Systematic review: Centre selection considerations, results of the meta-summary - subgroup analysis by type of intervention and RCT design.

Model interpretation: Out of 26 non-pharmacologic RCTs (27% of non-pharmacologic trials) which mentioned at least one consideration for centre selection pertaining to diversity and representativeness, 13 RCTs (14%) were concerned with diversity in terms of population characteristics, 13 RCTs (14%) mentioned diversity in terms of health service delivery and 12 RCTs (13%) referred to diversity in terms of centre setting.