A Pilot Study Embedding Clinical Pharmacists Within an Interprofessional Nephrology Clinic for the Initiation and Monitoring of Empagliflozin in Diabetic Kidney Disease

Abstract

Background:

The American Diabetes Association (ADA) recommends sodium-glucose cotransporter-2 (SGLT2) inhibitors as the second medication to be started, after metformin, for patients with chronic kidney disease (CKD). Sodium-glucose cotransporter-2 inhibitors may cause volume, blood pressure, and electrolyte disturbances; consequently, frequent monitoring and adjustments to other diabetes, blood pressure, and/or diuretic medications may be necessary.

Objective:

To evaluate the safety and efficacy of an interprofessional clinic model partnering nephrologists and pharmacists for the initiation and monitoring of SGLT2 inhibitors.

Methods:

A clinical pharmacist was embedded within the nephrology clinic to provide patient education, telephone follow-up, and to work collaboratively with the nephrologists. Diabetes, hypertension, and diuretic regimens were adjusted as needed after empagliflozin initiation. Diabetes regimens were adjusted to adhere to the 2019 ADA guidelines that promote agents with CKD and atherosclerotic cardiovascular disease benefit.

Results:

Fourteen patients were initiated on empagliflozin during the study period. Urine albumin-to-creatinine ratio (UACR) improved (mean % change −12% ± 61%); the mean percentage change was greater in patients with a higher baseline UACR. The mean change in hemoglobin A_1c was 0.3% ± 0.6%. Common adverse reactions were observed and improved over time; no serious adverse drug reactions occurred. Finally, empagliflozin initiation necessitated adjustments to diabetes, hypertension, and diuretic regimens in almost all patients (n = 13, 93%).

Conclusion:

The implementation of an innovative, interprofessional care model within a nephrology clinic for the initiation and monitoring of empagliflozin in patients with DKD demonstrated clinical benefit with minimal safety concerns.

Introduction

In developed countries, diabetic kidney disease (DKD) is the primary cause of end-stage renal disease (ESRD), accounting for 50% of cases.^1,2 DKD is characterized by elevated urinary albumin excretion in the presence or absence of a decreased estimated glomerular filtration rate (eGFR).³ Proteinuric DKD is associated with rapid eGFR decline and progression to ESRD.^4,5 The most common cause of death in patients with chronic kidney disease (CKD) is cardiovascular disease, and the proportion of cardiovascular deaths increases with declining eGFR.⁶ As the incidence of ESRD in the United States is projected to rise by 18% in the next 15 years, there have been calls for interventions focused on preventing the progression of CKD to ESRD.⁷ Glycemic control is one of the key factors in the prevention of DKD progression, and a novel class of antidiabetic medications, sodium-glucose cotransporter-2 (SGLT2) inhibitors, has emerged as a preferred treatment option in patients with comorbid diabetes and CKD.⁸ SGLT2 inhibitors lower blood glucose by reducing glucose and sodium reabsorption; they provide renoprotective effects by reducing glomerular hyperfiltration.¹ In clinical trials, SGLT2 inhibitors decreased the incidence and progression of albuminuria, slowed eGFR decline, and reduced the need for renal replacement therapy in the DKD population.^9–11 SGLT2 inhibitors also significantly decreased the risk of all-cause mortality in participants having DKD with an eGFR as low as 30 mL/min/1.73 m².^9–17

SGLT2 inhibitors are now recommended as the second medication to be started, after metformin, for diabetes management in patients with CKD, atherosclerotic cardiovascular disease (ASCVD), or heart failure (HF).^2,18–20 However, current guidelines and studies do not provide practical guidance for implementation in the real-world setting. A recent population-based study noted that patients seen by nephrologists are the most complex among different specialties²¹; in the real-world setting, patients with DKD have multiple comorbidities and a high medication burden.^22,23 SGLT2 inhibitors can cause volume, blood pressure, and electrolyte disturbances beyond their glycemic effects, and after initiation patients may require adjustments to their diabetes, blood pressure, and/or diuretic regimens.^{1,9,12,24–27} Appropriate patient education and close monitoring for adverse effects are extremely important,²⁸ and yet the current literature lacks guidance to support an implementation strategy of SGLT2 inhibitors, specifically the resources for patient education, management of other medications, and patient monitoring in the real-world setting.^29–31

To align patient care with current guidelines,^3,18,32 the outpatient nephrology clinic at the VA Boston Medical Center created an interprofessional clinic model utilizing nephrologists and pharmacists for initiation and monitoring of SGLT2 inhibitors in patients with DKD. Our aim was to assess the clinical, safety, and feasibility outcomes of our care model during the 3-month initiation and 3-month monitoring period (6 months total) of empagliflozin, our formulary SGLT2 inhibitor, in patients with DKD in a real-world setting.

Our rationale for the development of this collaborative clinic stems from evidence demonstrating that collaborative practice interventions utilizing pharmacists promotes patient education and closer monitoring, leading to improved clinical and health system outcomes in both diabetes and CKD.^33–39 Within our model, the nephrology providers are poised to manage changes in volume status or blood pressure, electrolyte derangements (ie, hyperkalemia), and metabolic syndromes that may occur within the first 3 months of empagliflozin initiation while clinical pharmacists are well positioned to conduct collaborative diabetes management including medication adjustments, monitoring, and patient education (Figure 1). Our clinical model was approved by the internal Research and Development Committee at the VA as a quality improvement initiative and was exempt from further Institutional Review Board oversight.

Figure 1.

Process flow diagram of interprofessional model for the initiation and monitoring of empagliflozin within an interdisciplinary nephrology clinic. ADEs indicates adverse drug events; BP, blood pressure; BG, blood glucose; DM, diabetes mellitus; HTN, hypertension.

Methods

Patients

Patients were targeted for inclusion if they had a history of type 2 diabetes mellitus (T2DM), urine albumin-to-creatinine ratio (UACR) >300 mg/g at least once within the past 12 months prior to screening, and an eGFR >35 mL/min/1.73 m² at the time of visit (Figure 2). The inclusion criteria intended to select complex patients at high risk for kidney disease progression. The UACR cutoff was chosen based on criteria in published trials that featured participants with an eGFR as low as 30 mL/min/1.73 m² and demonstrated the greatest benefit in proteinuria reduction and eGFR preservation among participants with UACR >300 mg/g.^9–12 After candidates were identified, the pharmacist completed an additional previsit screening via review of the electronic medical record (EMR) to select patients who were likely to adhere to instructions and have low risk for common side effects of SGLT2 inhibitors (Supplemental Table 1).

Figure 2.

Participant enrollment.

Abbreviations: T2DM, type 2 diabetes mellitus; UACR, urine albumin to creatinine ratio; eGFR, estimated glomerular filtration rate; SGLT2i, sodium-glucose cotransporter-2 inhibitor; LUTS, lower urinary tract symptoms.

^aThis patient had severe LUTS indicated by an International Prostate Symptom Score (I-PSS) of 20 out of 35.

^bThis patient had a Montreal Cognitive Assessment score of 22 out of 30 indicating mild cognitive impairment.

Study Design and Clinic Workflow

Prior to each clinic session, the interprofessional team conducted a huddle where pharmacists and nephrology providers discussed eligible empagliflozin candidates with scheduled clinic visits that day. A candidate patient was seen initially by the nephrology provider who introduced the patient to the potential benefits and risks of empagliflozin initiation. If the patient was interested in initiating empagliflozin, the patient subsequently had an in-person visit with the pharmacist in the same clinic space. The pharmacist completed diabetes and blood pressure assessments as well as a patient interview, with a guided note template, to collect the following key information:

1. current diabetes and blood pressure medication regimen;

2. home monitoring of blood sugar and blood pressure;

3. hypoglycemic events in the last year and hypoglycemic management; and

4. medication management (ie, adherence, assistance with medications, etc)

The pharmacist then provided the patient with a written clinical plan outlining empagliflozin use and a description of the mechanism of action in relation to side effects and monitoring parameters. The written clinical plan also noted any changes made to the patient’s current diabetes, hypertension, or diuretic regimens; the date of the patient’s next in-person nephrology clinic visit; and the date and time of the pharmacist’s first follow-up phone call with the patient. The written clinical plan was reviewed verbally and given to the patient, and the patient was provided the opportunity to ask questions. Finally, the pharmacist documented the clinical assessment and plan in a templated note in the EMR. The nephrology provider prescribed empagliflozin 10 mg by mouth once daily and any additional prescriptions or monitoring devices recommended during the visit. The patient’s primary care provider and diabetes management team were notified via secure e-mail that the patient had been initiated on empagliflozin and would receive intensive pharmacist follow-up for a 6-month period. The e-mail included the plan for pharmacist phone follow-up and contact information for the pharmacist and attending nephrologist for any questions or concerns that might arise.

Two weeks after initiation, the clinical pharmacist completed a telephone follow-up assessment to monitor the efficacy and safety of empagliflozin and documented the assessment in the EMR. If the patient was tolerating the regimen and blood glucose and blood pressure readings were appropriate, no changes were made. If the clinical pharmacist determined that adjustments were required, these recommendations were discussed with the nephrology provider; the nephrology provider then placed prescription orders for any diabetes, hypertension, or diuretic medications as necessary. The pharmacist coordinated the mailing of prescriptions as well as a new written clinical plan each time medication changes were made. A second telephone follow-up adhering to the same process was performed by the pharmacist at 6 weeks. The pharmacist performed telephone follow-up every 2 to 4 weeks thereafter as necessary based on clinical judgement. Once the patient was on a stable diabetes regimen for at least 1 month without adverse events (ie, hypoglycemia, hypotension, polyuria), the pharmacist discontinued telephone follow-up calls. Instead, the pharmacist conducted in-clinic follow-up as needed during scheduled nephrology clinic visits which occurred every 3 to 6 months. Additional in-clinic visits, earlier than 3 months, were arranged if needed. In-clinic follow-up consisted of a visit with the nephrology provider for assessment of kidney disease, laboratory monitoring, and a visit with the pharmacist for medication reconciliation as well as diabetes management. We collected data for every patient over at least 6 months; however, some patients may have required intensive pharmacist follow-up beyond 6 months due to further required changes to their diabetes regimens (Supplemental Table 1).

Outcomes

The primary outcome was efficacy of empagliflozin and included changes to renal parameters such as UACR and eGFR and changes to the hemoglobin A_1c (HbA_1c). Secondary outcomes included medication changes to diabetes, hypertension and/or diuretic regimens, safety, and feasibility. Safety outcomes included the incidence of hypoglycemia, hypotension, polyuria, genitourinary infection, acute renal failure (ARF), bone fractures, amputations, ketoacidosis, and/or ketonuria. Hypoglycemia was defined as a blood glucose level <70 mg/dL or symptoms of hypoglycemia requiring self-treatment per patient self-report. Hypotension was defined as a blood pressure <90/60 mm Hg or symptoms of hypotension at baseline which included incidence within the past month per patient self-report. The definition of ARF was based on the Standardized MedDRA Queries or an increase in serum creatinine greater than or equal to 0.5 mg/dL and an increase in blood urea nitrogen greater than or equal to 10 mg/dL.⁹ Laboratory monitoring was completed at the time of empagliflozin initiation, at 3 months, and at 6 months (±1 month) after empagliflozin initiation. Feasibility was determined by time spent by the pharmacist during initial and follow-up visits as well as the number of patients who were on a stable regimen at 6 months post empagliflozin initiation. All data were collected through chart review of clinical notes, VA prescription refill data, and laboratory results. A Fisher’s exact test was performed to compare the incidence of adverse effects at 3 and 6 months in those with an eGFR <45 mL/min/1.73 m² compared to ≥45 mL/min/1.73 m².

Results

Patients

Fourteen patients met the inclusion criteria and were initiated on empagliflozin (Table 1). Patients had a heavy comorbidity burden, polypharmacy,⁴⁰ and received outpatient medications from multiple prescribers.

Table 1.

Baseline Characteristics.^a

Age, mean years ± SD	67 ± 8
Male, n (%)	14 (100)
Number of medications, mean ± SD	17 ± 6
Number of outpatient prescribers in the last year, mean ± SD	14 ± 6
Incidence of hyperkalemiab	8 (57)
Comorbid conditions, n (%)
Hypertension	14 (100)
Dyslipidemia	14 (100)
Clinical ASCVD	8 (57)
10-year ASCVD risk score > 10%c	6 (43)
Obese (body mass index > 30 kg/m2)	7 (50)
Atrial fibrillation	4 (29)
Heart failure	2 (14)
Patients on a diabetes medication, n (%)	13 (93)
Insulin, basal only	5 (36)
Insulin, basal-bolus regimen	4 (29)
Metformin	8 (57)
Sulfonylurea	4 (29)
DPP-4 inhibitor	1 (7)
GLP-I receptor agonist	1 (7)
Patients on an antihypertensive medication, n (%)	14 (100)
Beta blocker	11 (79)
ACE inhibitor/ARB	9 (64)
Calcium channel blocker	8 (57)
Direct vasodilator (hydralazine)	1 (7)
Alpha blocker	1 (7)
Patients on a diuretic medication, n (%)	6 (43)
Loop diuretics	3 (21)
Thiazide diuretics	3 (21)

ACE, Angiotensin converting enzyme inhibitor; ARB, Angiotensin receptor blocker; ASCVD, atherosclerotic cardiovascular disease; DPP-4, dipeptidyl peptidase-4; GLP-1, glucagon-like peptide-1; SD, standard deviation.

^an = 14.

^bDefined as a serum potassium >5.5 mEq/L, baseline included incidence within the past 5 years.

^cCalculated using the American College of Cardiology ASCVD Risk Estimator Plus or Pooled Cohort Equation, as appropriate, for those who did not have a history of clinical ASCVD. All of the 6 patients without clinical ASCVD had an ASCVD risk score >10%.^38,39

Renal Parameters

At 6 months, the mean UACR improved on average from 1838 mg/g to 1138 mg/g (mean%change −12% ± 61%). Almost half of the included patients (n = 6, 43%) achieved greater than 30% reduction in UACR at both 3 and 6 months after empagliflozin initiation. The mean % change was greater in patients with a higher baseline UACR. The eGFR at 6 months declined slightly in half of the patients (n = 7, 50%) by an average of −2 ± 8mL/min/1.73 m² at 3 months and an average of −0 ± 12 mL/min/1.73 m² at 6 months. Three patients had an eGFR decrease to below 35 mL/min/1.73 m² within the initial 3-month monitoring period; however, the eGFRreturned to above 35mL/min/1.73m² in all 3 patients following adjustments to other implicated medications (ie, diuretics or other antihypertensives; Table 2).

Table 2.

Results: Changes From Baseline After 3 and 6 Months of Empagliflozin Therapy.^a

	Baseline	3 Months	Mean % Change ± SD, Baseline to 3 Months	6 Months	Mean % Change ± SD, Baseline to 6 Months
Renal Results: UACR
UACR, mg/g, n (%)
<300b	3 (21)	4 (29)	−11 ± 24	5 (36)	−23 ± 43
300–999	4 (29)	5 (36)	−6 ± 42	3 (21)	−4 ± 85
1000–3499	5 (36)	4 (29)	−4 ± 88	6 (43)	+3 ± 52c
>3500	2 (14)	1 (7)	−43 ± 20	0 (0)	-
Overall mean UACR, ± SD	1838 ± 2013	1270 ± 1236	−12 ± 60	1138 ± 1023	−12 ± 61
Achieved reduction > 30%, n (%)	-	6 (43)	-	6 (43)	-
Achieved reduction > 50%, n (%)	-	4 (29)	-	5 (36)	-
Renal Results: eGFR
eGFR, mL/min/l.73 m2, n (%)
>60	4 (28.5)	3 (21)	−5 ± 14	2 (14)	+13 ± 9
45–60	4 (28.5)	5 (36)	−3 ± 4	7 (50)	−1 ± 14
35–44	6 (43)	6 (43)	+0 ± 5.0	5 (36)	−3 ± 3
Mean eGFR, mL/min/1.73 m2, ± SD	56 ± 21	54 ± 22	−2 ± 8	55 ± 25	+0 ± 12
Glycemic results
Hemoglobin A1c (%), n (%)
< 7.0	9 (65)	8 (57)	+0.2 ± 0.4	5 (36)	+0.2 ± 0.4
7.1–8.0	3 (21)	4 (29)	−0.2 ± 0.3	6 (43)	+0.4 ± 0.6
8.1–9.1	2 (14)	2 (14)	−0.1 ± 0.1	3 (21)	+0.5 ± 0.7
Mean hemoglobin A1c (%), ± SD	7.0 ± 0.8	7.1 ± 0.7	+0.1 ± 0.4	7.4 ± 0.9	+0.3 ± 0.6
At or below hemoglobin A1c goal, n (%)d	11 (79)	11 (79)	-	8 (57)	-
Average weight, kg, mean ± SD	95.7 ± 15.6	89.9 ± 14.1	−2.7 ± 2.2	92.1 ± 16.1	−3.6 ± 4.5
Adverse Events	Baseline	3 Months		6 Months
History of hypoglycemia, n (%)e	3 (21)	5 (36)	-	4 (29)	-
Incidence of hypotension, n (%)f	5 (33)	6 (43)	-	2 (14)	-
Incidence of polyuria, n (%)	2 (14)	7 (50)	-	4 (29)	-
Incidence of acute renal failure, n (%)g	2 (14)	2 (14)	-	1 (7)	-

Abbreviations: UACR, urine albumin-to-creatinine ratio; eGFR, estimated glomerular filtration rate; ACE inhibitor, Angiotensin converting enzyme inhibitor; ARB, Angiotensin receptor blocker; SD, standard deviation.

^an = 14.

^bFor study inclusion, patients were required to have a UACR >300 mg/g in the past year; however, some patients may have subsequently updated laboratory test results and therefore had a UACR <300 at baseline due to other agents that lower UACR (ie, ACE inhibitors or ARBs).

^cOne patient experienced an acute kidney injury at 6 months due to over-diuresis (+103% increase in UACR), thus skewing the data.

^dDerived from the American Diabetes Association 2019 guidelines.^2,16

^ePer self-report, a blood glucose <70 mg/dL or symptoms of hypoglycemia requiring self-treatment, baseline included incidence within the past year.

^fPer self-report, a blood pressure <90/60 mm Hg or symptoms of hypotension, baseline included incidence within the past month.

^gAs defined by the KDIGO Guidelines,³⁶ baseline included incidence within the past year.

Glycemic Control

After 3 months of empagliflozin therapy, the same number of patients had an HbA_1c at goal compared to baseline^2,18 (n = 11, 79%), with an overall mean change in HbA_1c of +0.1% ± 0.4%. However, at 6 months, less patients had an HbA_1c at or below goal (n = 8, 57%). Patients reported an average weight loss of 3.6 kg (range: +7 to −12 kg) over 6 months (Table 2).

Medication Changes

Post empagliflozin initiation, almost all patients required adjustments to their diabetes, hypertension, or diuretic regimens (n = 13, 93%) over the 6-month period. Multiple adjustments took place over the course of the 6 months including uptitration and downtitration of various medications. There were 9 changes to noninsulin diabetes medications. All patients on insulin at baseline required dose adjustments; the total daily dose of insulin was reduced on average by 35% at 6 months. Diabetic medication changes were predominantly deprescribing of agents with high hypoglycemia risk.^2,18 There were 9 changes to antihypertensive agents and 4 changes to diuretics over 6 months. After 3 months of empagliflozin therapy, patiromer, a medication to decrease serum potassium, was discontinued in 2 of 2 patients who were taking the medication at baseline, and they were able to remain off of this medication at 6 months (Supplemental Table 2).

Adverse Drug Reactions

At 3 months, 5 (36%) patients reported hypoglycemia, 6 (43%) reported hypotension, and 7 (50%) reported polyuria. The incidence of all 3 decreased at 6 months to 4 (29%), 2 (14%), and 4 (29%). ARF occurred in 2 (14%) patients within 3 months. The first case of ARF occurred in a patient with an eGFR greater than 60 mL/min/1.73 m² in the setting of a pneumonia infection. The second occurred in a patient with an eGFR between 35 to 44 mL/min/1.73 m² as a result of overdiuresis, prompting the need for a diuretic dose reduction. ARF occurred in 1 (7%) patient within 6 months; this incidence of ARF occurred following a bilateral nephrostomy tube exchange. In all patients who experienced ARF, kidney function returned to baseline. There were no significant differences in the incidence of hypoglycemia, hypotension, polyuria, or ARF in those with eGFR <45 mL/min/1.73 m² compared to those with eGFR >45 mL/min/1.73 m² at 3 or 6 months (P values all >.05). Genital infections, bone fractures, amputations, ketoacidosis, and ketonuria were not observed for any of the included patients (Table 2).

Feasibility

On average, the pharmacists spent 101 minutes per patient for the initial in-clinic visit: 20 minutes for previsit safety screening, 36 minutes spent in clinic with the patient, and 45 minutes for documentation and provider communication. Subsequent telephone follow-up encounters (average 2.3 calls per patient, range 1–4 calls) required less time (average 27 minutes spent with patient and 30 minutes for documentation and communication). In-clinic follow-up required an average of 56 minutes per patient: 36 minutes with patient and 20 minutes for documentation and communication. For all pharmacist activities over the initial 3-month period, each patient required an average of 4.8 hours of pharmacist time (total time for n = 14 patients, 67 hours over 3 months). At 3 months post empagliflozin initiation, 9 (64%) patients were on a stable diabetes regimen without hypoglycemia and did not require additional intensive pharmacist phone follow-up beyond that time point. This increased to 10 (71%) patients at 6 months.

Discussion

To our knowledge, our model is the first to describe specific strategies for initiation and monitoring of SGLT2 inhibitors in the DKD population outside clinical trials. Our interprofessional model provided safe and effective empagliflozin initiation in a group of real-world patients with DKD and complex medical conditions. Our model aligns with the 2019 American Diabetes Association (ADA) guidelines that recommend SGLT2 inhibitors as the second medication to be added, after metformin, in patients with diabetes and established ASCVD or CKD.^2,18 In our cohort, 57% had clinical ASCVD,^41,42 and the remaining 43% had an estimated 10-year ASCVD risk score >10%^41,42; all patients had CKD. The need for changes to diabetes, hypertension, and diuretic regimens during the 6-month period following empagliflozin initiation emphasized the importance of systematic clinical monitoring. Based on our observations, we attempt to provide insights into best practices for how to safely initiate and monitor SGLT2 inhibitors in this vulnerable population.

A 6-month time period may not have been long enough to detect meaningful improvements in HbA_1c as seen in other models.^37–39 In line with the ADA guidelines, our model promoted initiation of agents with cardiovascular (CV) or renal benefit and discontinuation of agents without benefit, even in patients with an HbA_1c at or below goal.^2,18 Particularly, we were able to deprescribe sulfonylureas, which are the second most commonly prescribed diabetes medication in CKD⁴³ but have a high risk of hypoglycemia,⁴⁴ lack renal benefits, and confer increased cardiovascular mortality.^45,46 Prescription of empagliflozin also led to discontinuation of bolus insulin, further reducing the risk of hypoglycemia.^18,47 While patients required individualized insulin dose reductions depending on their baseline insulin requirements and blood glucose control, the results seen in our cohort suggest a preemptive 20% to 30% dose reduction in total daily insulin requirements may help prevent hypoglycemia when initiating empagliflozin.

UACR reduction greater than 30% in a 2-year period is associated with 20% to 30% relative risk reduction of incident ESRD and more than 1% absolute risk reduction in 10-year risk of ESRD.⁴⁸ We can reasonably assume that our observation of proteinuria reduction in a short period of time will likely improve future renal outcomes. Clinical trials also demonstrate that patients with an eGFR between 30 and 45 mL/min/1.73 m² experience similar renal and cardiac benefits without an increase in adverse events compared to those without kidney disease. Our cohort included 6 patients with an eGFR between 35 and 44 mL/min/1.73m² at baseline; these patients had the greatest magnitude of UACR reduction in the 6-month period without significant differences in the incidence of adverse effects.

In line with results from clinical trials, our patients required changes to their antihypertensive medications including dose reductions and discontinuation.^10,49 SGLT2 inhibitors cause natriuresis resulting in reduced blood pressure. SGLT2 inhibitors also induce diuretic synergy with loop diuretics resulting in potential overdiuresis. In our patients, the mean reduction in loop diuretics at 6 months was 77% (range −50% to −100%). We propose close monitoring and consideration for an empiric 50% dose reduction in loop diuretics to minimize the risk of overdiuresis.⁵⁰ The diuretic synergy of loop diuretics and SGLT2 inhibitors may be beneficial in those with some degree of diuretic resistance and warrants further investigation.⁵⁰ Also noteworthy, empagliflozin decreases UACR without raising serum potassium. This may enable reinitiation and tolerability of an ACE inhibitor or ARB for further proteinuria reduction and also may allow for deprescribing of patiromer in patients with hyperkalemia.^9,12

All patients tolerated empagliflozin throughout the 6-month initiation and monitoring period. The most common adverse event was polyuria which may be anticipated and managed with close monitoring and proper patient education.²⁸ Many patients who experienced polyuria anecdotally reported that the polyuria was transient and manageable. Due to the glucose-dependent mechanism of action of SGLT2 inhibitors, there is little to no risk for hypoglycemia when used as mono-therapy; however, in patients on other diabetes medications, including agents with a high risk of hypoglycemia, the risk increases.¹ In our cohort, we observed an increase in the incidence of hypoglycemia at 3 and 6 months, highlighting the importance of prudent blood sugar monitoring and downtitration of antidiabetic agents with a high risk of hypoglycemia, both activities completed by the clinical pharmacist in our model. Four (80%) of 5 patients who experienced hypoglycemia were also on insulin. Most cases of hypoglycemia were singular, asymptomatic blood glucose readings below 70 mg/dL observed in the early stages of empagliflozin initiation while implementing insulin dose reductions. There was a small increase in the incidence of hypotension which subsequently prompted reductions or discontinuation of blood pressure medications; this was in line with the current literature.²⁹ Our results generally support the safety of using empagliflozin in patients with eGFR between 35 and 44 mL/min/1.73m² while emphasizing the need for careful monitoring.

At 6 months, 4 (29%) patients continued to require pharmacist follow-up for diabetes management, all due to insulin titration. We did not conduct a cost–benefit analysis for our model. However, we propose the following financial considerations. We spent approximately 67 hours of pharmacist time to initiate 14 patients with DKD on empagliflozin, which costs $63 per hour (national average hourly pay rate of clinical pharmacists in 2017) 67 hours = $4221. According to the EMPA-REGOUTCOME [ESRD] trial, the number needed to treat was 14 for the renal composite outcome.^9,12 We argue that by conservative estimation, we could prevent one CKD3 to CKD4 progression, which translates to an annual cost saving of $9300 in 2010 US dollars,⁵¹ in addition to the potential for other cost savings by preventing hypoglycemia requiring emergency department visits, HF admissions, and incident ESRD.⁷ Our model defines the role and demonstrates the feasibility of clinical pharmacists in the CKD clinic and suggests the great potential for cost savings by preventing the progression of DKD to ESRD.⁷ These types of initiatives are being called for on a national level within the nephrology community.⁷

Our pilot study was limited by a small sample size and short follow-up duration which limits the ability to draw long-term conclusions on the cardiovascular and renal protective effects of SGLT2 inhibitors.¹⁶ We carefully screened patients for study inclusion to reduce the risk of SGLT2 adverse effects. Future studies should be conducted in larger populations with comparator groups to study scalability and comparative efficacy. Finally, a formal cost–benefit analysis of our care model has yet to be defined.

Nonetheless, the complexity of caring for patients with DKD was underscored by our small group of patients who required intensive follow-up and multiple medication changes to ensure patient safety.²⁹ The patients in our pilot may represent the most suitable candidates for SGLT2 initiation by the nephrologist in the CKD clinic. The nature of our collaborative model supported the successful initiation of SGLT2 inhibitors, allowing for frequent follow-up and monitoring which may have been too time intensive for nephrology providers alone. We believe that nephrologists and pharmacists offer a unique partnership for SGLT2 inhibitor initiation in patients with DKD, monitoring for medication adverse effects and performing subsequent adjustments to diabetes, hypertensive, and diuretic regimens. We hope that our results provide insights for guiding how to safely initiate and monitor SGLT2 inhibitors in patients who may benefit greatly from the renal and cardiovascular protective effects.

Conclusion

The implementation of an innovative, interprofessional care model within a nephrology clinic for the initiation and monitoring of empagliflozin over 6 months in patients with DKD demonstrated clinical benefit with minimal safety concerns. This model is a feasible approach for providers interested in partnering with pharmacists to align with current diabetes guidelines and improve patient care. While SGLT2 inhibitors remain highly efficacious in preventing and improving renal outcomes, they remain complex medications that require thorough monitoring and follow-up to ensure safe use.