Skip to main content
Figure 1 | BMC Medical Research Methodology

Figure 1

From: Multivariable risk prediction can greatly enhance the statistical power of clinical trial subgroup analysis

Figure 1

Panel A shows how overall treatment benefit (net RRR is a function of [treatment benefit] – [treatment harm]) varies as a function of pretreatment risk (as estimated by the control event rate [CER] of an RCT) when a treatment decreases pre-treatment risk by 50% but at a cost of 0.003 treatment-related adverse events per treatment-year. As a result, lower risk patients are harmed by treatment and higher risk patients benefit from treatment. Panel B demonstrates two statistical phenomena: (1) that statistical power can be greatly enhanced by combining risk factors (RF's) into a risk index, and (2) statistical power is greatest when the study population includes more low risk patients

Back to article page