Power calculation in matched case-referent studies: application and accuracy of the asymptotic power function

Although the asymptotic power function for matched case-referent (case-control) studies has been given, it is not widely applied. One reason for this is that the parameters in the power function do not correspond directly to the usual description of the research situation. When designing a matched case-referent study, the potentially available relevant information concerns the joint distribution of the determinant and matching variate, and the illness rate in relation to those factors, in the study base. Application of the asymptotic power function using these inputs is described. The asymptotic power function is then assessed by use of the Monte Carlo method and is found to be accurate. An example of the application of the asymptotic power function is given.