Comparative accuracy of diagnosis by collective intelligence of multiple physicians vs individual physicians.
Timely and accurate diagnosis is a prerequisite for safe and high-quality treatment. This study used data from the Human Diagnosis Project (Human Dx, an online case-solving platform) to examine diagnostic accuracy among individual physicians compared to groups of physicians (collective intelligence). Physicians can enter cases onto the platform or solve cases that others have entered. The more physicians involved in solving a given case, the more likely that the correct diagnosis would be identified. Groups of physicians across specialties outperformed individual subspecialists even for subspecialty-relevant cases. The authors advocate for testing the use of collective intelligence for diagnosis in clinical settings. A related editorial discusses how teaching diagnosis has evolved and the possibility of using collective intelligence to improve diagnostic accuracy. In a previous PSNet interview, Shantanu Nundy, Director of Human Dx, discussed his work with the project.