Propensity score matching in the evaluation of drug therapy management programs: an illustrative analysis of a program for patients with hepatitis C virus.

This retrospective cohort analysis demonstrates the application of propensity score methods in the Be In Charge (BIC) program, a drug therapy management program for patients with hepatitis C. Data were drawn from BIC participant records and program utilization and from a longitudinal database of administrative claims for pharmacy and medical services. Eligible patients must have received peginterferon alfa-2b (PEG-IFN alfa-2b) and ribavirin, but analyses evaluated only PEG-IFN alfa-2b use; BIC enrollees were matched with patients not enrolled in BIC (controls). Adherence was measured on the basis of the number of injections dispensed and the proportion of patients for whom an average of at least 1 injection per week was dispensed during follow-up. BIC subjects refilled 1.2 more injections than did controls (P < .001) within 12 weeks, 2.7 more (P < .001) within 24 weeks, and 6.7 more (P < .001) within 48 weeks. BIC enrollees were more likely than controls to refill the indicated dose (1 injection per week) within 12 weeks (72% vs 64%, P < .001), 24 weeks (52% vs 41%, P < .001), and 48 weeks (22% vs 13%, P = .002) of initiation. These data suggest that the BIC program may significantly improve adherence to PEG-IFN alfa-2b.