Predictors of relapse following a stepwise psychopharmacotherapy regime in patients with depressive disorders.

Abstract Background Real world predictors of relapse following routine treatment for depression remain under-researched. We sought to investigate this in an outpatient clinical sample with depressive disorders receiving stepwise pharmacotherapy based on early clinical decision-making, applying a naturalistic 24-month prospective design. Methods Patients were recruited at a University hospital in South Korea from March 2012 to April 2017. After 3-week antidepressant monotherapy (N = 1262), next treatment steps (1, 2, 3, and 4 or over) with alternative strategies (switching, augmentation, combination, and mixtures of these approaches) were administered based on measurements and patient preference at 3-week points in the acute treatment phase (3, 6, 9, and 12 weeks) (N = 1246). For those who responded [Hamilton Depression Rating Scale (HAMD) score of≤14] (N = 937), relapse (HAMD>14) was identified every 3 months from 6 to 24 months (N = 816). Predictors of relapse were evaluated using multi-variate Cox proportional hazards models. Results Four independent relapse predictors were identified: higher number of previous depressive episodes, higher anxiety at baseline, higher number of treatment steps, and poor medication adherence. In particular, treatment Step 4 was significantly associated with relapse compared to treatment Step 1, 2, and 3 after adjustment for relevant covariates. Limitation Withdrawal syndromes after discontinuing psychotropic drugs, known to confound the determination of relapse, were not evaluated. The study was conducted at a single site, which maximised consistency but may limit generalizability. Conclusions Predictors of relapse reported from more restricted trial or cohort samples were replicated in this long-term naturalistic prospective design.