They virtually always have the same conclusion: there is some information to suggest efficacy, but all studies are fatally flawed by bias or other errors, so further large scale studies with better methodologies need be done.