Online Diagnostic Tool Yields Accurate Diagnoses, Study Finds
An Internet-based diagnostic tool had a 95% accuracy rate in 100 actual diagnoses, according to a study by U.K. researchers in the May issue of the Archives in Disease in Childhood, Medscape reports. To test the ISABEL system's accuracy, junior physicians collected data on patients at two teaching hospitals and two large general hospitals; the system displays potential diagnoses based on patient symptoms entered into the tool. A patient's diagnosis was collected from discharge summaries and compared to the diagnosis generated by ISABEL for the same case. The system gave the correct diagnoses in 83 of 87 cases. However, the final diagnosis in 13 cases was not specific because the data used by the system is based on textbooks, which can contain vague diagnoses, according to Medscape. Researchers also tested ISABEL in 99 hypothetical cases and found that the system had a 91% accuracy rate. The study found that ISABEL could be particularly useful for junior physicians who often lack the "clinical wisdom and knowledge" needed for diagnoses. The ISABEL system has more than 9,000 registered users (Tokarski, Medscape, 6/6).
This is part of the California Healthline Daily Edition, a summary of health policy coverage from major news organizations. Sign up for an email subscription.