Bayesian inference for asymptomatic COVID-19 infection rates

TL;DR

This paper introduces Bayesian methods to analyze COVID-19 asymptomatic infection rates, providing a more flexible approach than traditional meta-analysis, especially when data heterogeneity exists.

Contribution

It develops and applies three Bayesian techniques that allow for more nuanced data pooling and inference in COVID-19 studies, addressing limitations of standard meta-analyses.

Findings

Bayesian methods reveal heterogeneity in study effects.

Pooling data may be inappropriate when true effects differ.

Methodology applicable to other COVID-19 outcomes.

Abstract

To strengthen inferences meta analyses are commonly used to summarize information from a set of independent studies. In some cases, though, the data may not satisfy the assumptions underlying the meta analysis. Using three Bayesian methods that have a more general structure than the common meta analytic ones, we can show the extent and nature of the pooling that is justified statistically. In this paper, we re-analyze data from several reviews whose objective is to make inference about the COVID-19 asymptomatic infection rate. When it is unlikely that all of the true effect sizes come from a single source researchers should be cautious about pooling the data from all of the studies. Our findings and methodology are applicable to other COVID-19 outcome variables, and more generally.