[From medical statistics to medical stochastics].

Stochastics is a science that studies mathematical models trying to predict how a given phenomenon may evolve over time because of chance. It is necessary to abandon the binary thinking of "it works or it doesn't work" and rely on the logic of probability and randomness, which represent the measures of uncertainty people are destined to confront in their life and health choices. In making decisions shared with the patient, the physician must consider the baseline risk, the risk ratio, the variation in absolute risk, the competing risk within the patient, but also the number of patients needed to treat to determine the benefit (NNT) or harm (NNH) in the "theoretical" patient of clinical research, the "average" patient of guidelines, and the "imagined" patient of administered medicine. The NNT indicates the magnitude of the random benefit in the "real" patient, while the NNH indicates the harm. These two elements can represent the interface between statistical and stochastic medicine for simplified shared decision-making in "real" patients.