Modification of the computational procedure in Parker and Bregman's method of calculating sample size from matched case-control studies with a dichotomous exposure.

Several authors have addressed the problem of calculating sample size for a matched casecontrol study with a dichotomous exposure. The approach of Parker and Bregman (1986, Biometrics 42, 919-926) is, in our view, one of the most satisfactory, since it requires specification of quantities that are often easily available to the investigator. However, its recommended implementation involves a computational approximation. We show here that the approximation performs poorly in extreme situations and can be easily replaced with a more exact calculation.