Caro et al criticize our management of the heterogeneity between studies in our systematic review of antiarrhythmic agents for maintaining sinus rhythm after AF.1
We pooled together studies including different types and duration of AF in part because the majority of primary studies did the same, but mainly because our target population was the general population of patients with AF. We consider this ability of a meta-analysis to evaluate a whole population that includes different subgroups as a strength, with the limitation that results are not necessarily applicable to every specific subpopulation, as we explained. Of our planned subgroup analysis, only persistent AF could be analyzed separately, and no difference with the overall results was found.
Lafuente-Lafuente C, Mouly S, Mahe I, Bergmann J. Ignoring Study Design Heterogeneity in Meta-analyses—Reply. Arch Intern Med. 2006;166(18):2046-2047. doi:10.1001/archinte.166.18.2046-b