How many meta-analyses does it take to settle a question?

Psychological treatments (PTs) are used as adjuncts to cardiac care. This issue of Psychosomatic Medicine provides a meta-analysis by Rutledge et al. (3) on the effects of PT and cardiac rehabilitation on depression and cardiac outcomes, and the journal recently published a systematic review and meta-regression on a similar topic by Dickens et al. (4). This editorial compares the results from these two meta-analyses and discusses the problems associated with combining different types of PT and other treatments, dose-response effects, floor effects, collapsing across outcomes, and therapist qualifications. PTs have mixed but generally positive effects on reducing mortality and cardiac outcomes, but it remains a challenge explaining how such beneficial outcomes can be achieved by relatively small effects on well-being (typical effect sizes: d = 0.2-0.3). Randomized controlled trials are needed on timing of PT, patients with cardiac problems who will benefit most from PT, and the mechanisms by which PT improves cardiac outcomes.