Arm-level outcome dataset for a Bayesian analysis of spinal manipulation and natural history in low back pain (131 trials, 133 analytic units)

This dataset contains the arm-level outcome data used in two companion analyses of low back pain (LBP): a Bayesian model of natural history (Paper I) and a Bayesian hierarchical network analysis comparing spinal manipulative therapy (SMT) with other treatments and sham, incorporating natural history (Paper II). Contents. One Excel workbook with two sheets. Sheet1 holds 714 arm-time rows, one per treatment arm per follow-up time, from 131 parent trials and 133 analytic units published between 1974 and 2026. It covers 337 physical treatment arms, 22,695 enrolled patients and 54,137 scheduled patient assessments. Follow-up runs from under 1 week to 156 weeks. Each row records: study identifiers (trial and analytic-unit IDs) study year and quality ratings (PEDro score and risk of bias) clinical phase (acute, chronic, or mixed) treatment description and six-class treatment coding sham and control indicators follow-up time in weeks baseline and follow-up scores with SDs and SEs sample sizes derived standardized mean differences, with the method used (continuous or probit) natural-history predictions and residuals audit fields documenting corrections, with as-filed values retained alongside corrected ones Audit_Status reconciles the inventory: 59 acute and 75 chronic analytic units, 3 contributing to both phases, and 1 not assigned to a phase. It also documents exclusions and special cases. Two secondary syntheses (Jenks 2022, Moorman and Newell 2022) are excluded from analysis and kept only for provenance. Sturion 2020 is retained pre-crossover only, and Kinalski 1989 is retained but not assigned to a phase. Use. This file is the analytic input for the R and JAGS models described in Papers I and II. Variable definitions are given in the accompanying codebook [or: in the Methods of Papers I and II]. Citation. If you use these data, please cite this record and the companion papers.