD stages	1	495	Risk Ratio (M‐H, Random, 95% CI)	0.99 [0.59, 1.65]
3.7 Hypoglycaemia	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.7.1 All CKD stages	1	495	Risk Ratio (M‐H, Random, 95% CI)	1.52 [1.06, 2.18]
3.8 HbA1c [%]	1	478	Mean Difference (IV, Random, 95% CI)	‐0.41 [‐0.70, ‐0.12]
3.8.1 CKD stages 1‐2	1	247	Mean Difference (IV, Random, 95% CI)	‐0.27 [‐0.51, ‐0.03]
3.8.2 CKD stages 3‐5	1	231	Mean Difference (IV, Random, 95% CI)	‐0.57 [‐0.86, ‐0.28]
3.9 Change in HbA1c [%]	1	478	Mean Difference (IV, Random, 95% CI)	‐0.27 [‐0.70, 0.16]
3.9.1 CKD stages 1‐2	1	247	Mean Difference (IV, Random, 95% CI)	‐0.06 [‐0.27, 0.15]
3.9.2 CKD stages 3‐5	1	231	Mean Difference (IV, Random, 95% CI)	‐0.50 [‐0.78, ‐0.22]

Comparison 4. GLP‐1 receptor agonists versus insulin.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 All‐cause death	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.1.1 CKD stages 3‐5	1	576	Risk Ratio (M‐H, Random, 95% CI)	0.76 [0.28, 2.11]
4.2 All‐cause death (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
4.2.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	1.02 [0.39, 2.67]
4.3 Cardiovascular death	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.3.1 CKD stages 3‐5	1	576	Risk Ratio (M‐H, Random, 95% CI)	1.02 [0.31, 3.33]
4.4 Cardiovascular death (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
4.4.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	1.02 [0.31, 3.36]
4.5 Kidney failure	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.5.1 CKD stages 3‐5	1	576	Risk Ratio (M‐H, Random, 95% CI)	1.02 [0.19, 5.50]
4.6 Severe hypoglycaemia	2	668	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.36, 1.22]
4.6.1 CKD stages 3‐5	1	576	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.38, 1.30]
4.6.2 All CKD stages	1	92	Risk Ratio (M‐H, Random, 95% CI)	0.20 [0.01, 4.05]
4.7 Myocardial infarction	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.7.1 CKD stages 3‐5	1	576	Risk Ratio (M‐H, Random, 95% CI)	1.02 [0.31, 3.33]
4.8 Heart failure	1		Risk Ratio (IV, Random, 95% CI)	Subtotals only
4.8.1 CKD stages 3‐5	1	576	Risk Ratio (IV, Random, 95% CI)	1.53 [0.06, 37.32]
4.9 Change in eGFR [mL/min]	2	657	Mean Difference (IV, Random, 95% CI)	3.18 [0.34, 6.02]
4.9.1 CKD stages 3‐5	1	576	Mean Difference (IV, Random, 95% CI)	1.65 [‐0.06, 3.36]
4.9.2 All CKD stages	1	81	Mean Difference (IV, Random, 95% CI)	4.55 [3.50, 5.60]
4.10 Change in eGFR [%]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
4.10.1 All CKD stages	1	81	Mean Difference (IV, Random, 95% CI)	5.83 [4.38, 7.28]
4.11 eGFR loss	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.11.1 CKD stages 3‐5	1	576	Risk Ratio (M‐H, Random, 95% CI)	0.72 [0.45, 1.16]
4.12 Change in creatinine clearance [mL/min]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
4.12.1 CKD stages 3‐5	1	576	Mean Difference (IV, Random, 95% CI)	1.20 [0.07, 2.33]
4.13 Urinary albumin‐creatinine ratio [%]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
4.13.1 All CKD stages	1	81	Mean Difference (IV, Random, 95% CI)	‐28.40 [‐37.55, ‐19.25]
4.14 Urinary albumin‐creatinine ratio [mg/g]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
4.14.1 All CKD stages	1	81	Mean Difference (IV, Random, 95% CI)	‐607.81 [‐727.04, ‐488.58]
4.15 Urinary albumin excretion rate [%]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
4.15.1 All CKD stages	1	81	Mean Difference (IV, Random, 95% CI)	‐29.71 [‐37.20, ‐22.22]
4.16 Urinary albumin excretion rate [mg/24 h]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
4.16.1 All CKD stages	1	81	Mean Difference (IV, Random, 95% CI)	‐159.67 [‐293.45, ‐25.89]
4.17 Change in body weight [kg]	2	630	Mean Difference (IV, Random, 95% CI)	‐1.68 [‐3.72, 0.36]
4.17.1 CKD stages 3‐5	1	549	Mean Difference (IV, Random, 95% CI)	‐0.60 [‐1.43, 0.23]
4.17.2 All CKD stages	1	81	Mean Difference (IV, Random, 95% CI)	‐2.68 [‐2.96, ‐2.40]
4.18 Change in body weight [kg]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
4.18.1 CKD stages 3‐5	1		Mean Difference (IV, Random, 95% CI)	‐3.80 [‐4.80, ‐2.80]
4.19 Body weight increase	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.19.1 CKD stages 3‐5	1	576	Risk Ratio (M‐H, Random, 95% CI)	0.57 [0.30, 1.10]
4.20 Serious adverse events	2	668	Risk Ratio (M‐H, Random, 95% CI)	0.74 [0.40, 1.37]
4.20.1 CKD stages 3‐5	1	576	Risk Ratio (M‐H, Random, 95% CI)	0.81 [0.60, 1.09]
4.20.2 All CKD stages	1	92	Risk Ratio (M‐H, Random, 95% CI)	0.25 [0.03, 2.15]
4.21 Withdrawal due to adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.21.1 All CKD stages	1	92	Risk Ratio (M‐H, Random, 95% CI)	1.20 [0.39, 3.66]
4.22 Hypoglycaemia	2	668	Risk Ratio (M‐H, Random, 95% CI)	0.69 [0.52, 0.90]
4.22.1 CKD stages 3‐5	1	576	Risk Ratio (M‐H, Random, 95% CI)	0.73 [0.64, 0.82]
4.22.2 All CKD stages	1	92	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.26, 0.95]
4.23 Acute kidney injury	2	583	Risk Ratio (M‐H, Random, 95% CI)	2.03 [0.44, 9.47]
4.23.1 CKD stages 1‐2	1	7	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
4.23.2 CKD stages 3‐5	1	576	Risk Ratio (M‐H, Random, 95% CI)	2.03 [0.44, 9.47]
4.24 Change in HbA1c [%]	2	630	Mean Difference (IV, Random, 95% CI)	0.08 [‐0.23, 0.39]
4.24.1 CKD stages 3‐5	1	549	Mean Difference (IV, Random, 95% CI)	‐0.10 [‐0.34, 0.14]
4.24.2 All CKD stages	1	81	Mean Difference (IV, Random, 95% CI)	0.22 [0.12, 0.32]
4.25 Change in HbA1c [%]	4		Mean Difference (IV, Random, 95% CI)	Subtotals only
4.25.1 CKD stages 3‐5	4		Mean Difference (IV, Random, 95% CI)	‐0.86 [‐1.59, ‐0.13]
4.26 Change in HbA1c [mmol/mol]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
4.26.1 CKD stages 3‐5	1	549	Mean Difference (IV, Random, 95% CI)	‐1.10 [‐3.68, 1.48]
4.27 Ischaemic stroke	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.27.1 CKD stages 3‐5	1	576	Risk Ratio (M‐H, Random, 95% CI)	3.56 [0.19, 68.65]

Comparison 5. GLP‐1 receptor agonists versus SGLT2 inhibitors.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
5.1 All‐cause death	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
5.1.1 CKD stages 1‐2	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
5.2 Cardiovascular death	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
5.2.1 CKD stages 1‐2	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
5.3 Withdrawal due to adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
5.3.1 CKD stages 1‐2	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
5.4 HbA1c [%]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
5.4.1 CKD stages 1‐2	1	80	Mean Difference (IV, Random, 95% CI)	‐0.90 [‐1.65, ‐0.15]
5.5 Body weight [kg]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
5.5.1 CKD stages 1‐2	1	80	Mean Difference (IV, Random, 95% CI)	7.43 [4.41, 10.45]
5.6 BMI [kg/m2]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
5.6.1 CKD stages 1‐2	1	80	Mean Difference (IV, Random, 95% CI)	1.53 [0.63, 2.43]

Comparison 6. GLP‐1 receptor agonist versus another GLP‐1 receptor agonist.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
6.1 All‐cause death	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
6.1.1 All CKD stages	1	196	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.03, 13.52]
6.2 Cardiovascular death	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
6.2.1 All CKD stages	1	196	Risk Ratio (M‐H, Random, 95% CI)	1.10 [0.05, 26.58]
6.3 Withdrawal due to adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
6.3.1 All CKD stages	1	196	Risk Ratio (M‐H, Random, 95% CI)	1.38 [0.61, 3.12]
6.4 Acute kidney injury	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
6.4.1 All CKD stages	1	196	Risk Ratio (M‐H, Random, 95% CI)	Not estimable

Comparison 8. Sensitivity analysis removing studies with overall high risk of bias: GLP‐1 receptor agonists versus placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
8.1 All‐cause death	8	17837	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.74, 0.98]
8.1.1 CKD stages 1‐2	2	11594	Risk Ratio (M‐H, Random, 95% CI)	0.79 [0.68, 0.91]
8.1.2 CKD stages 3‐5	4	5936	Risk Ratio (M‐H, Random, 95% CI)	0.90 [0.68, 1.19]
8.1.3 All CKD stages	3	307	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.11, 6.36]
8.2 Cardiovascular death	7	17777	Risk Ratio (M‐H, Random, 95% CI)	0.84 [0.68, 1.05]
8.2.1 CKD stages 1‐2	2	11594	Risk Ratio (M‐H, Random, 95% CI)	0.78 [0.65, 0.94]
8.2.2 CKD stages 3‐5	4	5936	Risk Ratio (M‐H, Random, 95% CI)	0.89 [0.60, 1.33]
8.2.3 All CKD stages	2	247	Risk Ratio (M‐H, Random, 95% CI)	0.41 [0.02, 9.85]
8.3 Change in body weight [kg]	2		Mean Difference (IV, Random, 95% CI)	‐0.65 [‐1.10, ‐0.21]
8.3.1 CKD stages 1‐2	1		Mean Difference (IV, Random, 95% CI)	‐0.50 [‐0.89, ‐0.11]
8.3.2 CKD stages 3‐5	2		Mean Difference (IV, Random, 95% CI)	‐0.89 [‐1.72, ‐0.06]
8.4 HbA1c [%]	2	357	Mean Difference (IV, Random, 95% CI)	‐0.99 [‐2.95, 0.97]
8.4.1 CKD stages 1‐2	1	80	Mean Difference (IV, Random, 95% CI)	‐2.01 [‐2.58, ‐1.44]
8.4.2 CKD stages 3‐5	1	277	Mean Difference (IV, Random, 95% CI)	‐0.01 [‐0.20, 0.18]
8.5 Change in HbA1c [%]	3		Mean Difference (IV, Random, 95% CI)	‐0.69 [‐0.85, ‐0.52]
8.5.1 CKD stages 1‐2	1		Mean Difference (IV, Random, 95% CI)	‐0.50 [‐0.68, ‐0.32]
8.5.2 CKD stages 3‐5	3		Mean Difference (IV, Random, 95% CI)	‐0.76 [‐0.90, ‐0.63]

8.1. Analysis.

Comparison 8: Sensitivity analysis removing studies with overall high risk of bias: GLP‐1 receptor agonists versus placebo, Outcome 1: All‐cause death

8.2. Analysis.

Comparison 8: Sensitivity analysis removing studies with overall high risk of bias: GLP‐1 receptor agonists versus placebo, Outcome 2: Cardiovascular death

8.3. Analysis.

Comparison 8: Sensitivity analysis removing studies with overall high risk of bias: GLP‐1 receptor agonists versus placebo, Outcome 3: Change in body weight [kg]

8.4. Analysis.

Comparison 8: Sensitivity analysis removing studies with overall high risk of bias: GLP‐1 receptor agonists versus placebo, Outcome 4: HbA1c [%]

8.5. Analysis.

Comparison 8: Sensitivity analysis removing studies with overall high risk of bias: GLP‐1 receptor agonists versus placebo, Outcome 5: Change in HbA1c [%]

Comparison 9. Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
9.1 All‐cause death	7	1012	Risk Ratio (M‐H, Random, 95% CI)	1.52 [0.41, 5.59]
9.1.1 CKD stages 1‐2	1	80	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
9.1.2 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	2.31 [0.42, 12.78]
9.1.3 Dialysis (HD/PD)	1	24	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
9.1.4 All CKD stages	3	307	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.11, 6.36]
9.2 All‐cause death (Hazard ratio)	2		Hazard Ratio (IV, Random, 95% CI)	1.42 [0.19, 10.88]
9.2.1 CKD stages 3‐5	2		Hazard Ratio (IV, Random, 95% CI)	1.42 [0.19, 10.88]
9.3 Cardiovascular death	6	952	Risk Ratio (M‐H, Random, 95% CI)	1.05 [0.18, 6.01]
9.3.1 CKD stages 1‐2	1	80	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
9.3.2 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	1.58 [0.19, 12.72]
9.3.3 Dialysis (HD/PD)	1	24	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
9.3.4 All CKD stages	2	247	Risk Ratio (M‐H, Random, 95% CI)	0.41 [0.02, 9.85]
9.4 Cardiovascular death (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
9.4.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	0.99 [0.06, 16.33]
9.5 3‐point MACE	2	2976	Risk Ratio (M‐H, Random, 95% CI)	0.75 [0.60, 0.94]
9.5.1 CKD stages 3‐5	1	939	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.58, 1.24]
9.5.2 All CKD stages	1	2037	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.53, 0.93]
9.6 3‐point MACE (Hazard ratio)	4		Hazard Ratio (IV, Random, 95% CI)	0.76 [0.67, 0.87]
9.6.1 CKD stages 1‐2	1		Hazard Ratio (IV, Random, 95% CI)	0.69 [0.56, 0.85]
9.6.2 CKD stages 3‐5	3		Hazard Ratio (IV, Random, 95% CI)	0.88 [0.72, 1.08]
9.6.3 All CKD stages	1		Hazard Ratio (IV, Random, 95% CI)	0.67 [0.50, 0.90]
9.7 4‐point MACE (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
9.7.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	1.01 [0.80, 1.28]
9.8 Kidney failure	2		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
9.8.1 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	0.79 [0.31, 2.02]
9.9 Composite kidney outcome (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
9.9.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	0.88 [0.73, 1.06]
9.10 Severe hypoglycaemia	2		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
9.10.1 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	2.94 [0.12, 71.46]
9.11 Myocardial infarction	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
9.11.1 CKD stages 3‐5	1	277	Risk Ratio (M‐H, Random, 95% CI)	2.94 [0.12, 71.46]
9.12 Fatal or nonfatal stroke	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
9.12.1 CKD stages 3‐5	1	324	Risk Ratio (M‐H, Random, 95% CI)	0.99 [0.06, 15.66]
9.13 Hospitalisation due to heart failure	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
9.13.1 CKD stages 3‐5	1	324	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.01, 8.02]
9.14 Withdrawal due to adverse events	6	988	Risk Ratio (M‐H, Random, 95% CI)	2.09 [1.06, 4.14]
9.14.1 CKD stages 1‐2	1	80	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
9.14.2 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	3.47 [1.86, 6.45]
9.14.3 All CKD stages	3	307	Risk Ratio (M‐H, Random, 95% CI)	1.10 [0.46, 2.61]
9.15 Withdrawal due to adverse events (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
9.15.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	4.65 [1.62, 13.31]
9.16 Acute kidney injury (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
9.16.1 CKD stages 3‐5	1	324	Hazard Ratio (IV, Random, 95% CI)	1.98 [0.18, 21.78]

9.1. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 1: All‐cause death

9.2. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 2: All‐cause death (Hazard ratio)

9.3. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 3: Cardiovascular death

9.4. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 4: Cardiovascular death (Hazard ratio)

9.5. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 5: 3‐point MACE

9.6. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 6: 3‐point MACE (Hazard ratio)

9.7. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 7: 4‐point MACE (Hazard ratio)

9.8. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 8: Kidney failure

9.9. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 9: Composite kidney outcome (Hazard ratio)

9.10. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 10: Severe hypoglycaemia

9.11. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 11: Myocardial infarction

9.12. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 12: Fatal or nonfatal stroke

9.13. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 13: Hospitalisation due to heart failure

9.14. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 14: Withdrawal due to adverse events

9.15. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 15: Withdrawal due to adverse events (Hazard ratio)

9.16. Analysis.

Comparison 9: Sensitivity analysis removing any very long study: GLP‐1 receptor agonists versus placebo, Outcome 16: Acute kidney injury (Hazard ratio)

Comparison 10. Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
10.1 All‐cause death	7	1012	Risk Ratio (M‐H, Random, 95% CI)	1.52 [0.41, 5.59]
10.1.1 CKD stages 1‐2	1	80	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
10.1.2 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	2.31 [0.42, 12.78]
10.1.3 Dialysis (HD/PD)	1	24	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
10.1.4 All CKD stages	3	307	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.11, 6.36]
10.2 All‐cause death (Hazard ratio)	2		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
10.2.1 CKD stages 3‐5	2		Hazard Ratio (IV, Random, 95% CI)	1.42 [0.19, 10.88]
10.3 Cardiovascular death	6	952	Risk Ratio (M‐H, Random, 95% CI)	1.05 [0.18, 6.01]
10.3.1 CKD stages 1‐2	1	80	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
10.3.2 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	1.58 [0.19, 12.72]
10.3.3 Dialysis (HD/PD)	1	24	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
10.3.4 All CKD stages	2	247	Risk Ratio (M‐H, Random, 95% CI)	0.41 [0.02, 9.85]
10.4 Cardiovascular death (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
10.4.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	0.99 [0.06, 16.33]
10.5 3‐point MACE	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
10.5.1 CKD stages 3‐5	1	939	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.58, 1.24]
10.6 3‐point MACE (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
10.6.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	0.84 [0.57, 1.24]
10.7 Kidney failure	2		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
10.7.1 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	0.79 [0.31, 2.02]
10.8 Severe hypoglycaemia	2		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
10.8.1 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	2.94 [0.12, 71.46]
10.9 Myocardial infarction	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
10.9.1 CKD stages 3‐5	1	277	Risk Ratio (M‐H, Random, 95% CI)	2.94 [0.12, 71.46]
10.10 Fatal or nonfatal stroke	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
10.10.1 CKD stages 3‐5	1	324	Risk Ratio (M‐H, Random, 95% CI)	0.99 [0.06, 15.66]
10.11 Hospitalisation due to heart failure	1		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
10.11.1 CKD stages 3‐5	1	324	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.01, 8.02]
10.12 Withdrawal due to adverse events	6	988	Risk Ratio (M‐H, Random, 95% CI)	2.09 [1.06, 4.14]
10.12.1 CKD stages 1‐2	1	80	Risk Ratio (M‐H, Random, 95% CI)	Not estimable
10.12.2 CKD stages 3‐5	2	601	Risk Ratio (M‐H, Random, 95% CI)	3.47 [1.86, 6.45]
10.12.3 All CKD stages	3	307	Risk Ratio (M‐H, Random, 95% CI)	1.10 [0.46, 2.61]
10.13 Withdrawal due to adverse events (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
10.13.1 CKD stages 3‐5	1		Hazard Ratio (IV, Random, 95% CI)	4.65 [1.62, 13.31]
10.14 Acute kidney injury (Hazard ratio)	1		Hazard Ratio (IV, Random, 95% CI)	Subtotals only
10.14.1 CKD stages 3‐5	1	324	Hazard Ratio (IV, Random, 95% CI)	1.98 [0.18, 21.78]

10.1. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 1: All‐cause death

10.2. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 2: All‐cause death (Hazard ratio)

10.3. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 3: Cardiovascular death

10.4. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 4: Cardiovascular death (Hazard ratio)

10.5. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 5: 3‐point MACE

10.6. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 6: 3‐point MACE (Hazard ratio)

10.7. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 7: Kidney failure

10.8. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 8: Severe hypoglycaemia

10.9. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 9: Myocardial infarction

10.10. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 10: Fatal or nonfatal stroke

10.11. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 11: Hospitalisation due to heart failure

10.12. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 12: Withdrawal due to adverse events

10.13. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 13: Withdrawal due to adverse events (Hazard ratio)

10.14. Analysis.

Comparison 10: Sensitivity analysis removing any very large study: GLP‐1 receptor agonists versus placebo, Outcome 14: Acute kidney injury (Hazard ratio)

Comparison 11. Sensitivity analysis removing studies with funding: GLP‐1 receptor agonists versus placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
11.1 HbA1c [%]	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
11.1.1 CKD stages 1‐2	1	80	Mean Difference (IV, Random, 95% CI)	‐2.01 [‐2.58, ‐1.44]

11.1. Analysis.

Comparison 11: Sensitivity analysis removing studies with funding: GLP‐1 receptor agonists versus placebo, Outcome 1: HbA1c [%]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

AMPLITUDE‐O 2021.

Study characteristics
Methods	Study design Parallel RCT Study time frame: May 2018 to April 2019 Duration of follow‐up: not clearly reported for only people with CKD and diabetes (overall median follow‐up period of 1.81 years (interquartile range, 1.69 to 1.98) Country: multinational (28 countries) Setting: multicentre (344 sites)
Participants	Study characteristics Inclusion criteria: T2DM with HbA1c > 7%; high CV risk; agreement to use contraception/ not donate germ cells during and for up to 5 weeks after the intervention period; signed informed consent Exclusion criteria: upper GI disease; pancreatitis or pancreatectomy; personal/family history of medullary thyroid cancer; active diabetic retinopathy; GLP‐1 receptor agonist or DPP‐4 inhibitor in last 3 months; eGFR < 25 mL/min/1.73 m2; elevated liver tests, high amylase, lipase, or calcitonin; short life expectancy Baseline characteristics Number (randomised/analysed): overall (2175/2175); intervention group 1 + 2 (1448/1448); control group (727/727); ITT Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group 1 Efpeglenatide: 4 mg, weekly injection (2 mg/week for 4 weeks, then 4 mg/week) Intervention group 2 Efpeglenatide: 6 mg, weekly injection (2 mg/week for 4 weeks, then 4 mg/week for 4 weeks, then 6 mg/week) Control group Placebo Co‐interventions A study drug was added to each participant's T2DM regimen
Outcomes	Primary outcomes Major adverse CV event: defined as nonfatal MI, nonfatal stroke or CV death Secondary outcomes Expanded composite outcome: MACE, coronary revascularisation or hospitalisation for unstable angina Renal composite outcome: new macroalbuminuria (i.e. UACR > 300 mg/g or 33.9 mg/mmol) + increased UACR of ≥ 30% from baseline, eGFR decreased by ≥ 40% for ≥ 30 days, or kidney failure (dialysis for ≥ 90 days, kidney transplant or eGFR < 15 mL/ min/1.73 m2 for ≥ 30 days) Adverse events and serious adverse events MI Stroke CV death Coronary revascularisation Hospitalisation for unstable angina Hospitalisation for heart failure All‐cause death HbA1c change Body weight change BP change UACR change
Notes	Additional information Funding source: Sanofi (Bridgewater, NJ, USA), who are also providing site management and data collection. Scientific leadership is provided by an independent international steering committee and primary data analysis will be conducted by the Population Health Research Institute Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: Web‐response system; no information on baseline characteristics was reported Note: authors contacted (31/8/2021), but they did not give further data Note: this study reported outcome data both as RR and HR, and some outcomes could report duplicate RoB judgements

AWARD‐7 2017.

Study characteristics
Methods	Study design Parallel RCT Study time frame: August 2012 to 30 November 2015 Duration of follow‐up: 26 weeks Country: multinational (9 countries) Setting: multicentre (99 sites)
Participants	Study characteristics Inclusion criteria: men and non‐pregnant women; ≥ 18 years; HbA1c ≥ 7.5% (58 mmol/mol) and ≤ 10.5% (91 mmol/mol); T2DM; participants with presumed DKD with or without hypertensive nephrosclerosis diagnosed with moderate or severe CKD; able and willing to perform multiple daily injections; BMI between 23 and 45 kg/m² Exclusion criteria: stage 5 CKD as defined by eGFR < 15 mL/min/1.73 m2 or having required dialysis; rapidly progressing kidney dysfunction likely to require KRT; history of a transplanted organ; T1DM; a CV condition within 12 weeks prior to randomisation: acute MI, NYHA class III or class IV heart failure, or cerebrovascular accident (stroke); acute or chronic hepatitis; serum calcitonin ≥ 35 pg/mL at screening visit; self or family history of medullary C‐cell hyperplasia Baseline characteristics Number (randomised/analysed): overall (577/576); intervention group 1(193/192); intervention group 2 (190/190); intervention group 3 (194/194) Mean age ± SD (years): intervention group 1 (64.7 ± 8.8); intervention group 2 (64.7 ± 8.6); intervention group 3 (64.3 ± 8.4) Sex (M/F): intervention group 1 (104/88); intervention group 2 (104/86); intervention group 3 (93/101)
Interventions	Intervention group 1 Dulaglutide: 0.75 mg/week Intervention group 2 Dulaglutide: 1.5 mg/week Intervention group 3 Insulin: glargine/day Co‐interventions All participants took insulin lispro
Outcomes	Primary outcomes Change from baseline in HbA1c (week 26) Secondary outcomes Percentage whose HbA1c was < 7.0% (53 mmol/mol) (week 26 and at EoT) Percentage whose HbA1c was < 8.0% (64 mmol/mol) (week 26 and at EoT) Change from baseline in 8‐point self‐monitored plasma glucose (SMPG) (week 26 and at EoT) Change from baseline in mean daily insulin lispro dose (week 26 and at EoT) Percentage with estimated average glucose < 8.5 mmol/L (week 26 and at EoT) Change from baseline in SCr (week 26 and at EoT) Change from baseline in eGFR (week 26 and at EoT) Change from baseline in estimated CrCl (week 26 and at EoT) Change from baseline in UACR (week 26 and at EoT) Percentage of participants with self‐reported hypoglycaemic events (week 26 and at EoT) Rate of hypoglycaemic event (week 26 and at EoT) Change from baseline in FBG (week 26 and at EoT) Change from baseline in body weight (week 26 and at EoT) Percentage with allergic/hypersensitivity reactions (week 26 and at EoT)
Notes	Additional information Funding source: Eli Lilly and Company, involved in the study design, data collection, data review, data analysis, and drafting of the report Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: interactive response system; baseline demographics, clinical characteristics, and concomitant medication use were similar between groups Note: this study reported outcome data both as RR and HR and some outcomes could report duplicate RoB judgements

AWARD‐PEDS 2022.

Study characteristics
Methods	Study design Parallel RCT Study time frame: December 2016 to December 2020 Duration of follow‐up: 26 weeks Country: multinational (9 countries) Setting: multicentre (46 sites)
Participants	Study characteristics Inclusion criteria: 10 to 18 years; T2DM; BMI > 85th percentile with a weight of at least 50 kg; HbA1c 6.5% to 11.0% (if the patient was taking metformin with or without basal insulin) or 6.5% to 9.0% (if the patient was treated with diet and physical exercise alone); some participants had CKD Exclusion criteria: T1DM; use of other glucose‐lowering drugs other than metformin and insulin; history of pancreatitis; personal or familiar history of multiple endocrine neoplasia type 2A/B, thyroid C‐cell hyperplasia or thyroid medullary carcinoma Baseline characteristics Number: overall/CKD (154/not reported) Mean age ± SD (years): not reported for CKD patients Sex (M/F): not reported for CKD patients
Interventions	Intervention group 1 Dulaglutide: 0.75 mg/week Intervention group 2 Dulaglutide: 1.50 mg/week Control group Placebo
Outcomes	Primary outcomes Change in HbA1c from baseline at week 26 Secondary outcomes Percentage of patient whose HbA1c was < 7.0% at week 26 Percentage of patient whose HbA1c was ≤ 6.5% at week 26 Change in FPG from baseline at week 26 Change in BMI from baseline at week 26 Safety analysis
Notes	Additional information Funding source: Eli Lilly Conflicts of interest: not reported Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Note: no data were available for people with CKD and diabetes. Authors were contacted, but they did not reply

Dekkers 2021.

Study characteristics
Methods	Study design Parallel RCT Study time frame: not reported Duration of follow‐up: 26 weeks Country: multinational (Europe and South Asia) Setting: multicentre
Participants	Study characteristics Inclusion criteria: European and South Asian patients with T2DM were, respectively, aged 18 to 70 and 18 to 75 years; HbA1c between at least 7.0 and 10.0% and at least 6.5 11.0%; systolic and diastolic blood pressures between lower than 150/85 mm Hg and lower 180/110 mm Hg; estimated glomerular filtration rate above > 60 mL/minute/1.73m2 and >30 mL/minute/1.73 m2 Exclusion criteria: history of heart failure (New York Heart Association Class III‐IV), history of coronary artery disease for European patients with T2DM, and acute coronary accident in the preceding 30 days for South Asian patients with T2DM Baseline characteristics Number (overall/diabetic nephropathy): intervention group (22/5); control group (28/7); these participants had diabetic nephropathy, defined as UACR at least 17 mg/g in men and at least 25 mg/g in women Mean age ± SD (years): not reported for CKD patients Sex (M/F): not reported for CKD patients
Interventions	Intervention group Liraglutide Control group Placebo Co‐interventions Standard care (metformin/sulphonylurea derivative/insulin)
Outcomes	Primary outcomes Stroke volume: 0 and 26 weeks Early deceleration peak / Atrial peak filling rate: 0 and 26 weeks Peak mitral annulus longitudinal motion: 0 and 26 weeks Left ventricular filling pressure (= early peak filling rate/peak mitral annulus longitudinal motion): 0 and 26 weeks Secondary outcomes HbA1c: 0, 8, 12, 16 and 26 weeks FBG: 0, 4, 8, 12, 16, 20, 26 weeks SBP: 0, 4, 8, 12, 16, 20, 26 weeks DBP: 0, 4, 8, 12, 16, 20, 26 weeks Albuminuria: 0 and 26 weeks Hypoglycaemic episodes: between week 0 and 26
Notes	Additional information Funding source: Novo Nordisk (Denmark) funded this investigator‐initiated study. Novo Nordisk had no role in the design of the study, data collection, data analysis, data interpretation, or writing of the report. This study was supported by the Dutch Kidney Foundation (innovation grant: IP11.56) Conflicts of interest: none Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Note: RoB assessment was not provided since no outcomes were reported

DUAL I 2014.

Study characteristics
Methods	Study design Phase 3, parallel RCT Study time frame: 23 May 2011 to 24 May 2012 Duration of follow‐up: 26 weeks Countries: Australia, Canada, Finland, Germany, Hungary, India, Ireland, Italy, Malaysia, Mexico, Russian Federation, Singapore, Slovakia, South Africa, Spain, Taiwan, Thailand, UK, USA Setting: multicentre (271 sites)
Participants	Study characteristics Inclusion criteria: ≥ 18 years; T2DM; HbA1c 7% to 10% (inclusive); BMI ≤ 40 kg/m², and who had been previously treated with metformin with or without pioglitazone for at least 90 days before screening; participants with CKD were included Exclusion criteria: patients treated with GLP‐1 receptor agonists, DPP4 inhibitors, or sulfonylureas within 90 days of screening Baseline characteristics Number: overall/CKD (1663/90) Mean age ± SD (years): not reported for CKD patients Sex (M/F): not reported for CKD patients
Interventions	Intervention group 1 IDegLira: once‐daily injections, started at 10 dose steps IDegLira consists of 100 units/mL insulin degludec and 3.6 mg/mL liraglutide Intervention group 2 Insulin degludec: 100 units/mL Intervention group 3 Liraglutide: started at 0.6 mg/day and increased by 0.6 mg/week to a maximum of 1.8 mg/day Co‐interventions Based on pre‐breakfast self‐monitored blood glucose measurements (mean from three consecutive days), doses of IDegLira and insulin degludec were titrated individually twice/week to achieve a pre‐breakfast plasma glucose of 4 to 5 mmol/L (72 to 90 mg/dL) by use of an algorithm
Outcomes	Primary outcomes Mean change from baseline in HbA1c at week 26 Secondary outcomes Mean change from baseline in body weight at week 26 Number of hypoglycaemic episodes at week 26 Change from baseline in incremental area under the curve 0 to 4 hours derived from the glucose at week 26 Concentration profile during meal test at week 26 Mean actual daily insulin dose at week 26
Notes	Additional information Funding source: Novo Nordisk. The funder was responsible for trial design, product supply, monitoring, data collection, surveillance of insulin titration, safety surveillance, statistical analysis, data interpretation, and review of the report for medical accuracy Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: interactive voice or web system; no information on baseline characteristics was reported Note: 90 patients had an eGFR < 60 included in the analysis Note: authors contacted but they did not reply

DUAL II 2014.

Study characteristics
Methods	Study design Phase 3, parallel RCT Study time frame: 28 November 2011 to 4 October 2012 Duration of follow‐up: 26 weeks Countries: Bulgaria, Switzerland, Denmark, Hungary, India, Slovenia, USA Setting: multicentre (75 centres)
Participants	Study characteristics Inclusion criteria: ≥ 18 years; inadequately controlled T2DM (HbA1c 7.5% to 10.0%, inclusive); BMI ≥ 27 kg/m², treated for ≥ 90 days with basal insulin at a stable dose (20 to 40 units/day) in combination with metformin with or without sulfonylurea or glinides. Patients with CKD were also included Exclusion criteria: use of any drug (except for basal insulin, metformin, sulfonylurea and glinides), which in the investigator’s opinion could interfere with glucose level (e.g. systemic corticosteroids); treatment with GLP‐1 receptor agonists (e.g. exenatide, liraglutide), DPP4 inhibitors and/or thiazolidinediones within 90 days prior to screening Baseline characteristics Number: overall/CKD (413/27) Mean age ± SD (years): not reported for CKD patients Sex (M/F): not reported for CKD patients
Interventions	Intervention group 1 IDegLira: 100 units/mL insulin degludec and 3.6 mg/mL liraglutide Intervention group 2 Insulin degludec Co‐interventions At randomisation, participants discontinued all glucose‐lowering drugs except for metformin (kept at pre‐trial dose and frequency) and transferred from current basal insulin to IDegLira or IDeg Doses of IDeg and IDegLira were adjusted biweekly according to a predefined titration algorithm, based on self‐measured pre‐breakfast FPG
Outcomes	Primary outcomes Change from baseline in HbA1c at week 26 Secondary outcomes Change in body weight at week 26
Notes	Additional information Funding source: Novo Nordisk A/S Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: central interactive voice/web system; baseline characteristics were similar between treatment groups and representative of a population with type 2 diabetes inadequately controlled on their current treatment Note: 27 patients had an eGFR < 60 included in the analysis Note: authors contacted, but they were not able to provide further data

DUAL III 2017.

Study characteristics
Methods	Study design Phase 3, parallel RCT Study time frame: 29 August 2012 to 11 March 2014 Duration of follow‐up: 26 weeks Countries: Australia, France, Hungary, Slovakia, USA Setting: multicentre (81 centres)
Participants	Study characteristics Inclusion criteria: ≥ 18 years; HbA1c of 7.0% to 9.0% (53 to 75 mmol/mol, both inclusive); BMI ≤ 40 kg/ m²; treatment with daily GLP‐1 receptor agonist at maximum dose according to local label or documented maximum tolerated dose in combination with a stable daily dose of metformin or 90 days or more prior to screening visit Exclusion criteria: treatment with any oral antidiabetic drugs except for metformin, pioglitazone, and sulfonylurea within 90 days prior to screening Baseline characteristics Number: overall/CKD (438/28) Mean age ± SD (years): not reported for CKD patients Sex (M/F): not reported for CKD patients
Interventions	Intervention group 1 IDegLira: 100 units/mL insulin degludec and 3.6 mg/mL liraglutide; starting dose of IDegLira was 16 dose steps (16 U insulin degludec and 0.6 mg liraglutide) Intervention group 2 GLP‐1 receptor agonist treatment continued their pre‐trial treatment schedule without making any changes Liraglutide (Victoza) was administered once daily using a 6.0 mg/mL solution provided in a 3 mL prefilled pen Exenatide (Byetta) was administered twice daily using a 250 mg/mL solution provided in 1.2 mL or 2.4 mL prefilled pens Co‐interventions All previous oral antidiabetic drugs (metformin, pioglitazone, sulfonylurea) were continued at pre‐trial doses in both groups unless there was a safety concern. Oral antidiabetic drug dose reduction was allowed for safety reasons (including hypoglycaemic events) based on the judgement of the investigator
Outcomes	Primary outcomes Change in HbA1c from baseline at week 26 Secondary outcomes Responders achieving a pre‐defined target: HbA1c < 7.0% (53 mmol/mol) at week 26 Responders achieving a pre‐defined target: HbA1c ≤ 6.5% (48 mmol/mol) at week 26 Change from baseline in body weight at week 26 Change from baseline in FPG at week 26 Number of severe or minor hypoglycaemic episodes at week 26 Number of adverse events at week 26
Notes	Additional information Funding source: Novo Nordisk. The funder was responsible for trial design, product supply, monitoring, data collection, surveillance of insulin titration, safety surveillance, statistical analysis, and data interpretation, and reviewing the report for medical accuracy Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: central interactive voice/web system; similar baseline characteristics were reported Note: 28 patients had eGFR < 60 Note: authors contacted but they did not reply Note: RoB assessment was not provided since no outcomes were reported

DUAL IV 2014.

Study characteristics
Methods	Study design Phase 3, parallel RCT Study time frame: 29 August 2012 to 23 October 2013 Duration of follow‐up: 26 weeks Countries: Bulgaria, Canada, Germany, India, Israel, Turkey, USA Setting: multicentre (77 sites)
Participants	Study characteristics Inclusion criteria: ≥ 18 years; T2DM; HbA1c levels of 53 to 75 mmol/mol (7.0% to 9.0%, both inclusive); BMI ≤ 40 kg/m2; previously treated with a stable daily dose of sulphonylureas (≥ half of the maximum approved dose according to local label) metformin (≥ 1500 mg or maximum tolerated dose) for at least 90 days before screening. Participants with CKD were also reported Exclusion criteria: treated with insulin‐ and GLP‐1 receptor agonist naive, treated with any antidiabetic agent other than sulphonylureas or metformin for ≤ 90 days before screening Baseline characteristics Number: overall/CKD (435/46) Mean age ± SD (years): not reported for CKD patients Sex (M/F): not reported for CKD patients
Interventions	Intervention group IDegLira: 100 units/mL insulin degludec and 3.6 mg/mL liraglutide; 10 dose steps Control group Placebo: 10 dose steps Co‐interventions Sulphonylureas and metformin were maintained at pre‐trial dose and frequency unless there was a safety concern. Counselling regarding a healthy lifestyle and exercise was provided for all participants during the trial Doses of IDegLira or placebo were adjusted twice /week according to a predefined titration algorithm, based on the mean fasting pre‐breakfast self‐monitored blood glucose measurements, from 3 consecutive days, aiming to achieve a mean pre‐breakfast blood glucose concentration of 4.0 to 6.0 mmol/l (72 to 108 mg/dL)
Outcomes	Primary outcomes Change in HbA1c at 26 weeks Secondary outcomes Responders achieving pre‐defined target: HbA1c below 7.0% (53 mmol/mol) at 26 weeks Responders achieving pre‐defined target: HbA1c below or equal to 6.5% (48 mmol/mol) at 26 weeks Change from baseline in FPG at 26 weeks Change from baseline in body weight number of treatment‐emergent (confirmed) at 26 weeks Hypoglycaemic episodes number of adverse events at 26 weeks
Notes	Additional information Funding source: Novo Nordisk. The funder was responsible for trial design, product supply, monitoring, data collection, surveillance of insulin titration, safety surveillance, statistical analysis, data interpretation, and review of the report for medical accuracy Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: central interactive voice/web system; treatment groups were well matched with respect to demographics and baseline characteristics Note: 46 patients had eGFR < 60 Note: authors contacted, but they did not reply Note: RoB assessment was not provided since no outcomes were reported

DUAL V 2016.

Study characteristics
Methods	Study design Phase 3, parallel RCT Study time frame: 20 September 2013 to 4 November 2014 Duration of follow‐up: 26 weeks Countries: Argentina, Australia, Greece, Hungary, Mexico, Russia, Slovakia, Spain, South Africa, USA Setting: multicentre (number of sites not reported)
Participants	Study characteristics Inclusion criteria: ≥ 18 years; T2DM; HbA1c level of 7% to 10% (inclusive); taking a stable dose of glargine (total daily dose, 20 to 50 U, inclusive, allowing individual fluctuations of ± 10% for at least 56 days prior to screening), with stable daily dosing of metformin (≥ 1500 mg or maximum tolerated dose), BMI ≤ 40; able to adhere to the protocol; participants with CKD were also reported Exclusion criteria: any use of oral antidiabetic agents (except for metformin) within 90 days prior to visit 1 (screening); current use of any drug (except metformin and insulin glargine) or anticipated change in concomitant medication, which in the investigator's opinion could interfere with the glucose metabolism (e.g. systemic corticosteroids); previous and/or current treatment with any insulin regimen other than basal insulin; previous and/or current treatment with GLP‐1 receptor agonists Baseline characteristics Number: overall/CKD (557/34) Mean age ± SD (years): not reported for CKD patients Sex (M/F): not reported for CKD patients
Interventions	Intervention group 1 IDegLira: 100 units/mL insulin degludec and 3.6 mg/mL liraglutide, initiated degludec/liraglutide at 16 dose steps (16 U of degludec/0.6 mg of liraglutide) Intervention group 2 Insulin degludec Co‐interventions Each treatment titrated to the same fasting glucose target. Target‐driven titration was performed twice weekly based on the mean of 3 previous daily self‐monitored pre‐breakfast blood glucose measurements. If this mean was above or below the 4.0 to 5.0 mmol/L target, patients were to respectively increase or decrease their dose by 2 dose steps or 2 U
Outcomes	Primary outcomes Change from baseline in HbA1c at week 26 Secondary outcomes Change from baseline in body weight at week 26 Number of treatment‐emergent confirmed hypoglycaemic episodes at week 26
Notes	Additional information Funding source: Novo Nordisk. Novo Nordisk was involved in the study design and protocol development, provided logistical support, and obtained the data, which was evaluated jointly by the authors and the sponsor Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: central interactive voice/web system; the treatment groups were comparable at baseline with respect to demographics and characteristics Note: 34 patients had an eGFR < 60 included in the analysis Note: authors contacted, but they were not able to provide further data

ELIXA 2018.

Study characteristics
Methods	Study design Phase 3, parallel RCT Study time frame: June 2010 to February 2015 Duration of follow‐up: the mean duration of exposure to study medication was 690 days in the lixisenatide group and 712 days in the placebo group ‐ last follow‐up week 108 Country: Argentina, Australia, Austria, Belarus, Belgium, Belgium, Bulgaria, Canada, Chile, China, Colombia, Denmark, Ecuador, Estonia, Finland, France, Georgia, Germany, Guatemala, India, Israel, Italy, Japan, Republic of Korea, Latvia, Lithuania, Mexico, Netherlands, Norway, Panama, Peru, Philippines, Poland, Portugal, Romania, Russian Federation, Serbia, South Africa, Spain, Sweden, Switzerland, Taiwan, Tunisia, Turkey, Ukraine, United Arab Emirates, UK, USA Setting: multicentre (number of sites not reported)
Participants	Study characteristics Inclusion criteria: T2DM; having an acute coronary event within 180 days before screening Exclusion criteria: < 30 years; percutaneous coronary intervention within the previous 15 days; coronary‐artery bypass graft surgery for the qualifying event; planned coronary revascularisation procedure within 90 days after screening; eGFR < 30 mL/min/1.73 m²; HbA1c < 5.5% or > 11.0%; an inability to provide written informed consent Baseline characteristics Number (randomised/analysed): overall/CKD (4642/1398) Mean age ± SD (years): not reported for CKD patients Sex (M/F): not reported for CKD patients
Interventions	Intervention group Lixisenatide (SC): once/day from 10 to 20 μg Control group Volume‐matched placebo Co‐interventions Glycaemic control was managed by the investigators in accordance with local clinical practice guidelines by the adjustment of concomitant glucose‐lowering agents or the addition of new antidiabetic medications, except other incretin therapies
Outcomes	Primary outcomes Time to first occurrence of primary CV event: CV death, non‐fatal MI, non‐fatal stroke or hospitalisation for unstable angina (median follow‐up of 25 months) Secondary outcomes Time to first occurrence of CV event: CV death, non‐fatal MI, non‐fatal stroke, hospitalisation for unstable angina or hospitalisation for heart failure (median follow‐up of 25 months) Time to first occurrence of CV event: CV death, non‐fatal MI, non‐fatal stroke, hospitalisation for unstable angina, hospitalisation for heart failure or coronary revascularisation procedure (median follow‐up of 25 months) Percent change from baseline in UACR at week 108
Notes	Additional information Funding source: Sanofi. The funder of this study was involved in the study design, data collection, data review, data analysis, and drafting of the report Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: Interactive Voice Response System and the Interactive Web Response System; similar baseline characteristics were reported Note: 1398 patients had an eGFR < 60 included in the analysis Note: authors were contacted, but they did not reply Note: This study reported outcome data both as RR and HR, and some outcomes could report duplicate RoB judgements

EXSCEL 2017.

Study characteristics
Methods	Study design Phase 3, parallel RCT Study time frame: 18 June 2010 to 24 April 2017 Duration of follow‐up: 3.2 years (median) Country: multinational (35 countries) Setting: multicentre (687 centres)
Participants	Study characteristics Inclusion criteria: adults with T2DM (HbA1c 6.5 to 10.0% (48 to 96 mmol/mol)) Exclusion criteria: history of 2 or more episodes of severe hypoglycaemia (defined as hypoglycaemia for which a patient received third‐party assistance) during the preceding 12 months; kidney failure or an eGFR at entry of < 30 mL/min/1.73 m2; a personal or family history of medullary thyroid carcinoma or multiple endocrine neoplasia type 2; a baseline calcitonin level > 40 ng/L, or previous treatment with a GLP‐1 receptor agonist Baseline characteristics Number (randomised/analysed): intervention group (7326/7326); control group (7365/7365) Mean age ± SD (years): not reported Sex (M/F): overall (9116/5575)
Interventions	Intervention group Exenatide: 2 mg once/week Control group Placebo Co‐interventions There was no run‐in period Open‐label medications for diabetes were permitted
Outcomes	Primary outcomes Primary efficacy outcome MACE events Primary safety outcome MACE events Secondary outcomes Secondary efficacy outcome all‐cause death Secondary efficacy outcome CV death Secondary efficacy outcome MI Secondary efficacy outcome of stroke Secondary efficacy outcome hospitalisation for acute coronary syndrome Secondary efficacy outcome hospitalisation for heart failure
Notes	Additional information Funding source: Amylin Pharmaceuticals, a wholly owned subsidiary of AstraZeneca Information on allocation concealment and balance/imbalance in baseline characteristics: interactive voice‐response system; similar baseline characteristics was reported Conflicts of interest: reported Note: there were some inconsistencies in data among publications Note: This study reported outcome data both as RR and HR and some outcomes could report duplicate RoB judgements

FLOW 2023.

Study characteristics
Methods	Study design Parallel RCT Study time frame: June 2019 to May 2021 Duration of follow‐up: median follow‐up 3.4 years Country: multinational (28 countries) Setting: multicentre (387 centres)
Participants	Study characteristics Inclusion criteria: T2DM; eGFR ≥ 50 and ≤ 75 mL/min/1.73 m2 and UACR > 300 and < 5000 mg/g or eGFR ≥ 25 to < 50 mL/min/1.73 m2 and UACR > 100 and < 5000 mg/g Exclusion criteria: known or suspected hypersensitivity to trial product(s) or related products; pregnancy, breastfeeding or intention to become pregnant; child‐bearing potential and not using a highly effective contraceptive method; participation in any clinical trial of an approved or non‐approved investigational medicinal product within 30 days before screening; any disorder which, in the investigator’s opinion, might jeopardise a subject’s safety or compliance with the protocol; congenital or hereditary kidney diseases or autoimmune kidney diseases; MI, stroke, hospitalisation for unstable angina or TIA within 60 days prior to the day of screening; NYHA class IV; or planned coronary, carotid or peripheral artery revascularisation; use of any GLP‐1RA within 30 days prior to screening; personal or first‐degree relative(s) with a history of multiple endocrine neoplasia type 2 or medullary thyroid cancer; chronic or intermittent haemodialysis or peritoneal dialysis within ≤ 90 days; uncontrolled and potentially unstable diabetic retinopathy or maculopathy; combination use of ACEi and ARB Baseline characteristics Number (randomised/analysed): intervention group (1767/1766); control group (1766/1766); ITT Mean age ± SD (years): intervention group (66.6 ± 9.0); control group (66.7 ± 9.0) Sex (M/F): intervention group (1248/519); control group (1216/550)
Interventions	Intervention group Semaglutide (SC): 1.0 mg once/week Control group Placebo Co‐interventions RAAS blockade
Outcomes	Primary outcomes Time to first occurrence of a composite endpoint (assessed at EoT) Onset of kidney failure: initiation of chronic KRT (dialysis or kidney transplantation) or persistent eGFR < 15 mL/min/1.73 m2 for at least 4 weeks Death from kidney failure CV death Onset of persistent ≥ 50% reduction in eGFR versus baseline Secondary outcomes Annual rate of change in eGFR (total eGFR slope) (assessed at EoT) CV MACE (assessed at EoT) Time to occurrence of all‐cause death (assessed at EoT) Supportive secondary endpoints (assessed at EoT) Time to occurrence of each of the individual components of the primary composite endpoint and of the confirmatory secondary MACE endpoint (assessed at EoT) Time to first occurrence of major adverse limb events, a composite endpoint Acute limb ischaemia hospitalisation (assessed at EoT) Chronic limb ischaemia hospitalisation (assessed at EoT) Annual rate of change in eGFR (chronic eGFR slope) (Week 12 to EOT) Change in eGFR (baseline to week 12) Change in eGFR (cystatin C CKD‐EPI) (assessed at EoT) Relative change in UACR (assessed at EoT) Change in body weight (assessed at EoT) Change in HbA1c (assessed at EoT) Change in SBP (assessed at EoT) Change in DBP (assessed at EoT) Number of severe hypoglycaemic episodes (assessed at EoT) Exploratory endpoints change in EQ‐5D‐5L index score (assessed at EoT) Change in EQ‐5D‐5L VAS score (assessed at EoT)
Notes	Additional information Funding source: Novo Nordisk Information on allocation concealment and balance/imbalance in baseline characteristics: central interactive web system; no information on baseline characteristics was reported Conflicts of interest: reported Note: The primary study reported 3533 participants randomised, while previous studies reported 3534. We have used the highest number of participants available in this table Note: The sponsor provided the analysis Note: AKI included both AKI and failure. To avoid confusion in definition and/or reporting, this data has been reported only in the table of adverse events Note: The paper reported either KRT and GFR <15 mL/min, but only the first outcome has been extracted in this review as kidney failure to avoid confusion Note: We did not include major kidney disease events, defined as a composite of the onset of kidney failure (dialysis, transplantation, eGFR of < 15 mL/min/1.73 m2), at least a 50% reduction in the eGFR from baseline, or death from kidney‐related or CV causes

GetGoal‐O 2015.

Study characteristics
Methods	Study design Phase 3, parallel RCT Study time frame: June 2013 to February 2015 Duration of follow‐up: 24 weeks Countries: Australia, Bulgaria, Canada, Denmark, Germany, Norway, Peru, Poland, South Africa, Spain, Sweden, UK, USA Setting: multicentre (71 centres)
Participants	Study characteristics Inclusion criteria: diagnosis of T2DM inadequately controlled with a current antidiabetic treatment regimen; ≥ 70 years at the time of signing the informed consent; at least 3 months on the current antidiabetic treatment regimen; ability to be compliant and to complete study procedures, including self‐injection; permitted antidiabetic therapies were metformin, sulfonylurea, meglitinide, pioglitazone, and basal insulin Exclusion criteria: HbA1c ≤ 7% and > 10%; basal insulin therapy combined with either a sulfonylurea or Meglitinide; severe kidney impairment at visit 6 (week 21), amylase and/or lipase > 3 times the ULN at visit 6 (week 21); history of severe hypoglycaemia associated with unawareness of symptoms or leading to unconsciousness, coma, or seizure ≤ 6 months before screening Baseline characteristics Number: overall/CKD (350/97) Mean age ± SD (years): not reported for CKD patients Sex (M/F): not reported for CKD patients
Interventions	Intervention group Lixisenatide: 10 to 20 µg, once/day Control group Placebo: once/day Co‐interventions Not reported
Outcomes	Primary outcomes Absolute change in HbA1c from baseline to week 24 Secondary outcomes Change in 2‐hour photoplethysmography from baseline to Week 24 Change in average 7‐point self‐measured plasma glucose profiles from baseline to week 24 Change in body weight from baseline to week 24 Change in FPG from baseline to week 24 Percentage requiring rescue therapy during the 24‐week treatment period Change in plasma glucose excursions from baseline to week 24 Change in total daily basal insulin dose from baseline to week 24 Percentage with symptomatic and severe symptomatic hypoglycaemia The percentage with HbA1c reduction > 0.5% at week 24 did not experience documented (plasma glucose < 60 mg/dL) symptomatic hypoglycaemia
Notes	Additional information Funding source: Sanofi Conflicts of interest: principal investigators are not employed by the organisation sponsoring the study Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Note: 97 patients had an eGFR < 60 included in the analysis Note: abstract‐only publication

GetGoal studies 2017 ‐ pooled.

Study characteristics
Methods	Study design Nine phase 3, parallel RCTs (according to the inclusion/exclusion criteria, only 5 studies were eligible) Study time frame: overall from 2008 and 2011 Duration of follow‐up: 24 weeks (except for GetGoalMono, 12 weeks) Country: multinational (number of countries not reported) Setting: multicentre (number of sites not reported)
Participants	Study characteristics Inclusion criteria: participants with type 2 diabetes in the 9 lixisenatide trials were included in the pooled analysis Exclusion criteria: type 1 diabetes, pregnancy and other medical conditions Baseline characteristics Number: intervention group (759); control group (482) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group Lixisenatide: 10 µg/day for 1 week, followed by 15 µg/day for 1 week, then 20 µg/day up to the end of trial Control group Placebo Co‐interventions GetGoal‐P: pioglitazone and metformin GetGoal‐L‐Asia: insulin and sulphonylureas GetGoal‐L: insulin and metformin GetGoal‐S: sulphonylureas and metformin GetGoal‐Mono: none
Outcomes	Primary outcomes HbA1c at 12 or 24 weeks Secondary outcomes Placebo‐adjusted mean change, from baseline to week 24 with lixisenatide (week 12 for GetGoal‐Mono) 2‐hour postprandial plasma glucose at 12 or 24 weeks FPG at 12 or 24 weeks Basal insulin dose at 12 or 24 weeks Body weight at 12 or 24 weeks Adverse events at 12 or 24 weeks
Notes	Additional information Funding source: Sanofi Conflicts of interest: many authors reported CIs with pharmaceutical organisations Trial registration number: GetGoal‐M: NCT00712673 ‐ note that the exclusion criteria reported "renal impairment defined with creatinine > 1.4 mg/dL in women and creatinine > 1.5 mg/dL in men" and maybe it was not relevant for this review GetGoal‐F1: NCT00763451 ‐ note that the exclusion criteria reported "renal impairment defined with creatinine > 1.4 mg/dL in women and creatinine > 1.5 mg/dL in men" and maybe it was not relevant for this review GetGoal‐Mono: NCT00688701 GetGoal‐P: NCT00763815 ‐ note that the exclusion criteria reported "renal impairment defined with creatinine > 1.4 mg/dL in women and creatinine > 1.5 mg/dL in men (applicable only for patients with metformin treatment)" but was still included in this review GetGoal‐L‐Asia: NCT00866658 GetGoal‐M‐Asia: NCT01169779 ‐ note that the exclusion criteria reported "renal impairment defined with creatinine > 1.4 mg/dL in women and creatinine > 1.5 mg/dL in men" and maybe it was not relevant for this review GetGoal‐L: NCT00715624 ‐ note that the exclusion criteria reported "renal impairment defined with creatinine > 1.4 mg/dL in women and creatinine > 1.5 mg/dL in men (applicable only for patients with metformin treatment)" but was still included in this review GetGoal‐Duo1: NCT00975286 ‐ note that the exclusion criteria reported "renal impairment defined with creatinine > 1.4 mg/dL in women and creatinine > 1.5 mg/dL in men" and maybe it was not relevant for this review GetGoal‐S: NCT00713830 ‐ note that the exclusion criteria reported "renal impairment defined with creatinine > 1.4 mg/dL in women and creatinine > 1.5 mg/dL in men (applicable only for patients with metformin treatment)" but was still included in this review Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Note: participants with normal kidney function (eGFR > 90 mL/min) were not included in this review

GRADE 2022.

Study characteristics
Methods	Study design Parallel RCT Study time frame: July 2013 and August 2017 Duration of follow‐up: 5 years Country: USA Setting: multicentre (36 centres)
Participants	Study characteristics Inclusion criteria: diabetes that had been diagnosed within the previous 10 years; treated with at least 500 mg metformin/day, but no other glucose‐lowering medications; HbA1c of 6.8% to 8.5% (50.8 to 69.4 mmol/mol) at the time of randomisation (some patients reported CKD) Exclusion criteria: history of a major CV event in the year before randomisation; NYHA classification of III or higher; eGFR < 30 mL/min/1.73 m2 Baseline characteristics Number CKD < 60 mL/min: intervention group 1 (35); intervention group 2 (28); intervention group 3 (34); intervention group 4 (28) Moderate to severe albuminuria: intervention group 1 (191); intervention group 2 (201); intervention group 3 (213); intervention group 4 (195) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group 1 Glargine: administered daily at an initial dose of up to 20 U Intervention group 2 Glimepiride: increased from 1 to 2 mg to a maximum of 8 mg/day Intervention group 3 Liraglutide: maximum dose of 1.8 mg/day Intervention group 4 Sitagliptin: 100 mg/day Co‐interventions Metformin
Outcomes	Primary outcomes Hb level of 7.0% higher at EoT Secondary outcomes Hospitalisation at EoT MACE at EoT Death at EoT Chronic eGFR slope (change in eGFR between year 1 and trial end) Composite kidney disease progression outcome (albuminuria, dialysis, transplant, or death due to kidney disease) at EoT eGFR < 60 mL/min/1.73m2, 40% decrease in eGFR to < 60 mL/min/1.73m2 at EoT Doubling of UACR to ≥ 30 mg/g, and progression of KDIGO stage at EoT
Notes	Additional information Funding source: National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) of the National Institutes of Health Information on allocation concealment and balance/imbalance in baseline characteristics: central web system; not reported Conflicts of interest: reported Note: author contacted but they did not reply Note: RoB assessment was not provided since no outcomes were reported

Harmony Outcomes 2018.

Study characteristics
Methods	Study design Phase 4, parallel RCT Study time frame: 1 July 2015 to 12 March 2018 Duration of follow‐up: 1.6 years (median) Country: multinational (28 countries) Setting: multicentre (610 centres)
Participants	Study characteristics Inclusion criteria: men and women ≥ 40 years; T2DM and established disease of the coronary arteries (MI, at least 50% stenosis in one coronary artery or more, or previous coronary revascularisation), cerebrovascular (ischaemic stroke, at least 50% carotid artery stenosis, or a previous carotid vascular procedure), or peripheral arterial circulation (intermittent claudication and an ankle‐to‐brachial index < 0.9, non‐traumatic amputation, or a previous peripheral vascular procedure) who had a HbA1c > 7.0% (53 mmol/mol) Exclusion criteria: eGFR < 30 mL/min/1.73 m²; severe gastroparesis, previous pancreatitis or substantial risk factors for pancreatitis; personal or family history of medullary carcinoma of the thyroid or multiple endocrine neoplasia type 2; history of pancreatic neuroendocrine tumours; current use of a GLP‐1 receptor agonist Baseline characteristics Number (randomised/analysed): overall (2222/2222) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group Albiglutide: 30 mg; dose of treatment could be increased to 50 mg Control group Placebo Co‐interventions If the glycaemic goal (which was based on local guidelines), as determined by the investigator, was not met after an increase in the dose of the study medication, other glucose‐lowering medications could be adjusted or added (except for a GLP‐1‐receptor agonist)
Outcomes	Primary outcomes Time to first occurrence of MACE at baseline and months 8, 16, 24 and end of study (up to 2.7 years) Secondary outcomes Time to first occurrence of MACE or urgent revascularisation for unstable angina at baseline and months 8, 16, 24 and end of study (up to 2.7 years) Time to first occurrence of the individual components of the primary endpoint at baseline and months 8, 16, 24 and end of study (up to 2.7 years) Time to first occurrence of cardiovascular death or hospitalisation due to heart failure at baseline and months 8, 16, 24 and end of study (up to 2.7 years) All‐cause death at baseline and months 8, 16, 24 and end of study (up to 2.7 years) Time to initiation of insulin of more than 3 months duration for those subjects not treated with insulin at study start at baseline and months 8, 16, 24 and end of study (up to 2.7 years) The proportion of subjects achieving glycaemic control (HbA1c ≤ 7.0 % at final assessment) with no severe hypoglycaemic incidents and weight gain < 5 % of body weight at baseline and months 8, 16, 24 and end of study (up to 2.7 years)
Notes	Additional information Funding source: Astra‐Zeneca and Bristol‐Myers Squibb Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: central interactive voice/web system; no information on baseline characteristics was reported Note: 2222 patients had an eGFR < 60 included in the analysis Note: This study reported outcome data both as RR and HR and some outcomes could report duplicate RoB judgements

Idorn 2013.

Study characteristics
Methods	Study design Parallel RCT Study time frame: not reported Duration of follow‐up: 12 weeks Country: Denmark Setting: single centre (Copenhagen University Hospital)
Participants	Study characteristics Inclusion criteria: men and women aged 18 to 85 years with T2DM kidney failure either on chronic HD or chronic PD treatment lasting > 3 months at baseline Exclusion criteria: T1DM; chronic pancreatitis/previous acute pancreatitis; known or suspected hypersensitivity to trial product(s) or related products; treatment with oral glucocorticoids, calcineurin inhibitors, DPP4 inhibitors or other drugs, which in the investigator's opinion could interfere with glucose or lipid metabolism 90 days prior to screening; cancer (except basal cell skin cancer or squamous cell skin cancer) or any other clinically significant disorder which in the investigators' opinion could interfere with the results of the trial; inflammatory bowel disease; cardiac disease defined as: decompensated heart failure (NYHA class III‐IV) and/or diagnosis of unstable angina pectoris and/or MI within the last 6 months; BMI ≤ 18.5 kg/m2 or ≥ 50.0 kg/m2; females of childbearing potential who are pregnant, breast‐feeding, intend to become pregnant or are not using adequate contraceptive methods; clinical signs of diabetic gastroparesis; impaired liver function (transaminases > 2 times ULN); receipt of any investigational product 90 days prior to this trial; known or suspected abuse of alcohol or narcotics; screening calcitonin ≥ 50 ng/L; subjects with personal or family history of medullary thyroid carcinoma or a personal history of multiple endocrine neoplasia type 2 Baseline characteristics Number (randomised/analysed): intervention group (14/7); control group (10/9) Mean age ± SD (years): intervention group (66.6 ± 2.6); control group (65.9 ± 4.4) Sex (M/F): intervention group (12/2); control group (9/1)
Interventions	Intervention group Liraglutide (SC): 0.6 mg once/day titrated to maximum tolerable dose up to 1.8 mg for 12 weeks Control group Placebo: once/day for 12 weeks Co‐interventions Doses of baseline antidiabetic medication were individually adjusted in parallel with study medication according to prespecified treatment goals. To minimise the risk of hypoglycaemia, basal insulin dose was reduced by 20‐50% at randomisation and suIphonylureas were paused, while metformin was continued in unchanged doses
Outcomes	Primary outcomes Plasma liraglutide concentration (pmol/L) (week 12) Secondary outcomes Severe adverse events, glycaemic control, change in baseline insulin, body weight, hypoglycaemic episodes (divided into minor blood glucose < 3.1 mmol/L and no need of assistance) and major (blood glucose < 3.1 mmol/L and requiring assistance from third person) (week 12) Cardiovascular parameters: heart rate, BP, and prohormone brain natriuretic peptide concentration in plasma (week 12)
Notes	Additional information Funding source: Bo Feldt‐Rasmussen, Novo Nordisk A/S Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: no information on the allocation concealment was provided; similar baseline characteristics were reported

LEADER 2017.

Study characteristics
Methods	Study design Parallel RCT Study time frame: September 2010 to April 2012 Duration of follow‐up: median follow‐up of 3.8 years Country: multinational (32 countries) Setting: multicentre (410 sites)
Participants	Study characteristics Inclusion criteria: T2DM who had HbA1c ≥ 7.0% if they either had not received drugs for this condition previously or had been treated with one or more oral antihyperglycaemic agents or insulin (human neutral protamine Hagedorn, long‐acting analogue, or premixed) or a combination of these agents; aged ≥ 50 years with at least one cardiovascular coexisting condition (coronary heart disease, cerebrovascular disease, peripheral vascular disease, CKD of stage 3 or greater, or chronic heart failure of NYHA class II or III); or age ≥ 60 years with at least one cardiovascular risk factor, as determined by the investigator (microalbuminuria or proteinuria, hypertension and left ventricular hypertrophy, left ventricular systolic or diastolic dysfunction, or an ankle‐brachial index (the ratio of the SBP at the ankle to the SBP in the arm) of less than 0.9) Exclusion criteria: T1DM; use of GLP‐1–receptor agonists, DPP‐4 inhibitors, pramlintide, or rapid‐acting insulin; familial or personal history of multiple endocrine neoplasia type 2 or medullary thyroid cancer; acute coronary or cerebrovascular event within 14 days before screening and randomisation Baseline characteristics Number (randomised/analysed): intervention group (3048/3048); control group (3017/3017) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group Liraglutide: 1.8 mg Control group Placebo Co‐interventions Not reported
Outcomes	Primary outcomes Composite outcome in the time‐to‐event analysis of the first occurrence of death from cardiovascular causes, nonfatal (including silent) MI, or nonfatal stroke from randomisation to month 60 + 30 days Secondary outcomes An expanded composite cardiovascular outcome (death from cardiovascular causes, nonfatal MI, nonfatal stroke, coronary revascularisation, or hospitalisation for unstable angina pectoris or heart failure) from randomisation to month 60 + 30 days Death from any cause from randomisation to month 60 + 30 days A composite kidney and retinal microvascular outcome (nephropathy defined as the new onset of macroalbuminuria or a doubling of the SCr level and an eGFR ≤ 45 mL/min/1.73 m2), the need for continuous KRT, or death from kidney disease; retinopathy defined as the need for retinal photocoagulation or treatment with intravitreal agents, vitreous haemorrhage, or the onset of diabetes‐related blindness from randomisation to month 60 + 30 days HbA1c from randomisation to month 60 + 30 days
Notes	Additional information Funding source: Novo Nordisk and the National Institutes of Health. Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: central interactive voice/web system; similar baseline characteristics were reported Note: This study reported outcome data both as RR and HR and some outcomes could report duplicate RoB judgements

Leiter 2014.

Study characteristics
Methods	Study design Phase 3, parallel RCT Study time frame: 7 May 2010 to 30 May 2012 Duration of follow‐up: 52 weeks Country: multinational (15 countries) Setting: multicentre (134 centres)
Participants	Study characteristics Inclusion criteria: male and non‐pregnant, non‐lactating female patients ≥ 18 years of age with historical diagnoses of T2DM; baseline HbA1c between 7.0 and 10.0% (53 to 86 mmol/mol); BMI between 20 and 45 kg/m2; fasting C‐peptide level of ≥ 0.8 ng/mL (0.26 mmol/L); GFR ≥ 15 to < 90 mL/min/1.73 m2; HB ≥ 10 g/dL for male patients and ≥ 9 g/dL for female patients; normal levels of thyroid‐stimulating hormone or clinically euthyroid Exclusion criteria: malignant disease (except squamous cell or basal cell); history of diabetic gastroparesis; current ongoing symptomatic biliary disease or history of pancreatitis; significant GI surgery or surgery thought to significantly affect upper GI function; recent clinically significant cardiovascular and/or cerebrovascular disease; history of HIV infection, and acute symptomatic hepatitis B or C infection Baseline characteristics Number (randomised/analysed): intervention group 1 (254/249); intervention group 2 (253/246) Mean age ± SD (years): intervention group 1 (63.2 ± 8.37); intervention group 2 (63.5 ± 9.02) Sex (M/F): intervention group 1 (136/113); intervention group 2 (130/116)
Interventions	Intervention group 1 Albiglutide (SC): 30 mg once/week (with treatment‐masked up‐titration, if needed, to 50 mg weekly) Intervention group 2 Sitagliptin: 50 mg dosed based on the eGFR value at randomisation per the sitagliptin package insert Co‐interventions All patients continued to receive their prescribed oral antihyperglycaemic medication regimen (metformin, thiazolidinedione, sulphonylureas, or any combination of these oral antihyperglycaemic medications) for the duration of the study except patients with GFR < 60 mL/min/1.73 m2, who were washed off their background metformin Diet and exercise
Outcomes	Primary outcomes Change from baseline in HbA1c at week 26 Secondary outcomes HbA1c at week 26 FBG at week 26 Body weight at week 26 Proportion of patients who met prespecified HbA1c treatment targets at week 26 Time to hyperglycaemic rescue Adverse and severe adverse events (clinical laboratory parameters, vital sign measurements, ECG readings, and physical examinations) at week 26
Notes	Additional information Funding source: GlaxoSmithKline Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: interactive voice response system; similar baseline characteristics were reported Note: the majority of data were reported at 26 weeks

LIRA‐PRIME 2019.

Study characteristics
Methods	Study design Parallel RCT Study time frame: March 2016 to August 2017 Duration of follow‐up: 104 weeks followed by a 1‐week follow‐up period after the end of treatment Country: multinational (9 countries) Setting: multicentre (219 sites)
Participants	Study characteristics Inclusion criteria: male or female at least 18 years of age at the time of signing informed consent; subjects diagnosed (clinically) with T2DM equal to or above 90 days prior to the screening visit; stable daily dose of metformin as monotherapy equal to or above 1500 mg or maximum tolerated dose within 60 days prior to the screening visit; HbA1c 7.5& to 9.0 % (59 to 75 mmol/mol) (both inclusive) and measured within the last 90 days prior to the screening visit Exclusion criteria: female who is pregnant, breastfeeding or intends to become pregnant or is of childbearing potential and not using adequate contraceptive methods (adequate contraceptive measures as required by local regulation or practice); treatment with any medication for the indication of diabetes other than metformin in a period of 60 days before the screening visit; an exception is short‐term treatment (below or equal to 7 days in total) with insulin in connection with intercurrent illness; receipt of any investigational medicinal product within 30 days before the screening visit; any disorder, which in the investigator's opinion might jeopardise subject's safety or compliance with the protocol Baseline characteristics Number (randomised/analysed): overall (861/not reported) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group Liraglutide: initiated treatment with 0.6 mg/d, with subsequent dose escalation up to 1.8 mg/d or maximum tolerated dose Control group Oral antidiabetic drugs including alpha‐glucosidase inhibitors, dipeptidyl peptidase‐4 (DPP‐4) inhibitors, meglitinides, sodium‐glucose cotransporter‐2 inhibitors (SGLT‐2is), sulphonylureas and thiazolidinediones Co‐interventions Metformin Patients are to continue using the same antidiabetic drugs
Outcomes	Primary outcomes Time to inadequate glycaemic control to weeks 26 to 104 Secondary outcomes Time to premature treatment discontinuation (for any reason including inadequate glycaemic control): weeks 0 ‐104 Change in HbA1c: week 0, week 104/premature treatment discontinuation Achieve HbA1c ≤ 7.0% without weight gain: week 104/premature treatment discontinuation Achieve HbA1c ≤ 7.0% without treatment‐emergent severe hypoglycaemic episodes or BG‐confirmed symptomatic hypoglycaemic episodes: week 104/premature treatment discontinuation Achieve HbA1c ≤ 7.0% without weight gain and no treatment‐emergent severe hypoglycaemic episodes or blood glucose confirmed symptomatic hypoglycaemic episodes: week 104/premature treatment discontinuation Change FPG: week 104/premature treatment discontinuation Change in body weight: week 104/premature treatment discontinuation Change in BMI: week 104/premature treatment discontinuation Change in SBP and DBP: week 104/premature treatment discontinuation Number of severe hypoglycaemic episodes: weeks 0 to 104 Number of severe or BG‐confirmed symptomatic hypoglycaemic episodes: weeks 0 to 104 Number of documented symptomatic hypoglycaemic episodes: weeks 0 to 104 Number of serious adverse events: weeks 0 to 105 Number of AEs leading to permanent discontinuation of trial product: weeks 0 to 105 Change in creatinine, total bilirubin: week 104/premature treatment discontinuation Change in eGFR: week 104/premature treatment discontinuation
Notes	Additional information Funding source: Novo Nordisk A/S Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: interactive web response system; no information on baseline characteristics was reported Note: RoB assessment was not provided since no outcomes were reported

LIRA‐RENAL 2016.

Study characteristics
Methods	Study design Parallel RCT Study time frame: June 2012 to August 2013 Duration of follow‐up: 26 weeks Countries: multinational (France, Poland, Russian Federation, Ukraine, UK, USA) Setting: multicentre (78 sites)
Participants	Study characteristics Inclusion criteria: 18 to 80 years, previously diagnosed with T2DM; HbA1c 7% to 10% (53 to 86 mmol/mmol inclusive); on stable diabetes treatment for > 90 days before screening; the following background diabetes treatments were allowed: monotherapy or dual‐therapy combinations of metformin and/or sulphonylurea and/or pioglitazone, monotherapy with basal or premix insulin, or any combination of basal or premix insulin with metformin and/or pioglitazone; moderate CKD (eGFR 30 to 59 mL/min/1.73 m2) > 90 days before screening (confirmed at screening); BMl of 25 to 45 kg/m2 (inclusive) Exclusion criteria: hypoglycaemic unawareness and/or recurrent severe hypoglycaemia as judged by the investigator; impaired liver function (ALT ≥ 2.5 x ULN; history of chronic pancreatitis or idiopathic acute pancreatitis; NYHA Functional Classification IV heart failure; episode of unstable angina, acute coronary event, cerebral stroke/TIA, or other significant cardiovascular event within the past 180 days; SBP ≥ 180 mm Hg or a DBP ≥ 100 mm Hg; screening calcitonin value ≥ 50 ng/L; personal or family history of medullary thyroid carcinoma or multiple endocrine neoplasia syndrome type 2 Baseline characteristics Number (randomised/analysed): intervention group (140/140); control group (139/137) Mean age ± SD (years): intervention group (68 ± 8.3); control group (66.3 ± 8.0) Sex (M/F): intervention group (75/65); control group (65/72)
Interventions	Intervention group Liraglutide: initiated at 0.6 mg/d with subsequent weekly dose escalations of 0.6 mg/d until the maintenance dose of 1.8 mg/d was reached Control group Placebo Co‐interventions If the patient was on insulin with an HbA1c ≤ 8% (64 mmol/mol) at screening, the pre‐trial insulin dose was reduced by 20% at day 0 and kept fixed until the liraglutide/placebo dose escalation was complete. Titration to pre‐trial insulin dose was allowed at the discretion of the investigators. Patients were to maintain their background diabetes medications during the study and were allowed to reduce insulin or sulphonylureas doses if hypoglycaemic episodes occurred
Outcomes	Primary outcomes Change in HbA1c from baseline to week 26. Responder endpoints at week 26 for HbA1c < 7.0% (< 53 mmol/mol) and HbA1c < 7.0% (< 53 mmol/mol) with no hypoglycaemic episodes were determined at week 26 Secondary outcomes Change from baseline to week 26 in FBG, body weight, BMI, SBP and DBP, fasting lipids, and selected cardiovascular biomarkers, total prescribed daily insulin dose at week 26 Adverse events, change from baseline in eGFR, UACR, amylase, lipase and pulse rate at week 26 Hypoglycaemic episodes at week 26
Notes	Additional information Funding source: Novo Nordisk Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: telephone‐ or Web‐based randomisation system; similar baseline characteristics were reported Note: This study reported outcome data both as RR and HR and some outcomes could report duplicate RoB judgements

Liu 2022.

Study characteristics
Methods	Study design Parallel RCT Study time frame: December 2017 and March 2019 Duration of follow‐up: 12 weeks Country: China Setting: single centre (Affiliated Hospital of Hebei University)
Participants	Study characteristics Inclusion criteria:18 to 65 years; GLP‐1RA naive, and meeting the standard of DN diagnosis; voluntary participation with informed consent Exclusion criteria: non‐diabetes induced renal conditions; presence of serious acute/chronic complications of diabetes; pregnant or breastfeeding women; severe mental disorders; relevant contraindications and allergies Baseline characteristics Number (randomised/analysed): intervention group (42/42); control group (42/42) Mean age ± SD (years): intervention group (46.89 ± 11.26); control group (45.38 ± 10.34) Sex (M/F): intervention group (21/21); control group (22/20)
Interventions	Intervention group Liraglutide: 0.6 mg to 1.8 mg/d Control group No treatment Co‐interventions Conventional treatment was used in the control group, including diabetes diet, exercise education, oral antidiabetic drug or insulin, BP control, and treatment of lipid metabolism disorders
Outcomes	Primary outcome HbA1c Secondary outcomes Blood glucose BMI Kidney function Body weight UAER
Notes	Additional information Funding source: Baoding Science and Technology Plan Self‐Funded Project (2041ZF306); Key Scientific Research Fund Project of Affiliated Hospital of Hebei University (2016Z007) Trial registration number: not reported Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Conflicts of interest: not reported Note: author contacted, but they did not reply

Ma 2024.

Study characteristics
Methods	Study design Parallel RCT Study time frame: January 2020 to January 2022 Duration of follow‐up: 12 weeks Country: China Setting: single centre
Participants	Study characteristics Inclusion criteria: 18 to 75 years; BMI 25 to 40 kg/m2; T2DM and UACR 30 to 300 mg/g; HbA1c > 7%; no history of hypertension or there is history of hypertension but BP control is stable and ≤ 140/90 mm Hg; no other glucose‐lowering agent or only metformin Exclusion criteria: received liraglutide or dapagliflozin within 3 months before screening; liver or kidney function impairment; recurrent UTI before screening Baseline characteristics Number (randomised/analysed): intervention group 1 (40/40); intervention group 2 (40/40); control group (40/40) Mean age ± SD (years): intervention group 1 (50.48 ± 6.51); intervention group 2 (49.83 ± 8.61); control group (52.83 ± 9.84) Sex (M/F): intervention group 1 (17/23); intervention group 2 (18/22); control group (19/21)
Interventions	Intervention group 1 Liraglutide: 1.2 mg/d Intervention group 2 Dapagliflozin: 10 mg/d Control group Conventional hypoglycaemic treatment Co‐interventions Dietary guidance and appropriate exercise intervention
Outcomes	Primary outcomes Blood glucose indexes at EoT Secondary outcomes Safety analysis at EoT HbA1c at EoT FPG at EoT Weight at EoT BMI at EoT
Notes	Additional information Funding source: self‐funded project Conflicts of interest: none Trial registration number: not reported Information on allocation concealment and balance/imbalance in baseline characteristics: the allocation process ensured reducing potential bias; similar baseline characteristics were reported

Muskiet 2019.

Study characteristics
Methods	Study design Parallel RCT POST‐HOC Study time frame: not reported Duration of follow‐up: 52 weeks Countries: multinational (Finland, Sweden, the Netherlands) Setting: multicentre (3 sites)
Participants	Study characteristics Inclusion criteria: diagnosis of T2DM, but otherwise healthy, HbA1c between 6.6% and 9.5%, inclusive, BMI 25 kg/m2 to 40 kg/m2, treated with a stable dose of metformin for at least 2 months prior to screening Exclusion criteria: patients previously in a study using exenatide; treated with oral anti‐diabetic medications other than metformin within 2 months of screening (thiazolidinediones within 5 months of screening); treated with insulin within 3 months of screening Baseline characteristics Number: intervention group 1 (2); intervention group 2 (5); data are related only to patients with microalbuminuria. Overall data on CKD people were not available Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group 1 Exenatide: 10 µg twice/day Intervention group 2 Insulin glargine Co‐interventions Stable and ongoing metformin monotherapy
Outcomes	Primary outcomes Beta‐cell function after 52 weeks of therapy: baseline (week ‐2) and 52 weeks Change from baseline to week 52 in CrCl and UAE Secondary outcomes Beta‐cell function 4 weeks after cessation of therapy: baseline (week ‐2) and 56 weeks Change in HbA1c: week 0 and week 52 Change in FPG: 0 weeks and 52 weeks Seven‐point self‐monitored blood glucose measurements: 0 weeks and 52 weeks Change in body weight: 0 weeks and 52 weeks eGFR during the study period
Notes	Additional information Funding source: AstraZeneca Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Note: authors contacted (10/7/21) but they did not reply

Neff 2016.

Study characteristics
Methods	Study design Parallel RCT Study time frame: not reported Duration of follow‐up: 34 weeks Country: Ireland Setting: single centre (St Vincent's Healthcare Group)
Participants	Study characteristics Inclusion criteria: T2DM with HbA1c 42 to 75 mmol/mol (6% to 9%); male or female > 30 years; have a negative pregnancy test at screening (women of childbearing potential only); BMI ≥ 25kg/m2; on a RAS antagonist, at a stable dose, for at least 8 weeks before inclusion into the study; established microalbuminuria; eGFR ≥ 30 mL/min/1.73m2 Exclusion criteria: any cognitive impediment that preclude the patient from giving free and informed consent; patients on DDP4 inhibitors or thiazolidinedione treatment; patients with stage 4‐5 kidney disease, defined as an eGFR ≤ 30 mL/min/1.73m2; patients who have used a GLP‐1 agent in the last 6 months; female patients of childbearing potential who are pregnant, breastfeeding, or unwilling to practice an acceptable barrier and/or hormonal method of contraception or abstinence during participation in the study; previous pancreatitis; hypersensitivity to GLP‐1 analogues; proliferative diabetic retinopathy; any other contraindications, as per the SmPC for liraglutide; patients with any other clinical condition or prior therapy that, in the opinion of the investigator, would make the patient unsuitable for the study or unable to comply with the dosing requirements; concurrent treatment with an investigational drug or participation in another clinical trial; use of an investigational drug within 4 weeks or 5 half‐lives, whichever is longer, preceding the first dose of investigational medicinal product Baseline characteristics Number (randomised/analysed): intervention group (10/10); control group (10/8) Mean age ± SD (years): intervention group (65 ± 7); control group (64 ± 10) Sex (M/F): intervention group (7/3); control group (7/3)
Interventions	Intervention group Liraglutide: 0.6 mg Control group Standard diabetes care includes RAAS inhibitor or antagonist Co‐interventions Not reported
Outcomes	Primary outcomes MCP‐1: creatinine ratio in urine at 26 weeks Secondary outcomes UACR at 26 weeks UAER at 26 weeks Safety in all participants as measured by adverse event rate at 34 weeks
Notes	Additional information Funding source: University College Dublin Conflicts of interest: not reported Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Note: abstract only, but NCT was available to extract outcomes

Parker 2022.

Study characteristics
Methods	Study design Parallel RCT Study time frame: not reported Duration of follow‐up: 32 days Country: Germany, UK Setting: multicentre (7 sites)
Participants	Study characteristics Inclusion criteria: T2DM (HbA1c: ≥ 6.5 to ≤ 10.5%) and CKD stage G3 (eGFR ≥ 30 to < 60 mL/min/1.73m2), on insulin and/or oral therapy, with a BMI 25 to 45 kg/ m2 Exclusion criteria: not reported Baseline characteristics Number (randomised/analysed): intervention group (21/not reported); control group (20/not reported) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group Cotadutide (SC): once/day titrated up to 300 µg Control group Placebo Co‐interventions Not reported
Outcomes	Primary outcomes Glucose response via mixed meal tolerance test up to the EoT Secondary outcomes HbA1c up to the EoT Body weight up to the EoT eGFR up to the EoT UACR up to the EoT Adverse and serious adverse events up to the EoT
Notes	Additional information Funding source: sponsored by AstraZeneca. Conflicts of interest: VP, TH, YC, MPe, LH, PA and LJ are employees and shareholders of AstraZeneca. TH has stock or stock options with AstraZeneca, GW Pharma and Jazz Pharma Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Note: RoB assessment was not provided since no outcomes were reported

PIONEER 5 2019.

Study characteristics
Methods	Study design Phase 3a trial, parallel RCT Study time frame: September 2016 to September 2017 Duration of follow‐up: 26 weeks on treatment and 5 weeks post‐treatment (31 weeks total follow‐up) Country: multinational (Denmark, Finland, Israel, Poland, Russia, Sweden, the UK, and the USA) Setting: multicentre (88 sites)
Participants	Study characteristics Inclusion criteria: ≥ 18 years with T2DM (diagnosed ≥ 90 days before screening); HbA1c 7.0% to 9.5% (53 to 80 mmol/mol) and moderate renal impairment (eGFR 30 to 59 mL/min/1.73 m²); able to provide informed consent and be treated with stable doses of one of the following regimens for 90 days before screening: metformin (≥ 1500 mg or maximum tolerated dose), sulphonylurea (at least half of the maximum approved dose or maximum tolerated dose), or both; or basal insulin with or without metformin Exclusion criteria: rapidly progressing kidney disease (as judged by the investigator) or known nephrotic albuminuria (>2200 mg/24 h or > 2200 mg/g); family or personal history of multiple endocrine neoplasia type 2 or medullary thyroid carcinoma; history of malignant neoplasms within the past 5 years; history of pancreatitis; myocardial infarction, stroke, or admission to hospital for unstable angina or transient ischaemic attack within the past 180 days, or NYHA Class IV heart failure; or proliferative retinopathy or maculopathy (determined by fundus photography or dilated fundoscopy 90 days or fewer before randomisation and requiring acute treatment) Baseline characteristics Number (randomised/completed): intervention group (163/158); control group (161/156) Mean age ± SD (years): intervention group (71 ± 8); control group (70 ± 8) Sex (M/F): intervention group (83/80); control group (73/88)
Interventions	Intervention group Semaglutide (oral): initiated at a 3 mg dose, then escalated to 7 mg at 4 weeks, and 14 mg at 8 weeks Control group Placebo for 26 weeks Co‐interventions: Participants were required to continue background glucose‐lowering medication throughout the trial. Those receiving metformin and sulphonylureas were required to maintain the same dose level and frequency as at trial entry; those receiving basal insulin were recommended to have the dose decreased by 20% after random assignment to the treatment group to minimise the risk of hypoglycaemic episodes. Up‐titration of basal insulin (to a dose not exceeding that at randomisation) was permitted in weeks 10 to 16, after the maximum dose of oral semaglutide was reached. Patients with persistent and unacceptable hyperglycaemia were offered treatment intensification with rescue medication, prescribed at the investigator’s discretion and as an add‐on to assigned treatment, in accordance with international guidelines
Outcomes	Primary outcomes Change in HbAc: change from baseline to week 26 Secondary outcomes Body weight HbA1c target of 7.0% and 6.5% achieved HbA1c < 7.0% without treatment‐emergent symptomatic hypoglycaemia that was severe or confirmed by blood glucose concentration (< 3.1 mmol/L (56 mg/dL)), without HbA1c decrease of 1.0 percentage points or more with weight loss of 3% or more from baseline Weight loss of 5% or more and 10% or more Change from baseline to week 26 in fasting plasma glucose BMI HRQoL (SF‐36 version 2 Health Survey (acute version), and the status version of the Diabetes Treatment Satisfaction Questionnaire) Adverse events Number of symptomatic hypoglycaemic episodes that were severe or confirmed by blood glucose concentration (< 3.1 mmol/L (56 mg/dL)), or both Severe hypoglycaemia was defined as the participant requiring third‐party assistance to administer corrective treatment (American Diabetes Association classification) eGFR UACR
Notes	Additional information Funding source: Novo‐Nordisk A/S Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: central web system; no information on baseline characteristics was reported Note: This study reported outcome data both as RR and HR and some outcomes could report duplicate RoB judgements

PIONEER 6 2019.

Study characteristics
Methods	Study design Phase 3, parallel RCT Study time frame: 17 January 2017 to 25 September 2018 Duration of follow‐up: 69 weeks (mean) Country: multinational (Algeria, Argentina, Brazil, Canada, Denmark, Germany, India, Israel, Italy, Malaysia, Mexico, Netherlands, Poland, Romania, South Africa, Spain, Taiwan, Thailand, Turkey, UK, USA) Setting: multicentre (214 sites)
Participants	Study characteristics Inclusion criteria: ≥ 50 years and established CV disease or CKD, or if ≥ 60 years had cardiovascular risk factors only Exclusion criteria: treatment with any GLP‐1 receptor agonist, DPP4 inhibitors, or pramlintide within 90 days before screening; NYHA class 4 heart failure; planned coronary‐artery, carotid‐artery, or peripheral‐artery revascularisation; MI, stroke, or hospitalisation for unstable angina or TIA within 60 days before screening; long‐term or intermittent HD or PD, or severe kidney impairment (GFR < 30 mL/min/1.73 m2); and proliferative retinopathy or maculopathy resulting in active treatment Baseline characteristics Number (randomised/analysed): intervention group (1120/1120); control group (1125/1125) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group Semaglutide: up to 14 mg Control group Placebo Co‐interventions Background therapy with glucose‐lowering agents to maintain/optimise glycaemic control according to standard of care in local or international guidelines is encouraged during the trial at the investigator's discretion
Outcomes	Primary outcomes Time from randomisation to first occurrence of a MACE composite endpoint consisting of: CV death, non‐fatal MI or non‐fatal stroke Secondary outcomes From randomisation to first occurrence of an expanded composite CV endpoint consisting of: CV death, non‐fatal MI, non‐fatal stroke, unstable angina requiring hospitalisation or hospitalisation for heart failure Time from randomisation to first occurrence of each of the individual components in the expanded composite CV endpoint Time from randomisation to first occurrence of a composite endpoint consisting of: all‐cause death, non‐fatal MI or non‐fatal stroke
Notes	Additional information Funding source: Novo Nordisk Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: central interactive voice/web system; no information on baseline characteristics was reported Note: not all participants with CKD were included into the analyses (we included only CKD stages 3‐5): authors contacted but they did not reply Note: This study reported outcome data both as RR and HR and some outcomes could report duplicate RoB judgements

REWIND 2019.

Study characteristics
Methods	Study design Parallel RCT Study time frame: 22 July 2011 to 21 August 2019 Duration of follow‐up: 26 weeks Country: multinational (24 countries) Setting: multicentre (371 sites)
Participants	Study characteristics Inclusion criteria: previous/new T2DM; HbA1c ≤ 9.5%; stable dose of 0, 1 or 2 oral glucose‐lowering drugs ± basal insulin for ≥ 3 months; BMI ≥ 23 kg/m2; if age ≥ 50 years, at least 1 of: prior MI, prior ischaemic stroke, coronary revascularisation ≥ 2 years earlier, carotid or peripheral revascularisation ≥ 2 months earlier, unstable angina hospitalisation, image proven myocardial ischaemias, or percutaneous coronary intervention; if age ≥ 55 years, any of the above or at least 1 of: documented myocardial ischaemias by stress test or imaging, > 50% coronary, carotid or lower extremity artery stenosis, ankle‐brachial index < 0.9, eGFR persistently < 60 mL/min/1.73 m2; hypertension with left ventricular hypertrophy; or persistent albuminuria; if age ≥ 60 years, any of the above or at least 2 of: any tobacco use, use of lipid‐modifying therapy or a documented untreated LDL cholesterol ≥ 3.4 mmol/L (130 mg/dL) within the past 6 months, HDL cholesterol < 1.0 mmol/L (40 mg/dL) for men and < 1.3 mmol/L (50 mg/dL) for women or triglycerides ≥ 2.3 mmol/L (200 mg/dL) within the past 6 months, use of ≥ 1 blood pressure drug or untreated SBP ≥ 140 mm Hg or DBP ≥ 95 mm Hg, or waist‐to‐hip ratio > 1.0 (men) and > 0.8 (women); run‐in adherence to study drug = 100% Exclusion criteria: uncontrolled diabetes; severe hypoglycaemia in preceding year; coronary or cerebrovascular event in preceding 2 months or plans to revascularize; eGFR < 15 mL/min/1.73 m2; gastric bypass or emptying abnormality; prior pancreatitis/concordant symptoms; liver disease or ALT ≥ 3.0 × ULN; family history of/or C‐cell hyperplasia or medullary thyroid cancer or MEN 2A or 2B or calcitonin value ≥ 20 pg/mL; unwilling to stop GLP‐1 receptor agonist or DPP‐4 inhibitor or weight loss drug; cancer within prior 5 years; pregnant or not using reliable birth control; life expectancy < 1 year Baseline characteristics Number (randomised/analysed): intervention group (3516/3516); control group (3587/3587) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group Dulaglutide (SC): 1.5 mg once per week Control group Matching placebo Co‐interventions Investigators were similarly advised and periodically reminded to optimise the use of cardioprotective measures, including treatment of lipids and BP, use of antiplatelet agents, and promotion of healthy lifestyles according to applicable guidelines
Outcomes	Primary outcomes First occurrence of the composite of either CV death or non‐fatal MI or non‐fatal stroke Secondary outcomes Each component of the primary composite CV outcome A composite clinical microvascular outcome comprising retinal or kidney disease Hospitalisation for unstable angina Heart failure requiring hospitalisation or an urgent heart failure visit All‐cause death
Notes	Additional information Funding source: Eli Lilly and Company. Scientists employed by the funder were on the steering committee and contributed to trial design, trial implementation, and data interpretation. Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: central web system; no information on baseline characteristics was reported Note: data reported in the analyses were related to patients with CKD stages 3‐5 Note: authors contacted but they did not reply Note: This study reported outcome data both as RR and HR and some outcomes could report duplicate RoB judgements

Selvarajah 2024a.

Study characteristics
Methods	Study design Parallel RCT Study time frame: August 2020 to April 2022 Duration of follow‐up: 26 weeks Country: multinational (8 countries) Setting: multicentre (65 sites)
Participants	Study characteristics Inclusion criteria: 18 to 79 years; BMI ≥ 25 kg/m2 (>23 kg/m2 for patients enrolled in Japan); T2DM with HbA1c 6.5% to 12.5%; CKD with eGFR 20‐90 mL/min/1.73 m2 and UACR > 50 mg/g Exclusion criteria: requirement for KRT or kidney transplant; treatment with GLP‐1 receptor agonist within 30 days of screening; recent acute or subacute kidney function deterioration; recent CV event within 3 months of screening; basal calcitonin > 50 ng/L Baseline characteristics Number (randomised/analysed): intervention group 1 (52/52); intervention group 2 (49/49); intervention group 3 (51/51); intervention group 4 (45/45); control group (51/51) ‐ ITT Mean age ± SD (years): intervention group 1 (67.2 ± 7.3); intervention group 2 (65.7 ± 8.8); intervention group 3 (66.1 ± 7.4); intervention group 4 (67.0 ± 7.8); control group (69.5 ± 7.3) Sex (M/F): intervention group 1 (43/9); intervention group 2 (46/3); intervention group 3 (41/10); intervention group 4 (33/12); control group (38/13)
Interventions	Intervention group 1 Cotadutide: 100 ug/d Intervention group 2 Cotadutide: 300 ug/d Intervention group 3 Cotadutide: 600 ug/d Intervention group 4 Semaglutide: 1 mg/week Control group Placebo
Outcomes	Primary outcomes Change in UACR from baseline at week 14 Secondary outcomes Change in UACR from baseline at week 26 Change in HbA1c from baseline at week 14 and at week 26 Change in FPG from baseline at week 14 and at week 26 Change in hypoglycaemia from baseline at week 14 and at week 26 (as measured with CGM) Change in clinically significant hypoglycaemia from baseline at week 14 and at week 26 (as measured with CGM) Safety analysis during the study Quality of life (KDQOL‐36) at EoT
Notes	Additional information Funding source: AstraZeneca Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: information on allocation concealment was not reported; similar baseline characteristics were reported Note: this study has been split in two (Selvarajah 2024a for semaglutide; Selvarajah 2024b for cotadutide) to extract data from both GLP‐1 classes versus placebo

Selvarajah 2024b.

Study characteristics
Methods	Study design Parallel RCT Study time frame: August 2020 to April 2022 Duration of follow‐up: 26 weeks Country: multinational (8 countries) Setting: multicentre (65 sites)
Participants	Study characteristics Inclusion criteria: 18 to 79 years; BMI ≥ 25 kg/m2 (>23 kg/m2 for patients enrolled in Japan); T2DM with HbA1c 6.5% to 12.5%; CKD with eGFR 20‐90 mL/min/1.73 m2 and UACR > 50 mg/g Exclusion criteria: requirement for KRT or kidney transplant; treatment with GLP‐1 receptor agonist within 30 days of screening; recent acute or subacute kidney function deterioration; recent CV event within 3 months of screening; basal calcitonin > 50 ng/L Baseline characteristics Number (randomised/analysed): intervention group 1 (52/52); intervention group 2 (49/49); intervention group 3 (51/51); intervention group 4 (45/45); control group (51/51) ‐ ITT Mean age ± SD (years): intervention group 1 (67.2 ± 7.3); intervention group 2 (65.7 ± 8.8); intervention group 3 (66.1 ± 7.4); intervention group 4 (67.0 ± 7.8); control group (69.5 ± 7.3) Sex (M/F): intervention group 1 (43/9); intervention group 2 (46/3); intervention group 3 (41/10); intervention group 4 (33/12); control group (38/13)
Interventions	Intervention group 1 Cotadutide: 100 ug/d Intervention group 2 Cotadutide: 300 ug/d Intervention group 3 Cotadutide: 600 ug/d Intervention group 4 Semaglutide: 1 mg/week Control group Placebo
Outcomes	Primary outcomes Change in UACR from baseline at week 14 Secondary outcomes Change in UACR from baseline at week 26 Change in HbA1c from baseline at week 14 and at week 26 Change in FPG from baseline at week 14 and at week 26 Change in hypoglycaemia from baseline at week 14 and at week 26 (as measured with CGM) Change in clinically significant hypoglycaemia from baseline at week 14 and at week 26 (as measured with CGM) Safety analysis during the study Quality of life (KDQOL‐36) at EoT
Notes	Additional information Funding source: AstraZeneca Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: information on allocation concealment was not reported; similar baseline characteristics were reported Note: this study has been split in two (Selvarajah 2024a for semaglutide; Selvarajah 2024b for cotadutide) to extract data from both GLP‐1 classes versus placebo

SEMPA 2023.

Study characteristics
Methods	Study design Parallel RCT Study time frame: June 2019 to May 2021 Duration of follow‐up: 32 weeks Country: Denmark Setting: single centre
Participants	Study characteristics Inclusion criteria: T2DM with HbA1c ≥ 6.5%; age ≥ 50 years and established CV disease, heart failure or CKD (eGFR < 60 mL/min/1.73 m2); or age ≥ 60 years with high CV risk Exclusion criteria: eGFR < 45 mL/min/1.73 m2; treatment with GLP1RA, SGLT2 or DPP4 inhibitors within 30 days of screening; active cancer diagnosis other than basal cell carcinoma; liver disease; acute or chronic pancreatitis Baseline characteristics Number (randomised/analysed) CKD < 60 mL/min: treatment group 1 (1/not reported); treatment group 2 (1/not reported); treatment group 3 (5/not reported); control group (3/not reported) ‐ no other data on patients with all other stages of CKD was available Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group 1 Semaglutide: 1.0 mg/week Intervention group 2 Empagliflozin: 10 mg/d Intervention group 3 Semaglutide: 1.0 mg/week Empagliflozin: 10 mg/d Control group Placebo
Outcomes	Primary outcomes Change in kidney oxygenation at week 32 Change in arterial stiffness at week 32 Secondary outcomes Change in UACR from baseline at week 32 Change in eGFR from baseline at week 32
Notes	Additional information Funding source: Novo Nordisk Foundation, the Central Denmark Region Research Fund and the Danish Medical Associations Research Foundation. Conflicts of interest: not reported Information on allocation concealment and balance/imbalance in baseline characteristics: information on allocation concealment was not reported; similar baseline characteristics was reported Note: only some baseline data for people with CKD <60 ml/min were available and reported in the number of participants in this review. Note: authors contacted but they did not reply

Sivalingam 2024.

Study characteristics
Methods	Study design Parallel RCT Study time frame: October 2020 to February 2023 Duration of follow‐up: 26 weeks Country: Denmark Setting: single‐centre
Participants	Study characteristics Inclusion criteria: ≥ 18 years; T2DM; UACR > 100 mg/g; eGFR ≥ 30 mL/min/1.73 m2; treatment with maximal tolerated dose of an ACEi or ARB for a minimum of 4 weeks prior to randomisation Exclusion criteria: T1DM; cancer; or any other clinically significant disorder which could interfere with the results Baseline characteristics Number (randomised/analysed): intervention group (30/28); control group (30/26) Mean age ± SD (years): intervention group (70.5 ± 6.8); control group (69.4 ± 9.1) Sex (M/F): intervention group (6/24); control group (7/23)
Interventions	Intervention group Semaglutide: 1.0 mg/week Control group Placebo Co‐intervention Empagliflozin: 25 mg/d
Outcomes	Primary outcomes Change in UACR from baseline at EoT Secondary outcomes Change in measured GFR from baseline at EoT Change in HbA1c from baseline at EoT Change in body weight from baseline at EoT
Notes	Additional information Funding source: Novo Nordisk Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: not reported

SOUL 2023.

Study characteristics
Methods	Study design Parallel RCT Study time frame: June 2019 and March 2021 Duration of follow‐up: up to 2 years Country: multinational (33 countries) Setting: multicentre (444 centres)
Participants	Study characteristics Inclusion criteria: ≥ 50 years with T2DM and evidence of ASCVD (coronary artery disease, cerebrovascular disease, symptomatic peripheral arterial disease) and/or CKD (eGFR < 60 mL/min/1.73 m2) Exclusion criteria: MI, stroke, or hospitalisation for unstable angina, or TIA within 60 days prior to screening; planned coronary, carotid or peripheral artery revascularisation; NYHA class IV HF; kidney failure requiring chronic or intermittent HD or PD; history of major surgical procedures involving the stomach that may affect drug absorption Baseline characteristics Number (randomised/analysed): overall (4082/not reported) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group Semaglutide: 14 mg once/day Control group Placebo Co‐interventions Standard of care
Outcomes	Primary outcomes Time from randomisation to first occurrence of a major adverse CV event (MACE; a composite of CV death, nonfatal MI or nonfatal stroke) Secondary outcomes CV death Kidney‐related death Persistent ≥ 50% reduction in eGFR Persistent eGFR < 15 mL/min/1.73 m2 Acute or chronic limb ischaemia hospitalisation
Notes	Additional information Funding source: Novo Nordisk Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Conflicts of interest: reported Note: only baseline data available Note: RoB assessment was not provided since no outcomes were reported

SUSTAIN‐6 2016.

Study characteristics
Methods	Study design Parallel RCT Study time frame: February 2013 through December 2013 Duration of follow‐up: 104 weeks Country: multinational (20 countries) Setting: multicentre (230 sites)
Participants	Study characteristics Inclusion criteria: T2DM and HbA1c ≥ 7% were eligible if they had not been treated with a glucose‐lowering agent or had been treated with no more than 2 oral antihyperglycaemic agents, with or without basal or premixed insulin; aged ≥ 50 years with established CVD (previous CV, cerebrovascular, or peripheral vascular disease), chronic heart failure (NYHA class II or III), or CKD of stage 3 or higher or aged ≥ 60 years with at least one cardiovascular risk factor Exclusion criteria: treatment with a DPP‐4 inhibitor within 30 days before screening or with a GLP‐1 receptor agonist or insulin other than basal or premixed within 90 days before screening; history of an acute coronary or cerebrovascular event within 90 days before randomisation; planned revascularisation of a coronary, carotid, or peripheral artery; long‐term dialysis Baseline characteristics Number (randomised/analysed): intervention group (579/not reported); control group (576/not reported) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group Semaglutide: 0.5 mg or 1.0 mg once/week for 104 weeks Control group Placebo Co‐intervention Not reported
Outcomes	Primary outcomes First occurrence of death from CV causes, nonfatal MI (including silent), or nonfatal stroke up to 104 weeks Secondary outcomes First occurrence of an expanded composite CV outcome (death from CV causes, nonfatal MI, nonfatal stroke, revascularisation (coronary or peripheral), and hospitalisation for unstable angina or heart failure), an additional composite outcome (death from all causes, nonfatal MI, or nonfatal stroke), the individual components of the composite outcomes, retinopathy complications, and new or worsening nephropathy up to 104 weeks Serious and non‐serious adverse events and hypoglycaemic episodes, which were defined as severe (according to American Diabetes Association criteria 6) or as confirmed on analysis of plasma glucose (with symptomatic hypoglycaemia defined as < 56 mg/dL (3.1 mmol/L) up to 104 weeks Neoplasm and pancreatitis events up to 104 weeks
Notes	Additional information Funding source: Novo Nordisk Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: central interactive voice/web system; similar baseline characteristics were reported Note: further details regarding the subpopulation of patients with CKD were required from the authors. Participants included had CKD stages 3‐5 (no other data were available) Note: This study reported outcome data both as RR and HR and some outcomes could report duplicate RoB judgements

SUSTAIN 8 2019.

Study characteristics
Methods	Study design Parallel RCT Study time frame: March 15 2017 to Nov 16 2018 Duration of follow‐up: 52 weeks Country: multinational (11 countries) Setting: multicentre (111 sites)
Participants	Study characteristics Inclusion criteria: ≥ 18 years with T2DM; HbA1c levels 7.0% to 10.5% (53 to 91 mmol/mol); on a stable daily dose of metformin (≥ 1500 mg or maximum tolerated dose) for at least 90 days before screening; some had an eGFR ≥ 60 mL/min/1.73 m² Exclusion criteria: known or suspected hypersensitivity to trial product(s) or related products; previous participation in this trial (participation is defined as signed informed consent); female who is pregnant, breastfeeding or intends to become pregnant or is of child‐bearing potential and not using an adequate contraceptive method (adequate contraceptive measure as required by local regulation or practice); participation in any clinical trial of an approved or non‐approved investigational medicinal product within 90 days prior to the day of screening; any disorder which in the investigator's opinion might jeopardise subject's safety or compliance with the protocol; subject with alanine aminotransferase above 2.5 x UNL; family or personal history of multiple endocrine neoplasia type 2 or medullary thyroid carcinoma (family is defined as a first degree relative); history or presence of pancreatitis (acute or chronic); history of diabetic ketoacidosis; any of the following: MI, stroke, hospitalisation for unstable angina or TIA within the past 180 days prior to the day of screening; subjects presently classified as being in NYHA Class IV; planned coronary, carotid or peripheral artery revascularisation known on the day of screening; renal impairment measured as eGFR below 60 ml/min/1.73 m2 as defined by Kidney Disease Improving global outcomes (KDIGO 2012) classification using isotope dilution mass spectrometry for SCr measured at screening; treatment with any medication for the indication of diabetes or obesity other than stated in the inclusion criteria within the past 90 days prior to the day of screening (however, short term insulin treatment for a maximum of 14 days prior to the day of screening is allowed); proliferative retinopathy or maculopathy requiring acute treatment (verified by fundus photography or dilated fundoscopy performed within the past 90 days prior to randomisation); presence or history of malignant neoplasms within the past 5 years prior to the day of screening (basal and squamous cell skin cancer and any carcinoma in‐situ is allowed); medical history of diabetes‐related lower limb amputations or signs of critical lower limb ischaemias, (e.g. skin ulcer, osteomyelitis, or gangrene) within the last 26 weeks prior to screening Baseline characteristics Number(randomised/analysed): intervention group 1 (107/not reported); intervention group 2 (119/not reported) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group 1 Semaglutide: 1 mg once/ day, reached after an 8‐week fixed dose‐escalation period Intervention group 2 Canagliflozin; 300 once/day Co‐interventions Rescue medication was offered to patients with confirmed FPG levels > 13.3 mmol/L (240 mg/dL) from week 8 to the end of week 13 and > 11.1 mmol/L (200 mg/dL) from week 14 to end of study treatment, or to patients with confirmed HbA1c > 8.5% (69.4 mmol/mol) from week 26 to end of study treatment. Choice of rescue medication was at the investigator’s discretion
Outcomes	Primary outcomes Change in HbA1c: week 0, week 52 Secondary outcomes Change in body weight (kg): week 0, week 52 Change in FPG: week 0, week 52 Change in SMPG (mean 7‐point profile): week 0, week 52 Change in SMPG (mean postprandial increment over all meals): week 0, week 52 Change in BMI: week 0, week 52 HbA1c < 7.0% (53 mmol/mol), American Diabetes Association target (yes/no): week 52 HbA1c reduction ≥ 1% (yes/no): week 0, week 52 Weight loss ≥ 3% (yes/no): week 0, week 52 Weight loss ≥ 5% (yes/no): week 0, week 52 Weight loss ≥ 10% (yes/no): week 0, week 52 HbA1c < 7.0% (53 mmol/mol) without severe or blood glucose‐confirmed symptomatic hypoglycaemia episodes and no weight gain (yes/no): week 0, week 52 HbA1c reduction ≥ 1% and weight loss ≥ 3% (yes/no): week 0, week 52 HbA1c reduction ≥ 1% and weight loss ≥ 5% (yes/no): week 0, week 52 HbA1c reduction ≥ 1% and weight loss ≥ 10% (yes/no): week 0, week 52 Total number of treatment‐emergent adverse events: weeks 0‐57 Change in creatinine: week 0, week 52 Change in eGFR: week 0, week 52 Change in albumin: week 0, week 52 Total number of treatment‐emergent severe or blood glucose‐confirmed symptomatic hypoglycaemic episodes: week 0, week 57 Participants with treatment‐emergent severe or blood glucose‐confirmed symptomatic hypoglycaemic episodes: week 0, week 57
Notes	Additional information Funding source: Novo Nordisk Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: central interactive voice/web system; similar baseline characteristics were reported Note: authors contacted (10/7/21) but they did not reply Note: RoB assessment was not provided since no outcomes were reported

Tuttolomondo 2021.

Study characteristics
Methods	Study design Parallel RCT Study time frame: April 2017 to April 2019 Study duration: 9 months Country: Italy Setting: single centre
Participants	Study characteristics Inclusion criteria: men and women ≥ 50 years with established or newly detected T2DM whose HbA1c level ≤ 9.5% or less on stable doses of up to two oral glucose‐lowering drugs with or without basal insulin therapy Exclusion criteria: T1DM; previous GLP‐1 receptor agonist treatment; treatment with more than half of the sulphonylurea maximum dose at screening; chronic systemic glucocorticoid use; gastric emptying abnormality; eGFR value < 15 mL/min/1.73 m²; cancer in the last 5 years; severe hypoglycaemia in the last year; a life expectancy < 1 year; coronary or cerebrovascular event within the last 2 months;plans for revascularisation Baseline characteristics Number (randomised/analysed): intervention group (16/not reported); control group (14/not reported) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group Dulaglutide (SC): 1.5 mg Control group No antidiabetic treatment Co‐interventions Traditional antidiabetic treatment (metformin or metformin plus sulphonylurea or basal insulin plus metformin)
Outcomes	Primary outcomes Arterial stiffness (pulse wave velocity and augmentation index), evaluated at baseline and at 3 and 9 months Endothelial function (reactive hyperaemia index), evaluated at baseline and at 3 and 9 months Secondary outcomes FPG, evaluated at baseline and at 3 and 9 months HbA1c, evaluated at baseline and at 3 and 9 months Adverse events, including hypoglycaemia, were assessed during the study period
Notes	Additional information Funding source: none Conflicts of interest: none Information on allocation concealment and balance/imbalance in baseline characteristics: central interactive voice/web system information on allocation concealment was reported; similar baseline characteristics were reported Note: authors contacted (10/7/21) but they did not reply Note: we have included only participants with micro‐ or macroalbuminuria. No other data on CKD patients in general were available Note: RoB assessment was not provided since no outcomes were reported.

van der Aart‐van 2023.

Study characteristics
Methods	Study design Cross‐over RCT Study time frame: January 2019 to April 2021 Duration of follow‐up: 6 weeks (first phase) Country: the Netherlands Setting: single centre
Participants	Study characteristics Inclusion criteria: ≥ 18 years with T2DM; eGFR > 30 mL/min/1.73 m2 and a UACR > 3.5 mg/mmol and ≤ 100 mg/mmol (> 30.9 and ≤ 884 mg/g) Exclusion criteria: CV disease (MI, angina pectoris, percutaneous transluminal coronary angioplasty, coronary artery bypass grafting, stroke, heart failure) for less than 6 months before inclusion; uncontrolled BP (SBP/DBP > 160/100 mm Hg); rapid progression of kidney function decline (eGFR change > 30% in the 6 months before inclusion) and use of an SGLT2 inhibitor, GLP‐1 or DPP4i for less than 6 weeks prior to screening Baseline characteristics Number (randomised/analysed): overall (20/not reported) Mean age ± SD (years): overall (70.5 ± 5.1) Sex (M/F): overall (16/4)
Interventions	Intervention group 1 Dapagliflozin: 10 mg/d Intervention group 2 Exenatide: 2 mg/week Intervention group 3 Dapagliflozin: 10 mg/d Exenatide: 2 mg/week Co‐interventions Not reported
Outcomes	Primary outcomes Percentage change in UACR at baseline and EoT Secondary outcomes Body weight was measured at each study visit HbA1c, lipids, SCr, potassium, albumin, Hb and HCT during the study at baseline and EoT Adverse events at EoT
Notes	Additional information Funding source: AstraZeneca Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Conflicts of interest: reported Note: author contacted but no further data were available Note: RoB assessment was not provided since no outcomes were reported

von Scholten 2017.

Study characteristics
Methods	Study design Cross‐over RCT Study time frame: April to October 2015 and completed in April 2016 Duration of follow‐up: 12 weeks first phase Country: Denmark Setting: single centre
Participants	Study characteristics Inclusion criteria: T2DM (WHO criteria), HbA1c ≥ 48 mmol/mol (6.5%), persistent albuminuria (≥ 30 mg/g in at least 2 out of 3 consecutive morning spot urine samples); receiving stable RAS‐blocking treatment Exclusion criteria: diagnosis of clinical heart failure; eGFR ≤ 30 mL/min/1.73 m2 Baseline characteristics Number (randomised/analysed): overall (32/not reported) Mean age ± SD (years): overall (66 ± 7) Sex (M/F): overall (27/5)
Interventions	Intervention group Liraglutide: 0.6 mg/d for 7 days, escalated to 1.2 mg for 7 days and lastly to 1.8 mg/d for the remaining 10 weeks Control group Placebo: once/d Co‐interventions Eligible patients were enrolled in a 2‐week, single‐blind run‐in period during which all patients received placebo. Participants were then randomised A diet/exercise regimen and background AHA therapy (if applicable) were continued Following completion of the initial 26‐week treatment period, patients entered a 26‐week placebo‐controlled extended treatment period (phase B, where they continued with their assigned randomised treatment from phase A); the aim of phase B was to gather additional data on the safety and longer‐term efficacy of ertugliflozin
Outcomes	Primary outcomes UAER at baseline and EoT Secondary outcomes Measured GFR at baseline and EoT Safety at baseline and EoT
Notes	Additional information Funding source: unrestricted grant from Novo Nordisk A/S, with no influence over the trials conduct or analysis Conflicts of interest: reported Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Note: RoB assessment was not provided since no outcomes were reported

Wajdlich 2024.

Study characteristics
Methods	Study design Cross‐over RCT Study time frame: not reported Duration of follow‐up: 1‐day first period Country: Poland Setting: single‐centre
Participants	Study characteristics Inclusion criteria: 18 to 80 years; T2DM; eGFR < 30 mL/min/1.73 m2 or > 60 mL/min/1.73 m2; stable kidney function Exclusion criteria: T1D; history of AKI; proteinuria > 2 g/d; HbA1c > 10%; history of pancreatitis; liver failure; pregnancy; heart failure NYHA III‐IV; SBP > 180 mmHg and/or DBP > 120 mmHg Baseline characteristics Number (randomised/analysed): overall (34/34) Mean age ± SD (years): overall (66 ± 2.7) Sex (M/F): overall (18/16)
Interventions	Intervention group Semaglutide: 1.2 mg Control group Placebo
Outcomes	Primary outcomes BP at 24 h Secondary outcomes Hormonal levels (atrial natriuretic protein, renin, aldosterone) at 24 h
Notes	Additional information Funding source: Medical University of Lodz Conflicts of interest: not reported Trial registration number: not reported Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Note: Authors contacted but they did not reply

Wang 2020a.

Study characteristics
Methods	Study design Parallel RCT Study time frame: March 2016 to April 2019 Duration of follow‐up: 24 weeks Country: China Setting: multicentre (4 general hospitals)
Participants	Study characteristics Inclusion criteria: ≥ 18 years; eGFR ≥ 30 mL/min/1.73 m2; macroalbuminuria (UAER > 0.3 g/24 h); T2DM Exclusion criteria: not reported Baseline characteristics Number (randomised/analysed): intervention group 1 (46/43); intervention group 2 (46/38) Mean age ± SD (years): intervention group 1 (55.9 ± 8.9); intervention group 2 (56.2 ± 8.0) Sex (M/F): intervention group 1 (35/11); intervention group 2 (29/17)
Interventions	Intervention group 1 Exenatide: twice/d for 24 weeks, initiated at 5 μg twice/d, titrated up to 10 μg twice/d. after 4 weeks, and then maintained at 10 μg twice/d Intervention group 2 Insulin lispro for 24 weeks Co‐interventions Insulin glargine
Outcomes	Primary outcomes UAER percentage change from baseline after 24 weeks Secondary outcomes Adverse events, including hypoglycaemia from baseline after 24 weeks Change in eGFR during the follow‐up from baseline after 24 weeks
Notes	Additional information Funding source: AstraZeneca Conflicts of interest: none Information on allocation concealment and balance/imbalance in baseline characteristics: allocation concealment mechanism prevented the selection bias whereby the allocation sequence was maintained by a researcher who did not participate in the clinical work; similar baseline characteristics was reported

Zhang 2012a.

Study characteristics
Methods	Study design Parallel RCT Study time frame: not reported Duration of follow‐up: 16 weeks Country: China Setting: single centre
Participants	Study characteristics Inclusion criteria: T2DM and microalbuminuria Exclusion criteria: T1DM Baseline characteristics Number (randomised/analysed): intervention group 1 (13/not reported); intervention group 2 (18/not reported) Mean age ± SD (years): not reported Sex (M/F): not reported
Interventions	Intervention group 1 Exenatide for 16 weeks Intervention group 2 Glimepiride for 16 weeks Co‐interventions Not reported
Outcomes	Primary outcomes BMI up to week 16 Secondary outcomes FPG up to week 16 2‐hour postprandial plasma glucose up to week 16 HbA1c up to week 16
Notes	Additional information Funding source: not reported Conflicts of interest: not reported Trial registration number: not reported Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Note: RoB assessment was not provided since no outcomes were reported

Zhou 2019a.

Study characteristics
Methods	Study design Parallel RCT Study time frame: January 2015 to August 2018 Duration of follow‐up: 24 weeks Country: China Setting: single centre (the first affiliated hospital of Soochow University)
Participants	Study characteristics Inclusion criteria: adults with diabetic kidney disease in early stages of CKD Exclusion criteria: acute and chronic infections; take drugs regularly; major surgery; women who are pregnant or breastfeeding; malignant tumours; severe trauma Baseline characteristics Number (randomised/analysed): intervention group 1 (40/40); intervention group 2 (40/40) Mean age ± SD (years): intervention group 1 (55.68 ± 12.21); intervention group 2 (56.42 ± 11.85) Sex (M/F): intervention group 1 (22/18); intervention group 2 (24/16)
Interventions	Intervention group 1 Exenatide (SC): 5 µg twice/d Intervention group 2 Olmesartan Metformin: 5 g Co‐interventions Diet and exercise
Outcomes	Primary outcomes Blood glucose, assessed before and after treatment Secondary outcomes Kidney function, assessed before and after treatment Serum advanced glycosylation end products assessed before and after treatment FBG HbA1c SCr UACR eGFR Urine microalbuminuria
Notes	Additional information Funding source: the National Natural Science Foundation of China (N81700589) Conflicts of interest: not reported Trial registration number: not reported Information on allocation concealment and balance/imbalance in baseline characteristics: not reported Note: Non‐English, Chinese paper. Authors contacted, but they did not reply. Data was reported on GFR: since unit and variance were not clearly reported, data were not extracted in this review

ACEi: angiotensin‐converting‐enzyme inhibitor; AKI: acute kidney injury; ARB: angiotensin receptor blocker; BMI: body mass index; BP: blood pressure; CKD: chronic kidney disease; CrCl: creatinine clearance; CV: cardiovascular; DBP: diastolic BP; DKD: diabetic kidney disease; DM: diabetes mellitus; DPP‐4: Dipeptidyl peptidase‐4; ECG: electrocardiograph; eGRF: estimated GFR; EoT: end of treatment; EQ‐5D‐5L: EuroQol 5 Dimension 5 Level; FBG: fasting blood glucose; FPG: fasting plasma glucose; GFR: glomerular filtration rate; GI: gastrointestinal; GLP‐1: glucagon‐like peptide 1; Hb: haemoglobin; HbA1c: glycated Hb; HCT: haematocrit; HD: haemodialysis; HDL: high‐density lipoprotein; HR: hazard ratio; HRQoL: Health‐related quality of life; ITT: intention to treat; KDIGO: Kidney Disease Improving Global Outcomes; KRT: kidney replacement therapy; LDL: low‐density lipoprotein; MACE: Major Adverse Cardiovascular Events; MI: myocardial infarction; NYHA: New York Heart Association; PD: peritoneal dialysis; RAAS: renin–angiotensin–aldosterone system; RAS: renin‐angiotensin system; RCT: randomised controlled trial; RoB: risk of bias; RR: relative risk; SBP: systolic BP; SC: subcutaneous; SCr: serum creatinine; SGLT2: sodium‐glucose cotransporter‐2; SMPG: self‐measured plasma glucose; TIA: transient ischaemic attack; T1DM: type 1 DM; T2DM: type 2 DM: UACR: urinary albumin:creatinine ratio; UAER: urinary albumin excretion rate; ULN: upper limit of normal; VAS: visual analogue scale

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Ahren 2013	Wrong population: in GetGoal‐M (NCT00712673) the exclusion criteria reported "renal impairment defined with creatinine >1.4 mg/dL in women and creatinine >1.5 mg/dL in men"
Bolli 2014	Wrong population: GetGoal‐F1 reported that people with kidney impairment were excluded
Rosenstock 2014	Wrong population: patients with renal impairment defined with SCr >1.4 mg/dL in women and SCr >1.5 mg/dL in men were excluded

SCr: serum creatinine

Characteristics of ongoing studies [ordered by study ID]

ChiCTR2200059717.

Study name	Effect of ultra‐short‐term insulin therapy combined with once‐weekly semaglutide in clinical remission in middle‐aged and elderly patients with type 2 diabetes mellitus: an open‐label, randomised controlled trial
Methods	Study design Parallel RCT Duration of follow‐up: not reported Country: China Setting: single centre Recruitment status: not yet recruiting
Participants	Study characteristics Inclusion criteria: 45 to 75 years ; HbA1c 7.5‐10.0%; course of diabetes < 5 years; not taking hypoglycaemic drugs or only taking metformin; BMI 20 to 32.5 kg/m2 Exclusion criteria: acute complications of diabetes, severe chronic complications (proliferative retinopathy, diabetic nephropathy stage V); various types of anaemia that affect HbA1c levels; fasting C‐peptide < 1.2 ng/mL (1 ng/ml = 0.33 nmol/L); severe heart, liver and kidney diseases; uncontrolled severe infectious diseases; use of drugs that affect blood sugar, such as glucocorticoids, growth hormones, antipsychotics, etc.; cognitive impairment and cannot cooperate with research and follow‐up; combined with malignant tumours and other diseases, life expectancy < 5 years Planned enrolment: 186
Interventions	Intervention group 1 Semaglutide: 0.25 mg to 1.0 mg Intervention group 2 Insulin glargine: adjusted according to FBG, the maximum dose is less than or equal to 40 U/d
Outcomes	Primary outcomes FBG at EoT 2 hours postprandial blood glucose at EoT HbA1c at EoT Proportion of discontinuing hypoglycaemic drugs at EoT Proportion of adding other hypoglycaemic drugs at EoT Secondary outcomes Glucose below target range time at EoT Time when glucose is above target range at EoT Time glucose is in target range at EoT Islet ß‐cell function at EoT Insulin sensitivity at EoT Incidence of hypoglycaemia at EoT BMI at EoT Atherosclerotic heart disease at EoT Diabetic nephropathy at EoT
Starting date	Planned starting date: 5 May 2022 Date of registration: 9 May 2022
Contact information	Zhao Jing +86 15205197656
Notes	Funding source: sponsored

ChiCTR2400080751.

Study name	A randomized, controlled trial to evaluate the improvement of oxidative stress levels and renal function in patients with diabetic kidney disease treated with semaglutide for 28 weeks
Methods	Study design Parallel RCT Duration of follow‐up: 28 weeks Country: China Setting: single centre Recruitment status: completed
Participants	Study characteristics Inclusion criteria: T2DM with diagnosed DKD (UACR > 30 mg/g or eGFR 30 to 60 mL/min/1.73 m2); 18 to 75 years; HbA1c 7.0% to10.0%; stable treatment with ≤ 3 oral hypoglycaemic agents for > 3 months Exclusion criteria: calcitonin > 50 ng/L; personal or familial history of thyroid medullary carcinoma or MEN type 2; abnormal liver function; known significant non‐DKD Planned enrolment: 30
Interventions	Intervention group 1 Semaglutide: 0.5 mg/week Intervention group 2 Insulin detemir: 0.2 up to 2 U/d
Outcomes	Primary outcomes SCr at EoT eGFR at EoT uACR at EoT Secondary outcomes HbA1c at EoT Weight at EoT BMI at EoT FBG at EoT
Starting date	Date of first enrolment: 27 March 2023 Date of registration: 6 February 2024
Contact information	Tian Shasha: +86 198 3451 6507
Notes	Funding source: Chu Hsien‐I Memorial Hospital of Tianjin Medical University

NCT05536804.

Study name	A study of tirzepatide (LY3298176) in participants with overweight or obesity and chronic kidney disease with or without type 2 diabetes (TREASURE‐CKD)
Methods	Study design Parallel RCT Study time frame: not reported Duration of follow‐up: one year Country: multinational (Austria, Canada, Denmark, Mexico, the Netherlands, USA) Setting: multicentre (28 sites) Recruitment status: currently recruiting
Participants	Study characteristics Inclusion criteria: BMI ≥ 27 kg/m² at screening; diagnosed CKD, eGFR ≥ 30 to ≤ 60 mL/min/1.73 m² or eGFR ≥ 30 to ≤ 75 ml/min/1.73 m² if UACR >30 mg/g; receiving an ACEi or ARB that is considered the maximal appropriate dose by the investigator for treatment of CKD or hypertension (unless the participant has low BP or hypotension), participants without diabetes have HbA1c < 6.5% at screening, participants with T2DM have been diagnosed at least 180 days prior to screening, have HbA1c ≥ 7.0% to ≤ 10.5% at screening Exclusion criteria: self‐reported change in body weight > 5 kg within 90 days prior to screening; prior or planned surgical treatment for obesity; have or plan to have endoscopic and/or device‐based therapy for obesity or have had device removal within the last 180 days; eGFR < 30 mL/min/1.73 m²; history of unstable or rapidly progressing kidney disease according to investigator judgment; known clinically significant gastric emptying abnormality (e.g. severe diabetic gastroparesis or gastric outlet obstruction, or chronically take drugs that directly affect GI motility); have had a history of chronic or acute pancreatitis Planned enrolment: 140
Interventions	Intervention group Tirzepatide Control group Placebo
Outcomes	Primary outcomes Change from baseline in kidney oxygenation in participants with or without T2DM: baseline, week 52 Secondary outcomes Change from baseline in kidney oxygenation in participants with T2DM (bold MRI): baseline, week 52 Change from baseline in kidney oxygenation in participants without t2d (bold MRI): baseline, week 52 Per cent change from baseline in body weight: baseline, week 52 Per cent change from baseline in renal sinus fat content (MRI): baseline, week 52 Per cent change from baseline in renal fat content (MRI proton density fat fraction): baseline, week 52 Per cent change from baseline in renal blood flow (phase‐contrast MRI): baseline, week 52 Change from baseline in apparent diffusion coefficient MRI baseline, week 52 Change from baseline in GFR iohexol clearance mL/min/m²: baseline, week 52 Change from baseline in 24‐hour UAE mg/24 h: baseline, week 52 Per cent change from baseline in UACR: baseline, week 52
Starting date	Starting date: 8 February 2023 Date of registration: 13 September 2022
Contact information	Phone Number:1‐317‐615‐4559 Email: clinical_inquiry_hub@lilly.com
Notes	Funding source: Eli Lilly and Company

NCT06182891.

Study name	Renoprotective effects of dulaglutide in patients with type 2 diabetic nephropathy
Methods	Study design Parallel RCT Study time frame: October 2021 to October 2023 Duration of follow‐up: 56 weeks Country: China Setting: single‐centre Recruitment status: completed
Participants	Study characteristics Inclusion criteria: ≥ 18 years; T2DM; diabetic nephropathy with UACR ≥ 30 mg/g Exclusion criteria: T1DM; UTI Planned enrolment: 300
Interventions	Intervention group 1 Dulaglutide Intervention group 2 Other hypoglycaemic agents not including GLP‐1 receptor agonists
Outcomes	Primary outcomes Change in UACR at 3, 6, and 12 months Secondary outcomes Change in cystatin C at 3, 6, and 12 months Adverse events at 3, 6, and 12 months
Starting date	Starting date: 30 October 2021 End date: 30 October 2023
Contact information	Study Director: Renhao Wang
Notes	Funding source: The Affiliated Hospital of Xuzhou Medical University

NCT06555146.

Study name	Semaglutide's effect on renal hemodynamics and function in patients with type 2 diabetes mellitus and nephropathy (Sema‐RMA)
Methods	Study design Cross‐over RCT Duration of follow‐up: 5 weeks Country: Denmark Setting: single‐centre Recruitment status: not yet recruiting
Participants	Study characteristics Inclusion criteria: 20 to 60 years; T2DM with HbA1c 6.5% to 8.5% or < 6.5%; FPG > 7.0 mmol/L or > 11.0 mmol/L after 2 h at oral glucose tolerance test; eGFR 30 to 60 mL/min/1.73 m2; UACR ≥ 30 mg/g Exclusion criteria: kidney transplant; treatment for kidney failure with dialysis; heart failure NYHA III‐IV; dysregulated arterial hypertension; liver failure Planned enrolment: 15
Interventions	Intervention group Semaglutide: 0.5 mg Control group Placebo
Outcomes	Primary outcomes Total blood flow at EoT Secondary outcomes eGFR at EoT BP at EoT
Starting date	Planned starting date: August 2024
Contact information	ali.asmar@regionh.dk; peter.soerensen@regionh.dk +4530516279
Notes	Funding source: Bispebjerg Hospital

NL‐OMON46820.

Study name	An open‐label randomised cross‐over study to evaluate the albuminuria lowering effect of dapagliflozin, exenatide and their combination in patients with type 2 diabetes
Methods	Study design Cross‐over RCT Duration of follow‐up: 45 weeks Country: the Netherlands Setting: single centre Recruitment status: completed
Participants	Study characteristics Inclusion criteria: T2DM; HbA1c > 6.5%; eGFR > 30 mL/min/1.73 m2; UACR 3.5‐100 mg/mmol; ≥ 18 years Exclusion criteria: CV disease within 6 months for inclusion; uncontrolled hypertension (< 160/100 mm Hg); malignancy; use of SGLT‐2i, GLP‐1RA, or DPP‐4i Planned enrolment: 20
Interventions	Intervention group 1 Exenatide: 2 mg/week Intervention group 2 Dapagliflozin: 10 mg/d Intervention group 3 Exenatide: 2 mg/week Dapagliflozin: 10 mg/d
Outcomes	Primary outcomes Albuminuria at EoT Secondary outcomes BP at EoT Weight at EoT
Starting date	Starting date: 21 January 2019 Date of registration: 28 March 2018
Contact information	Martini Ziekenhuis: Van Swietenplein 1 Groningen 9728 NT NL
Notes	Funding source: Lokale subsidies; Astra Zeneca (beschikbaar stellen studiemedicatie); Antaros Medical (kosten MRI's); Astra Zeneca; Antaros Medical

OTID 2024.

Study name	Obesity Treatments to Improve Type 1 Diabetes (OTID): a randomized controlled trial of the combination of glucagon‐like peptide 1 analogues and sodium‐glucose cotransporter 2 inhibitors‐protocol for Obesity Treatments to Improve Type 1 Diabetes (the OTID trial)
Methods	Study design Parallel RCT Duration of follow‐up: 26 weeks Recruitment status: completed Country: Kuwait Setting: single centre Recruitment status: completed
Participants	Study characteristics Inclusion criteria: 21 to 65 years; BMI ≥ 25 kg/m2; T1DM for at least 1 year; CKD stages 1 to 4 Exclusion criteria: diabetes other than type 1; having been treated with SGLT‐2i or GLP‐1RA within 3 months of screening; weight change > 5% within 3 months of screening; eGFR ≤ 15 mL/min/1.73 m2 Planned enrolment: 60
Interventions	Intervention group 1 Liraglutide: up to 3.0 mg/d or semaglutide up to 1.8 mg/week Intervention group 2 Dapagliflozin: 5 to 10 mg/d Intervention group 3 Liraglutide: up to 3.0 mg/d or semaglutide up to 1.8 mg/week Dapagliflozin: 5 to 10 mg/d Intervention group 4 Liraglutide: up to 3.0 mg/d or semaglutide up to 1.8 mg/week Dapagliflozin: 10 mg/d Intensive lifestyle approach Control group Usual standard care
Outcomes	Primary outcomes Change in body weight at week 26 Secondary outcomes Change in HbA1c at week 26 Change in renal biochemistry at week 26 Change in UACR at week 26
Starting date	Starting date: 25 May 2022 Date of registration: 25 May 2022
Contact information	Kavita Narula k.narula@nhs.net
Notes	Funding source: Ministry of Health Kuwait and the Kuwait Foundation for Advancement of Science. The funder has no role in the design of this study and will not have any role during the execution, analyses, interpretation of the data or submission of the outcome

REMODEL 2022.

Study name	REMODEL 2022
Methods	Study design Parallel RCT Duration of follow‐up: 12 months Country: multinational (Canada, Denmark, France, Italy, Poland, South Africa, Spain) Setting: multicentre (31 sites) Recruitment status: completed
Participants	Study characteristics Inclusion criteria: Male or female; ≥ 18 years at the time of signing informed consent; diagnosed with T2DM ≥ 180 days prior to the day of screening; HbA1c ≤ 9.0%; for non‐biopsy population: SCr‐based eGFR ≥ 30 and ≤ 75 mL/min/1.73 m2; for biopsy sub‐population: SCr‐based eGFR ≥ 40 and ≤ 75 mL/min/1.73 m2; UACR ≥ 20 and < 5000 mg/g; treatment with maximum labelled or tolerated dose of a RAAS blocking agent including an ACEi or ARB unless such treatment is contraindicated or not tolerated; treatment dose must be stable for at least 28 days prior to screening Exclusion criteria: Use of any GLP‐1RA within 30 days prior to screening; prior solid organ transplant or awaiting solid organ transplant; MI, stroke, hospitalisation for unstable angina pectoris or TIA within 180 days prior to the day of screening; presence or history of malignant neoplasms (other than basal or squamous cell skin cancer, in situ carcinomas of the cervix, or in situ prostate cancer) within 5 years prior to the day of screening; congenital or hereditary kidney diseases including polycystic kidney disease, autoimmune kidney diseases including glomerulonephritis or congenital urinary tract malformations; uncontrolled and potentially unstable diabetic retinopathy or maculopathy verified by a fundus examination performed within the past 90 days prior to screening or in the period between screening and Visit 2; pharmacological pupil‐dilation is a requirement unless using a digital fundus photography camera specified for non‐dilated examination; treatment with systemic anti‐inflammatory or immunosuppressant drugs within 90 days prior to screening; ny contraindication for MRI according to standard checklist used in clinical routine, including claustrophobia or metallic foreign bodies, metallic implants, internal electrical devices, or permanent makeup/tattoos that cannot be declared MR compatible; combination use of an ACEi and an ARB Planned enrolment: 106
Interventions	Intervention group Semaglutide: gradually increased from 0.25 to 1.0 mg over 8 weeks for 12 months Control group Placebo
Outcomes	Primary outcomes Change in kidney oxygenation (cortex), blood oxygenation‐level dependent MRI: baseline (week 0) to EoT (week 52) Change in kidney oxygenation (medulla), BOLD MRI(R2): baseline (week 0) to EoT (week 52) Change in global kidney perfusion (MRI): baseline (week 0) to EoT (week 52) Change in kidney inflammation (cortex), T1 mapping (MRI): baseline (week 0) to EoT (week 52) Change in kidney inflammation (medulla), T1 mapping (MRI): baseline (week 0) to EoT (week 52) Secondary outcomes Change in glomerular basement membrane width (kidney biopsy): baseline (week 0) to EoT (week 52) Change in apparent diffusion coefficient (cortex) (MRI): baseline (week 0) to EoT (week 52) Change in apparent diffusion coefficient (medulla) (MRI): baseline (week 0) to EoT (week 52) Change in mean renal artery resistive index (MRI): baseline (week 0) to EoT (week 52) Change in mean arterial flow (MRI): baseline (week 0) to EoT (week 52) Change in natriuresis (urinary sodium excretion): baseline (week 0) to EoT (week 52) Change in UAE: baseline (week 0) to EoT (week 52) Change in kidney function (CrCl): baseline (week 0) to EoT (week 52)
Starting date	Starting date: 28 April 2021 Date of registration: 26 April 2021
Contact information	clinicaltrials@novonordisk.com
Notes	Funding source: Novo Nordisk A/S

RESET1 2024.

Study name	Reducing cardiometabolic risk with semaglutide in type 1 diabetes (RESET1): Study protocol of a phase 2 double‐blinded randomised placebo‐controlled trial
Methods	Study design Parallel RCT Study time frame: March 2024 to June 2026 Duration of follow‐up: 26 weeks Country: Australia Setting: single centre Recruitment status: currently recruiting
Participants	Study characteristics Inclusion criteria: 25 to 70 years; T1DM of duration at least 2 years; BMI ≥ 25 kg/m2; HbA1c ≥ 7%; at least one CV risk factor (microalbuminuria, hypertension or antihypertensive treatment, hyperlipidaemia or lipid‐lowering therapy, current smoking) Exclusion criteria: treatment with a GLP‐1 receptor agonist, metformin or SGLT2i in the last 3 months; eGFR < 45 mL/min/1.73 m2; abnormal liver function; history of pancreatitis; history of medullary thyroid cancer or MEN type 2 Planned enrolment: 76
Interventions	Intervention group Semaglutide 1.0 mg/week Control group Placebo
Outcomes	Primary outcomes Change in arterial stiffness from baseline at week 26 Secondary outcomes Change in weight from baseline at week 26 Change in HbA1c from baseline at week 26
Starting date	Starting date: 12 March 2024 Date of registration: 7 December 2023
Contact information	Ruth Frampton: RESET1@garvan.org.au
Notes	Funding source: Diabetes Australia Research Trust; Garvan Research Foundation

ACEi: angiotensin‐converting‐enzyme inhibitor; ARB: angiotensin receptor blocker; BMI: body mass index; BP: blood pressure; CKD: chronic kidney disease; CrCl: creatinine clearance; CV: cardiovascular; DKD: diabetic kidney disease; DM: diabetes mellitus; DPP4i: Dipeptidyl peptidase‐4 inhibitor; eGFR: estimated GFR; EoT: end of treatment; FBG: fasting blood glucose; GFR: glomerular filtration rate; GLP‐1RA: glucagon‐like peptide 1 receptor antagonist; Hb: haemoglobin; HbA1c: glycated Hb; MI: myocardial infarction; MRI: magnetic resonance imaging; NYHA: New York Heart Association; RAAS: renin–angiotensin–aldosterone system; RCT: randomised controlled trial; SCr: serum creatinine; SGLT2i: sodium‐glucose cotransporter‐2 inhibitor; T1DM: type 1 DM; T2DM: type 2 DM; TIA: transient ischaemic attack; UACR: urinary albumin‐creatinine ratio; UAE: urinary albumin excretion

Differences between protocol and review

Data on HR was reported in a separate table to improve readability of the results.

New outcomes have been added, specifically:

1. Fatal MI

2. Fatal stroke

3. Ischaemic stroke

4. UACR

5. Albumin excretion

6. Coronary revascularisation

Non‐fatal or fatal MI, non‐fatal or fatal stroke, non‐fatal peripheral arterial events, and serious adverse events have been moved from the primary to the secondary outcomes. Cardiovascular death has been added to the SoF table. We have now included all relevant outcomes according to those reported in the 2022 KDIGO guideline to make our findings consistent with the current guideline (KDIGO 2022).

Contributions of authors

1. Draft the protocol: PN, SG, DT, TT, GFMS

2. Study selection: PN, DT, GP

3. Extract data from studies: PN, DT, GP

4. Enter data into RevMan: PN

5. Carry out the analysis: PN, DT, GP

6. Interpret the analysis: PN, SG, DT, GP, TT, GFMS

7. Draft the final review: PN, SG, DT, GP, TT, GFMS

8. Disagreement resolution: SP

9. Update the review: PN, SG, GFMS

Sources of support

Internal sources

• None, Other

  No sources of support

External sources

• None, Other

  No sources of support

Declarations of interest

1. Patrizia Natale: has declared that they have no conflict of interest.

2. Suetonia Green: has declared that they have no conflict of interest.

3. David Tunnicliffe: has declared that they have no conflict of interest.

4. Giovanni Pellegrino: has declared that they have no conflict of interest.

5. Tadashi Toyama: Mitsubishi Tanabe Pharma Corporation (Independent Contractor ‐ Other); Mitsubishi Tanabe Pharma Corporation (Independent Contractor ‐ Consultant).

6. Giovanni FM Strippoli: has declared that they have no conflict of interest.

New