A non-randomized, double-blind study of 521 patients with heart failure found that the combination of data from minute ventilation and physical activity sensors could predict 34 percent of heart failure events, according to a study published in the European Journal of Heart Failure.

"Monitoring early signs of clinical deterioration could allow physicians to adjust medical treatment for patients at risk of acute heart failure decompensation," researchers write in the abstract. "To date, several strategies using different surrogate measures of clinical status emerged, but none has yet been proven to predict clinical events."

The algorithm uses the combined data from minute ventilation, which measures the lung ventilation process per minute, and physical activity sensors to predict heart failure events in patients who have implanted cardiac resynchronization therapy with defibrillation (CRT-D) devices. Patients were implanted with Sorin Group's Paradym CRT device.

The average follow-up time after patients started the study was about a year and a half and in that time around 25 percent or 130 patients experienced a heart failure event. Based on this, researchers reported that the sensitivity to predict an event was 34 percent and that the false positive rate was 2.4 patients per year.

Another digital health company focused on predicting heart failure, Perminova announced in February that it is developing a necklace for clinical use that can track various vital signs including thoracic fluid levels, which is an early predictor of congestive heart failure. This predictor of congestive heart failure isn’t currently monitored by most connected sensors.