Natural-history evidence for acute and chronic low back pain: analysis data

This dataset contains the natural-history evidence analyzed in “The Natural History Trajectory of Acute and Chronic Low Back Pain: A Bayesian Meta-Analysis” (Menke, Figueredo, Peñaherrera-Aguirre, 2026). Natural history was estimated as an explicit counterfactual: the course of low back pain expected without the treatments evaluated in the companion network meta-analysis. All data were extracted from published cohort studies; no individual-level patient data are included. Analysis files P1_acute_analysis_rows.csv: 26 observations from 8 independent acute cohorts, on a common standardized scale, as entered in the acute saturating-trajectory model. P1_chronic_analysis_rows.csv: 15 observations from 12 independent chronic cohorts (proportion with a favorable course, sample size, follow-up in weeks, outcome family, and uncertainty floor), as entered in the chronic logit model. Source and registry files nh_trajectory_registry.csv: reviewed registry of candidate cohorts and the role assigned to each (primary evidence, validation only, or source map). nh_trajectories.csv: extracted trajectory data from candidate cohorts, including cohorts not used in the primary analysis. nh_recovery_probabilities.csv: directly reported recovery proportions. nh_common_scale_expanded.csv: additional harmonized rows. README.txt describes each file. The R/JAGS code for the analyses is archived separately at https://doi.org/10.5281/zenodo.23655011.