Bayesian methodology: an alternative to regular medical practice

The Bayes theorem provides a formula to calculate the probability of an event to occur conditioned by the occurrence of an anterior one (conditioned probability). In medicine it has been applied to calculate the probability of suffering a disease when having a positive result in a given test. This formula emphasizes the importance of prevalence of a disease (or a priori probability of the positive predictive value of a diagnostic test). The novelty of applying the bayesian methodology in clinical practice results from taking into consideration previous external information (or "a priori probability"), and to calculate how it is modified by the evidence (or "verisimilitude") provided by certain empirical tests, to obtain a new probability conditioned by the empirical evidence (or "a posteriori probability"). It also allows to perform sequential analysis (repeated observation of a given event a number of times not fixed in advance) and to incorporate the subjective probabilities to the reasoning. Some authors have proposed the use of bayesian methodology in research studies, such as clinical trials. Nevertheless, this methodology does not adapt well to this kind of reasoning which is hypothetical-deductive.