News

The reporting on antibody testing focusses on “sensitivity” and “specificity” but lacks the context of how those attributes can be used.  Specifically, results for an individual patient needs to include the consideration of the prevalence.  This video is about this context.

The positive predictive value of these tests is quite poor if the prevalence is low (5%).  It improves if other risk factors are considered.  Say, for example, one considers a population of people that live with a person that have been diagnosed with COVID.  The prevalence in that population will certainly be much higher than the general population.

Learn more about what MIT is doing to fight COVID-19 pandemic.

 

This video was inspired after reading this New York Times article.