Diagnostic reasoning

Abstract

Research in cognitive science, decision sciences, and artificial intelligence has yielded substantial insights into the nature of diagnostic reasoning. Many elements of the diagnostic process have been identified, and many principles of effective clinical reasoning have been formulated. Three reasoning strategies are considered here: probabilistic, causal, and deterministic. Probabilistic reasoning relies on the statistical relations between clinical variables and is frequently used in formal calculations of disease likelihoods. Probabilistic reasoning is especially useful in evoking diagnostic hypotheses and in assessing the significance of clinical findings and test results. Causal reasoning builds a physiologic model and assesses a patient's findings for coherency and completeness against the model; it functions especially effectively in verification of diagnostic hypotheses. Deterministic reasoning consists of sets of compiled rules generated from routine, well-defined practices. Much human problem solving may derive from activation and implementation of such rules. A deeper understanding of clinical cognition should enhance clinical teaching and patient care.