HeartMan DSS: A decision support system for self-management of congestive heart failure

• Published in: Expert systems with applications Vol. 186; p. 115688
• Main Authors: Bohanec, Marko, Tartarisco, Gennaro, Marino, Flavia, Pioggia, Giovanni, Puddu, Paolo Emilio, Schiariti, Michele Salvatore, Baert, Anneleen, Pardaens, Sofie, Clays, Els, Vodopija, Aljoša, Luštrek, Mitja
• Format: Journal Article
• Language: English
• Published: New York Elsevier Ltd 30.12.2021; Elsevier BV
• Subjects: Algorithms; Applications programs; Cell phones; Congestive heart failure; Customization; Decision models; Decision support system; Decision support systems; Disease management; Heart failure; Machine learning; Nutrition; Optimization; Patients; Personal health system; Physical exercise; Physicians; Physical exercise; Personal health system; Decision models; Congestive heart failure; Decision support system
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2021.115688

•The HeartMan DSS provides personalized support to congestive heart failure management.•The DSS recommendations address physical health, psychological support, and environment.•Physical health aspects include physical exercise, nutrition and medication therapy.•Employed methods are varied and include expert modelling, machine learning and optimization.•A clinical trial indicated improved medical parameters and positive patients’ response. Congestive heart failure is a chronic medical condition that affects about 2% of the adult population. Even though it cannot be cured, it can be relieved by a proper, long-term, complex and personalized disease management. In this paper we present the HeartMan Decision Support System (DSS), aimed at supporting individual patients in their uptake of well-established clinical guidelines (i.e., both medication and behaviour based) for disease management. The HeartMan DSS is a central component of the wider HeartMan mobile-health platform that employs mobile phones, wristband sensors and a web application for communication with patients, their physicians and caregivers. The DSS itself provides recommendations for (1) managing patient’s physical health in terms of exercise, nutrition, medications and self-monitoring, (2) psychological support, and (3) managing environmental parameters. The DSS employs a variety of methods: rule-based decision models and adaptable workflows developed using literature and in collaboration with medical experts, classification models developed by machine learning from data, and optimization algorithms. Taken together, they provide a comprehensive, personalized and user-friendly disease management platform. The system was evaluated in a clinical proof-or-concept trial, involving 56 patients in four hospitals. The results confirmed that the system was successful in improving self-care behaviour, decreased patients' levels of depression and anxiety, and improved the overall predicted 1-year mortality risk.