Abstract 24: Diabetic Ketoacidosis in Patients with Type 2 Diabetes Treated with Sodium Glucose Co-transporter 2 Inhibitors versus Other Antihyperglycemic Agents: An Observational Study of Four US Administrative Claims Databases

Abstract

Introduction: Diabetic ketoacidosis (DKA) is a serious acute metabolic complication of diabetes. Rare DKA events have occurred in patients taking sodium glucose co-transporter 2 inhibitors (SGLT2i). This study evaluated the risk of DKA in patients with type 2 diabetes mellitus (T2DM) taking SGLT2i versus other antihyperglycemic agents (AHAs) in clinical practice. Methods: This study, per protocol reviewed and approved by the European Medicines Agency, identified patients from 4 large US claims databases using broad and narrow definitions of T2DM; the broad definition captured all patients with a T2DM diagnosis and the narrow definition was intended to exclude T1DM misclassified as T2DM. DKA was identified from diagnosis codes in inpatient or emergency room claims. Eligible new users of SGLT2i and 7 groups of AHA comparators were matched (1:1) on exposure propensity scores (PS) to adjust for imbalances in baseline covariates. Cox proportional hazard models conditioned on PS-matched pairs were used to estimate hazard ratios (HR) of DKA risk for new users of SGLT2i versus other AHAs. P values were calibrated using negative control outcomes to address potential residual bias. Pooled HRs were calculated when I 2 was <40% across 4 databases. Results: The number of new users of SGLT2i in each database ranged from 11,141 to 152,728 using the broad T2DM definition and from 7,779 to 130,708 using the narrow definition. Across databases, the unadjusted incidence rates of DKA (events per 1000 patient-years) ranged from 2.75 to 8.84 with SGLT2i and 1.38 to 15.82 with other AHAs using the broad T2DM definition and from 1.15 to 3.91 with SGLT2i and 0.75 to 7.94 with other AHAs using the narrow definition. Using the broad T2DM definition, a significantly increased risk of DKA was observed among new users of SGLT2i versus 5 groups of other AHAs; when using the narrow definition, an increased risk of DKA with SGLT2i was observed only compared with sulfonylureas ( Figure ). Conclusion: In this claims database study, an increased risk of DKA was observed for new users of SGLT2i versus new users of several non-SGLT2i AHAs when T2DM was defined broadly. When T2DM was defined narrowly to exclude possible misdiagnosed T1DM patients, an increased risk of DKA with SGLT2i was observed compared to sulfonylureas.