The fundamentals of Evidence-Based Medicine (EBM) are the clinical experience, the application of best evidences from research and the consideration of patient expectations. It enabled significant progresses in the management of diseases with a low or multifactorial causality. But it has also led to unintended negative consequences, partly related to conflicts of interest. The objective of this article is to bring the attention back to the scientific rigor that must sustain the medical practice, namely in the occurrence : 1) formulating a question that addresses all the elements of an individual clinical situation; 2) exploring the literature systematically; 3) estimating the degree of confidence in the conclusions of clinical trials. EBM provides intuitive tools to address some uncomfortable concepts of biostatistics and to identify the biases and the embellished data that invalidate many studies. However, it is difficult to decide of the care of a single patient from observations issued from the comparison of'heterogeneous groups. Personalized medicine should help to overcome this difficulty and should facilitate clinical decision making by targeting the patients who are most likely to benefit from an intervention without much inconvenience.