Personalized follow-up strategies with learning effects for disease monitoring

Abstract

Effective follow-up strategies are crucial for managing patients’ risks of adverse outcomes (AOs) and associated costs. Current literature on follow-up strategy design primarily focuses on healthcare providers’ perspectives, often overlooking the significant role of patient learning behaviors in enhancing follow-up effectiveness during their healthcare journey. This paper investigates the impacts of two types of learning behaviors on follow-up strategy design. By employing the ‘virtual age’ and ‘learning parameters’, we assess the impact of follow-up services and learning behaviors on AO risks. A unified optimization model, based on patient heterogeneity, is then constructed to analyze the trade-off between follow-up services, AO risks, and the impact of patient learning behaviors. Formulated as a mixed integer nonlinear programming problem, the model is solved to determine the optimal frequency and timing of follow-up services over a planned horizon for heterogeneous patient groups. A case study focusing on pediatric type 1 diabetes mellitus patients demonstrates that learning behaviors can effectively control medical service costs while enhancing disease monitoring efficacy.