[Evidence-based medicine in oncology: do the results of trials reflect clinical reality?].

Evidence-based medicine (ebm) is the answer to the postulate to grade the basis of scientific knowledge in medical care and to protect it against proceedings of unjustifiable arbitrariness. The ranking of controlled clinical trials, the evaluation of publications, meta-analyses, and references to "levels of evidence" in medical guidelines are well established. This is not inconsistent with the fact that many diagnostic and therapeutic measures are not evidence-based and that, even in reputable scientific journals, marketing intentions come into conflict with evidence-based facts. The demand for implementing ebm is furthermore an unsustainable ethical pretension as long as ebm itself is not evidence-based. In many cases better results from ebm are not supported by outcome studies. Health services research which, amongst others, evaluates implementation of study results under everyday conditions should be seen as an essential part of ebm. In oncology, cancer registries contribute to this type of transparency. Cancer registries show to what extent ebm is established as an encouraging future programme for the daily cancer health-care delivery and whether ebm exists as a barely realisable parallel world of promising controlled clinical trials.