Active surveillance of drug safety in healthcare data: Sequential monitoring of bacterial and serious urinary tract infection risk in sodium‐glucose cotransporter 2 inhibitor users

INTRODUCTION: Active surveillance of adverse events using healthcare data is emerging as complementary to the monitoring of spontaneous reports and stand-alone pharmacoepidemiologic studies. The risk of urinary tract infections (UTIs) was listed as a special warning for sodium-glucose cotransporter-2 inhibitors (SGLT2is) when marketed in Europe in 2012. This study tested the feasibility of active surveillance, continuously re-evaluating the UTI risk following approval. METHOD: This sequential cumulative cohort study emulated surveillance of UTI among type 2 diabetes patients initiating SGLT2is using Danish healthcare data (2007-2022). Outcomes were bacterial UTI-related prescriptions and UTI-related hospital contacts. Dipeptidyl peptidase-4 inhibitors (DPP-4is) and glucagon-like peptide-1 receptor agonists (GLP-1 RAs) were active comparators. Annual incidence rates (IRs) and propensity score weighted hazard ratios (HRs) were estimated. RESULT: The 198 613 eligible initiation episodes were included. Among SGLT2is, the UTI-related prescription IRs were in the range 140.9-200.9 per 1000 person years, declining steadily after 2014. IRs of hospital contacts ranged between 11.8 and 21.7 per 1000 person years, initially decreasing but rising again after 2018. The earliest conclusive risk estimate was observed in 2016, indicating no increased UTI risks of interest associated with SGLT2i use. HRs for bacterial UTI-related prescriptions remained under 1 over time in both comparisons, while the HR for serious UTI events varied with follow-up time. CONCLUSION: This is the first study demonstrating the feasibility of continuous surveillance for evaluating real-world UTI risk among SGLT2i users without aiming to establish causality. The IR of serious events was low, and stable HRs indicate no increased risk from 2016.