Evaluating Clinical Significance of Treatment Outcomes in Studies of Resistant Major Depression

Background: The efficacy of antidepressant interventions can be examined in terms of statistical significance, a standardized ef-fect size, or measures of clinical significance. An intervention typically needs to deliver at least a moderate size effect (i.e., Cohen’s d>0.4 or>10 % difference in response rates) in order to be considered clinically significant. In this presentation, the efficacy of TMS in a large, registration quality study is calibrated against the performance of other relevant strategies for treatment resistant depression. Methods: Results of the recent TMS trial (i.e., change on the HAMD and MADRS and response and remission rates for the active and sham arms) were tallied and compared against the outcomes reported for the three levels of interventions studied for citalopram non-responders in Sequenced Treatment Alternatives to Relieve Depression (STAR*D), published and unpublished results of PBO-controlled studies of olanzapine-fluoxetine combination (OFC) and aripiprazole augmentation, and electroconvulsive therapy (ECT). Results: Standardized effect sizes and Numbers Needed to Treat values observed in the recent TMS trial compare favorably with the outcomes observed for the range of standard strategies studied in STAR*D and to both OFC and aripiprazole augmentation. The magnitude of the effect of TMS is approximately one half that of ECT. Conclusions: The efficacy of TMS for patients who have not benefited from at least one course of antidepressant therapy appears to be comparable to both standard strategies now in use and atypical antipsychotic/antidepressant combinations. Although clearly less effective than ECT, TMS has clinically significant benefits and might be considered earlier in treatment algorithms