A new analysis in JAMA Network Open suggests pooling the diagnoses of multiple physicians into a ranked list could be an effective approach to improving diagnostic accuracy.
The researchers analysed data from the Human Diagnosis Project (Human Dx) to compare the accuracy of diagnosis by the collective intelligence of multiple clinicians and medical students versus individual doctors. As part of Human Dx, users create cases from their own clinical practice and identify the intended diagnosis or the differential diagnoses. Respondents then independently generate ranked differential diagnoses for a case and are notified if they are correct or incorrect. Groups ranged in size from two to nine individuals.
The researchers found that collective intelligence was associated with increasing diagnostic accuracy from 62.5 per cent (95% CI 60.1%-64.9%) for individual physicians up to 85.6 per cent (95% CI 83.9%-87.4%) for groups of nine.
“If this phenomenon is replicable outside of a digital context, it implies that collective intelligence could be broadly applicable, including in low-resource settings,” the authors said. "However, the additional benefit of using the collective intelligence approach would need to be weighed against the time and workload necessary to implement in practice,” they added, saying further evaluation of the approach in a real-world setting is needed.