“…Such postrandomization exclusions and attrition can contribute to systematic loss to follow-up and, in the reviewed trials, increased the risk of bias associated with the already small and underpowered samples (12, 29, 38, 41-43, 45, 46). Furthermore, publication bias may have been present because several trials were not published in full (12,27,29,30,34), and most of these trials reported null findings according to the available abstracts (27,29,30). Most of the trials adequately blinded the participants and the assessors of outcomes to minimize the risk of bias.…”