The goal of the investigators proposed research is to improve diagnosis of patients presenting to the ED with dizziness, some of whom are misdiagnosed with potentially grave medical consequences. The prevailing diagnostic paradigm for the evaluation of the dizzy patient is based upon a 'pathophysiologic' approach. This approach begins a search for etiology with the assumption that the quality of symptoms (vertigo, presyncope, imbalance, or nonspecific dizziness) reflects the underlying pathophysiologic mechanism (vertigo = vestibular, presyncope = cardiovascular, imbalance = neurologic, and nonspecific = psychiatric). Although this assumption often holds true, the 'pathophysiologic' approach mandates a thorough etiologic search in each organ system, not only the one suggested by symptom quality. This strategy is well suited to the referral clinic setting where it was developed, but poorly suited to the time-pressured environment of the ED, where the high index of illness severity demands effective triage rather than diagnostic certainty. The candidate hypothesizes that: 1) potentially serious misdiagnoses of dizzy patients are uncommon but not rare events in the ED and may result from an over-reliance on the diagnostic importance on symptom quality; 2) a novel 'triage' approach to diagnosis would reduce misdiagnoses and improve outcomes in an 'in vitro' computer model of the diagnostic approach to dizziness; and 3) a clinical decision-support system based on this approach would reduce misdiagnoses under simulated patient conditions. To test these hypotheses, the candidate has designed three specific aims to: 1) measure the frequency, potential severity, and possible cause of misdiagnosis of dizzy patients in the ED (by gathering extensive case data on each ED dizzy patient and referencing ED physician [EP] diagnoses against those of a multidisciplinary expert panel); 2) design a computerized decision model to test a new 'triage' approach to diagnosis (by comparing 'in vitro' simulations of the two diagnostic approaches using hypothetical case scenarios); and 3) 'pilot' a web-based decision support system to reduce misdiagnosis of simulated ED dizzy patients (by comparing EP performance on a video-case-based examination with or without the use of the decision support system, using a randomized trial design). Results of this study will form the foundation for subsequent research into the effectiveness of error-reduction interventions among dizzy patients. The research career award candidate has devoted himself to acquiring the clinical and research skills required to complete this project and launch a successful career as an independent investigator. He has garnered the support and enthusiasm of both his clinical department and a large, multidisciplinary team that will enable him to complete the stated objectives. This research project and the research paradigms derived from it will form the nucleus of a career devoted to research in medical decision-making, causes of diagnostic errors, and methods to prevent them.
Showing the most recent 10 out of 14 publications
