Developing a dynamic modeling framework for surveillance, prediction, and real-time resource allocation to reduce health disparities during Covid-19 and future pandemics

项目摘要 黑人、西班牙裔和农村美国人死于冠状病毒病的可能性是2019年的两倍(新冠肺炎)。这些健康 在整个大流行期间，由于缺乏获得基本资源的机会，加剧了差距。这样的不平等并不是 这是新冠肺炎独有的。在过去的一个世纪里，新出现的传染病显著地延续了健康差距。 在服务不足的社区。导致这些差异的相互关联的途径，包括异质性疾病 流行病学、社会人口学特征以及治疗的可获得性和接受性仍未得到充分研究。移动医疗 诊所(MHC)是通过及时提供干预措施来缩小健康差距的有效和多功能工具 医疗服务不足的人群。然而，无法有效地识别高风险社区并确定其优先顺序 给MHC决策者带来了艰巨的挑战，并导致了次优的分配策略。以帮助改善 提高这些实地干预措施的效率，减少新冠肺炎期间和未来大流行期间的健康差距，我们的 一项提案寻求开发一个建模工具包，以改善服务不足地区的传染病监测和预测 并将向高风险社区实时提供基本资源列为优先事项。我们的创新， 多层次建模框架将利用统计模型、机器学习、基于间隔和基于代理 通过1)建立用于传染病监测的实时数据系统馈送来缩小健康差距的模型 和对服务不足社区的疾病流行病学的估计2)确定高危人群以供分配 基本资源，3)评估社会人口学和临床特征之间的复杂相互作用，传染性 疾病流行病学、可改变的健康障碍和干预吸收，以改进应急计划 新冠肺炎大流行和未来的卫生紧急情况，以及4)建立建模工具包，以便为交付提供信息 向服务不足的社区实时提供必要的资源。这将通过实时集成 传染病结果数据、人口统计、社会经济和临床特征、疫苗迟疑调查、 社区层面的背景因素，以及医疗保健的结构性障碍的数据，以估计#年的关键投入参数 动态仿真建模框架。我们提出的框架将被推广到其他传染性疾病 模型投入将取决于疾病和地点，以便迅速转化为其他公共卫生问题。 为了演示我们工具包的实用性，我们的建模框架将重点放在新冠肺炎移动疫苗接种的交付上 向南卡罗来纳州(南卡罗来纳州)服务不足的人群提供诊所。我们的提案将通过制定 用于疾病监测和了解服务不足地区传染病流行病学的建模基础设施 最终改善向最需要的人及时提供基本资源的情况。新冠肺炎声称 截至2022年2月，近100万美国人死亡，400多万人住院。这一点的利用 公共卫生决策者的工具包可以防止未来数以千计的新冠肺炎死亡。通过对输入进行调整 数据来源，我们的建模框架可以很容易地翻译到其他传染病和地理区域，并具有 有可能在未来的大流行中拯救更多的生命。

Project Summary Black, Hispanic, and rural Americans are twice as likely to die from Coronavirus Disease 2019 (Covid-19). These health disparities have been fueled by inadequate access to essential resources throughout the pandemic. Such inequities are not unique to Covid-19. Over the past century, emerging infectious diseases have significantly perpetuated health disparities in underserved communities. The interconnected pathways leading to these disparities, including heterogeneous disease epidemiology, sociodemographic characteristics, and treatment access and uptake, remain understudied. Mobile health clinics (MHC) are an effective and versatile tool for reducing health disparities through timely delivery of interventions to medically underserved populations. However, the inability to effectively identify and prioritize high-risk communities has posed daunting challenges for MHC decision makers and has led to suboptimal allocation strategies. To help improve the efficiency of these field-level interventions and reduce health disparities during Covid-19 and future pandemics, our proposal seeks to develop a modeling toolkit to improve infectious disease surveillance and prediction in underserved populations and prioritize the delivery of essential resources to high-risk communities in real time. Our innovative, multilevel modeling framework will utilize statistical models, machine learning, compartment-based and agent-based models to reduce health disparities through 1) establishing a real-time data system feed for infectious disease surveillance and estimation of disease epidemiology in underserved communities 2) identifying at-risk populations for allocation of essential resources, 3) evaluating the complex interplay between sociodemographic and clinical characteristics, infectious disease epidemiology, modifiable health barriers, and intervention uptake in order to improve emergency planning during the Covid-19 pandemic and future health emergencies, and 4) establishing a modeling toolkit to inform delivery of essential resources to underserved communities in real-time. This will be accomplished through real-time integration of infectious disease outcome data, demographic, socioeconomic, and clinical characteristics, vaccine hesitancy surveys, community-level contextual factors, and data on structural barriers to health care for estimation of key input parameters in the dynamic simulation modeling framework. The framework we propose will be generalizable to other infectious diseases, where model inputs will be disease and location dependent for swift translation to other public health problems. To demonstrate the utility of our toolkit, our modeling framework will focus on delivery of Covid-19 mobile vaccination clinics to underserved populations in South Carolina (SC). Our proposal will improve pandemic planning by developing the modeling infrastructure for disease surveillance and understanding of infectious disease epidemiology in underserved communities, ultimately improving timely delivery of essential resources to those of greatest need. Covid-19 has claimed nearly 1 million American lives and has hospitalized over 4 million individuals through February 2022. Utilization of this toolkit by public health decision makers can prevent thousands of future Covid-19 deaths. Through adaptation of input data sources, our modeling framework is easily translatable to other infectious diseases and geographic regions and has potential to save many more lives in future pandemics.