LAUSR

BACKGROUND Early improvement to antidepressant drugs predicts remission, but the predictive value of early improvement to electroconvulsive therapy (ECT) is still unclear. The main aim of this study was to determine the optimal definition of early improvement in the treatment of major depressive disorder (MDD) with ECT, by analyzing its value as a predictor of remission. METHODS A naturalistic study was conducted in 87 adult MDD patients treated with acute ECT. ROC curves were generated to identify the best time point (week 1 or 2 of treatment) predicted remission. Sensibility, specificity, and predictive values were calculated for several definitions of early improvement previously proposed in the literature (20%, 25%, 30%, and 35%). Complementary, time to remission was analyzed and a logistic regression model was performed to further characterize the impact of the optimal definition of early improvement on remission while adjusting for other clinically relevant variables. RESULTS A 30% reduction in the HAM-D score after 2 weeks was identified as the optimal definition of early improvement, with acceptable sensitivity (76%), and specificity (67%). Complementary analyses provided further support for this definition showing a shorter time to remission and a significant effect adjusted for the rest of variables. LIMITATIONS Relatively small sample size, ECT restricted to brief bilateral treatment. CONCLUSIONS Early improvement, defined as a 30% of reduction in the HAM-D21 score at week 2, is a good predictor of remission in MDD patients treated with bilateral ECT, with potential clinical impact. Lack of early improvement could indicate a need for further changes in treatment strategies.