JMIR Publications

ABSTRACT

Background:

Telemonitoring strategies in heart failure (HF) patients are often based on manual readings and interpretation of various parameters by healthcare professionals (HCPs). Automated multiparameter prediction models (MPMs) have the potential to improve early recognition of decompensated HF and to reduce the workload for both HCPs and patients. To reduce costs and facilitate large-scale implementation, these models should preferably be based on non-invasive measurements, with user-friendly devices.

Objective:

This exploratory study aimed to evaluate whether a MPM, using various parameters from a wrist-worn device supplied with a photoplethysmography sensor and a triaxial accelerometer, contributes to the detection of decompensated HF and death in patients with unstable HF.

Methods:

Patients who were admitted to the hospital with acute decompensated HF, regardless of etiology or left ventricular ejection fraction, were instructed to wear a research-grade wrist-worn device from the moment of discharge. The device measured heart rate (HR)-related data, respiration rate (RR), activity, and sleep. Participants were instructed to wear the device 24 hours a day for three consecutive months. A decision-tree-based MPM was developed based on the collected data and compared to four alternative prediction models. The combined endpoint of interest was hospital readmission due to decompensated HF, decompensated HF treated at the outpatient clinic by increasing loop diuretic dose, or death due to HF.

Results:

17 patients participated in the study (median age 77 (interquartile range 70-84) years, 53% male). During follow-up, the device wearing compliance was 78% (55-81%). The activity-related parameters (energy expenditure and activity counts) performed best with respect to data quality: 72% and 79% of the data were of high quality, respectively. Concerning HR, 46% of the data was of high quality, whereas only 14% of the RR data was of high quality. Sleep data was lacking in 99% of the time during follow-up. The most optimal prediction model for the detection of the combined endpoint showed a specificity of 98% and a sensitivity of 3% in the two weeks prior to an event (area under the curve: 0.5).

Conclusions:

A MPM using a non-invasive wrist-worn device, measuring HR, RR, activity and sleep, showed high specificity, but low sensitivity for the prediction of decompensated HF and HF-related mortality. An important reason for the low sensitivity could be low data quality (specifically HR, RR, sleep) in a real-life setting in this elderly HF population. Future studies are needed to address this issue and include larger populations. Clinical Trial: Netherlands Trial Register (NL9038)