Primary adherence to medications for the treatment of major depressive disorder: A descriptive study using real-world data.

Abstract

Background: Primary adherence (PA) to a medication occurs when a patient fills a newly prescribed medication within a predefined window. Failing to fill a necessary medication can have adverse downstream clinical consequences and increased costs. The extent of PA to major depressive disorder (MDD)-related medications in the United States is not well understood.

Objective: To describe overall primary adherence rates to MDD-related medications, both the initial prescription and at each subsequent switch in the patient's treatment journey.

Methods: We used deidentified electronic health record (EHR) and claims data from the Optum Labs Data Warehouse, a national, longitudinal, real-world data asset containing data on more than 350 million patients. Within an index period (1/1/2021 to 12/31/2022), a PA event was recorded when a medication fill occurred within 30 days after the initial index prescription of a new medication class, where new classes were defined as having no prescription for that class in the preceding 180 days. A medication switch was defined as a prescription for a different medication than the immediately preceding prescribed medication. Patient characteristics and PA outcomes were summarized by (1) medication class, (2) counts of total medication switches during the index period, and (3) order of medication switches.

Results: Overall, PA to the index medication was 75%. PA did not vary substantively across medication classes, ranging from 74% (tricyclic class) to 77% (atypical class). Approximately 15% of the sample switched MDD-related medications during the index period, and among this sample, PA to all medications decreased as the number of switches increased, declining from 59% for 1 prescription switch to less than 37% for 4 switches. Among patients with at least 1 switch, PA declined as the number of switches increased from the first switch (70%) to the fourth switch (57%).

Conclusions: PA to initial prescriptions is relatively high although adherence declined with medication switching. This pattern represents a critical clinical vulnerability, which can be addressed through approaches such as education, medication regimen management, and follow-up visits with care providers to promote PA, increase treatment efficacy, and avoid adverse long-term outcomes. Future research should identify barriers to adherence during medication optimization and develop targeted interventions to sustain treatment persistence.