Automated Medical Interviewing for Diagnostic Decision Support in the Emergency Department. Inclusive Dates: 09/01/08 - 08/31/10.

摘要

Misdiagnosis is frequent, particularly in the emergency department (ED). The goal of this research is to develop and validate a workflow-sensitive, computer-based, diagnostic decision support system for the ED that reduces critical misdiagnosis of patients. Our overall research mission is to improve the accuracy and timeliness of medical diagnosis in frontline healthcare settings, particularly the ED. Misdiagnosis is common. Roughly 5% of hospital autopsies disclose Class I errors (where accurate diagnosis could have led to discharge alive), and diagnostic errors result in the preventable deaths of 40-80,000 annually in the US. The ED is a high-risk site for dangerous misdiagnosis. It is the hospital site with the greatest risk of medical errors, the majority of which are misdiagnoses, many of which are lethal or morbid. While many factors likely contribute to high rates of misdiagnosis in the ED, cognitive factors play a large role, and diagnostic decision support tools offer a possible solution. The long-range goal of this particular research program is to develop and validate a computer-based, diagnostic decision support system for the ED that reduces critical misdiagnosis (stroke, myocardial infarction, etc.) of patients with undifferentiated medical symptoms (dizziness, chest pain, etc.). To succeed, the system must fit within the ED clinical workflow--an issue which has been largely ignored in development of similar systems, resulting in failure to adopt. To this end, we envision a patient-centered (e.g., kiosk-type) system for Automated Medical Interviewing for Diagnostic Decision Support (AMIDDS) capable of symptom-specific, adaptive diagnostic history-taking from patients in a pre-encounter setting such as the ED waiting area. The goal of this project was to rigorously develop and test an AMIDDS prototype capable of accurately, reliably, and efficiently identifying key elements of patients dizziness symptoms as a prelude to diagnostic decision support to prevent critical misdiagnosis in this patient population. Our Specific Aims were to: (1) validate that a Complaint Module can accurately discern which ED patients are dizzy and (2) validate that a Dizziness Module can accurately elicit key aspects of a dizziness symptom history.