Measuring Antidepressant Primary Adherence Among Insured Patients With Major Depressive Disorder Using Linked Electronic Health Record and Administrative Claims Data

Major depressive disorder (MDD) is challenging to treat. Antidepressants are prescribed to manage symptoms; the first step toward effective treatment is for patients to fill their new prescription (primary adherence). This study describes antidepressant primary adherence (APA). Patients with MDD were identified from deterministically linked electronic health record (EHR) and claims data in the Optum Research Database from 01Jan2018—28Feb2019. Patients had: a selective serotonin reuptake inhibitor, serotonin and norepinephrine reuptake inhibitor, or tricyclic antidepressant EHR prescription record (first record date = index date); continuous health insurance coverage 2 years before (baseline) and 6 months beginning on index (follow-up); and no baseline index antidepressant claims or prescription records. APA was defined as a pharmacy claim (fill) for the index antidepressant during follow-up. Logistic regression predicted APA by age ( < 40 versus ≥40), sex, race, index antidepressant class, and baseline antidepressant utilization. 3,314 patients met eligibility criteria (mean age 54 years; 67% female; 13% non-white). 73% were naïve to all 3 antidepressant classes. 60% of patients achieved APA, overwhelmingly filling the index antidepressant on the index date. 1% of patients filled a prescription for a non-index antidepressant. Antidepressant-naive patients had lower APA (57% versus 71% for antidepressant-experienced patients, p<0.001). Adjusted odds ratios (ORs) from the regression showed significantly lower APA for non-white (OR 0.56, p<0.001) and antidepressant-naïve (OR 0.51, p<0.001) patients, and for index SNRI versus SSRI (OR 0.79, p=0.015). Linked EHR and claims data provide an opportunity to measure primary adherence in a real-world setting by comparing prescriptions written with pharmacy claims.