Telemonitoring in worsening or newly diagnosed heart failure: a real-world matched cohort study with Swedish registry controls

Background Telemonitoring programs may reduce heart failure (HF) morbidity by detecting early signs of worsening, enabling proactive remote patient management; they have also been successful in clinical trials 1,2. However, several successful interventions require high levels of technical and professional support. Also, lack of integration into clinical workflows may hamper usability. Evidence for its real-world usability therefore remains limited. Purpose To evaluate whether a telemonitoring follow-up program in HF is associated with lower risk of the composite outcome all-cause mortality or HF hospitalization, and whether it affects HF healthcare utilization. Methods Consecutive HF patients with new or worsening HF were recruited for telemonitoring (June 2020–July 2023) for 6 months (“cases”) and followed for 2.5 years (n=278). The control group consisted of 2,775 individuals from the national Swedish HF registry, matched for age, sex, LVEF, duration of HF, calendar time, and hospital admission <1 year prior to baseline. The study was conducted during real-world implementation of telemonitoring for HF. Patients provided daily monitoring of weight, blood pressure, and pulse rate by wireless transmission, integrated into the HF team workflow, and clinically evaluated by an HF nurse during office hours. Follow-up data were collected from medical records and national registers. The primary endpoint was time to first event of all-cause mortality or HF hospitalization. Hazard ratios (HR) were estimated using inverse probability of treatment weighting (IPTW)–weighted Cox regression in the intention-to-treat (ITT) population. Sensitivity analyses included a per-protocol (PP) analysis (excluding cases terminated due to poor compliance) and a propensity score (PS)–matched cohort (270 cases matched 1:3 to 810 controls). Counts of HF admissions and HF hospital days were analyzed using IPTW-weighted Poisson regression with offset for follow-up time. Results In ITT, the primary endpoint occurred in 30/278 (10.8%) cases and 435/2775 (15.7%) controls (event rates 4.5 vs 6.8 per 100 person-years), corresponding to an HR of 0.75 (95% CI 0.47–1.21; p=0.24). Results were consistent in PP (HR 0.76 (95% CI 0.47–1.23)) and PS-matched analyses (HR 0.78 (95% CI 0.52–1.17)). All-cause mortality showed a borderline reduction (HR 0.57 (95% CI 0.31–1.04; p=0.06)), while time to first HF hospitalization did not differ (HR 0.99 (95% CI 0.49–1.98; p=0.97)). The rate of HF hospital days among admitted patients was substantially lower in cases (rate ratio 0.62 (95% CI 0.50–0.78; p<0.0001)) (Table, Figure). Conclusion In this real-world implementation, telemonitoring was not associated with a statistically significant reduction in the composite of all-cause mortality or HF hospitalization. The most robust finding was reduced HF hospital days, suggesting lower healthcare burden with earlier discharge among patients who require admission.