Assessment of Adherence to and Persistence on Disease-Modifying Antirheumatic Drugs (DMARDs) in Patients With Rheumatoid Arthritis

OBJECTIVE: Biologic disease-modifying antirheumatic drugs (DMARDs) are efficacious for treating rheumatoid arthritis (RA). However, measurements of relative effectiveness, including treatment adherence and persistence, are lacking. We evaluated adherence and persistence during new episodes of use of traditional and biologic DMARDs. METHODS: Using Tennessee Medicaid databases (1995-2004), we assembled a retrospective cohort of patients diagnosed with RA, and identified new episodes of use for 12 DMARD regimens. We evaluated persistence through survival analyses, and adherence within episodes through the medication possession ratio. A risk score was included in the analyses to account for measured confounders. RESULTS: We identified 14,932 patients with RA; 6018 patients had 10,547 episodes of new use of DMARDs. Considering methotrexate as the reference and after adjustment for measured confounders, episodes of new use of sulfasalazine [adjusted hazard ratio (aHR) = 1.59; 95% confidence interval (CI) = 1.47-1.72] and infliximab alone (aHR = 1.37, 95% CI = 1.09-1.73) were more likely to be discontinued; and new episodes of etanercept (aHR = 0.82, 95% CI = 0.73-0.92) and methotrexate + adalimumab (aHR = 0.63, 95% CI = 0.48-0.84) were less likely to be discontinued. Compared with methotrexate, adherence was higher for leflunomide, infliximab, etanercept, and adalimumab and lower for sulfasalazine and all combined therapies. CONCLUSIONS: We developed an approach to assess persistence on and adherence to the most common DMARD therapies. In this large cohort, persistence and adherence to leflunomide and most biologic DMARD therapies were at least comparable to methotrexate. Adherence was lower for sulfasalazine and all combined therapies.