The focus of this research is on increasing the understanding of the cognitive processes involved in information acquisition and clinical decision making in medicine. This will be investigated by using theories of probability (specifically, Bayes' theorem) and heuristics (cognitive strategies) and experimental procedures previously developed in these areas. The primary aim of this research is to examine the way in which physicians seek diagnostically relevant information and the relationship of information selection to the selection of a most likely diagnosis. The specific objectives of this research are (1) to examine the effects of how a problem is framed on the selection of diagnostic information in patient cases, (2) to investigate the cognitive heuristics used in the selection of diagnostic information, (3) to study the threshold value of information that influences the selection of diagnostic information, (4) to examine the relationship between selection of information and a diagnosis, and (5) to identify educational strategies for training medical students and physicians to (a) select optimally diagnostic information for the systematic evaluation of disease candidates in the differential diagnosis list and (b) become aware of potential biases in diagnostic evaluation that may result from the selection of non-optimal information. A series of experiments involving medical students, beginning house officers, and practicing physicians will be conducted to study these issues. In addition, a computerized simulation model of the information acquisition, evaluation, and diagnostic process in medical decision making will be developed. The data will be analyzed by comparing physicians' selections of diagnostic data and most likely diagnosis for patient cases to optimal decisions based on the competing hypotheses heuristic and Bayes' theorem. These optimal decisions dictate that each piece of diagnostic data be evaluated across all competing diseases in the differential diagnosis list. Both framing effects and threshold effects of the problem will be studied to determine how they influence physicians' judgments and cognitive strategies. Results of these experiments will have significant implications in training medical students and physicians in information management and problem solving. Insight into efficient heuristics (strategies) for collecting and evaluating information will be obtained. Non-optimal strategies and biases in judgments that may result will also be identified.
Showing the most recent 10 out of 11 publications
