Meta-Analyses Involving Cross-Over Trials: Methodological Issues

This methodological paper provides a corrected description of the Becker-Balagtas approach for combining results from cross-over trials with binary outcomes alongside parallel-design trials in meta-analyses. The authors present a revised formulation of the marginal odds ratio estimation method and demonstrate its implementation with a worked example comparing brand-name and generic anti-epileptic drugs, providing practical guidance for researchers conducting meta-analyses that include both cross-over and parallel trial designs.