Boosting medical diagnostics by pooling independent judgments.
Diagnostic error remains a significant source of preventable patient harm. Because bias on the part of health care providers can contribute to diagnostic errors, improving the medical decision-making process may serve as an error prevention strategy, possibly through the use of collective intelligence. This study sought to better understand the conditions under which collective intelligence might augment medical judgment. Using large data sets consisting of more than 140 doctors and 20,000 diagnoses, the investigators determined that when providers have similar individual diagnostic accuracy rates, pooling their judgments improves decision accuracy. However, if accuracy levels vary across providers, pooling independent judgment leads to worse diagnostic outcomes. The authors suggest that similar diagnostic accuracy should be a key condition when creating groups for the purpose of improving medical diagnosis. A previous WebM&M commentary discussed a case of diagnostic error.