Use of Linear Models to Analyze Physicians' Decisions

Abstract

Linear models of judgment are powerful tools for studying medical decision making. The recent increase in applications of these models to medicine reflects more available computing resources and the parallel development of clinical prediction rules derived from multivariate analysis of patient data. Psychological research into expert and novice decision making shows that linear models derived from judges' decisions usually predict future decisions more accurately than either the judge or a mechanical application of the judge's stated policies. Studies of medical decision making have shown similar results, as well as marked variation among experts in how they appear to use clinical information. Cognitive feedback, which is feedback to the learner of the judgment model derived from previous decisions, is highly effective for teaching complex judgment tasks. Many technical problems remain to be mastered in constructing linear models of medical judgment. These include how to select the correct variables, how to provide a selection of variables broad enough to accommodate individual variations in strategy, how to model intercorrelated variables, and how to characterize and aggregate individual strategies. Despite the methodologic challenges, linear models remain a powerful method for studying how physicians combine multiple items of imperfect information to make a judgment. These techniques may provide important insights into variation in physician judgments. In addition, they hold promise in teaching the appropriate integration of complex data in the day-to-day practice of medicine.

Factors Affecting Antibiotic Prescribing for Acute Respiratory Infection by Emergency Physicians

“…We derived the weights for the clinical variables using the method of judgment analysis [15] [16]. Participants did not explicitly state whether or not they were influenced by a variable.…”

Role of specific similarity in a medical diagnostic task.

“…These models have been extensively applied to research on medical diagnostic tasks (see Slovic, Rorer, & Hoffman, 1971, for an early example of a multiple-regression--policy-capturing approach to medical diagnosis, and Wigton, 1988, for a recent review; see Schwartz, Gorry, Kassirer, & Essig, 1973, for an Preparation of this article was supported by a grant from the Ontario Ministry of Health to Geoffrey R. Norman and Lee R. Brooks. We thank Donald Rosenthal of the McMaster University Medical Centre for his consultation on dermatology.…”

The correlation of feature identification and category judgments in diagnostic radiology

“…In fact, independence between feature identification and diagnosis is assumed in prominent models of medical decision making. Bayesian decision models (see, e.g., Fischhoff & Beyth-Marom, 1983), regression models (Slovic, Rorer, & Hoffman, 1971;Wigton, 1988), and computer-based decision aids (Guppy et al, 1989) normally focus on the way in which the available evidence is combined to produce a diagnostic decision, without allowing for the possibility that provisional diagnostic decisions can influence decisions about what features are present. These assumptions may be viable in domains such as laboratory medicine, where a plausible argument for independence of features and diagnoses might be made, but these models are also a basis for much research in such perceptually rich and ambiguous specialties as radiology (Lusted, 1968;Slovic et al, 1971).…”