Causal Inference
A multi-state Causal Framework for Estimating the Effects of Treatment Regimes over the Course of Chronic Disease Ming Ding* Ming Ding UNC Chapel Hill
The development of chronic disease is a long-term process involving multiple endpoints, and limited methods can assess the health benefits of treatment regimes over the disease course. Existing multi-state Cox models are not easy to use for comparing the effectiveness of treatment regimes. To address this, a discrete-time split-state framework has been proposed [1], which divides disease states into substates by conditioning on past history. As this framework is both “memoryless” and “memorable”, the time-specific transition rates can be synthesized into summary measures, multimorbidity-adjusted life year (MALY) and substate-specific life year (SSLY) [2]. In this abstract, building on this framework, we propose to investigate the causal effects of static and dynamic treatment regimes over the disease course, under the assumptions of constant baseline confounders and instantaneous effects of interventions on transition rates. Our method identifies the optimal treatment regime that generates the most benefits using MALY, and elucidates the mechanisms through which treatment regimes affect disease progression using SSLY. In the application, we identified the optimal body mass index (BMI) intervention using data from the Atherosclerosis Risk in Communities (ARIC) study, where the disease course was modeled in healthy, at metabolic risk, coronary heart disease, heart failure, and mortality states. Compared to Regime 1, the MALY was 1.80 years higher under Regime 4 (Figure), and the estimated SSLY shows that the gain in MALY for Regime 4 is due to prolonged life expectancy in all substates except the healthy state. In summary, our method provides a framework to evaluate the health benefits of treatment regimes over the disease course and has the potential to improve the precision prevention of chronic diseases. This abstract has a preprint online.[3]
[1] Ding et al. BMC Med Res Methodol. 2025;25(1):54.
[2] Ding et al. BMC Med Res Methodol. In revision
[3] Ding. https://doi.org/10.1101/2025.07.25.25332203
