Smartwatch predicts heart failure hospitalizations

Sun 12 April 2026
Wearables in health
News

Smartwatches could play a significant role in the early detection of health deterioration in patients with heart failure. This is according to new research from the University Health Network and the University of Toronto. The study demonstrates that data from consumer wearables, such as the Apple Watch, can detect changes in health status days or even weeks before medical intervention is required.

Heart failure is one of the most common chronic conditions worldwide, affecting an estimated 64 million people. It is also a major cause of hospital admissions and healthcare costs. In current practice, doctors rely primarily on consultations and measurements taken in the hospital. However, these snapshots provide only limited insight into patients’ daily functioning.

Wearables offer an alternative by enabling continuous monitoring in the home setting. In the study, recently published in Nature Medicine, 217 patients with heart failure were monitored for three months while they went about their normal lives. The smartwatch continuously collected data on heart rate, physical activity, and oxygen saturation, among other metrics.

AI translates data into clinical insights

The collected data were analyzed using an AI model developed and validated by researchers. This model calculated daily cardiorespiratory fitness, an important measure of the coordination between the heart and lungs. Notably, these AI-driven measurements closely matched traditional exercise tests conducted in the hospital.

A key finding is that a 10 percent or greater decline in this fitness was associated with a more than threefold increase in the risk of an unplanned hospital admission or emergency treatment. As such, the wearable functions as an early warning system.

According to the researchers, the innovation lies primarily in the use of real-world data. Instead of measurements taken under controlled conditions, the study examines how patients actually move and function in their daily environment. This provides a more dynamic and realistic picture of their health status.

Patient-centered care

The results highlight the potential of wearables to transform care for heart failure patients from reactive to proactive. By detecting deterioration earlier, healthcare providers can intervene more quickly with, for example, medication adjustments or additional monitoring. This can prevent hospitalizations and improve quality of life.

In addition, this technology can contribute to better access to care. Patients in rural or hard-to-reach areas, in particular, can benefit from continuous remote monitoring without frequent hospital visits.

The researchers emphasize that further validation and integration into clinical workflows are necessary. The ultimate goal is an accessible, virtually continuous monitoring system that provides healthcare providers with real-time insight into patients’ health and enables timely intervention. As such, wearable technology is an important building block for future-proof, data-driven care.

Wearable heart diagnosis

Last year, researchers at the University of Tampere developed a new method that enables smartwatches to detect congestive heart failure, expanding their diagnostic potential beyond atrial fibrillation. The approach, published in Heart Rhythm O2, analyzes intervals between consecutive heartbeats using advanced time-series analysis.

By examining complex patterns in heart rate variability, the method can distinguish between healthy individuals, atrial fibrillation and heart failure. In tests using international ECG datasets, it achieved an accuracy of around 90 percent.

Traditionally, diagnosing heart failure requires costly imaging techniques such as cardiac ultrasound. This new approach could enable affordable, accessible screening using consumer devices like smartwatches. Researchers believe this innovation may support earlier detection, improved treatment outcomes and broader adoption of digital health and self-monitoring tools.