36-OR: Comparative Effectiveness of SGLT2 Inhibitors vs. GLP-1 Agonists

Abstract

This retrospective cohort study compared sodium-glucose co-transporter-2 (SGLT2) inhibitor and glucagon-like peptide-1 (GLP-1) agonist therapeutic drug classes on cardiovascular (CV) outcomes, treatment persistence/discontinuation, healthcare utilization and costs. Using medical and pharmacy claims data from the Humana Research Database, new users of SGLT2 inhibitors or GLP-1 agonists from January 2015 to June 2017 were identified. Eligible patients were 19-89 years, enrolled in a Medicare Advantage Prescription Drug or fully-insured commercial health plan, with at least 12 months of continuous enrollment pre- and post-index date (i.e., date of first new prescription). Propensity score matching (1:1) was used to balance baseline characteristics. Cox proportional hazards models were used to compare CV outcomes and treatment discontinuation. Healthcare resource utilization and costs were compared using Wilcoxon signed rank test. A total of 5,507 patients were included in each treatment group. Patients indexed to use GLP-1 agonists were more likely to discontinue treatment (hazard ratio 1.15, 95% confidence interval 1.10-1.21) and more likely to encounter an inpatient hospitalization (14.4% vs. 11.9%, P<0.001) or emergency department visit (27.4% vs. 23.5%, P <0.001) compared to patients indexed on SGLT2 inhibitors. The risk of primary composite CV outcome (myocardial infarction or stroke or mortality) and secondary CV outcome (heart failure or death) was similar across cohorts. The average per-person per-month total cost ($1,796 vs. $1,617, P<0.001), medical ($904 vs. $834, P<0.001) and pharmacy costs ($891 vs. $783, P<0.001) were higher for the GLP-1 agonist cohort compared to the SGLT2 inhibitor cohort. Although differences in extra-glycemic CV benefits were not established, these findings may be useful to providers and payers, as this real-world data showed SGLT2 inhibitors were associated with longer treatment persistence and reduced healthcare utilization and costs relative to GLP-1 agonists. Disclosure I. Poonawalla: None. A. Bowe: Employee; Self; Humana. M. Tindal: None. Y.A. Meah: None. P. Schwab: Employee; Self; Humana. Employee; Spouse/Partner; Humana.