Interpreting medical risk at the bedside using fuzzy measures

Medical risk is assessed when clinical decisions are made for prediction in the individual patient. These involve usually diagnosis, choice of diagnostic testing or treatment. Evidence-based medicine is the accepted method by which medical risk is quantified. Evidence based medicine relies on scientific truth being defined from systemic observation of patients, the information from which is statistical, but also based on probabilities. Translation of the results of large double blind randomized trials to the individual patient who has different context and genetic makeup from any other patient presents a dilemma for the physician. Methods and Results: fuzzy measures of matching based in fuzzy subsethood allow the individual patient to be matched to the average patient, or any other patient, of a clinical trial. The measure of match is applied to the trial statistic in order to come up with a prediction for the patient's clinical course. The same fuzzy measure of matching is able to compare the results of experimental groups when expressed as members of a fuzzy set. In this fashion the results of large clinical trials can be compared to the statistical ldquosignificancerdquo of the trial. The fuzzy prediction measure, F Pred, has embedded within it a measure of different unknown conditions of the subjects being compared. Conclusion: fuzzy measures based in fuzzy subsethood allow for the comparison of individuals and groups while accounting for different conditions. The measures provide non probabilistic information for decision making at the bedside.