See article by Zhou et al I n this issue of JCP, Zhou and colleagues 1 review and integrate placebo-controlled efficacy trials of medications for treatment-resistant depression (TRD) to compare efficacy in a meta-analysis. They conclude that, among 11 augmentation options for TRD, aripiprazole and quetiapine have the most robust evidence for efficacy, with the caveats that these treatments carry substantial risks of adverse events and no long-term data are available. In the absence of direct comparisons, this exercise highlights the formidable challenges that clinicians face when making decisions. While Zhou and colleagues 1 provide an excellent analysis of the available data, we will argue that these findings are of limited use for most people receiving and providing care for TRD. As stated by Tricoci and colleagues in commenting on guidelines in cardiology, "the current system generating research is inadequate to satisfy the information needs of caregivers and patients in determining benefits and risks of drugs, devices, and procedures. " 2(p837) It is not just in psychiatry that we lack evidence for most clinical decisions. We need a new research paradigm beyond meta-analyses of efficacy studies.