Randomised controlled trial of problem solving treatment, antidepressant medication, and combined treatment for major depression in primary care

This paper reports no significant differences between the study
groups, but then interprets this as equal efficacy.

If problem solving was effective, why wasn't it better than
antidpressives when combined with them? Lack of evidence of difference
isnt the same as evidence for there being no difference.

Given that lack of depression is a fixed outcome, the workers could
have looked at response times perhaps - tolerability seems to be an issue
in any event. This paper is interesting, but not evidence. Not quite.