Figure 1From: Assessing subgroup effects with binary data: can the use of different effect measures lead to different conclusions?UK Hip trial: results for any treatment in strong-suspicion and moderate-suspicion subgroups. Points on the red diagonal line indicate lack of effect; points on the other lines indicate the same risk difference (RD), risk ratio for harm (RR(H)), risk ratio for benefit (RR(B)) or odds ratio (OR) as in the strong-suspicion subgroup.Back to article page