Bayes' Theorem, COVID-19, and Screening Tests
2020
This article reviews the implications of increased testing for COVID-19 using reverse transcriptase polymerase chain reaction (rRT-PCR) through the application of Bayes’ Theorem for three hypothetical, stylized case scenarios.
The scenarios involve three patients with a low, moderate, and high pre-test probability of COVID-19 infection. The category of low probability would include "asymptomatic individuals in a presumed low prevalence environment" and might vary from 10 to 20%. The category of moderate probability would include "individuals with cough and fever in a city/jurisdiction with known cases of COVID-19" and might vary from 40 to 60%. The category of high probability would include patients "with fever, cough, shortness of breath and with a known close contact with confirmed COVID-19" and might be estimated to be 80 to 90%.
For each of these individuals, the implications of a "positive" and "negative" result are quantified using Bayes’ Theorem and discussed. The examples include the use of likelihood-ratio form of Bayes and nomograms.
Bayes' Theorem, COVID-19, and Screening Tests
Source:
Chan GM. Bayes' Theorem, COVID-19, and Screening Tests. American Journal of Emergency Medicine 2020; 38 (10): 2011-2013. https://doi.org/10.1016%2Fj.ajem.2020.06.054