PubMed, the Internet portal of biomedical and life sciences literature, indexed an interesting article, entitled ”How much evidence is needed in the case of fatal diseases and when?” (Z Arztl Fortbild Qualitatssich. 2006;100(7):551-5). Author is Antes G, from the Deutsches Cochrane Zentrum, Institut fur Medizinische Biometrie und Medizinische Informatik, Albert-Ludwigs-Universitat Freiburg. Spectacular cases of healing of usually fatal diseases gain much attention and put pressure on physicians to apply treatments which have not been sufficiently approved in clinical studies. But often these cases are poorly documented. Also, the number of spontaneous remissions in cancer is greatly overestimated. Promising individual observations must be confirmed by case series in order to collect information which can help future patients. Efforts are necessary to organize well planned studies, especially on rare diseases. To access the abstract of the article, click here.