ANALYSIS OF CASE-CROSSOVER DESIGNS

The case-crossover design provides a means to study the effects of tra nsient exposures on the risk of acute illness, for example, the effect s of drinking alcohol on the immediate risk of a heart attack. Only ca ses are required by the design, since each case is effectively its own control; what a case was doing at the time of an acute event is compa red with what the case would have been doing usually. Maclure has desc ribed an approach based on the Mantel Haenszel method of analysis. It is shown here how the analysis of case-crossover designs can be achiev ed by a method of maximum likelihood. The method is quite general and, in principle, can be used to analyse the joint effects of many transi ent exposures. For binary exposures the Mantel Haenszel approach is an approximate solution to the likelihood equations. In practice, case-c rossover designs are limited by the information available on each case 's 'usual' behaviour. Extracting such information requires in-depth qu estioning, but, in principle, it can be obtained. To do so requires ca reful questionnaire design. The approach is illustrated by analysis of 24 hour alcohol consumption and the risk of myocardial infarction. Th e problem with this analysis is how to estimate the probability of wha t a case would 'usually' have been doing from information on drinking frequency.