A Computer-Based Approach to Developing Diagnostic Rules
MetadataShow full item record
Purpose: In 2015, the National Academy of Medicine published a report revealing that diagnostic error may be America’s third leading cause of death and responsible for the majority of paid medical malpractice claims. Medical education researchers are now looking to the learning sciences for theories that might support improvements in the diagnostic performance of tomorrow’s health care providers. One such theory, called “Dual-Process Theory”, suggests that people utilize two distinct approaches to diagnostic reasoning: pattern recognition and analytical reasoning. To date, researchers have paid little attention to how we reason analytically. Dual-process theorists suggest that analytical reasoning is, in part, predicated upon a clinician’s knowledge of diagnostic rules. These rules encompass knowledge in the form of experientially-based, statistically-framed estimates of the frequency with which a given disease is associated with each of its characteristic findings. The purpose of this project is to produce a computer-based training tool which supports learners in how to analytically reason via the acquisition and application of conditional probability (CP) derived diagnostic rules. Methods: This tool will have four components: 1) a display of CP derived diagnostic rules associated with signs and symptoms most likely to be linked to a given clinical presentation 2) a set of interactive tools enabling learners to identify which of those rules are most robust in ruling in/out the various differentials, 3) a set of practice cases where learners are given multiple opportunities to apply these CP derived rules, and 4) interactive screen prompts designed to guide the students in developing a cognitive strategy to apply high-yield rules to diagnose a multitude of test cases. Results: The exhibitor will demonstrate a tool which: 1) displays disease by sign/symptom CPs, 2) enables their rearrangement (by history & physical, breadth by feature strength, and depth by selected differential) as the basis for formulating diagnostic rules, and 3) functions which support the construction of a diagnostic strategy. Conclusions: After completion of the described educational tool, the authors will execute an IRB approved study involving students in a year 2 systems course, and a treatment/control research design.