OBJECTIVE: To review published clinical trials of the pharmacologic management of refractory depression. DATA SOURCES: MEDLINE was searched for relevant articles published from 1983 to 1990. The bibliographies of review articles were searched for additional references. Studies of nonpharmacologic treatments, such as electroconvulsive therapy, were not included. STUDY SELECTION: Eleven studies were found that did not contain obvious digressions from several methodologic assessment criteria (adapted from the McMaster guidelines for the evaluation of clinical trials). Further scrutiny by a nonblind reviewer resulted in the selection of four reports that were considered acceptable. An assessment by a second reviewer, blind as to author, results and journal name, confirmed this judgement. DATA EXTRACTION: Data describing response to the treatments were extracted by a single (nonblind) reviewer. Post-hoc power estimates and 95% confidence intervals were calculated whenever possible. DATA SYNTHESIS: The efficacy of augmenting an antidepressant regimen with lithium carbonate, triiodothyronine or reserpine was not supported by findings from the clinical trials reviewed. However, many trials with negative results lacked adequate statistical power to exclude the possibility of the drug's efficacy. The use of a monoamine oxidase inhibitor was supported by the one study that met the review's methodologic criteria. However, this study was not conducted under double-blind conditions. CONCLUSION: The generally recommended strategies for the pharmacologic treatment of refractory depression are not supported by methodologically sound studies.
OBJECTIVE: To review published clinical trials of the pharmacologic management of refractory depression. DATA SOURCES: MEDLINE was searched for relevant articles published from 1983 to 1990. The bibliographies of review articles were searched for additional references. Studies of nonpharmacologic treatments, such as electroconvulsive therapy, were not included. STUDY SELECTION: Eleven studies were found that did not contain obvious digressions from several methodologic assessment criteria (adapted from the McMaster guidelines for the evaluation of clinical trials). Further scrutiny by a nonblind reviewer resulted in the selection of four reports that were considered acceptable. An assessment by a second reviewer, blind as to author, results and journal name, confirmed this judgement. DATA EXTRACTION: Data describing response to the treatments were extracted by a single (nonblind) reviewer. Post-hoc power estimates and 95% confidence intervals were calculated whenever possible. DATA SYNTHESIS: The efficacy of augmenting an antidepressant regimen with lithium carbonate, triiodothyronine or reserpine was not supported by findings from the clinical trials reviewed. However, many trials with negative results lacked adequate statistical power to exclude the possibility of the drug's efficacy. The use of a monoamine oxidase inhibitor was supported by the one study that met the review's methodologic criteria. However, this study was not conducted under double-blind conditions. CONCLUSION: The generally recommended strategies for the pharmacologic treatment of refractory depression are not supported by methodologically sound studies.