CROSSOVER TRIALS WITH A BINARY RESPONSE - A POWERFUL METHOD DESPITE THE CARRYOVER EFFECT

The two-period crossover trial has the evident advantage that, by use of within-patient comparisons, the usual large between-patient variabi lity is not used as a measure to compare treatments. A prerequisite, h owever, is that the order of the treatments does not substantially inf luence the outcome of the treatment. Crossover studies with a binary r esponse (such as yes/no or present/absent), although widely used for i nitial screening of new compounds, have not previously been studied fo r such order effects, This study uses a mathematical model based on st andard statistical tests to study to what extent such order effects, h ere identical to carryover effects, may reduce the power of detecting a treatment effect. It is concluded that, despite large carryover effe cts, the crossover study with a binary response remains a powerful met hod and that testing for carryover effects makes sense only if the nul l hypothesis of no treatment effect cannot be rejected.