PhD defence X. (Xinyu) Li

Towards better health economic decision-making on precision medicine in type 2 diabetes mellitus

Type 2 Diabetes is a leading global health challenge, contributing significantly to disability-adjusted life years and healthcare costs. Despite the availability of effective treatments, many patients fail to achieve optimal disease management, largely due to the heterogeneity of diabetes and the limitations of current clinical approaches. This thesis seeks to improve health economic decision-making in diabetes care by enhancing the understanding of diabetes subgroups and refining health economic models to support more precise and cost-effective treatment strategies.Using real-world data from multiple cohorts, this research examined the benefits of data-driven diabetes subgroups in predicting health outcomes and guiding treatment decisions. Findings indicate that k-means clustering-based subgroups, such as Severe Insulin-Resistant Diabetes (SIRD) and Severe Insulin-Deficient Diabetes (SIDD), differed in terms of metabolic profiles, complication risks, and treatment responses. However, alternative ways to define subgroups using clinical markers such as HbA1c levels and cardiovascular risk, were more useful than k-means clustering to identify patients who benefit most from early intervention.To address limitations in economic modeling, this thesis also evaluated current diabetes models and proposed methodological improvements. A cohort-level model, the MICADO model, was validated and applied to assess the cost-effectiveness of SGLT2 inhibitors in routine care, highlighting discrepancies between trial data and real-world populations. Finally, a meta-regression tool was introduced to systematically estimate risk parameters for this type of models.This work contributes to the development of methods for the economic evaluation of precision medicine strategies and more efficient resource allocation in diabetes care.