On the validity of meta-analyses: exhaustivity must be warranted, exclusion of duplicate patients too

2010 Journal of clinical epidemiology 63;3 (342-343)