We did not use the methodological quality as an
exclusion criterion. Given the heterogeneity of the study
designs, populations and types of intervention and the
paucity of included documents, meta-analysis was neither
feasible nor appropriate. The results were presented in a
quantitative narrative synthesis.