OBJECTIVES: We assessed the real-world adherence of patients treated with sitagliptin between 2014 and 2016 over a 2-year period using electronic drug dispensing records.

METHODS: This study was carried out on a 1/97^th sample of the French National Healthcare System Claims Databases (i.e. the EGB database). Beneficiaries with at least 3 dispensations of sitagliptin from 01/01/2014 to 31/12/2016 were followed for 2 years, or until death or treatment withdrawal. Adherence was assessed through the Medication Possession Ratio (MPR) - with a maximum set at 100% - and the Proportion of Days Covered (PDC). The following parameters have been defined: period covered by a dispensation; grace period (i.e. uncovered period without considering discontinuation): 7 days; minimum timeframe to consider a withdrawal: 180 days.

RESULTS: A total of 4,406 patients were included in the study population, aged 68.1 ± 42.1 years. The median exposure time available for the study population was 728 days [Q1-Q3: 394-730; mean: 572 ± 230]. During that period, 86% of patients (n = 3,803) had at least one discontinuation, with an average number of 5 ± 4 discontinuations per patient and an average duration of 22 ± 17 days. Median adherence was estimated (i) at 94% according to the MPR [Q1-Q3: 82-100], i.e. 76% of adherent patients (MPR ≥ 80%), and (ii) at 86% [Q1-Q3 : 75-92] according to the PDC, i.e. 68% of adherent patients (PDC ≥ 80%).

CONCLUSIONS: Almost 3/4 of patients are adherent to sitagliptin with MPR, and 2/3 with PDC. According to the literature, the MPR method would lead to an overestimation of adherence due to double counting of days covered by two dispensations; external validation is necessary to verify these hypotheses. In addition, our results raise the question of frequent discontinuations in this pathology where daily glycaemic control plays a major role in preventing complications.