AI-Enabled Monitoring of Peripheral Edema May Improve Prevention of Heart Failure Events

A home monitoring device powered by artificial intelligence (AI) reduced recurrent heart failure events and readmissions compared to the standard of care, providing more actionable information than conventional monitoring, according to preliminary findings from the HEARTFELT study, presented at the 2026 Annual Meeting of the American Association of Heart Failure Nurses in San Diego, California. 

Monitoring the progression of leg edema, a hallmark clinical sign of heart failure, is an integral part of heart failure care. In the home setting, remote monitoring of heart failure symptoms typically requires patient engagement or the collection of data with the help of implanted devices. However, the burden of daily tasks related to monitoring may result in fatigue and may reduce adherence among high-risk patients with highly symptomatic disease, leaving them vulnerable to future heart failure events and hospitalizations. A fully passive, AI-enabled remote monitoring device that is currently being tested in a multicenter clinical trial in the United Kingdom may help address these barriers and improve symptom reporting by requiring a low level of patient participation.

The HEARTFELT Study, a single-blind, randomized crossover trial, was designed to assess the effectiveness of autonomous remote monitoring of peripheral edema in patients with heart failure in comparison with conventional remote monitoring. The AI-enabled system relies on a small, noninvasive sensor, which is installed into patients’ homes and used to measure the volume of the patients’ feet and lower legs as they walk past the sensor. AI technology is used to capture 360-degree imagery, tracking the patients’ movements up to 1,000 times daily. If the device detects that peripheral edema is increasing, it sends alerts to the clinical team, which may lead to a phone call or review to check symptoms and adjust treatment, as needed.

An interim analysis was conducted after 210 patients with chronic heart failure, prior decompensation, and poor adherence to daily weighing underwent two consecutive 3-month treatment periods, during which they were assigned to either standard monitoring or standard monitoring plus automated device alerts. The analysis showed that the use of automated alerts was associated with a 47% reduction in recurrent heart failure events compared with standard monitoring (HR 0.53). No safety concerns emerged during the first 6 months of the study. Participants who experienced heart failure events were older, had a higher symptom burden, and demonstrated lower adherence to self-management practices than those who did not. The investigators reported that participants had an overall low adherence to daily weighing, with only 40% of patients identifying daily weighing as important for heart failure management before enrollment in the study. Overall, participants preferred passive monitoring over scales, wearable devices, or implantable systems. 

A separate analysis of data collected from the first 78 participants recruited in the trial, who were assigned to alternating periods of standard care and AI-assisted monitoring, showed that the AI-enabled remote monitoring device was effective and did not raise any safety concerns over 6 months of use. Patients assigned to remote monitoring had lower heart failure-related hospitalization and mortality rates than those who underwent conventional monitoring, with no device-related complications reported in either group. Moreover, the AI-powered device proved superior in terms of data availability compared with conventional monitoring methods. 

The research team aims to enroll up to 1,600 patients with randomized crossover between the intervention and control groups at 6 months, with a long-term follow-up period of up to 4 years after the main trial concludes. The trial is specifically designed to target patients who struggle with adherence or have a history of inconsistent self-monitoring, as well as those with a high symptom burden, such as recurrent peripheral edema. The authors pointed out that the AI-powered technology has the potential to address a major unmet need in heart failure care by identifying worsening congestion before patients experience decompensation, particularly in individuals who struggle with traditional self-monitoring approaches. While heart failure care generates large amounts of clinical data, AI-enabled devices can help transform this information into actionable clinical insights.