GeMTC (gemtc R package)
GeMTC is a free software tool that helps researchers compare multiple medical treatments at once by combining data from different studies, making the process faster and less error-prone.
At a glance
Use when
Comparing multiple treatments using both direct and indirect evidence in a Bayesian framework; conducting simulation studies involving many network meta-analyses; when automation and reproducibility in R are desired
Avoid when
Dealing with highly inconsistent networks requiring inconsistency models; when using non-Bayesian methods or frequentist frameworks; when users lack R programming experience
Inputs
Structured dataset of clinical trial results comparing multiple treatments, including treatment arms, outcomes, and study design
Outputs
Bayesian estimates of relative treatment effects, including posterior distributions, odds ratios, and rankings (e.g., SUCRA)
How it works
GeMTC is an R package that automates the generation of Bayesian random effects consistency models for network meta-analysis (mixed treatment comparisons). It enables automated model specification, including selection of basic parameters, trial baselines, priors, and starting values, eliminating the need for manual coding in general-purpose MCMC software like WinBUGS. The method supports analysis from structured data files and has been validated against five published network meta-analyses.
- HTA domains
- Clinical Effectiveness
- Categories
- Evidence SynthesisIndirect Comparisons
- Assumptions
- Consistency between direct and indirect evidence; homogeneous variance across trials; exchangeable random effects; proper specification of priors and model structure
- Strengths
- Automates model generation, reducing manual effort and coding errors; integrates with R for reproducibility; supports simulation studies; uses Bayesian framework for flexible inference
- Limitations
- Limited to consistency models; assumes homogeneous variance; requires familiarity with R and Bayesian methods; may not support complex network structures or inconsistency models out of the box
- Also known as
- gemtc, GeMTC R package
Questions this answers
- › How effective is treatment A compared to treatments B, C, and D when not all have been directly compared in trials?
- › How can multiple treatments be simultaneously compared using both direct and indirect evidence?
- › How can network meta-analysis models be generated automatically from structured data?
- › How can errors in network meta-analysis data input be minimized?
- › How can simulation studies involving many network meta-analyses be efficiently conducted?
- › How does a Bayesian consistency model handle heterogeneity and indirect comparisons in a treatment network?
References & sources
Similar by meaning
Beta record. Generated from the primary source via AI extraction and independent audit, pending final human review.