[Paper Review] meta4diag: Bayesian Bivariate Meta-analysis of Diagnostic Test Studies for Routine Practice

Jingyi Guo, Andrea Riebler|arXiv (Cornell University)|Dec 19, 2015

Statistical Methods and Bayesian Inference24 references20 citations

TL;DR

This paper introduces meta4diag, an R package that enables fast, user-friendly Bayesian bivariate meta-analysis of diagnostic test studies using integrated nested Laplace approximations (INLA) instead of Markov chain Monte Carlo (MCMC). It supports flexible prior specification, including penalized complexity (PC) priors, and delivers accurate posterior estimates of sensitivity, specificity, and summary receiver operating characteristic (SROC) curves without MCMC sampling, making it suitable for routine clinical and epidemiological research.

ABSTRACT

This paper introduces the \proglang{R} package \pkg{meta4diag} for implementing Bayesian bivariate meta-analyses of diagnostic test studies. Our package \pkg{meta4diag} is a purpose-built front end of the \proglang{R} package \pkg{INLA}. While \pkg{INLA} offers full Bayesian inference for the large set of latent Gaussian models using integrated nested Laplace approximations, \pkg{meta4diag} extracts the features needed for bivariate meta-analysis and presents them in an intuitive way. It allows the user a straightforward model-specification and offers user-specific prior distributions. Further, the newly proposed penalised complexity prior framework is supported, which builds on prior intuitions about the behaviours of the variance and correlation parameters. Accurate posterior marginal distributions for sensitivity and specificity as well as all hyperparameters, and covariates are directly obtained without Markov chain Monte Carlo sampling. Further, univariate estimates of interest, such as odds ratios, as well as the SROC curve and other common graphics are directly available for interpretation. An interactive graphical user interface provides the user with the full functionality of the package without requiring any \proglang{R} programming. The package is available through CRAN \url{https://cran.r-project.org/web/packages/meta4diag/} and its usage will be illustrated using three real data examples.

Motivation & Objective

• To address the computational inefficiency and convergence issues of frequentist and MCMC-based methods in small-sample, sparse diagnostic meta-analyses.
• To provide a user-friendly interface for researchers without advanced R programming skills to perform Bayesian bivariate meta-analysis.
• To implement a flexible prior specification framework, especially the penalized complexity (PC) prior, to incorporate expert knowledge while maintaining robustness.
• To deliver accurate posterior distributions for sensitivity, specificity, and hyperparameters without MCMC sampling, leveraging the INLA method.
• To support common diagnostic metrics such as SROC curves, likelihood ratios, and diagnostics odds ratios directly within the output.

Proposed method

• The package is built on the INLA (Integrated Nested Laplace Approximation) R package to enable fast Bayesian inference without MCMC sampling.
• It implements the binomial-normal model for sensitivity and specificity using exact binomial likelihood, avoiding continuity corrections for zero cells.
• The model uses a bivariate normal distribution for the random effects of logit-transformed sensitivity and specificity, capturing their inherent correlation.
• Users can specify prior distributions for variance components and correlation using both conventional and penalized complexity (PC) priors.
• The PC prior framework allows intuitive prior specification based on probability contrasts, with hyperparameters interpretable as expected deviations from a baseline.
• An interactive graphical user interface (GUI) enables model specification, prior exploration, and result visualization without writing R code.

Experimental results

Research questions

• RQ1 Can a Bayesian bivariate meta-analysis of diagnostic test studies be performed efficiently and accurately without relying on MCMC sampling?
• RQ2 How does the performance of penalized complexity (PC) priors compare to conventional priors in terms of bias, mean-squared error, and coverage in bivariate meta-analysis?
• RQ3 Can a user-friendly, interactive GUI be effectively integrated into a specialized Bayesian meta-analysis package to improve accessibility for non-expert users?
• RQ4 To what extent does the binomial-normal model outperform the normal-normal model in terms of statistical efficiency and robustness when data are sparse?
• RQ5 Can the package reliably produce standard diagnostic metrics such as SROC curves, likelihood ratios, and diagnostics odds ratios directly from the posterior distributions?

Key findings

• The meta4diag package successfully implements fast Bayesian bivariate meta-analysis using INLA, eliminating the need for time-consuming MCMC sampling.
• The use of penalized complexity (PC) priors enables intuitive, robust, and informative prior specification that performs at least as well as conventional priors in simulation studies.
• The binomial-normal model with exact binomial likelihood outperforms the normal-normal approximation in terms of bias, mean-squared error, and coverage, especially in sparse data settings.
• The interactive GUI allows users to explore prior distributions and model settings in real time, with immediate updates to results including SROC curves and forest plots.
• Posterior marginal distributions for sensitivity, specificity, and all hyperparameters are computed accurately and directly, without requiring post-processing or custom code.
• The package provides ready-to-use outputs such as SROC curves, diagnostics odds ratios, and likelihood ratios, with corresponding R code visible for reproducibility.