Background: Rates of relapse and predictive relapse factors were studied over more than 4 years in a sample of Spanish outpatients with DSM-III-R criteria for unipolar major depressive episode.
Methods: A final sample of 139 outpatient was followed monthly in a naturalistic study. The Structured Clinical Interview for DSM-III-R was used. Phases of evolution were recorded using the Hamilton Depression Rating Scale, applying the Frank criteria. Survival analysis, Kaplan-Meier product limit and proportional hazards models were used.
Results: A higher rate of relapses was observed in the partial remission group (91.4%) compared to the complete remission one (51.3%). The four factors with predictive relapse value were: "partial remission versus complete remission", "the intensity of clinical symptoms", "the age" and "the number of previous depressive episodes". The existence of partial remission was the most powerful predictive factor.
Limitations: The decreasing sample size during the follow-up and the difficulty in warranting the treatment compliance.
Conclusions: At medium term, relapse rates for a major depressive episode are high. Partial remission after a depressive episode seems to be an important predictive factor for relapses in a 4-year follow-up.
Clinical relevance: Not reaching complete remission is a strong risk factor for relapses in a 4-year follow up study.