Assessment of risk of bias in included studies—We assessed methodological
quality in the three dimensions initially described by Chalmers 1989: namely the control for
selection bias at entry (the quality of random allocation, assessing the generation and
concealment methods applied); the control of selection bias after entry (the extent to which
the primary analysis included every person entered into the randomized cohorts); and the
control of bias in assessing outcomes (the extent to which those assessing the outcomes were
kept unaware of the group assignment of the individuals examined).
The two review authors independently assessed risk of bias for each study using the criteria
outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins
2006). Any disagreement was resolved by discussion.