[From symptom to therapy: clinical reasoning].

The path from a symptom to the diagnosis and appropriate treatment begins with hypotheses and is characterized by uncertainties and probabilities. Hypotheses are developed by comparing clinical findings with internalized prototypic images of diseases. Heuristic, i.e. associative rather then logic thinking, is used to cut down the complexity of a clinical case to a size more easy to handle. Hypotheses are evaluated with clinical findings, laboratory tests and radiological images and are either refuted or eventually accepted, which leads to the diagnosis. Tests have a sensitivity and specificity, which are used to calculate the likelihood ratio of a specific test. This likelihood ratio allows to go from a given pretest probability to a posttest probability with the help of Bayes' theorem. Although daily clinical practice is hardly feasible with these principles, it is worthwhile to become conscious of these mechanisms, to understand our thinking and to prevent recurring mistakes and misinterpretations.