Abstract
During recent decades we have seen an encouraging development of trial methodology. Less effort, however, has been devoted to the application of findings of trials to the practice of medicine. We cannot expect results of controiled trials alone to determine standard therapy, for clinical judgements are also required. Both beneficial and adverse effects are often only detectable on the group level. There are generally very few indications that drug effects differ qualitatively in subgroups of patients. It is important not only to prove or disprove effects, but also to find out how to implement a proper therapy for a majority of eligible patients. Four examples from treatment after myocardial infraction show that drug usage may not always be based upon results of clinical trials. The adoption of Swedish guidelines for thrombolytic therapy in acute myocardial infarction is taken as an example of an apparently optimal application of trial results in practice.