Objectives: To assess persistence on SSRIs (most prescribed antidepressants) and associated healthcare costs in a naturalistic setting.
Methods: For this retrospective cohort study based on a US reimbursement claims database, all adults with a claim for a SSRI (citalopram, escitalopram, fluoxetine, paroxetine or sertraline) related to a diagnosis of depression were included. Patients should have had no previous reimbursement for any antidepressant within the previous 6 months. Non-persistence was defined as failing to renew prescription within 30 days in the 6-month period after the index date.
Results: In the 45,481 patients included, persistence decreased from 95.5% at 1 month, to 52.6% at 2 months, 37.6% at 3 months and 18.9% at 6 months. Among factors associated with higher 6-month persistence were age 18-34 years, physician's specialty, treatment with escitalopram, absence of abuse history and psychotropic prescription history. During the 6-month after index date, healthcare costs tended to be higher in non-persistent than in persistent patients although not significantly (RR=1.05, adjusted p=0.055).
Conclusion: Despite some limitations associated with the use of computerized administrative claims data (residual unmeasured confounding), these results highlight a generally low persistence rate at 6 months. Special attention should be given to persistence on treatment, with consideration of potential antidepressant impact.