Studies in Health Technology and Informatics, 2017. 244 (The Practice of Patient Centered Care: Empowering and Engaging Patients in the Digital Era): pp. 18-22.
Benis A, Harel N, Barkan R, Sela T, Feldman B.
Abstract
HMOs record medical data and their interactions with patients. Using this data we strive to identify sub-populations of healthcare customers based on their communication patterns and characterize these sub-populations by their socio-demographic, medical, treatment effectiveness, and treatment adherence profiles. This work will be used to develop tools and interventions aimed at improving patient care. The process included: (1) Extracting socio-demographic, clinical, laboratory, and communication data of 309,460 patients with diabetes in 2015, aged 32+ years, having 7+ years of the disease treated by Clalit Healthcare Services; (2) Reducing dimensions of continuous variables; (3) Finding the K communication-patterns clusters; (4) Building a hierarchical clustering and its associated heatmap to summarize the discovered clusters; (5) Analyzing the clusters found; (6) Validating results epidemiologically. Such a process supports understanding different communication-channel usage and the implementation of personalized services focusing on patients’ needs and preferences.