Adherence and persistence in patients with type 2 diabetes mellitus newly initiating canagliflozin, dapagliflozin, dpp-4s, or glp-1s in the United States

Objective: Sodium-glucose co-transporter 2 inhibitors were first approved in the US in 2013; therefore, real-world (RW) studies describing outcomes are limited. This retrospective study evaluated adherence and persistence among patients initiating canagliflozin (CANA), dapagliflozin (DAPA), GLP-1 agonists (GLP-1s), and DPP-4 inhibitors (DPP-4s) over a 12-month follow-up from a US managed care perspective. Methods: Patients newly initiating CANA, DAPA, GLP-1s, or DPP-4s from February 1, 2014-June 30, 2014 were identified from the QuintilesIMS PharMetrics Plus Database. The first fill defined the index date/drug. Patients were required to have a T2DM diagnosis (ICD-9-CM 250.x[0,2]) and 12 months of continuous enrollment pre- and post-index (follow-up). Main outcome measures were adherence (proportion of days covered, PDC; medication possession ratio, MPR) and persistence on index therapy. PDC or MPR 0.80 was considered adherent. Patients were considered persistent until evidence of discontinuation (gap 90 days between two subsequent index therapy prescriptions). Kaplan-Meier (KM) analysis assessed time to discontinuation, while a Cox proportional hazards model (PHM) evaluated risk of discontinuation. Logistic regression models evaluated the likelihood of non-adherence. Results: The final sample consisted of 23,702 patients (6,546 CANA, 3,087 DAPA, 6,273 GLP-1s, and 7,796 DPP-4s; 56% male, and mean [SD] age=55 [9.1] years). Mean PDC ranged from 0.56 (GLP-1), to 0.71 (CANA), with 33-56% adherent, respectively; MPR results were similar. Fifty-two per cent (GLP-1) to 68% (CANA) were persistent over the follow-up. CANA patients had the longest time to discontinuation. In regression analyses, compared to CANA 100mg, DAPA, DPP-4, and GLP-1 patients had a significantly higher likelihood of non-adherence and a significantly higher risk of discontinuation. CANA 300mg patients had a significantly lower likelihood of non-adherence and a significantly lower risk of discontinuation compared to CANA 100mg. Conclusions: Adherence and persistence were significantly better with CANA (100mg and 300mg) compared to DAPA, GLP-1s, and DPP-4s in the RW setting.