SilverCrucible analysis code and model outputs: Bayesian natural history and network meta-analysis of spinal manipulation for low back pain

R and JAGS code and saved model outputs for two companion analyses of low back pain: a Bayesian natural-history model (Paper I) and a Bayesian hierarchical network meta-analysis of spinal manipulative therapy and other treatments versus sham, incorporating natural history (Paper II). SilverCrucible_code_v1.zip contains the 53 analysis scripts, the shared configuration file, and a README with requirements and run instructions. The analysis dataset is published separately (doi:10.5281/zenodo.23609810); download @131trials_133units.xlsx into the unzipped folder and run 0_RUN_ALL.R to reproduce all models, tables, and figures. SilverCrucible_model_outputs_v1.zip contains the 12 saved model objects (.rds) from the production run, so tables and figures can be rebuilt without refitting the MCMC models. Data_dictionary_131trials_133units.xlsx defines every column of the analysis dataset. Production software: R 4.6.1, JAGS 4.3.2, rjags 4.17, coda 0.19-4.1.