Diagnostic Test Accuracy Meta-Analysis
Diagnostic Test Accuracy (DTA) Meta-Analysis involves aggregating estimates of sensitivity and specificity across studies using a bivariate meta-analysis model. Two models are commonly recognized in the literature: 1) the bivariate meta-analysis model, 2) the hierarchical summary receiver-operating-characteristic (HSROC) curve model. More about these models can be found in the Cochrane Handbook for DTA Meta-Analysis.
Shiny App
The Bayesian DTA Meta-Analysis Shiny App provides an easy-to-use interface to implement a Bayesian bivariate meta-analysis model with or without a perfect reference test.
Markdown scripts
The Bayesian Bivariate markdown script is a R Markdown script that conducts Bivariate Meta-Analysis to evaluate the accuracy of an index test (test which is under evaluation) under the assumption that the reference test is perfect.
The Bayesian Bivariate Latent Class markdown script is a R Markdown script that conducts Bivariate Meta-Analysis to evaluate the accuracy of an index test (test which is under evaluation) against a reference test in the context where the latter is not perfect
Both scripts run with minimal input from the user and will produce an output including parameter estimates, convergence diagnostic tools and more. See the READ ME! file for more details. A toy example dataset from Butler-Laporte et al (2021) is providided (see READ ME! file for reference)
Peer-reviewed articles
The following articles illustrate use of DTA meta-analysis models.