Clinical trials and haemophilia: does the Bayesian approach make the ideal and desirable good friends?

When the disease is rare and/or the outcome is uncommon the trial design does not warrant precise and unbiased estimates due to a lack of power or the expected length of recruitment and observation periods. Is there any reliable method to control for bias and consequently achieve an advantage from estimates generated by different study designs? An interesting statistical approach suitable to solve this problem has been theorized by Thomas Bayes. A Bayesian analysis is aimed at answering the question 'How this trial will modify our belief about that treatment effect?' In summary, the Bayesian approach can be defined as the explicit and quantitative use of any kind of external evidence in the design, analysis, and interpretation of an experimental trial. The results of a Bayesian analysis is the 95% credible interval in which we believe the estimate to lie with a probability of 95%, or the estimate of the probability that the quantity of interest is less than a specific value. The principal advantages of the Bayesian approach are that it allows to directly make probability statements about quantities of interest; it allows to easily make predictive statements, conditional on the current state of knowledge; it enables evidence from a variety of sources to be taken into account within a coherent modelling framework; it requires the investigator to explicit prior beliefs and demands. Exemplifications of the advantages of the Bayesian approach will be given discussing some papers published in Haemophilia.