OBJECTIVES: BUGSnet is a published, open-source R package to facilitate the execution of network meta-analyses (NMAs) and generation of outputs that satisfy the statistical elements of prominent best-practice NMA guidelines and checklists, including the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist, the ISPOR-AMCP-NPC guidelines, and NICE Decision Support Unit (DSU) Technical Support Document (TSD) 2. The first version of BUGSnet lacked support for contrast-based or shared parameter models (i.e. models including both contrast and arm-based data). The aim of the present project was to implement support for such models.

METHODS: BUGSnet was expanded to support contrast-based and shared parameter models by adding two new functions, nma.model.contrast and nma.model.shared, with each function generating the corresponding Just Another Gibbs Sampler (JAGS) code to run the analyses. The original nma.model function was retained for backwards compatibility with existing BUGSnet scripts. Contrast-based analyses are specified using “NA” for the first or “reference” arm of each trial and odds or hazard ratios or mean differences for all other arms. Results from contrast-based and shared parameter analyses were validated against previous analyses of interventions for Parkinson’s disease conducted by the NICE DSU.

RESULTS: BUGSnet version 1.1.0 now includes support for conducting contrast-based and shared parameter NMAs. The results of analyses conducted using the new contrast-based and shared parameter models closely matched those conducted by the NICE DSU in TSD 2.

CONCLUSIONS: The present project addressed one of the most notable limitations of the first version of BUGSnet, thereby broadening the range of applications to include synthesizing evidence from trials reporting contrast and/or arm-based data. The alignment of BUGSnet outputs with the requirements of best-practice NMA guidelines from multiple prominent healthcare stakeholders makes the package a compelling tool for conducting NMAs suitable for inclusion in peer-reviewed publications or health technology appraisal processes.