Researchers at Boston Children’s Hospital are working on a system that incorporates local microbiology data with two easy questions patients can ask themselves to predict whether or not they should see a doctor.

We need more innovation like this, especially in terms of infectious diseases. Knowing the local burden of a given infection can dramatically improve diagnostic testing performance and inform whether or not such testing should be conducted [1]. Unfortunately, infection data is all too often siloed in hospital and clinic microbiology lab data.

Remember from your basic epidemiology that positive predictive value is dependent on the prevalence of the disease. So, if we know there is a high prevalence of strep throat in the community based on local infection data, the predictive value of simple questions and even rapid strep tests increases. ↩