Home Outpatient Monitoring and Engagement to Predict HF Exacerbation

The HOME PREDICT HF study looks at new ways to predict hospitalizations for heart failure. We will use a set of devices at home and surveys to collect information about patient's health. This study uses the Eureka app, a new study app developed by the University of California, San Francisco. The study is designed to happen remotely, using this application on a patient's smartphone, so that is as convenient as possible to participate.

HOME PREDICT HF is single center, prospective, unblinded, randomly assigned training and validation observational cohorts to develop machine learning algorithms from an in-home suite of sensors in order to predict 3-month heart failure hospitalization and/or emergency department visits. Study population includes adults presenting with a diagnosis of reduced ejection fraction (LVEF <= 40%), NYHA class II-IV) who have had a hospitalization for HF in the previous 6 months. The study objectives include (1) To collect observational data from multiple sensors, patient-reported outcomes, and medical record data to develop (train) machine-learning algorithms (2) To validate trained algorithms in a separate validation cohort (3) To collect data to inform the design of a future intervention study. The primary outcome is Ninety-day heart failure hospitalization/emergency department visit for heart failure.