Sodium-Glucose Co-Transporter 2 Inhibitors and the Risk of Genitourinary Infections at HbA1c ≥10%: A Population Health-Based Retrospective Review

Abstract

Background: Sodium-glucose co-transporter 2 inhibitors (SGLT2i) are first-line treatment for type 2 diabetes. Evidence has shown a 3- to 5-fold increase in the risk of genitourinary infections with their use due to inhibition of renal glucose reabsorption, resulting in glucosuria. Increased glucosuria is thought to increase the risk of genitourinary infections at a greater degree in patients with a significantly elevated HbA1c (≥10%), and initiation of SGLT2i is often delayed in these patients. While a limited body of evidence exists indicating that A1c level is not an independent risk factor for SGLT2i-induced genitourinary infection, pragmatically this concern remains a barrier to SGLT2i utilization. Objective: Evaluate the real-world genitourinary (GU) infection rate in patients receiving SGLT2i with a baseline HbA1c ≥10% compared to patients with a baseline HbA1c <10%. Methods: This retrospective cohort study evaluated data from 5542 adult patients treated between January 2013 and January 2023, who were prescribed an SGLT2i. Data collected included sex, age, race/ethnicity, renal function, date of SGLT2i start, number of SGLT2i orders, name and dose of SGLT2i, HbA1c, and a predetermined set of diagnosis codes related to bacterial and fungal genitourinary infections. The primary outcome was the overall GU infection rate after SGLT2i initiation within groups of baseline HbA1c of ≥10% and <10%, and the secondary outcome was total GU infections within these same groups. Results: The primary outcome was equivalent between those with HbA1c <10% and HbA1c ≥10% (0.0064 ± 0.0565 vs 0.0030 ± 0.0303 infection per month [mean ± standard deviation]; P < 0.0001 for both lower and upper bounds). There was no statistically significant difference in total GU infections between the same groups (0.027 ± 0.21 vs 0.015 ± 0.14, P = 0.11). Female gender and prior recurrent infection were associated with increased GU infection after SGLT2i. Conclusion and relevance: A baseline HbA1c ≥ 10% was not significantly associated with an increased risk of GU infection following the initiation of SGLT2i compared to those with a baseline HbA1c of <10%.