Non-randomised comparison of continuous (24/7) versus weekday telecare in a 3-month telemonitoring programme for primary care patients with heart failure

Background/Introduction Telemedical support is intended to improve access to care, which may be particularly important in areas with limited healthcare infrastructure. However, the choice of telecare organisation that balances resource use and clinical benefit is not straightforward. Purpose To compare outcomes between two telecare support models delivered with a 3-month telemonitoring programme in primary care patients with heart failure (HF): continuous (24/7) versus weekday telecare. Methods This non-randomised comparative study included primary care patients with HF enrolled in a 3-month telemonitoring programme supported by either continuous (24/7) telecare (n=426) or weekday telecare (n=251). Hospitalisations were categorised as 1) HF-related admissions and 2) all admissions defined as cardiovascular admissions or gastrointestinal bleeding. Outcomes were analysed as the proportion of patients with ≥1 admission during the 12 months before and the 12 months after monitoring. Two-year all-cause mortality was compared between groups. Results Groups were similar in sex and NYHA class. Patients supported by continuous telecare were older (median age 72 years, IQR [66-77] vs 69 years, IQR 63-74; p=0.002) and differed markedly in place of residence: the weekday group was predominantly recruited from towns/cities of 20–100,000 inhabitants (65.0%), whereas the continuous telecare group predominantly lived in rural areas or towns ≤20,000 inhabitants (61.5%) (p<0.001). Baseline use of mineralocorticoid receptor antagonists and oral anticoagulants was more frequent in the weekday group (both p<0.01). During the 3-month monitoring period, hospitalisation patterns were similar between the continuous and weekday telecare models. For HF urgent hospitalisations, the between-model ratio (weekday/continuous) was similar in the year before monitoring but lower in the year after monitoring, indicating fewer patients with urgent HF admissions in the weekday model after the intervention. The difference-in-differences analysis supported a greater before-to-after reduction in the weekday model (Table 1). For HF all admissions and for cardiovascular/gastrointestinal bleeding admissions, before-to-after changes did not differ significantly between models (Table 1). Two-year all-cause mortality was low and did not differ between models (Figure 1). Conclusions Continuous telecare was provided to an older, predominantly rural/small-town cohort, yet 12-month all-cause mortality did not differ from weekday telecare. The absence of a sustained reduction in admissions with a time-limited 3-month round-the-clock service suggests this higher-risk group may require longer-term, system-level support to achieve the admission reductions seen with weekday telecare.