Modeling MRSA in the Community

描述(申请人提供)：该项目的目标是创建一种新的基于试剂的耐甲氧西林金黄色葡萄球菌(MRSA)模型。MRSA是一种耐药细菌，最常导致皮肤感染，但可导致任何器官的严重和致命感染。在过去十年中，耐甲氧西林金黄色葡萄球菌呈指数级传播，是对公共卫生的严重威胁。新的MRSA菌株在社区中迅速进化，在近50年前首次出现的医疗环境中与MRSA相互作用。MRSA具有许多独特的功能，这些功能代表了基于代理的建模的挑战和机遇。我们的模型将允许代理人根据疾病状况和风险感知调整他们的行为。该模型将包括代表与传染病传播和控制相关的人口和医疗保健系统的行为特征的基于理论和经验的变量。这些变量将代表与疾病传播相关的社交网络、与信息传输相关的社交网络以及遵循健康建议的个人倾向差异。这个项目汇集了具有独特跨学科专业知识的研究人员：流行病学、MRSA、社会科学、贝叶斯统计学、基于代理的建模、高性能计算和公共卫生。具体地说，我们将开发一个灵活的基于代理的模型，逐步扩大到芝加哥大都市区的人口。第一阶段将包括一条横跨芝加哥南侧和邻近郊区的走廊，该地区的MRSA一直是一个严重的问题，并已在社区和医疗保健环境中进行了广泛研究。在我们研究团队在MRSA方面的专业知识的指导下，我们的模型将捕获环境的特征和个人的特征，这些特征在MRSA流行病学中特别突出。我们将开发变量来表示天生的个体MRSA风险、特定位置的MRSA传播概率和杂合体(存在MRSA的无生命物体)的污染。使用该模型，我们将检验有关导致MRSA扩散的因素的假设。我们将确定哪些临床、公共卫生和机构措施可能对疫情产生最大影响。拟议工作的一个重要组成部分将是研究模型结果的变化(即模型的不确定性)，以及模型参数、网络规范和其他变量的变化对这种变化的影响。要做到这一点，我们将使用芝加哥大学和阿贡国家实验室的高性能计算能力来运行单个模型的配置数千次。 公共卫生相关性：计算模型已被用于了解不同类型的传染病控制干预措施的有效性。在这个项目中，我们模拟了耐甲氧西林金黄色葡萄球菌(MRSA)的传播和控制，这是一种耐药细菌，现在美国每年死于这种细菌的人比死于艾滋病的人还多。我们的模型将引入代表感染相关行为变化的变量，将有助于确定感染控制和治疗的最佳实践。

DESCRIPTION (provided by applicant): The goal of this project is to create a new agent-based model of methicillin resistant Staphylococcus aureus (MRSA), an antibiotic-resistant bacterium that most commonly causes skin infections but can cause serious and fatal infections of any organ. During the past decade, MRSA has spread exponentially and is a serious public health threat. New strains of MRSA have rapidly evolved in the community, interacting with MRSA in the healthcare environment, which first appeared almost 50 years ago. MRSA has a number of distinctive features that represent challenges and opportunities for agent-based modeling. Our model will allow agents to adapt their behavior dependent on disease conditions and perceptions of risk. The model will include theoretically based and empirically derived variables representing behavioral features of the population and the healthcare system that are relevant to infectious disease transmission and control. These will include variables representing social networks relevant for disease transmission, social networks relevant for information transfer and individual variation in propensity to follow health recommendations. This project brings together investigators with unique interdisciplinary expertise: epidemiology, MRSA, social sciences, Bayesian statistics, agent-based modeling, high performance computing, and public health. Specifically, we will develop a flexible agent-based model scaling up in stages to the population of the Chicago metropolitan area. The 1st stage will include a corridor across the south side of Chicago and adjoining suburbs, an area in which MRSA has been a serious problem and has been extensively studied both in community and healthcare settings. Guided by our research team's expertise in MRSA, our model will capture features of the environment and characteristics of the individual that are particularly salient for MRSA epidemiology. We will develop variables to represent innate individual MRSA risk, location-specific MRSA transmission probabilities, and contamination of fomites (inanimate objects on which MRSA is present). Using the model, we will test hypotheses about factors contributing to MRSA spread. We will determine which clinical, public health and institutional measures are likely to have the greatest impact on the epidemic. An important component of the proposed work will be to study variation in model outcomes (i.e., model uncertainty), and the effect of changes in model parameters, network specifications, and other variables on this variation. To do this, we shall use high-performance computing capabilities at the University of Chicago and Argonne National Laboratory to run individual model configurations thousands of times. Public Health Relevance: Computational models have been used to understand the effectiveness of different types of control interventions for infectious diseases. In this project, we model the spread and control of methicillin resistant Staphylococcus aureus (MRSA), an antibiotic-resistant bacterium that now kills more people each year in the U.S. than AIDS. Our models, which will introduce variables representing variation in infection-related behavior, will help identify best practices for infection control and treatment.