CAREER: Advancing Disease Modeling for Decision Making in Healthcare

这项教师早期职业发展计划（Career）资助将通过整合个人层面和人口层面的传染病模型来支持数据驱动的医疗保健决策，为国家健康和福利的进步做出贡献。该项目旨在建立疾病建模的统一理论，以纳入可能影响疾病传播，进展和恢复的流行趋势的人类行为的变化。这项研究将深入了解如何利用这些模型来提高公共卫生服务的质量和成本效益。与临床医生和卫生政策制定者的伙伴关系将确保这些模型解决政策关切并促进跨学科合作和数据共享。该项目的教育目标是促进临床合作者和学生的STEM教育，包括那些来自代表性不足和少数族裔社区的学生。本研究研究状态空间聚集方法，以统一个体水平的模拟与疾病建模的区室模型。目标是在单一和重复的决策环境中优化这些模型，并将这些方法集成到成本效益分析中。这项工作的动机是需要包含行为和群体异质性的可处理疾病模型，这需要数据驱动的方法将个体聚集到可解释的群体中，同时尊重潜在疾病过程的动态。了解如何将模拟结果投射到低维空间，并弥合建模中个人和群体级别表示之间的差距，将允许这些方法更好地为政策提供信息。具体来说，我们将研究这些模型的结构特性、平衡和过渡动力学，目的是利用它们简化疾病动力学的优化和马尔可夫决策过程框架。然后，可以将这一领域的创新纳入成本效益和概率敏感性分析，以便对可行的政策空间进行更有效和彻底的评估。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This Faculty Early Career Development Program (CAREER) grant will contribute to the advancement of national health and welfare by integrating individual-level and population-level models of infectious diseases to support data-driven healthcare decision making. This project aims to build a unifying theory of disease modeling to incorporate variation in human behavior that can affect epidemic trends in disease transmission, progression, and recovery. The research will provide insights into how these models can be used to improve both quality and cost-effectiveness of public health services. Partnership with clinicians and health policy makers will ensure that the models address policy concerns and advance interdisciplinary collaboration and data sharing. The educational goals of the project aim to advance STEM education among clinical collaborators and students, including those from underrepresented and minority communities.This research studies state space aggregation methods to unify individual-level simulations with compartmental models for disease modeling. The objectives are to optimize these models in single and repeated decision contexts, and to integrate these methods in cost-effectiveness analysis. This work is motivated by the need for tractable disease models that incorporate behavior and population heterogeneity, which require data-driven methods to aggregate individuals into interpretable groups while respecting the dynamics of the underlying disease process. Understanding how simulation outcomes can be projected to a lower-dimensional space and bridging the gap between individual and group-level representation in modeling will allow for these methods to better inform policy. Specifically, structural properties, equilibria, and transition dynamics of these models will be studied with the goal of using them to simplify optimization and Markov decision process frameworks on disease dynamics. Innovations in this area can then be integrated into cost-effectiveness and probabilistic sensitivity analysis to provide more efficient and thorough evaluation of the feasible policy space.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.