[NNT and NNH: statistics and stochastics in the evaluation of added therapeutic value].

Any pharmacological, invasive, surgical health intervention should have an added therapeutic value as well as the requirements of quality, safety, efficacy, to be considered as a medical device based on scientific evidence and of clinical utility for the patient. The intervention should be shared between the doctor and the patient who should have rigorous but simple tools to decide on the best therapy to undertake. Assessment of relative risk reduction is commonly used in the scientific literature to quantify both statistical and clinical significance. The reduction of the relative risk is independent of the baseline risk and is useful for comparing the results of trials conducted on populations at different risk levels. A biased reading of relative risk reduction can be used to emphasize the magnitude of the benefit for market innovation. The absolute risk reduction, on the other hand, is proportional to the magnitude of the baseline risk and is a more useful parameter for physicians and patients to understand the extent of benefit and harm, especially if the parameter is expressed in terms of the number needed to treat (NNT) or to harm (NNH). The mode of scientific communication is important for doctors' choices and patients' trust. Even true data can be fake in communication and perception by resorting to verbose paraphrases. Presenting the results of clinical trials in stochastic as well as statistical terms is useful for doctors and patients to verify whether it is worth practicing a certain treatment whose success sometimes has the probability of winning the lottery, also considering side effects and adverse events. One of the most important challenges in precision medicine will be understanding the relationship between probability and randomness.