OBJECTIVES: To use group-based trajectory modeling (GBTM) to identify clusters of patients with multiple sclerosis (MS) with similar patterns of adherence to once- or twice-daily oral disease-modifying drugs (DMDs) over time.

METHODS: Eligibility criteria for this IQVIA™ RWD Adjudicated Claims - US data study were: ≥2 MS diagnoses (ICD-9 CM code 340.xx and ICD-10 CM code G35) between 7/1/2013–12/31/2016, ≥1 once-/twice-daily oral DMD claim between 7/1/2013–12/31/2016 (index), continuous 1-year eligibility with commercial insurance before/after index, no baseline oral DMD, and age 18-64 years. Using the GBTM approach, individuals who followed similar longitudinal progressions of adherence (proportion of days covered [PDC] over time) were identified and “clustered” together. The optimal number of groups best representing the heterogeneity in developmental trajectories was selected considering Extract Log-Likelihood, Akaike Information Criterion, Bayesian information criterion, and Bayes factor.

RESULTS: A total of 7,689 patients met eligibility criteria; mean (SD) age 45.2 (10.2) years, 75.9% female. GBTM revealed three distinct groups of adherence over time: an ‘Immediately Non-Adherent’ group (PDC ≤0.2 by Month 3), a ‘Gradually Non-Adherent’ group (PDC >0.2 at Month 3 but ≤0.4 by Month 7), and an ‘Adherent’ group (PDC ≥0.8) comprising 14.9%, 19.5%, and 65.6% of patients, respectively. Factors predictive of membership in the Immediately Non-Adherent vs. Gradually Non-Adherent group were higher age, female sex, West vs. Northeast US, chronic lung disease, twice- (vs. once-) daily dose, and switching. Factors predictive of membership in the Gradually Non-Adherent vs. Adherent group were lower age, female sex, South/Midwest/Northeast vs. West US, depression, switching, and 90-day pre-index relapse requiring an emergency room visit.

CONCLUSIONS: This trajectory analysis in patients with MS receiving once-/twice- daily oral DMDs revealed three distinct longitudinal patterns of adherence associated with specific patient characteristics and healthcare utilization patterns. Patients in different groups may warrant different types of clinical interventions to address nonadherence.