[Mixing results of prospective and retrospective studies by objective decision analysis: evaluating probability by healthy human minds?].

The major shortcomings of formal decision analysis derive from the complexity of the clinical problem, behavioral qualities of the investigator and the method itself. It seems difficult to translate study results into a probability estimate for the problem in an individual patient. Therefore, probability assessments from the medical literature, one's own clinical experience and data sources should be combined. Decision analysis is not itself clinical reality but provides a model for thinking explicitly about the problem.