Improving cardiac rehabilitation patient adherence via personalized interventions

Abstract

OBJECTIVES: Despite documented benefits and physicians' recommendations to participate in cardiac rehabilitation (CR) programs, the average dropout rate remains between 12-56%. This study's goal was to demonstrate that using personalized interventions can significantly increase patient adherence. METHOD: Ninety-five patients (ages 18-90) eligible for the CR program were randomly recruited and received personalized interventions using the Well-Beat system. Adherence levels were compared to those of a historical control group. The Well-Beat system provided Sheba CR Health Care Provider (HCP) guidelines for personalized patient-therapist dialogue. The system also generated ongoing personalized text messages for each patient sent twice a week and related each patient's dynamic profile to their daily behavior, creating continuity, and reinforcing the desired behavior. RESULTS: A significant increase in patient adherence to the CR program: Three months after initiation, 76% remained active compared to the historical average of 24% in the matched control group (log-rank p-value = 0.001). CONCLUSIONS: Using an Artificial Intelligence (AI)-based engine that generated recommendations and messages made it possible to improve patient adherence without increasing HCP load, benefiting all. Presenting customized patient insights to the HCP and generating personalized communications along with action motivating text messages can also be useful for remote care.