Introduction

2026-09-15

metaROC Overview

The metaROC package provides a suite of model implementations for the meta-analysis of diagnostic test accuracy (DTA) studies, with a particular focus on multiple thresholds per study with meta-analysis of receiver operating characteristic (ROC) curves. This includes, among others, generalized linear mixed models (GLMM), linear mixed models (LMM), copula-based models, and accelerated failure time (AFT) models.

metaROCcontains three main functionalities: (1) estimation of meta-analytic models, (2) visualization of estimated meta-analysis results, (3) end-to-end framework for conducting simulation experiments for meta-analysis of DTA studies.

Model estimation

The Application Guide provides an overview of model estimation, both for models that consider a single diagnostic threshold per study (single threshold model, STM) and for models that consider multiple diagnostic thresholds per study (multiple thresholds model, MTM). Additionally, there are individual vignettes for each implemented model:

Single thresholds models (STM):

• SROC model (Moses et al. 1993)
• Bivariate LMM (Reitsma et al. 2005)
• Bivariate GLMM (Chu and Cole 2006), (Chu et al. 2010)
• SROC Lehmann model (Holling et al. 2012)
• beta copula model (Nikoloulopoulos 2015)

Multiple threshold models (MTM):

• logit LMM (Steinhauser et al. 2016)
• logit GLMM (Hoyer and Kuss 2018)
• non-parametric SROC model (Martinez-Camblor 2017)
• AFT model (Hoyer et al. 2018), (Hoyer and Kuss 2020)
• Bayesian multinomial model (Jones et al. 2019)
• semi-parametric global rank-based model (Frömke et al. 2022)
• discrete GLMM (Stoye et al. 2024)
• survival copula model

Models that combine aggregated and individual participant data:

• AFT model

Additional models are planned to be added in the future.

Visualization

Both visualization of real and simulated data is supported, as well as visualization of estimated models, e.g., using summary ROC (SROC) curves. The Plotting Guide illustrates the different options implemented in metaROC.

Simulation

The Simulation Guide introduces the simulation framework implemented in metaROC. Simulation can be conducted either by simulating meta-analysis data directly from a meta-analysis model or by setting up parameters similar to real-world conditions that result in model misspecification for all implemented models.

References

Chu, Haitao, and Stephen R Cole. 2006. “Bivariate Meta-Analysis of Sensitivity and Specificity with Sparse Data: A Generalized Linear Mixed Model Approach.” Journal of Clinical Epidemiology 59 (12): 1331–32. https://doi.org/10.1016/j.jclinepi.2006.06.011.

Chu, Haitao, Hongfei Guo, and Yijie Zhou. 2010. “Bivariate Random Effects Meta-Analysis of Diagnostic Studies Using Generalized Linear Mixed Models.” Medical Decision Making 30 (4): 499–508. https://doi.org/10.1177/0272989X09353452.

Frömke, Cornelia, Mathia Kirstein, and Antonia Zapf. 2022. “A Semiparametric Approach for Meta-Analysis of Diagnostic Accuracy Studies with Multiple Cut-Offs.” Research Synthesis Methods 13 (5): 612–21. https://doi.org/10.1002/jrsm.1579.

Holling, Heinz, Walailuck Böhning, and Dankmar Böhning. 2012. “Meta-Analysis of Diagnostic Studies Based Upon SROC-Curves: A Mixed Model Approach Using the Lehmann Family.” Statistical Modelling 12 (4): 347–75. https://doi.org/10.1177/1471082X1201200403.

Hoyer, Annika, Stefan Hirt, and Oliver Kuss. 2018. “Meta-Analysis of Full ROC Curves Using Bivariate Time-to-Event Models for Interval-Censored Data.” Research Synthesis Methods 9: 62–72. https://doi.org/10.1002/jrsm.1273.

Hoyer, Annika, and Oliver Kuss. 2018. “Meta-Analysis for the Comparison of Two Diagnostic Tests to a Common Gold Standard: A Generalized Linear Mixed Model Approach.” Statistical Methods in Medical Research 27 (5): 1410–21. https://doi.org/10.1177/0962280216661587.

Hoyer, Annika, and Oliver Kuss. 2020. “Meta-Analysis of Full ROC Curves with Flexible Parametric Distributions of Diagnostic Test Values.” Research Synthesis Methods 11 (2): 301–13. https://doi.org/10.1002/jrsm.1395.

Jones, Hayley E, Constantine A Gatsonsis, Thomas A Trikalinos, Nicky J Welton, and AE Ades. 2019. “Quantifying How Diagnostic Test Accuracy Depends on Threshold in a Meta-Analysis.” Statistics in Medicine 38 (24): 4789–803. https://doi.org/10.1002/sim.8301.

Martinez-Camblor, Pablo. 2017. “Fully Non-Parametric Receiver Operating Characteristic Curve Estimation for Random-Effects Meta-Analysis.” Statistical Methods in Medical Research 26 (1): 5–20. https://doi.org/10.1177/0962280214537047.

Moses, Lincoln E, David Shapiro, and Benjamin Littenberg. 1993. “Combining Independent Studies of a Diagnostic Test into a Summary ROC Curve: Data-Analytic Approaches and Some Additional Considerations.” Statistics in Medicine 12 (14): 1293–316. https://doi.org/10.1002/sim.4780121403.

Nikoloulopoulos, Aristidis K. 2015. “A Mixed Effect Model for Bivariate Meta-Analysis of Diagnostic Test Accuracy Studies Using a Copula Representation of the Random Effects Distribution.” Statistics in Medicine 34 (29): 3842–65. https://doi.org/10.1002/sim.6595.

Reitsma, Johannes B, Afina S Glas, Anne WS Rutjes, Rob JPM Scholten, Patrick M Bossuyt, and Aeilko H Zwinderman. 2005. “Bivariate Analysis of Sensitivity and Specificity Produces Informative Summary Measures in Diagnostic Reviews.” Journal of Clinical Epidemiology 58 (10): 982–90. https://doi.org/10.1016/j.jclinepi.2005.02.022.

Steinhauser, Susanne, Martin Schumacher, and Gerta Rücker. 2016. “Modelling Multiple Thresholds in Meta-Analysis of Diagnostic Test Accuracy Studies.” BMC Medical Research Methodology 16: 97. https://doi.org/10.1186/S12874-016-0196-1.

Stoye, Ferdinand V, Claudia Tschammler, Oliver Kuss, and Annika Hoyer. 2024. “A Discrete Time-to-Event Model for the Meta-Analysis of Full ROC Curves.” Research Synthesis Methods 15 (6): 1031–48. https://doi.org/10.1002/jrsm.1753.