[After surgery or thanks to surgery?].

In this issue of the Dutch Journal of Medicine (NTvG), Saadatmand and colleagues report the results of a large retrospective study on breast cancer patients in the Netherlands. One of their conclusions is that breast-conserving therapy results in better survival compared to breast amputation. Here we argue that retrospective studies are not designed to draw any conclusions with regard to outcome. Confounding by indication is the main reason why differences in outcome may not be causally connected in the context of retrospective data. In the case of different options for breast cancer surgery, many randomised trials have firmly concluded that survival is not influenced by type of surgery. Despite the association found in this retrospective analysis, this study does not alter that conclusion.