Patients who don’t respond well to medication for depression are often referred for psychotherapy. But there has been little evidence that this actually works.A recent trial attempted to assess whether this method could help patients. Researchers assigned a group of 808 patients, all of whom had been chronical- ly depressed for more than two years, to one of five antidepressants based on an algorithm (possibilities included sertraline, escitalopram, bupropion, venlafaxine, or mirtazapine).Those who did not respond to the medication by 12 weeks (491 patients total) had their med- ication regimens optimized according to a predefined algorithm. In this algorithm (referred to as “MEDS” in the paper) patients with no response to the first medication were switched to a different antidepressant, while those with a partial response received augmentation with either bupropion or venlafaxine.These patients were then randomly assigned to one of three conditions: MEDS alone, MEDS plus supportive therapy, or MEDS plus a type of cognitive behavioral thera- py. Outcomes were measured after 12 weeks on the Hamilton depression scale, and there was no significant difference in either remission rates (average of 15%) or partial response rates (22.5%).Thus, sur- prisingly, adding psychotherapy offered no notable advantages over medication by itself (Kocsis JH et al., Arch Gen Psychiatry 2009;66:1178–1188).