Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Bivariate random-effects meta-analysis and the estimation of between-study correlation.

Richard D Riley1, Keith R Abrams, Alexander J Sutton

  • 1Centre for Medical Statistics and Health Evaluation, School of Health Sciences, University of Liverpool, Shelley's Cottage, Brownlow Street, Liverpool, L69 3GS, UK. richard.riley@liv.ac.uk

BMC Medical Research Methodology
|January 16, 2007
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Evaluating the Impact of Different Natural History Modeling Methods on Cost-Effectiveness Decisions: A Case Study in Duchenne Muscular Dystrophy.

MDM policy & practice·2026
Same author

Variation in Tier 3 Specialist Weight Management Services Within the National Health Service (NHS): Insights From Four NHS Sites in England and Wales.

Clinical obesity·2026
Same author

Estimating the impact of cancer diagnosis on life expectancy by stage at diagnosis: population-based estimates for a range of cancer sites in England.

BMJ oncology·2026
Same author

Mapping the Use of Real-World Evidence Across the EU Health Technology Assessment Regulation: Methodological Considerations, Challenges, and Opportunities for Harmonization.

Journal of market access & health policy·2026
Same author

Cost-Effectiveness of Once Weekly Rezafungin for the Treatment of Invasive Candidiasis in the United Kingdom.

Mycoses·2026
Same author

Simulation-based assessment of a Bayesian M-spline survival model with flexible baseline hazard and time-dependent effects.

BMC medical research methodology·2026

Bivariate random-effects meta-analysis (BRMA) improves evidence synthesis by jointly analyzing multiple endpoints and their correlations. This study shows BRMA offers advantages over separate univariate analyses, particularly for diagnostic markers and surrogate outcomes.

Area of Science:

  • Biostatistics
  • Medical Research Methodology
  • Statistical Modeling

Background:

  • Evidence synthesis often involves multiple endpoints, necessitating methods that can jointly analyze these outcomes.
  • Multivariate meta-analysis utilizes correlations between endpoints, offering a more comprehensive synthesis than separate univariate analyses.
  • Estimating the between-study correlation (rhoB) is crucial for multivariate random-effects meta-analysis.

Purpose of the Study:

  • To assess maximum likelihood estimation for normal and generalized bivariate random-effects meta-analysis (BRMA) models.
  • To compare the estimation properties of BRMA with traditional univariate random-effects meta-analyses (URMAs).
  • To investigate the estimation of between-study correlation (rhoB) in applied examples and simulations.

Main Methods:

Related Experiment Videos

  • Evaluation of maximum likelihood estimation for normal and generalized bivariate random-effects meta-analysis (BRMA) models.
  • Application of BRMA to two case studies: a diagnostic marker and a surrogate outcome.
  • Simulation study comparing BRMA with two separate univariate random-effects meta-analyses (URMAs).

Main Results:

  • The normal BRMA model estimated rhoB at -1 in applied examples, attributed to maximum likelihood estimator boundary truncation.
  • Simulations indicated this boundary estimation occurs with small study numbers or large within-study variation, leading to inflated between-study variance estimates.
  • Despite boundary issues, BRMA did not introduce systematic bias in pooled estimates and yielded conservative standard errors; BRMA generally outperformed URMAs, especially with missing data.
  • A generalized BRMA model demonstrated superior performance for meta-analysis of proportions compared to normal BRMA and generalized URMAs, though convergence issues with rhoB estimation were noted.

Conclusions:

  • Bivariate random-effects meta-analysis (BRMA) models provide significant advantages over separate univariate syntheses.
  • The study highlights the benefits of BRMA in both normal and generalized modeling frameworks.
  • Understanding the estimation of between-study correlation (rhoB) is key to effectively applying BRMA.