A vine copula mixed effect model for trivariate meta-analysis of diagnostic test accuracy studies accounting for disease prevalence

TL;DR

This paper introduces a vine copula mixed effect model for trivariate meta-analysis of diagnostic test accuracy studies, accounting for disease prevalence, and demonstrates its advantages over traditional models through real data re-analyses.

Contribution

It extends bivariate copula models to a trivariate setting, including disease prevalence, and shows improved data fit and flexibility over existing models.

Findings

Vine copula models outperform traditional GLMM in data fit.

The model captures asymmetric tail dependence in data.

Computational feasibility is maintained despite increased complexity.

Abstract

A bivariate copula mixed model has been recently proposed to synthesize diagnostic test accuracy studies and it has been shown that is superior to the standard generalized linear mixed model (GLMM) in this context. Here we call trivariate vine copulas to extend the bivariate meta-analysis of diagnostic test accuracy studies by accounting for disease prevalence. Our vine copula mixed model includes the trivariate GLMM as a special case and can also operate on the original scale of sensitivity, specificity, and disease prevalence. Our general methodology is illustrated by re-analysing the data of two published meta-analyses. Our study suggests that there can be an improvement on trivariate GLMM in fit to data and makes the argument for moving to vine copula random effects models especially because of their richness including reflection asymmetric tail dependence, and, computational feasibility despite their three-dimensionality.