Remote Cardiac Rehabilitation Cuts Rehospitalization Risk, Major Meta-Analysis Finds

For decades, cardiac rehabilitation has been one of medicine’s most underused prescriptions. Major international guidelines recommend it as a cornerstone of secondary cardiovascular prevention, yet real-world participation in traditional center-based programs hovers at a mere 20 to 30 percent, crippled by geographic barriers, resource limitations, and the crushing financial and time costs of repeated hospital visits. Now, the largest and most comprehensive synthesis of remote cardiac rehabilitation to date suggests that the home may be just as good as the clinic — and in one crucial respect, possibly better. A team of Chinese researchers has pooled data from 45 randomized controlled trials, published in Frontiers in Digital Health, and found that rehabilitation delivered through telemonitoring, mobile apps, and video consultations significantly improves quality of life and cuts the risk of cardiac rehospitalization by roughly 30 percent, all without increasing deaths or adverse events.

The meta-analysis, led by Hui Pei and Qiang Xiao of The Second Affiliated Hospital of Shandong First Medical University, was designed to settle a fragmented evidence base. Previous reviews of digitally delivered cardiac rehabilitation had focused largely on short-term physiological gains — six-minute walk distances, peak oxygen uptake, adherence rates — while leaving the two questions that matter most to clinicians and health systems unanswered: is remote rehabilitation actually safe over the long term, and does it keep patients out of the hospital? To address this, the researchers searched PubMed, Embase, and the Cochrane Library from database inception to July 2026, supplemented by grey literature and manual reference screening, ultimately identifying 622 records that narrowed to 45 randomized controlled trials after rigorous full-text assessment.

Methodologically, the review followed PRISMA guidelines and the Cochrane Handbook, with two independent review teams extracting data and assessing risk of bias using the Cochrane Risk of Bias Tool. The team applied random-effects models as the primary analytical framework, using the DerSimonian and Laird method to estimate between-study variance, and ran fixed-effects sensitivity analyses to test robustness. Heterogeneity was assessed with the Cochrane Q test and I² statistic, publication bias with funnel plots and Egger’s regression test, and — in a step that separates this analysis from most of its predecessors — trial sequential analysis was used to determine whether the cumulative evidence had reached the required information size needed for definitive conclusions.

On quality of life, the results were strikingly scale-dependent. Across 12 trials, remote cardiac rehabilitation significantly improved scores on the MacNew Heart Disease Health-related Quality of Life Questionnaire (mean difference 0.17, P < 0.00001) and the Minnesota Living with Heart Failure Questionnaire (mean difference −5.38, P = 0.03), while the generic SF-36 overall score remained statistically comparable between groups. The psychological picture was more nuanced. SF-36 mental health scores showed significant improvement at 6 months (mean difference 1.73, P = 0.03) but not at 12 months, whereas the MacNew emotional score suggested early psychological benefits that were sustained through 12-month follow-up (mean difference 0.35, P < 0.0001). The Hospital Anxiety and Depression Scale revealed no significant difference between the two delivery models.

Physical function, however, told a story of cumulative, time-dependent benefit. At 6 months, differences between remote and center-based rehabilitation were minimal to modest on both the SF-36 and MacNew physical functioning domains. But beyond 12 months, three trials confirmed that the therapeutic efficacy of remote rehabilitation expanded substantially, with highly significant improvements over center-based care on both scales (SF-36: mean difference 1.48, P = 0.0002; MacNew: mean difference 0.35, P < 0.00001). The researchers attribute this to the cumulative benefits of remote exercise adherence and habit formation — the digital program keeps patients moving long after a hospital-based course would have ended, and the gains compound.

Safety was the question that had dogged the field. Shifting rehabilitation into the home inherently reduces face-to-face medical supervision, deprives clinicians of routine physical examinations, and compromises emergency responsiveness during acute exercise-induced events. Yet the meta-analysis of 17 trials found no statistically significant increase in major adverse cardiovascular events (risk ratio 0.98, P = 0.80), all-cause mortality (risk ratio 1.02, P = 0.91), or all-cause adverse events (risk ratio 1.12, P = 0.36) compared with conventional care. The favorable cardiovascular safety profile establishes remote rehabilitation not merely as a feasible alternative, but as an effective strategy for cardiac functional recovery that does not trade safety for convenience.

The headline finding — the one with the most immediate clinical and economic implications — was cardiac rehospitalization. Across 21 studies, remote rehabilitation was associated with a statistically significant 30 percent reduction in the risk of cardiac rehospitalization (risk ratio 0.70, 95% confidence interval 0.52 to 0.93, P = 0.01), while all-cause readmission showed only a marginal, non-significant trend (risk ratio 0.91, P = 0.08). Exploratory subgroup analyses revealed where the benefit concentrated: within the hardware-integrated digital modality — programs incorporating wearable activity trackers, Bluetooth-enabled physiological monitors, or tele-ECG internet-of-things systems — the risk ratio fell to 0.59 (P = 0.02), whereas software-only programs relying on standalone apps did not reach significance (risk ratio 0.77, P = 0.51). The 6-month observation subgroup showed an even more dramatic reduction (risk ratio 0.38, P < 0.00001), and high-risk patients displayed a non-significant numerical protective trend (risk ratio 0.59 versus 0.74, P = 0.15).

The hardware finding is the most hypothesis-generating element of the entire analysis. The researchers suggest that continuous physiological telemonitoring enables prompt adaptations to exercise regimens — real-time data streams allow clinicians to deliver timely medical interventions and dynamic prescription updates, reinforcing proactive secondary prevention in ways that a phone call or an app cannot. The inherently low event rates in low-to-intermediate-risk cohorts create a statistical floor effect that masks therapeutic gains, whereas high-risk patients offer a larger margin for clinical optimization. But the authors are careful: these exploratory estimates did not reach formal statistical significance for subgroup interaction and cannot establish high-risk patients as preferential beneficiaries. The marked risk reduction in the 6-month subgroup may reflect trial-level characteristics rather than an intrinsically optimal therapeutic window.

Trial sequential analysis added an important caveat to the enthusiasm. The cumulative Z-curve for cardiac rehospitalization crossed both conventional and prespecified monitoring boundaries — a robust result — but neither analysis reached its required information size (5,615 for cardiac rehospitalization and 4,334 for all-cause readmission). Current evidence, the authors conclude, remains statistically non-definitive, warranting further adequately powered confirmatory trials. Statistical heterogeneity remained predominantly low (I² < 50%), and Egger’s tests confirmed no significant publication bias for either hospitalization outcome, lending credibility to the pooled estimates even as they fall short of definitive proof.

The economic picture emerged as three distinct profiles. In Belgium, a 24-week comprehensive telerehabilitation program significantly decreased average medical cost per patient from €2,720 to €2,156 (P = 0.01), yielding a highly favorable incremental cost-effectiveness ratio of −€21,707 per quality-adjusted life year — a dominant, cost-saving strategy. Australia and the Netherlands corroborated the pattern, with savings of Aus $1,590 and up to €3,160 per patient respectively, while middle-income settings showed direct implementation costs of $61 versus $138 for center-based care. Poland’s TELEREH-HF trial demonstrated cost-effectiveness within national thresholds despite increased expenditures. But Australia’s ToDo-CR trial showed the third profile: high setup outlays not offset by reduced healthcare utilization produced an unacceptably high ICER of Aus $93,395.50, and the TeleClinical Care pilot required Aus $6,028 per avoided readmission, with cost-saving potential only when enrollment scaled beyond 237 patients. The authors propose a stepped-care model — front-loading intensive psychological and physical support during the early post-discharge window, then tapering to low-intensity remote maintenance — as the path to translating clinical and economic potential into long-term health-system sustainability, particularly for low- and middle-income countries seeking to bridge the chronic shortage of specialized cardiac rehabilitation facilities in rural areas.

Subject of Research: Long-term quality of life, rehospitalization risks, and economic viability of remote cardiac rehabilitation: an updated systematic review and meta-analysis

Article Title: Long-term quality of life, rehospitalization risks, and economic viability of remote cardiac rehabilitation: an updated systematic review and meta-analysis

Article References: Pei, H., Li, H., Hu, M., Liu, B., Hu, D., Li, Y., & Xiao, Q. (2026). Long-term quality of life, rehospitalization risks, and economic viability of remote cardiac rehabilitation: an updated systematic review and meta-analysis. Frontiers in Digital Health, 8, Article 1935832. https://doi.org/10.3389/fdgth.2026.1935832

Image Credits: AI Generated

DOI: 10.3389/fdgth.2026.1935832

Keywords: Long-term, quality, life, rehospitalization, risks, economic, viability, remote, cardiac, rehabilitation, updated, systematic