Abstract
Introduction
Type 2 diabetes is a leading cause of chronic kidney disease (CKD). Individuals with both conditions have increased risk of poor cardiorenal outcomes and mortality. The rapidly evolving landscape for CKD-protective therapies in type 2 diabetes currently includes sodium-glucose cotransporter 2 inhibitors (SGLT2i) and glucagon-like peptide-1 receptor agonists (GLP-1 RA), both of which demonstrate cardiorenal outcome benefits. As part of the FOUNTAIN platform (ClinicalTrials.gov ID: NCT05526157; EUPAS ID: EUPAS48148), this study aimed to better understand changes in patient characteristics and treatment patterns corresponding with updates to clinical guideline recommendations and drug labeling and the emergence of new CKD-protective therapies such as finerenone in the US in 2021–2022.
Methods
An observational real-world data study assessed patient characteristics and drug utilization in separate SGLT2i and GLP-1 RA new-user cohorts of adults with CKD and type 2 diabetes in an earlier (1 January 2012–30 June 2021) and a later (9 July 2021–30 September 2023) period using Optum’s de-identified Clinformatics® Data Mart Database (Optum® CDM).
Results
Compared with the earlier period new users, later period new users in both cohorts were older, had more severe CKD, used less intensive type 2 diabetes medication, and had better metabolic control; SGLT2i new users more frequently had no type 2 diabetes therapy before the index date and greater congestive heart failure prevalence; and GLP-1 RA new users had increased SGLT2i use and decreased insulin use.
Conclusions
These findings inform and contextualize future studies assessing cardiorenal outcomes for these and additional treatments, including finerenone, for individuals with CKD and type 2 diabetes.
Graphical Abstract
Supplementary Information
The online version contains supplementary material available at 10.1007/s13300-025-01825-5.
Keywords: Renal insufficiency, chronic; Chronic kidney disease; Diabetes mellitus, type 2; Drug utilization; Glucagon-like peptide-1 receptor agonists; Sodium-glucose cotransporter 2 inhibitors; FOUNTAIN
Key Summary Points
| Why carry out this study? |
| To better understand how clinical characteristics and treatment patterns for adults with chronic kidney disease (CKD) and type 2 diabetes changed correspondingly with landscape changes in the US. |
| What did the study ask? |
| How did sodium-glucose cotransporter 2 inhibitor (SGLT2i) or glucagon-like peptide-1 receptor agonist (GLP-1 RA) treatment patterns and new user baseline characteristics change between an earlier (1 January 2012–30 June 2021) and a later (9 July 2021–30 September 2023) period. |
| What were the study outcomes? |
| Differences in treatment patterns and patient characteristics, both common and unique, between medication cohorts were observed in the later versus earlier period. |
| What has been learned from the study? |
| Better understanding of these changes can inform and contextualize future studies assessing cardiorenal outcomes for individuals with CKD and type 2 diabetes. |
Digital Features
This article is published with digital features, including a Graphical Abstract to facilitate understanding of the article. To view digital features for this article go to https://doi.org/10.6084/m9.figshare.30580040.
Introduction
Chronic kidney disease (CKD) poses a considerable and growing burden on global health and is a leading cause of mortality worldwide [1]. The treatment of patients with CKD is multidisciplinary, primarily aiming to slow kidney disease progression to avoid dialysis or transplantation and secondarily to reduce the significant cardiovascular disease burden accompanying CKD. Type 2 diabetes is a leading cause of CKD, and individuals with both conditions have an increased risk of kidney failure, cardiovascular disease outcomes, and mortality [2–4]. The landscape for CKD-protective therapies used in type 2 diabetes is evolving rapidly; in addition to established first-line treatments such as renin–angiotensin–aldosterone system inhibitors (RAASi; e.g., angiotensin-converting enzyme inhibitor(s) [ACEi], or angiotensin receptor blockers [ARB]), currently available treatments include sodium-glucose cotransporter 2 inhibitors (SGLT2i), glucagon-like peptide-1 receptor agonists (GLP-1 RA), and nonsteroidal mineralocorticoid receptor antagonists (ns-MRA), such as finerenone [5–8].
After the initial introduction of SGLT2i in 2013 [9] and following demonstration of kidney and cardiovascular outcome benefits for patients with type 2 diabetes, in 2021 the US Food and Drug Administration (FDA) provided the first indication that SGLT2i reduced the risk of kidney disease progression in patients with CKD [10, 11]. More recently, the first GLP-1 RA was approved by the FDA to reduce risk of worsening kidney disease among adults with type 2 diabetes and CKD [6, 12, 13]. In 2022, updated Kidney Disease: Improving Global Outcomes (KDIGO) clinical practice guidelines recommended SGLT2i as first-line therapy for patients with estimated glomerular filtration rate (eGFR) ≥ 20 mL/min/1.73 m2 (previously ≥ 30 mL/min/1.73 m2) and GLP-1 RA as an additional therapy option for patients with insufficient glycemic control as part of a comprehensive risk-reduction approach for people with type 2 diabetes and CKD [5, 6, 14]. Among emerging CKD-protective treatments, finerenone was approved by the FDA in 2021 and was also recommended in the 2022 KDIGO guidelines as first-line therapy for patients with type 2 diabetes who are at high risk of kidney disease progression and cardiovascular events [6, 15].
With updated clinical guideline recommendations and changes to drug labeling in 2020–2022, as well as the 2021 approval of finerenone—the only new drug indicated for CKD associated with type 2 diabetes approved in this time period—it is of interest to understand how patterns of SGLT2i and GLP-1 RA use and characteristics of patients prescribed these medications in the US may have changed. Better understanding of these potential changes provides important context for assessing the success of guideline implementation and can inform predictions of the evolving clinical care as well as laying the foundation for future real-world effectiveness studies. Performed as part of the FOUNTAIN (FinerenOne mUlti-database NeTwork for evidence generAtIoN) platform [16], the aims of this study were to describe treatment changes over time in patients with type 2 diabetes and CKD initiating treatment with either an SGLT2i or a GLP-1 RA and to describe changes in the baseline characteristics in these cohorts with the evolution of the treatment landscape, specifically before and after FDA approval of finerenone in 2021.
Methods
Study Design and Setting
An observational cohort study was conducted using real-world data from Optum’s de-identified Clinformatics® Data Mart Database (Optum® CDM). Optum® CDM is derived from a database of administrative health claims for members of large commercial and Medicare Advantage health plans (UnitedHealth Group Inc., Eden Prairie, MN, US; see Electronic Supplementary Material [ESM] Table S1). The study assessed drug utilization and temporal changes in treatment options (e.g., SGLT2i, GLP-1 RA) in adults with CKD and type 2 diabetes regarding two time periods: an earlier period (1 January 2012–30 June 2021) and a later period after the US approval of finerenone (9 July 2021–30 September 2023), with the latter period representing updated treatment guidelines, changes to CKD drug labeling (e.g., SGLT2i), and the availability of new CKD-protective medications (ESM Fig. S1). This study specifically assessed the later period; methods and results for SGLT2i and GLP-1 RA cohorts in the earlier period have been previously published, although the findings from the earlier period are presented herein to contextualize the results and trends in the later period [17, 18]. The study was deemed not to constitute research involving human subjects according to 45 Code of Federal Regulations 46.102(f) and deemed exempt from board oversight and from full review by the RTI International (Research Triangle Park, NC, US) institutional review board. Optum® CDM data are compliant with the Health Insurance Portability and Accountability Act of 1996, and the study was performed in accordance with the Helsinki Declaration of 1964 and its later amendments. No permission was required for access to or use of Optum® CDM. Patient consent for participation and patient consent for publication were not applicable.
Study Population
The population eligible for inclusion was adults (aged ≥ 18.0 years) with both CKD (stage 2–4 or stage unspecified) and type 2 diabetes who initiated either an SGLT2i or GLP-1 RA in the later period (9 July 2021 [i.e., the FDA approval date of finerenone] to 30 September 2023) with at least 12 months of continuous enrollment before medication initiation (ESM Fig. S2). New users of SGLT2i or GLP-1 RA were defined as patients with an outpatient prescription dispensing (“dispensing”) for any medication in either drug class and no dispensing in the previous 12 months for any other medication in that class. Separate, medication-specific, new-user cohorts were created; these were nonmutually exclusive such that an individual patient could be included in both if both an SGLT2i and GLP-1 RA were newly initiated during the study period. The index dispensing was the first eligible, new-use dispensing during the study period, and the date of this dispensing was the index date. Patients with type 1 diabetes, kidney cancer, or kidney failure (eGFR < 15 mL/min/1.73 m2, stage 5 CKD, receipt of maintenance dialysis, and/or kidney transplant) on or before the index date were excluded.
Variables
Demographic and Clinical Characteristics
Demographic variables and baseline clinical characteristics available at the index date included age, sex, smoking status, obesity (by diagnosis code or body mass index ≥ 30), comorbidities before or on the index date, markers of type 2 diabetes and/or CKD severity, and medications other than glucose-lowering drugs (GLDs) used ≤ 180 days before or on the index date. Diabetes and other medical conditions were defined in Optum® CDM by recorded diagnoses from medical claims (ESM Table S1). CKD was defined using diagnostic codes (CKD stage 2–4 or stage unspecified), two eGFR measurements separated by 90–540 days corresponding with CKD stage 3 or 4 (15–59 mL/min/1.73 m2), or urine albumin-to-creatinine (uACR) measurements (> 30 mg/g) [19]. The patient was classified as CKD stage unspecified in the absence of eGFR, uACR, and stage-specific or stage-unspecified diagnosis codes.
Exposures and Outcomes
Exposures to the index medication were identified from outpatient dispensing records, where medication classes were defined by National Drug Codes based on Anatomical Therapeutic Chemical (ATC)-defined medication classes (ESM Table S2). Current-use periods for a medication were defined as from the day after the index date to the end of presumed supply for consecutive dispensings plus a grace period of 30 days. For each medication cohort, the use of other drugs of interest (i.e., SGLT2i, ACEi/ARB, GLP-1 RA, steroidal MRA [sMRA]) relative to the index medication was classified into three distinct time periods relative to the index date: within 90 days before, on, and within 90 days after, respectively. Indication data were not available. Study-defined categories for the index medication were monotherapy (only drug of interest initiated), combination therapy (simultaneous initiation with another drug of interest), add-on therapy (initiation in addition to existing drug of interest), switched-to therapy (replacement of drug of interest), simultaneous add-on and switched-to therapy, and nonevaluable index therapy. Treatment utilization outcomes were discontinuation (date following last day of current use), switches, and add-ons during each study period. Patients were followed until the first of study end, disenrollment, development of kidney failure or kidney cancer, or death (ESM Fig. S2).
Statistical Analyses
Analyses were performed using SAS version 9.4 or higher (SAS Institute, Inc., Cary, NC, US). All analyses were descriptive, and no comparative analyses of outcomes between the SGLT2i and GLP-1 RA cohorts were performed. Categorical variables were described with counts and percentages, and continuous variables were described with means, standard deviations (SDs), medians, first and third quartiles, and first and 99th percentiles. Treatment changes over time were assessed at 90, 180, 270, and 365 days following the index date. The following treatment states were assessed: (1) treated with index medication; (2) untreated with index medication; (3) death; and (4) lost to follow-up, end of study, or censored. Sankey diagrams were generated for visualization of the treatment states over time. Temporal changes in baseline characteristics between the later and earlier time periods were evaluated separately for the SGLT2i and GLP-1 RA cohorts using the standardized mean difference (SMD) in the level of each covariate, with SMDs of 0.2, 0.5, and 0.8 considered to correspond to small, medium, and large differences, respectively [20].
Results
Later Period Baseline Demographic and Clinical Characteristics
SGLT2i Cohort
After applying all inclusion and exclusion criteria, the new-user cohort included 94,080 patients with a mean age of 73.1 (SD 8.9) years, of whom 47.0% were female, 56.1% were white, 46.0% had obesity, and 26.0% were current smokers (Table 1). The most common baseline comorbidities were hypertension (94.6%) and hypercholesterolemia (88.9%), and common cardiovascular comorbidities included coronary heart disease (43.8%) and congestive heart failure (CHF; 35.9%). Statins (80.1%) and ACEi/ARB (33.1%/59.0%) were the most used medications (ESM Table S3), and the median time from the first recorded type 2 diabetes or CKD code was 5.1 years and 4.0 years, respectively. Most baseline characteristics in the later period were similar to those in the earlier period (SMD < 0.2), except that new users were older (mean [SD] age 73.1 [8.9] vs. 68.6 [10.1] years; SMD 0.48) (Table 1; Fig. 1a). Prevalence of comorbidities was generally higher in the later period, particularly for CHF (35.9% vs. 21.5%; SMD 0.32) (Table 1).
Table 1.
Baseline characteristics of new SGLT2i and GLP-1 RA users in the earlier period and later period
| Characteristic | SGLT2i | GLP-1 RA | ||||
|---|---|---|---|---|---|---|
| Earlier period (N = 56,219) [17] | Later period (N = 94,080) | SMD (later vs. earlier)a | Earlier period (N = 70,158) [18] | Later period (N = 72,816) | SMD (later vs. earlier)a | |
| Age group (years) at index date, n (%) | ||||||
| < 40 | 415 (0.7) | 193 (0.2) | – | 643 (0.9) | 334 (0.5) | – |
| 40–49 | 2254 (4.0) | 1270 (1.3) | – | 2977 (4.2) | 1970 (2.7) | – |
| 50–59 | 7556 (13.4) | 5478 (5.8) | – | 10,119 (14.4) | 7403 (10.2) | – |
| 60–69 | 17,099 (30.4) | 21,172 (22.5) | – | 22,577 (32.2) | 21,129 (29.0) | – |
| 70–79 | 21,872 (38.9) | 43,637 (46.4) | – | 26,480 (37.7) | 32,262 (44.3) | – |
| ≥ 80 | 7023 (12.5) | 22,330 (23.7) | – | 7362 (10.5) | 9718 (13.3) | – |
| Age at index date, years | ||||||
| Mean (SD) | 68.6 (10.1) | 73.1 (8.9) | 0.478 | 67.9 (10.1) | 70.0 (9.3) | 0.214 |
| Median (1st, 99th percentile) | 70 (41, 88) | 74 (47, 89) | – | 69 (40, 87) | 71 (43, 88) | – |
| Sex, n (%) | ||||||
| Male | 30,583 (54.4) | 49,832 (53.0) | − 0.029 | 33,652 (48.0) | 32,007 (44.0) | − 0.081 |
| Female | 25,633 (45.6) | 44,244 (47.0) | – | 36,502 (52.0) | 40,807 (56.0) | – |
| Unknown | N/A | 0 (0) | – | 4 (< 0.1) | 2 (< 0.1) | – |
| Calendar year of index date, n (%) | ||||||
| 2012–2017 | 15,363 (27.3) | N/A | – | 20,482 (29.2) | N/A | – |
| 2018 | 6219 (11.1) | N/A | – | 10,249 (14.6) | N/A | – |
| 2019 | 9818 (17.5) | N/A | – | 13,876 (19.8) | N/A | – |
| 2020 | 13,704 (24.4) | N/A | – | 14,866 (21.2) | N/A | – |
| 2021 b | 11,115 (19.8) | 13,594 (14.4) | – | 10,685 (15.2) | 10,354 (14.2) | – |
| 2022 | N/A | 40,803 (43.4) | – | N/A | 28,048 (38.5) | – |
| 2023 | N/A | 39,683 (42.2) | – | N/A | 34,414 (47.3) | – |
| Race, n (%) | ||||||
| Asian | 2752 (4.9) | 4058 (4.3) | − 0.028 | 2131 (3.0) | 1891 (2.6) | − 0.027 |
| Black | 9167 (16.3) | 16,533 (17.6) | 0.034 | 11,900 (17.0) | 12,181 (16.7) | − 0.006 |
| Hispanic | 10,582 (18.8) | 14,008 (14.9) | − 0.105 | 12,225 (17.4) | 10,311 (14.2) | − 0.090 |
| White | 31,362 (55.8) | 52,749 (56.1) | 0.006 | 41,030 (58.5) | 43,255 (59.4) | 0.019 |
| Other/unknown | 2356 (4.2) | 6732 (7.2) | 0.128 | 2872 (4.1) | 5178 (7.1) | 0.131 |
| BMI, n (%) | ||||||
| < 20 (underweight) | 156 (0.3) | 547 (0.6) | 0.046 | 136 (0.2) | 148 (0.2) | 0.002 |
| 20–24.9 (normal) | 1373 (2.4) | 3037 (3.2) | 0.047 | 1043 (1.5) | 1063 (1.5) | − 0.002 |
| 25–29.9 (overweight) | 4850 (8.6) | 9256 (9.8) | 0.042 | 4589 (6.5) | 5026 (6.9) | 0.014 |
| 30–39.9 (obese) | 11,902 (21.2) | 20,663 (22.0) | 0.019 | 15,039 (21.4) | 18,212 (25.0) | 0.085 |
| ≥ 40 (severely obese) | 5705 (10.1) | 9271 (9.9) | − 0.010 | 9711 (13.8) | 11,985 (16.5) | 0.073 |
| Unknown | 32,233 (57.3) | 51,306 (54.5) | − 0.056 | 39,640 (56.5) | 36,382 (50.0) | − 0.131 |
| Obesity, n (%) | ||||||
| Yes (by diagnosis or BMI ≥ 30) | 26,443 (47.0) | 43,308 (46.0) | − 0.020 | 37,234 (53.1) | 42,147 (57.9) | 0.097 |
| Smoking status, n (%) | ||||||
| Current smoker | 11,583 (20.6) | 24,436 (26.0) | 0.127 | 14,612 (20.8) | 16,157 (22.2) | 0.033 |
| Macrovascular complications, n (%) | ||||||
| Coronary heart disease | 19,663 (35.0) | 41,248 (43.8) | 0.182 | 23,175 (33.0) | 24,513 (33.7) | 0.013 |
| Cerebrovascular disease | 6854 (12.2) | 14,880 (15.8) | 0.105 | 8670 (12.4) | 9210 (12.6) | 0.009 |
| Peripheral vascular disease | 15,737 (28.0) | 32,551 (34.6) | 0.143 | 20,167 (28.7) | 22,714 (31.2) | 0.053 |
| Cardiovascular risk factors, n (%) | ||||||
| Hypertension | 52,558 (93.5) | 89,006 (94.6) | 0.047 | 65,828 (93.8) | 67,617 (92.9) | − 0.039 |
| Hypercholesterolemia | 50,291 (89.5) | 83,606 (88.9) | − 0.019 | 62,519 (89.1) | 63,825 (87.7) | − 0.046 |
| Congestive heart failure | 12,073 (21.5) | 33,790 (35.9) | 0.323 | 14,543 (20.7) | 16,772 (23.0) | 0.056 |
| Severe liver disease | 556 (1.0) | 961 (1.0) | 0.003 | 707 (1.0) | 665 (0.9) | − 0.010 |
| HIV infection | 297 (0.5) | 486 (0.5) | − 0.002 | 346 (0.5) | 396 (0.5) | 0.007 |
| Dementia | 1808 (3.2) | 4889 (5.2) | 0.099 | 2808 (4.0) | 2909 (4.0) | 0.000 |
| Chronic obstructive pulmonary disease | 10,300 (18.3) | 20,211 (21.5) | 0.079 | 13,891 (19.8) | 14,037 (19.3) | − 0.013 |
| Malignancy (other than kidney cancer and nonmelanoma skin cancers) | 7017 (12.5) | 14,657 (15.6) | 0.089 | 8347 (11.9) | 9563 (13.1) | 0.037 |
Earlier period (1 January 2012–30 June 2021) results were reported previously [17, 18]
BMI Body mass index, GLP-1 RA glucagon-like peptide-1 receptor agonists, HIV human immunodeficiency virus, N/A not available, NE not estimable, SD standard deviation, SGLT2i sodium-glucose cotransporter 2 inhibitors, SMD standardized mean difference
a“–” indicates SMD was not evaluated or was not applicable for given variable
bBy design, only 6 months of observation were included in 2021 for the later (9 July 2021–30 September 2023) and earlier (1 January 2012–30 June 2021) periods
Fig. 1.
SMDs by variable between the earlier and later periods. a SGLT2i cohorts. b GLP-1 RA cohorts. ACEi Angiotensin-converting enzyme inhibitor(s), ARB angiotensin receptor blockers, CKD chronic kidney disease, eGFR estimated glomerular filtration rate, HbA1c hemoglobin A1c, N number, SGLT2i sodium-glucose cotransporter 2 inhibitors, SMD standardized mean difference. Baseline covariates with an absolute SMD ≥ 0.2 are shown, with the exception of CKD stage covariates that met the 0.2 threshold and were classified by “diagnosis only” or “eGFR only”; these latter two covariates were omitted due to their observed rates of missingness. Earlier period (1 January 2012–30 June 2021) results were previously reported [17, 18]
GLP-1 RA Cohort
The new-user cohort in the later period included 72,816 patients with a mean age of 70.0 (SD 9.3) years, of whom 56.0% were female, 59.4% were white, 57.9% had obesity, and 22.2% were current smokers (Table 1). Similarly to the SGLT2i cohort, common comorbidities included hypertension (92.9%), hypercholesterolemia (87.7%), coronary heart disease (33.7%), and CHF (23.0%); the use of statins (78.3%) and ACEi/ARB (33.3%/55.2%) was common at baseline (ESM Table S3); and the median time from the first recorded type 2 diabetes or CKD code was 4.8 years and 3.7 years, respectively. Compared with the earlier period cohort, the later period cohort was older (mean [SD] age 70.0 [9.3] vs. 67.9 [10.1] years; SMD 0.21), and the prevalence of comorbidities was relatively similar (Table 1; Fig. 1b).
Later Period Baseline Markers of Type 2 Diabetes Severity and GLD Use
SGLT2i Cohort
The median Diabetes Severity Complications Index (DSCI) score [21] was 3 (ESM Table S4). Hemoglobin A1c (HbA1c) results were not available for 51.0% of the cohort; among patients with results, the highest HbA1c levels (> 9.0% [> 74.9 mmol/mol]) were observed for 8.6% of patients, and HbA1c values were > 7.0% (> 53.0 mmol/mol) for 29.4% of patients, which is the threshold for a diabetes diagnosis. The prevalence of hyperkalemia was 10.5%. Within 180 days before and including the index date, insulin use was reported for 29.1% of patients, and 63.2% had used either one or two medications in a GLD class other than SGLT2i, most often metformin (46.4%). Compared with the earlier period, a greater proportion of patients in the later period had HbA1c levels ≤ 7.0% (< 53.0 mmol/mol) (19.6% vs. 12.5%; SMD 0.20), and a smaller proportion had levels indicating lack of metabolic control (> 9% [> 74.9 mmol/mol]) (8.6% vs. 16.5%; SMD − 0.24). Additionally, a greater proportion of patients in the later period had no previous type 2 diabetes drug classes in the 180 days prior to and including the index date.
GLP-1 RA Cohort
The median DSCI score was 3, and among the 51.1% of patients with HbA1c results, 12.2% had the highest HbA1c levels (> 9.0% [> 74.9 mmol/mol]) and 33.9% had values > 7.0% (> 53.0 mmol/mol) (ESM Table S4). The prevalence of hyperkalemia was 7.6%. Insulin use was reported for 34.6% of patients ≤ 180 days before the index date, and 62.1% had used either one or two medications in a GLD class other than GLP-1 RA, most often metformin (50.4%). Compared with the earlier period, a greater proportion of new users in the later period had HbA1c levels ≤ 7% (17.2% vs. 10.3%; SMD 0.20) and a lower proportion had HbA1c levels indicating lack of metabolic control (> 9.0%) (17.8% vs. 12.2%; SMD, − 0.16). Additionally, use of insulin decreased in the later period compared with the earlier period (34.6% vs. 44.8%; SMD, − 0.21).
Later Period Baseline Markers of Kidney Dysfunction Severity
SGLT2i Cohort
CKD stage was captured mainly from eGFR laboratory values; notably, 36.6% of patients were missing eGFR results, and 29.0% had no recorded diagnosis codes for CKD stage in the year before the index date (ESM Table S5). Among patients with staging information and on the basis of eGFR result or diagnostic code, stage 3 CKD was the most common stage (52.1%; ESM Fig. S3). Assessment of uACR in the year before the index date was not recorded for 75.1% of patients; among patients with results, A2 was the most common CKD stage reported based on uACR (30–300 mg/g) (40.5%). Historical (> 366 days before index) or previous (91–365 days) use of ACEi/ARB was recorded for 81.9–89.2% of patients, and recent use (≥ 90 days) was recorded for 72.5% of patients. Historical or previous use of sMRA was observed in 10.9–14.0% of patients, and recent use was recorded for 9.5% of patients. Patients in the later period compared with the earlier period had more severe kidney dysfunction (stage 3: 52.1% vs. 26.3%; SMD 0.55) and had a longer minimum recorded duration of CKD (mean: 4.8 vs. 3.8 years; SMD 0.31).
GLP-1 RA Cohort
Similarly to the SGLT2i cohort, notable proportions of GLP-1 RA new users were missing results for eGFR (37.1%), CKD stage diagnosis codes (37.6%), and uACR (75.2%) (Supplemental Table S5). Stage 3 CKD was most common based on eGFR or diagnosis codes (44.0%; ESM Fig. S3), and stage based on uACR was most often A1 (43.2%). Historical or previous (79.7–87.1%) and recent use (71.2%) of ACEi/ARB was common; historical or previous sMRA use was recorded for 8.4–11.2% of patients, and recent use for 6.9% of patients. New users in the later period compared with those in the earlier period had more severe kidney dysfunction (stage 3: 44.0% vs. 27.6%; SMD 0.35). Similar to what was observed for the SGLT2i cohort, new users of GLP-1 RA in the later periods had a longer minimum recorded duration of CKD (mean: 4.6 years vs. 3.6 years; SMD 0.33). Additionally, new users in the later period were slightly more likely to have a historical (22.9% vs. 13.9%; SMD 0.23), previous (22.5% vs. 13.2%; SMD 0.25), and recent (19.9% vs. 11.3%; SMD 0.24) use of an SGLT2i.
Later Period Characteristics of Index Medication
SGLT2i Cohort
The median duration of initial exposure was 4.3 months, and at the index date, SGLT2i was most commonly prescribed as an add-on therapy to another medication of interest (54.9%), most often added to an ACEi/ARB (54.0%) (Table 2; Supplemental Fig. S4). Addition to a GLP-1 RA medication occurred for 9.0% of patients. SGLT2i were prescribed in a similar manner in the later and earlier periods, although when an SGLT2i was prescribed as an add-on therapy, it was less often added to an ACEi/ARB in the later period than the earlier period (54.0% vs. 64.9%; SMD − 0.23).
Table 2.
Classification of the index medication at the index date in the SGLT2i and GLP-1 RA cohorts
| Classification of index medication | SGLT2i | GLP-1 RA | ||||
|---|---|---|---|---|---|---|
| Earlier period (N = 56,219) [17] | Later period (N = 94,080) | SMD (later vs. earlier)a | Earlier period (N = 70,158) [18] | Later period (N = 72,816) | SMD (later vs. earlier)a | |
| Classification of index therapy at the index date, n (%) b | ||||||
| Monotherapy | 10,117 (18.0) | 16,483 (17.5) | − 0.012 | 14,535 (20.7) | 14,571 (20.0) | − 0.018 |
| Combination therapy | 1846 (3.3) | 3904 (4.1) | 0.046 | 1804 (2.6) | 1826 (2.5) | − 0.004 |
| Add-on | 32,392 (57.6) | 51,623 (54.9) | − 0.055 | 40,500 (57.7) | 38,882 (53.4) | − 0.087 |
| Switch | 3972 (7.1) | 6321 (6.7) | − 0.014 | 5244 (7.5) | 4810 (6.6) | − 0.034 |
| Add-on and switch | 1873 (3.3) | 3585 (3.8) | 0.026 | 2042 (2.9) | 3034 (4.2) | 0.068 |
| Indeterminate | 6019 (10.7) | 12,164 (12.9) | 0.069 | 6033 (8.6) | 9693 (13.3) | 0.151 |
| Index drug was an “add-on” to…, n (%) | ||||||
| Finerenone | N/A | 104 (0.1) | – | N/A | 105 (0.1) | – |
| SGLT2i | N/A | N/A | – | 5577 (7.9) | 8826 (12.1) | 0.139 |
| GLP-1 RA | 6801 (12.1) | 8478 (9.0) | − 0.101 | N/A | N/A | – |
| sMRA | 2873 (5.1) | 5770 (6.1) | 0.044 | 3318 (4.7) | 3276 (4.5) | − 0.011 |
| ACEi/ARB | 36,511 (64.9) | 50,777 (54.0) | − 0.225 | 44,708 (63.7) | 38,374 (52.7) | − 0.225 |
| Index drug was a “switch” to…, n (%) | ||||||
| Finerenone | N/A | 53 (0.1) | – | N/A | 39 (0.1) | – |
| SGLT2i | N/A | N/A | – | 2318 (3.3) | 2980 (4.1) | 0.042 |
| GLP-1 RA | 1854 (3.3) | 2470 (2.6) | − 0.040 | N/A | N/A | – |
| sMRA | 796 (1.4) | 1665 (1.8) | 0.028 | 847 (1.2) | 783 (1.1) | − 0.012 |
| ACEi/ARB | 4941 (8.8) | 6309 (6.7) | − 0.078 | 6147 (8.8) | 4601 (6.3) | − 0.093 |
| Duration of initial exposure episode after cohort entry, months | ||||||
| Mean (SD) | 10.2 (12.1) | 6.4 (5.8) | – | 9.4 (12.3) | 5.2 (5.1) | – |
| Median (1st, 99th percentile) | 5.4 (1, 55) | 4.3 (1, 25) | – | 4 (1, 57) | 3.2 (1, 24) | – |
| Days’ supply of index drug, days | ||||||
| Mean (SD) | 47.2 (27.8) | 50.5 (30.0) | 0.114 | 40.4 (23.4) | 40.0 (22.6) | − 0.015 |
| Median (1st, 99th percentile) | 30 (14, 90) | 30 (7, 100) | – | 30 (7, 90) | 28 (14, 90) | – |
| No. of dispensings for initial exposure episode after cohort entry | ||||||
| Mean (SD) | 6.6 (8.7) | 4.0 (4.2) | – | 7.1 (10.0) | 4.0 (4.3) | – |
| Median (1st, 99 th percentile) | 3 (1, 42) | 2 (1, 20) | – | 3 (1, 49) | 2 (1, 21) | – |
| No. of distinct “current use” periods for the index therapy, n (%) | ||||||
| 1 | 28,202 (50.2) | 66,520 (70.7) | – | 28,047 (40.0) | 43,432 (59.6) | – |
| 2 | 13,912 (24.7) | 18,264 (19.4) | – | 16,028 (22.8) | 15,668 (21.5) | – |
| 3 | 6748 (12.0) | 5662 (6.0) | – | 9629 (13.7) | 6637 (9.1) | – |
| 4 | 3421 (6.1) | 1949 (2.1) | – | 5798 (8.3) | 3061 (4.2) | – |
| 5+ | 3936 (7.0) | 1685 (1.8) | – | 10,656 (15.2) | 4018 (5.5) | – |
| No. of dispensings for the index therapy over the study period | ||||||
| Mean (SD) | 11.1 (12.0) | 5.5 (5.8) | – | 13.9 (14.7) | 6.7 (7.4) | – |
| Median (1st, 99th percentile) | 8 (1, 57) | 4 (1, 27) | – | 9 (1, 68) | 4 (1, 35) | – |
| No. of discontinuations (interruptions) of “current use,” n (%) | ||||||
| 0 | 28,202 (50.2) | 66,520 (70.7) | – | 28,047 (40.0) | 43,412 (59.6) | – |
| 1 | 13,912 (24.7) | 18,264 (19.4) | – | 16,028 (22.8) | 15,668 (21.5) | – |
| 2 | 6748 (12.0) | 5662 (6.0) | – | 9629 (13.7) | 6637 (9.1) | – |
| 3 | 3421 (6.1) | 1949 (2.1) | – | 5798 (8.3) | 3061 (4.2) | – |
| 4 | 1689 (3.0) | 830 (0.9) | – | 3701 (5.3) | 1597 (2.2) | – |
| 5+ | 2247 (4.0) | 855 (0.9) | – | 6955 (9.9) | 2421 (3.3) | – |
| No. of patients with an interruption of “current use” lasting ≥ 90 days, n (%) | 18,100 (32.2) | 17,228 (18.3) | – | 28,959 (41.3) | 20,032 (27.5) | – |
| Duration of total exposure to index therapy, months | ||||||
| Mean (SD) | 19.7 (17.6) | 10.3 (8.7) | – | 21.9 (20.0) | 10.7 (10.6) | – |
| Median (1st, 99th percentile) | 14.5 (2, 79) | 7.8 (NE, NE) | – | 16.1 (2, 89) | 7.5 (2, 54) | – |
| Other drug classes started during follow-up, n (%) | ||||||
| Finerenone | N/A | 967 (1.0%) | – | N/A | 409 (0.6%) | – |
| SGLT2i | N/A | N/A | – | 6815 (9.7%) | 5528 (7.6%) | – |
| GLP-1 RA | 7102 (12.6%) | 8146 (8.7%) | – | N/A | N/A | – |
| sMRA | 1519 (2.7) | 2810 (3.0) | – | 2193 (3.1) | 1282 (1.8) | – |
| ACEi/ARB | 1943 (3.5) | 2244 (2.4) | – | 2948 (4.2) | 1907 (2.6) | – |
| Duration of total follow-up, months | ||||||
| Mean (SD) | 21.9 (20.4) | 10.4 (7.1) | – | 23.8 (21.7) | 10.0 (7.2) | – |
| Median (1st, 99th percentile) | 15.8 (0, 83) | 8.8 (0, 26) | – | 17.8 (0, 99) | 8.1 (0, 26) | – |
| Administrative reason for end of follow-up, n (%) | ||||||
| End of study period | 40,752 (72.5) | 78,244 (83.2) | – | 47,696 (68.0) | 62,322 (85.6) | – |
| Disenrollment from the database or emigration from the database catchment area | 11,059 (19.7) | 9370 (10.0) | – | 14,605 (20.8) | 7268 (10.0) | – |
| Development of kidney failure during follow–up | 1245 (2.2) | 1925 (2.0) | – | 2864 (4.1) | 1198 (1.6) | – |
| Development of kidney cancer | 173 (0.3) | 202 (0.2) | – | 321 (0.5) | 135 (0.2) | – |
| Death | 2990 (5.3) | 4336 (4.6) | – | 4672 (6.7) | 1891 (2.6) | – |
Earlier period (1 January 2012–30 June 2021) results were previously reported [17, 18]
ACEi Angiotensin-converting enzyme inhibitors, ARB angiotensin receptor blockers, GLP-1 RA glucagon-like peptide-1 receptor agonists, N/A not applicable, NE not estimated, SD standard deviation, SGLT2i sodium-glucose cotransporter 2 inhibitor, sMRA steroidal mineralocorticoid receptor antagonist
a“–” indicates SMD was not evaluated or was not applicable for given variable
bIndex therapy was classified according to use of study-defined medications of interest
GLP-1 RA Cohort
The median duration of the initial exposure was 3.2 months, and at the index date GLP-1 RA was most prescribed as an add-on therapy (53.4%), most often to an ACEi/ARB (52.7%) (Table 2, Supplemental Fig. S4), similar to what was observed in the SGLT2i cohort. Addition to an SGLT2i medication occurred for 12.1% of patients. GLP-1 RA were prescribed in a similar manner in the later and earlier periods, although when prescribed as an add-on therapy, they were more often added to ACEi/ARB in the earlier than later period (52.7 vs. 63.7%; SMD − 0.23).
Later Period Treatment Changes During Follow-Up
SGLT2i Cohort
Among patients still under observation at 12 months, 55.0% were receiving treatment at 1 year (Fig. 2a). The largest proportional increase in the “no exposure” treatment state occurred between the 90-day and 180-day timepoints (16.0%). Thereafter, the proportion of patients with no treatment remained stable. At each timepoint, a small proportion of nonusers who remained under observation were found to change and become current users in both SGLT2i and GLP-1 RA cohorts. Treatment patterns in the later period were similar to those observed in the earlier period, where 50.0% of patients were receiving treatment at 1 year and the largest proportional increase in “no exposure” states occurred between the 90- and 180-day timepoints [17].
Fig. 2.
Treatment states at specific timepoints. a SGLT2i cohort. b GLP-1 RA cohort. Sankey diagrams display the proportion of the population at each timepoint in each of the treatment states for each data source. The connecting bars between timepoints show the proportion of the population that moved from one state to a different state at the next timepoint. These figures display proportions of the population over time, and a patient may move between treatment states over time (e.g., begin as “treated,” move to “untreated” at the next timepoint, then move back to “treated” at the next). If death occurred, the patient was placed in a separate category and remained in that state for each subsequent checkpoint. The height of the bar at each timepoint displays the relative size of the cohort remaining under observation at each timepoint. Patients who were lost to follow-up are not included in the percentage calculations at each timepoint; thus, the percentages sum to 100% for each timepoint. The percentages describe the patients still under observation at that timepoint. The sum of the bars for each timepoint may not equal 100% due to rounding. ACEi Angiotensin-converting enzyme inhibitors, ARB angiotensin receptor blockers, CKD chronic kidney disease, GLP-1 RA glucagon-like peptide-1 receptor agonists, GFR glomerular filtration rate, HbA1c glycated hemoglobin, SGLT2i sodium-glucose cotransporter 2 inhibitor
GLP-1 RA Cohort
Among patients still under observation at 12 months, 52.0% were receiving treatment at 1 year (Fig. 2b). The largest proportional increase in the “no exposure” treatment state occurred between the index and 90-day timepoints (20.0%), after which the proportion of patients with no treatment remained stable. The percentage of patients receiving treatment at 1 year was comparable to the earlier period (both 52.0%), with similar proportional timepoint changes also observed [18].
Discussion
This observational cohort study describes real-world utilization patterns and characteristics of patients dispensed an SGLT2i or GLP-1 RA in the US using data from Optum® CDM in the period following changes to the landscape for treatment and prevention of CKD in type 2 diabetes (later period, July 2021–September 2023) as well as temporal changes in utilization and baseline characteristics relative to the period prior (earlier period, January 2012–June 2021).
Clinical profiles and treatment patterns for new users of SGLT2i or GLP-1 RA with CKD and type 2 diabetes in the earlier period (2012–2021) have been previously reported [17, 18]. Although separate cohorts were derived from Optum® CDM for the present study, the study methodology and variables were consistent, allowing for comparison between periods. In the later period, new users in both SGLT2i and GLP-1 RA cohorts were older, use of medications for type 2 diabetes was less intensive, and metabolic control was better, and although more patients had renal function assessed, patients had more severe CKD (stage 3: 44.0–52.1% vs. 26.3–27.6%). In the SGLT2i cohort, having no therapy for type 2 diabetes before the index date was more common in the later period than in the earlier period (28.7% vs. 16.0%; SMD 0.31), corresponding with decreased use of GLDs. Additionally, the prevalence of CHF notably increased (35.9% vs. 21.5%; SMD 0.32). In the later period GLP-1 RA cohort, use of SGLT2i increased considerably (23.4% vs. 13.3%; SMD 0.26) and use of insulin decreased (34.6% vs. 44.8%; SMD − 0.21) compared with the earlier period.
Differences observed between earlier and later periods in the medication-specific cohorts are likely due to more recent available evidence regarding the efficacy and effectiveness of these therapies, evolution of guideline recommendations, and indication space—particularly for SGLT2i—occurring in 2020–2022. Notably, these findings are largely in line with those from a real-world temporal analysis similarly characterizing SGLT2i and GLP-1 RA new users in Japan between the same two time periods [22]. The demonstration of cardiorenal benefits with SGLT2i in clinical trials both for participants with CKD and type 2 diabetes and for those without [23–25] and the updated 2022 KDIGO guidelines recommending SGLT2i as first-line therapy for patients with CKD and type 2 diabetes regardless of glycemic control [6] have likely driven a shift toward their broader use for preventive cardiorenal protection and earlier implementation for patients with type 2 diabetes. Further, changes in the indication space may have driven differences between periods, such as the increased prevalence of CHF observed in the later period. This may reflect the indication of SGLT2i for heart failure in 2022 and the subsequent increased use of SGLT2i by physicians for individuals with CHF and reduced ejection fraction [26–28]. Although this study was not designed to assess these trends, these observations prompt several questions for future investigations, including physician adoption of treatments and treatment paradigm shifts. Treatments like SGLT2i and GLP-1 RA are recommended as part of a comprehensive approach to reduce cardiorenal outcome risk, often in addition to established treatments like RAASi in patients with type 2 diabetes. Accordingly, over half of patients in this study initiated SGLT2i or GLP-1 RA as an add-on treatment, most often to ACEi/ARB (> 50.0%), consistent with the frequent prescription of antihypertensive agents for patients with CKD and type 2 diabetes reported by other studies, including those reporting the earlier period findings [17, 18, 22]. Notably, 12.0% of patients initiated GLP-1 RA as an add-on to SGLT2i, with 9.0% initiating SGLT2i as an add-on to GLP-1 RA. Indeed, the combined use of these medications is suggested to be synergistic and, as reported in a recent meta-analysis, superior outcomes with the addition of a GLP-1 RA to SGLT2i can be observed in type 2 diabetes, although observations of cardiovascular and mortality outcomes are limited [29]. This finding, together with recent guideline updates, may explain the improved glycemic control observed in the later study period.
This study has several strengths, including assessment of a large, real-world dataset and the selection of study periods that represent the rapidly evolving treatment landscape for people with CKD and type 2 diabetes. The eligibility criteria aligned with the clinical criteria for medication prescription, with minimized exclusion criteria to maximize inclusion of patients with CKD and type 2 diabetes initiating these medications in real-world settings. Nonetheless, limitations, including those inherent to healthcare claims database studies, must be considered. For instance, the duration since or sequence of CKD and type 2 diabetes diagnoses is difficult to ascertain as individuals may have been diagnosed previously while on another health insurance plan or may have experienced gaps in coverage. Additionally, while both laboratory and prescription dispensing data are captured in Optum® CDM, indication information was not available and data were generated from healthcare delivery rather than for research purposes and may therefore be incomplete or missing. Following, it is not possible to distinguish the rationale for medication initiation—such as whether initial SGLT2i or GLP-1 RA prescriptions were for weight loss, glycemic control, or cardiovascular protection in this population—which may introduce uncertainty in understanding usage patterns. High degrees of missingness were observed for key variables, including HbA1c (48.9–51.0%), eGFR (36.6–37.1%), and uACR (75.1–75.2%); as this was a descriptive analysis, complex methods such as multiple imputation to account for missing data were not applied. However, missing data were represented as a separate category to mitigate variable misclassification and allow for separate assessment among patients with missing data. Additionally, undercapture of medications in Optum® CDM is possible if patients paid out of pocket or sought out-of-network care where a prescription was written. Moreover, certain assessments are limited by data availability, such as evaluations of CKD progress by measures of blood pressure, as these data are not widely available or reliable in real-world sources. Sex- and/or gender-stratified analyses were not conducted as part of this study and would be an important part of future studies assessing real-world treatment patterns. As patients included in the study may be seeking more frequent care and/or have at least 12 months of continuous enrollment, generalizability may be limited compared with those with less healthcare utilization or baseline history. Finally, as no direct, formal comparisons of the SGLT2i and GLP-1 RA cohorts were conducted, observations between medication cohorts are purely descriptive, and no causal conclusions can be drawn.
Conclusion
Understanding changes in clinical characteristics and treatment patterns over time among new users of SGLT2i or GLP-1 RA in the US as described in this observational study is important to inform and contextualize future studies assessing cardiorenal outcomes for these and additional treatments, including finerenone, for individuals with CKD and type 2 diabetes.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgments
Medical Writing, Editorial, and Other Assistance
Gabrielle Dardis, PhD, and Matt Fitzpatrick, PhD, of RTI Health Solutions, a subsidiary of RTI International, a nonprofit organization that conducts work for government, public, and private organizations, including pharmaceutical companies, provided medical writing support, and John Forbes of RTI Health Solutions provided medical editing support with funding from Bayer AG during manuscript development.
Author Contributions
Conceptualization: Catherine B. Johannes, Anam M. Khan, Ryan Ziemiecki, J. Bradley Layton, David Vizcaya, Nikolaus G. Oberprieler. Data curation: Craig I. Coleman. Formal analysis: Craig I. Coleman. Funding acquisition: Craig I. Coleman, Nikolaus G. Oberprieler. Investigation: Catherine B. Johannes, Anam M. Khan, Ryan Ziemiecki, J. Bradley Layton, David Vizcaya, Nikolaus G. Oberprieler. Methodology: Catherine B. Johannes, Anam M. Khan, Ryan Ziemiecki, J. Bradley Layton, David Vizcaya, Nikolaus G. Oberprieler. Project administration: Catherine B. Johannes, Craig I. Coleman, Anam M. Khan, Ryan Ziemiecki, J. Bradley Layton, Fangfang Liu, David Vizcaya, Nikolaus G. Oberprieler. Supervision: Catherine B. Johannes, Craig I. Coleman, Anam M. Khan, Ryan Ziemiecki, J. Bradley Layton, Fangfang Liu, David Vizcaya, Nikolaus G. Oberprieler. Validation: Ryan Ziemieck. Writing—review & editing: Catherine B. Johannes, Craig I. Coleman, Csaba P. Kovesdy, Anam M. Khan, Ryan Ziemiecki, J. Bradley Layton, Fangfang Liu, David Vizcaya, and Nikolaus G. Oberprieler.
Funding
This study was funded by Bayer AG. Bayer AG funded the journal’s Rapid Service and Open Access Fees. Authors affiliated with Bayer were involved in the study design, analyses, and development of this publication.
Data Availability
The dataset generated and/or analyzed during the current study is not publicly available due to legal and policy restrictions.
Declarations
Conflict of Interest
Fangfang Liu and Nikolaus G. Oberprieler are employees of Bayer, which funded this study. David Vizcaya was an employee of Bayer at the time the research was conducted. Catherine B. Johannes, Anam M. Khan, Ryan Ziemiecki, and J. Bradley Layton are or were employees of RTI Health Solutions, which received research funding for this study from Bayer. Craig I. Coleman has received grant funding and consulting fees from Bayer AG and AstraZeneca Pharmaceuticals. Csaba P. Kovesdy received consulting fees from Abbott, Akebia, Ardelyx, Astra Zeneca, Bayer, Boehringer Ingelheim, Cara Therapeutics, CSL Behring, CSL Vifor, GSK, Pharmacosmos, ProKidney, Renibus and Takeda.
Ethical Approval
This study was deemed not to constitute research involving human subjects according to 45 Code of Federal Regulations 46.102(f) and deemed exempt from board oversight and from full review by the RTI International institutional review board. Data from Optum’s de-identified Clinformatics® Data Mart Database (Optum® CDM) are compliant with the Health Insurance Portability and Accountability Act of 1996, and the study was performed in accordance with the Helsinki Declaration of 1964 and its later amendments. No permission was required for access to or use of Optum® CDM. Patient consent for participation and patient consent for publication were not applicable.
Footnotes
Employee at the time the research was conducted: Catherine B. Johannes, David Vizcaya.
Prior Presentation: This work was presented in part at the 2025 International Society for Pharmacoepidemiology (ISPE) Annual Meeting held 22–26 August in Washington, DC, USA.
References
- 1.Kovesdy CP. Epidemiology of chronic kidney disease: an update 2022. Kidney Int Suppl (2011). 2022;12(1):7–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Webster AC, Nagler EV, Morton RL, Masson P. Chronic kidney disease. Lancet. 2017;389(10075):1238–52. [DOI] [PubMed] [Google Scholar]
- 3.Deng Y, Li N, Wu Y, et al. Global, regional, and national burden of diabetes-related chronic kidney disease from 1990 to 2019. Front Endocrinol (Lausanne). 2021;12:672350. 10.3389/fendo.2021.672350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Levey AS, Schoolwerth AC, Burrows NR, Williams DE, Stith KR, McClellan W. Comprehensive public health strategies for preventing the development, progression, and complications of CKD: report of an expert panel convened by the Centers for Disease Control and Prevention. Am J Kidney Dis. 2009;53(3):522–35. 10.1053/j.ajkd.2008.11.019. [DOI] [PubMed] [Google Scholar]
- 5.American Diabetes Association. American Diabetes Association Professional Practice Committee. Chronic kidney disease and risk management: standards of care in diabetes—2024. Diabetes Care. 2024;47(Suppl 1):S219–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.KDIGO. Kidney Disease: Improving Global Outcomes Work Group (KDIGO) 2022 clinical practice guideline for diabetes management in chronic kidney disease. Kidney Int. 2022;2022(102):S1-127. [DOI] [PubMed] [Google Scholar]
- 7.Kintscher U, Edelmann F. The non-steroidal mineralocorticoid receptor antagonist finerenone and heart failure with preserved ejection fraction. Cardiovasc Diabetol. 2023;22(1):162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bayer HealthCare Pharmaceuticals Inc. KERENDIA (finerenone) tablets, for oral use. Highlights of prescribing information. 2022. https://labeling.bayerhealthcare.com/html/products/pi/Kerendia_PI.pdf. Accessed 6 Nov 2023.
- 9.Janssen Pharmaceuticals, Inc. INVOKANA (canagliflozin) tablets, for oral use. Highlights of prescribing information. 2020. https://www.accessdata.fda.gov/drugsatfda_docs/label/2020/204042s034lbl.pdf. Accessed 29 May 2025.
- 10.AstraZeneca. Press release: Farxiga approved in the US for the treatment of chronic kidney disease in patients at risk of progression with and without type-2 diabetes. 2021.https://www.astrazeneca.com/media-centre/press-releases/2021/farxiga-approved-in-the-us-for-ckd.html#!. Accessed
- 11.Palmer SC, Tendal B, Mustafa RA, et al. Sodium-glucose cotransporter protein-2 (SGLT-2) inhibitors and glucagon-like peptide-1 (GLP-1) receptor agonists for type 2 diabetes: systematic review and network meta-analysis of randomised controlled trials. BMJ. 2021;372:m4573. 10.1136/bmj.m4573. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sattar N, Lee MMY, Kristensen SL, et al. Cardiovascular, mortality, and kidney outcomes with GLP-1 receptor agonists in patients with type 2 diabetes: a systematic review and meta-analysis of randomised trials. Lancet Diabetes Endocrinol. 2021;9(10):653–62. [DOI] [PubMed] [Google Scholar]
- 13.Novo Nordisk. OZEMPIC (semaglutide) injection, for subcutaneous use. Highlights of prescribing information. 2025. https://www.accessdata.fda.gov/drugsatfda_docs/label/2025/209637s025lbl.pdf. Accessed 13 Feb 2025.
- 14.Shi Q, Nong K, Vandvik PO, et al. Benefits and harms of drug treatment for type 2 diabetes: systematic review and network meta-analysis of randomised controlled trials. BMJ. 2023;381:e074068. 10.1136/bmj-2022-074068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.US Food and Drug Administration. Center for Drug Evaluation and Research. Approval package for Application number: 215341Orig1s000. 2021. https://www.accessdata.fda.gov/drugsatfda_docs/nda/2021/215341Orig1s000Approv.pdf. Accessed 25 Apr 2025.
- 16.Oberprieler NG, Pladevall-Vila M, Johannes C, et al. FOUNTAIN: a modular research platform for integrated real-world evidence generation. BMC Med Res Methodol. 2024;24(1):224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Johannes CB, Ziemiecki R, Pladevall-Vila M, et al. Clinical profile and treatment adherence in patients with type 2 diabetes and chronic kidney disease who initiate an SGLT2 inhibitor: a multi-cohort study. Diabetes Ther. 2025;16(2):205–26. 10.1007/s13300-024-01671-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Pladevall-Vila M, Ziemiecki R, Johannes CB, et al. Clinical profile and treatment patterns in individuals with type 2 diabetes and chronic kidney disease who initiate a GLP-1 receptor agonist: a multinational cohort study. Diabetes Ther. 2025;16(5):931–54. 10.1007/s13300-025-01717-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Levey AS, Becker C, Inker LA. Glomerular filtration rate and albuminuria for detection and staging of acute and chronic kidney disease in adults: a systematic review. JAMA. 2015;313(8):837–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med. 2009;10(25):3083–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Glasheen WP, Renda A, Dong Y. Diabetes Complications Severity Index (DCSI)—update and ICD-10 translation. J Diabetes Complications. 2017;31(6):1007–13. 10.1016/j.jdiacomp.2017.02.018. [DOI] [PubMed] [Google Scholar]
- 22.Yano Y, Okami S, Kanegae H, et al. Transitional changes in medication-initiator cohort profiles in persons with chronic kidney disease and type 2 diabetes—a hospital-based cohort study in Japan. Diabetes Obes Metab. 2025;27(7):3714–24. [DOI] [PMC free article] [PubMed]
- 23.Heerspink HJL, Stefansson BV, Correa-Rotter R, et al. Dapagliflozin in patients with chronic kidney disease. N Engl J Med. 2020;8(15):1436–46. [DOI] [PubMed] [Google Scholar]
- 24.Packer M, Anker SD, Butler J, et al. Cardiovascular and renal outcomes with empagliflozin in heart failure. N Engl J Med. 2020;8(15):1413–24. [DOI] [PubMed] [Google Scholar]
- 25.The Empa-Kidney Collaborative Group, Herrington WG, Staplin N, et al. Empagliflozin in patients with chronic kidney disease. N Engl J Med. 2023;388(2):117–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Heidenreich PA, Bozkurt B, Aguilar D, et al. 2022 AHA/ACC/HFSA guideline for the management of heart failure: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2022;3(18):e895–1032. [DOI] [PubMed] [Google Scholar]
- 27.Varshney AS, Calma J, Kalwani NM, et al. Uptake of sodium-glucose cotransporter-2 inhibitors in hospitalized patients with heart failure: insights from the Veterans Affairs Healthcare system. J Card Fail. 2024;30(9):1086–95. 10.1016/j.cardfail.2023.12.018. [DOI] [PubMed] [Google Scholar]
- 28.Abdel Jawad M, Spertus JA, Ikeaba U, et al. Early adoption of sodium-glucose cotransporter-2 inhibitor in patients hospitalized with heart failure with mildly reduced or preserved ejection fraction. JAMA Cardiol. 2025;1(1):89–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Tuersun A, Hou G, Cheng G. Efficacy and safety of the combination or monotherapy with GLP-1 receptor agonists and SGLT-2 inhibitors in type 2 diabetes mellitus: an update systematic review and meta-analysis. Am J Med Sci. 2024;368(6):579–88. 10.1016/j.amjms.2024.07.011. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The dataset generated and/or analyzed during the current study is not publicly available due to legal and policy restrictions.