Reproducibility bundle: Remote monitoring for heart failure — a systematic review and meta-analysis of 14 trials stratified by intervention class

Harmonized dataset, complete analysis code, numerical results and figure-generation scripts for a systematic review and meta-analysis of randomized trials of remote monitoring in heart failure, stratified by a prespecified four-class intervention taxonomy (implantable hemodynamic sensor, CIED-based diagnostics, structured telehealth, wearable/external sensor). Fourteen trials (13 randomized, 1 single-arm; 11,828 participants) are included. The bundle regenerates every statistic and every figure reported in the manuscript and supplement. The analysis model is random-effects inverse-variance pooling on the log hazard-ratio scale with REML between-study variance and a Hartung–Knapp small-sample adjustment; health status is pooled as a mean difference on a harmonized 0–100 scale. Headline results: HF-specific hospitalization HR 0.70 (95% CI 0.57–0.86, k = 5); all-cause mortality HR 0.87 (95% CI 0.76–0.99, k = 11); patient-reported health status mean difference +3.88 points (95% CI +1.44 to +6.32, k = 7). Contents: data/ harmonized trial-level dataset with data dictionary; code/ primary four-class analysis, secondary invasive-versus-non-invasive re-partition, and all figure scripts; results/ primary and secondary estimates as JSON plus subgroup tables; figures/ regenerated PNG (300 dpi) and SVG. Requirements: Python 3.9 or later with NumPy, SciPy, pandas and matplotlib. Running meta_analysis_primary.py regenerates all primary, stratified and sensitivity estimates; the figure scripts import it, so plots and reported numbers share a single computation engine. MANIFEST.txt lists every file with its SHA-256 checksum. Registration: PROSPERO CRD420251109400. Reporting: PRISMA 2020. The dataset is trial-level aggregate data extracted from published reports. No individual-patient data are included.