Table 3.

Number and percentage of studies scoring positively on each of the credibility of subgroup analysis criteria.

Credibility of subgroup analysis criterion	Number (%) of studies
Is the subgroup variable a characteristic measured at baseline?	21/21 (100%)
Was the subgroup variable a stratification factor at randomisation?	4/17 (19%)*
Was the hypothesis specified a priori?	5/21 (24%)
Was the subgroup analysis one of a small number of subgroup analyses tested (≤5)?	10/21 (48%)
Was the test of interaction significant (interaction p < 0.05)?	17/21 (81%)
Was the significant interaction effect independent, if there were multiple significant interactions?	5/18 (24%)*
Was the direction of the subgroup effect correctly pre-specified?	4/21 (19%)
Was the subgroup effect consistent with evidence from previous studies?	19/21 (90%)
Was the subgroup effect consistent across related outcomes?	10/21 (48%)
Was there indirect evidence to support the apparent subgroups effect (biological rationale, laboratory tests, animal studies)?	13/21 (62%)

^*Note: A lower denominator reflects the fact that these criteria were not applicable to all of the studies evaluated, either because the study was not an RCT or because the study did not report a significant interaction.