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

Table 3 Table 3. Statistical power when risk factors are combined into a risk score

		Number of Risk Factors in the Risk Score
		1	3	6
		Statistical Power for Risk Index (AUROC) †
Risk Factor Effect Size (Odds Ratio)	1.5	.12 (.53)	.27 (.55)	.45 (.58)
	2.0	.21 (.54)	.61 (.59)	.83 (.65)
	3.0	.41 (.57)	.92 (.67)	.93 (.76)

* In each instance all risk factor have a prevalence of 25% and the 5-year CER for those without any risk factors is .75% (see Table 1 and Figure).
† Statistical subgroup comparisons tests the power to detect whether the treatment's relative benefit varies as a function of the risk score. The Area Under the Receiver Operator Characteristic (AUROC) curve is a measure of the overall predictiveness of a model for predicting a dichotomous outcome (i.e, event occurred vs. event did not occur).

ISSN: 1471-2288