Document resource
Introduction Early identification of congestion is central to heart failure (HF) management, yet in real-world practice, this often relies on patient-led behaviours, such as daily weighing, which are inconsistently performed, particularly among older, multimorbid patients with high treatment burden. This creates gaps in monitoring, delays clinical response, and may disproportionately affect those at highest risk of HF hospitalisation.Methods HF-TRACK ( NCT06334822) is a multicentre, randomised, controlled crossover trial evaluating an AI-enabled, passive peripheral oedema monitoring device used in the home. Participants completed two 3-month periods: standard care with daily weighing alone and an intervention combining daily weighing with automated oedema monitoring that generated alerts for clinical review.Results Data availability from daily weighing was low, with a median of 0.2 days of weight data per week. In contrast, automated oedema monitoring generated substantially more consistent data, with a median of 4.7 days per week (see figure 2). From a nursing perspective, limited data availability may delay recognition of clinical deterioration and increase reliance on reactive rather than preventive care.HF-related hospitalisations occurred less frequently during the intervention period (0.20 events per patient-year) compared with standard care (0.37 events per patient-year), p=0.11 for superiority (figure 2).Participants who experienced unschedduled events during follow-up were older, had greater comorbidity burden, and reported more difficulties with adherence than those without events (See table 1). Prior to enrollment, fewer had managed to weigh themselves before their most recent HF hospitalisation, despite reporting similar confidence in recognising and reporting congestion symptoms. Only 40% of participants recognised the importance of daily weighing prior to study entry, despite its inclusion in routine clinical education.At consent, participants who later experienced HF events reported greater overall HF symptom burden and more severe peripheral oedema. On EQ-5D, pain/discomfort, anxiety/depression, and mobility were the most affected domains across the cohort.Participants expressed a preference for passive monitoring, significantly favouring the AI-device over scales, wearables or implantables (all p<0.05). Feedback was largely positive, with most describing the device as easy to use, unobtrusive, and reassuring. Reported study burdens were primarily related to questionnaires; explicitly negative device feedback was rare (device taking space).Conclusion Passive, automated oedema monitoring was associated with a lower HF hospitalisation rate (p=0.11) and was preferred by patients (p=0.05), particularly those at higher risk and with greater adherence challenges. These findings suggest that reducing monitoring burden while improving data availability may be key to improving outcomes in real-world HF populations.Abstract 427 Figure 1Violin plots illustrating the differences in data availability between the AI-device and Bluetooth-connected weighting scalesAbstract 427 Figure 2Event rate per patient-year during the first six months of the HF-TRACK trail, based on pre-specified analysis