Email Post: Does differential drop-out explain the influence of study design on antidepressant response? A meta-analysis