Telemonitoring predicts in advance heart failure admissions

Heart failure (HF) is increasingly common and characterised by frequent admissions to hospital. To reduce the risk of HF hospitalisation (HFH), approaches as telemonitoring (TM) have been introduced. This study aimed to develop an algorithm for detecting patients at high risk of HFH, using daily collected physiological data (blood pressure, heart rate, weight) by non-invasive TM.

The analysis was based on home-TM data collected from a single centre as part of HF care. The prediction of HFH was considered as a signal processing and classification problem. Signal processing aimed to transform the signals to enhance the information relevant to HFH. We attempted to construct an algorithm that could identify such patterns and classify them as abnormal by assessing the predictive value of each of the monitored signals and their combinations using analysis of vectors (e.g. vectors of raw signal values, vectors of signals obtained by Multi-Resolution Analysis).

The best predictive results were achieved with the combined used of weight and diastolic BP. The highest predictive performance was achieved using 8-day TM data (area under the receiver operator characteristic curve (AUC) 0.82±0.02). Prediction based on 4-day TM data was slightly less accurate with an AUC of 0.77±0.01.

We have found that using an algorithm based on weight and diastolic blood pressure measured over 8days predicts heart failure admissions with a high degree of accuracy. The value of such an algorithm should be tested in clinical trials.