Efficacy stratification of digital health modalities in home-based cardiac rehabilitation for heart failure: a systematic review and meta-analysis

BACKGROUND: Digital health (DH) enables home-based cardiac rehabilitation (HBCR) for heart failure, but no prior systematic review has compared wearable-guided vs. interaction-guided DH modalities' efficacy on cardiorespiratory fitness, exercise capacity, and quality of life. METHODS: We systematically searched PubMed, Embase, Cochrane Library, and CNKI until April 2025 for RCTs. Two investigators independently extracted data, assessed the risk of bias using the Cochrane Risk of Bias Tool 2.0, and graded evidence (GRADE). RESULTS: Analysis of 20 RCTs (n = 4,652) showed DH-HBCR significantly improved 6-minute walk test (6MWT: MD = 21.92 m, 95% CI 12.27-31.58), peak oxygen uptake (VO₂peak: MD = 0.96 mL/kg/min, 95% CI 0.56-1.35), and quality of life (MLHFQ: MD = -7.21, 95% CI -11.40--3.02) versus control. Wearable-guided interventions provided superior gains in 6MWT (MD = 51.98 m, 95% CI 34.21-69.76; I² = 0%) and MLHFQ (MD = -11.14, 95% CI -14.55--7.72; I² = 0%), while interaction-guided interventions significantly improved VO₂peak (MD = 1.00 mL/kg/min, 95% CI 0.49-1.51; I² = 15.2%) more effectively. CONCLUSIONS: DH-mediated HBCR was associated with improved exercise capacity, cardiorespiratory fitness, and quality of life in patients with HF. Preliminary evidence suggests differential efficacy profiles between wearable-guided and interaction-guided modalities. Given the limitations of the included evidence, these findings should be considered hypothesis-generating. Future research should develop phenotype-stratified protocols and conduct head-to-head RCTs to validate these observations.