[Drug prescription and current data of the science: factors of the discrepancies]

Randomized clinical trial is the main source of evidence in medicine. Information from clinical trials, however, is limited because only a limited number of questions can be addressed with this approach and, in particular, the issue of drug effectiveness (in real life) remains unanswered. On the other hand, the literature about drug effects (through animal studies, model of effect, observational studies, etc) is very large, but provides little scientific evidence useful for drug prescription: most studies do not address relevant clinical questions or have methodologic problems. In practice, physicians are overwhelmed by thousands of new articles among which they have to select the few that are useful and valid for their practice. This task is difficult because physicians do not have enough time in their clinical activity and they are not trained for critical appraisal of the literature. Therefore, they are extremely dependent on the diffusion process that selects and synthesizes the information for them. Another important factor that has been shown to influence prescription is patient's opinion, as one of the main objective for physicians is patient's satisfaction. Some physicians also use prescription as a way to terminate a consultation or as a medical act independently from the drug which is prescribed (drug prescription being part of the physician-patient relationship). At least, the absence of real control over drug prescription is considered to be one of the main factors explaining the large discrepancy between numerous prescription guidelines (from consensus conference or expert opinions) and the reality.