Adherence, persistence, and treatment discontinuation with sitagliptin compared with sulfonylureas as add-ons to metformin: A retrospective cohort database study.

Bloomgarden, Zachary T; Tunceli, Kaan; Liu, Jinan; Brodovicz, Kimberly G; Mavros, Panagiotis; Engel, Samuel S; Radican, Larry; Chen, Yong; Rajpathak, Swapnil; Qiu, Ying; Brudi, Philippe; Fonseca, Vivian

BACKGROUND Data are limited regarding adherence to dipeptidyl peptidase-4 inhibitors. METHODS The present retrospective cohort study of a claims database involved adults with type 2 diabetes mellitus, continuous enrollment for 12 months before the first prescription of add-on sitagliptin (SITA) or a sulfonylurea (SU) to metformin (MET) monotherapy (index date), and ≥45 days of MET coverage ≤90 days before the index date. The SITA and SU users were matched on duration of follow-up and propensity score (PS). Logistic regression analysis incorporated age, gender, comorbidities, and concomitant medications as independent variables. RESULTS Approximately 99 % of SITA patients were PS matched, resulting in 14 807 well-balanced PS-matched SITA/SU pairs. Mean proportion of days covered (PDC) was significantly higher for SITA (vs SU) + MET after 1 year (P < 0.001). Adherence (PDC ≥80 %) to SITA (vs SU) + MET was 59.1 % (vs 55.9 %; P < 0.001) at 1 year and 52.6 % (vs 49.9 %; P = 0.007) at 2 years. Using logistic regression models including out-of-pocket expense (OPE) as a covariate, we found improved mean PDC and adherence for SITA (vs SU) + MET. Numbers of patients who continued to use SITA (vs SU) + MET were significantly higher after Years 1, 2, and 3 (all P < 0.05). CONCLUSIONS Users of SITA + MET had significantly higher mean PDC, adherence, and persistence than those on SU + MET. These trends were robust to model alterations and were more marked when accommodating OPEs.