DSM melancholic features are unreliable predictors of ECT response: a CORE publication.

OBJECTIVE: To determine the relationship between baseline melancholic features with outcomes in patients with major depressive disorder referred for electroconvulsive therapy (ECT). METHOD: In a multihospital (Consortium for Research in ECT) collaborative ECT study, SCID-1 interviews were obtained at study entry. Ratings of the 24-item Hamilton Rating Scale for Depression were obtained thrice weekly during the course of ECT, once during a subsequent treatment-free week, and periodically during 6-month continuation treatment with either bitemporal ECT or nortriptyline plus lithium (continuation pharmacotherapy). RESULTS: The evaluable sample was severely ill with a mean 24-item Hamilton Rating Scale for Depression score of 35.2 (+/-6.9). Of 489 patients, 63.6% (311) met DSM-IV criteria for melancholic features. During acute ECT, 62.1% of those with melancholic features remitted, as compared with 78.7% for those without melancholic features (P = 0.002). During medication continuation treatment (continuation pharmacotherapy), relapse rates were higher for those with melancholic features than for those without these features. Conversely, with continuation ECT, the rate of relapse was lower for those with, compared with those without, melancholic features. CONCLUSIONS: Ascertaining melancholic features by SCID-1 criteria does not identify depressed patients more likely to respond to ECT as had been anticipated from the literature. Melancholic features were associated with poorer treatment outcomes in acute ECT. Those with melancholic features were less likely to relapse with continuation ECT, but those with melancholic features were more likely to relapse with continuation pharmacotherapy. The limitations of the DSM-IV criteria for melancholia are discussed.