Personalised algorithms in heart failure telemedicine to improve self-management

This project aims to develop personalised algorithms within the already existing EMPOWER platform (24CARE) to improve prediction of worsening heart failure. 

Despite recent advances in care, chronic heart failure remains associated with high morbidity and mortality. In 2017, there are an estimated 230,100 patients diagnosed with heart failure in the Netherlands. The prevalence of heart failure is increasing and is associated with high morbidity and mortality. The health care costs for heart failure care amounted 937 million euros, in 2015.

Using personalised algorithms, heart failure nurses will be spending less time adjusting thresholds for biometric measurements (e.g. blood pressure, weight) and unnecessary (remote or physical) visits by patients. The patients will only contact the heart failure nurse when he/she has signs of worsening of heart failure or drug side effects. The expectation is that the frequency of (telephonic/video) consults with heart failure nurses will decrease due to calculating thresholds based on actual biometric measurements of the individual patient. Feedback of valid alerts to patients will increase their self-management and thereby empowering patients to prevent worsening of their heart failure. Patients manage their health in their home setting through prediction of worsening of heart failure and prevention through increasing awareness and early interventions. This may eventually lead to reduction in hospital visits and thereby costs.

In work package 1, proven time series modelling will be applied to the individual biometric series. In work package 2 the separate patient-level biometric measurements will be integrated into a single model. In work package 3, the algorithms are validated, using actually observed patient outcomes. The end result will be validated personalised algorithms ready to use in the EMPOWER platform.