Impact of neighborhood and workforce deprivation on diabetes outcomes in Veterans: a spatio-temporal analysis

对退伍军人医疗保健的预期影响：该项目将提出一个全面的地理空间 框架，以解决卓越战略3的VA蓝图：利用信息技术， 分析和医疗保健模型，以优化个人福祉和人口健康结果。 通过创建包含健康信息、劳动力生产率、 邻里剥夺，我们将开发一个全面的数据库，以检查多个方面的 糖尿病护理通过使用先进的地理信息系统和时空统计，我们将确定热点的高 疾病风险，贫困的社区资源，以及低VA劳动力能力。这些信息将改善 通过帮助退伍军人事务部决策者更好地将资源与成果不佳的地区相匹配，获得护理。最后通过 精确定位卫生支出过多的地区，VA可以制定降低成本的措施，以改善 退伍军人的健康，同时控制成本。 背景：糖尿病是美国第七大死亡原因，可导致严重的 并发症，并与增加的医疗费用有关。退伍军人的患病率估计显示， 不成比例的疾病负担，估计接近25%，相比之下，一般美国的8% 人口有证据表明，少数民族的糖尿病患病率更高， 结果，并发症的风险更高，与非西班牙裔白人相比，死亡率更高。这 在控制了患者层面的因素（如教育、收入、知识、健康）后， 识字和自我效能;提供者层面的因素，如偏见，沟通和信任;以及系统层面的因素， 例如，获得护理。很少有人注意到可能由区域差异造成的差异。 病人一级资源、社区一级资源和卫生人力资源的变化。 目的：本研究旨在确定和解释健康结果的空间和时间变化， 社区资源、VA劳动力能力和2型糖尿病患者的健康差异。目的 1将检查糖尿病结局的时空趋势，包括代谢控制，成本和死亡率。 目标2将开发一个新的时空邻里剥夺指数，并研究其与 糖尿病结果和种族差异。目标3将开发和验证一个新的地理劳动力 剥夺指数，以研究其与糖尿病结局和种族差异的关系。 方法：我们将建立一个队列的退伍军人与2型糖尿病接受住院或门诊 在2000年至2015年期间，通过链接来自VA的多个患者和管理文件， VHA国家患者护理和药房福利管理数据库，使用先前验证的VA 算法利用先进的地理信息系统和空间统计方法，我们将研究 2型糖尿病患者中的糖尿病结局。在目标1中，我们将开发一个灵活的贝叶斯 时空模型，以确定糖尿病相关结果高发的热点。在目标2和3中， 我们将使用时空潜在因素模型来开发新的社区和劳动力剥夺 指数，使我们能够调查社区资源可用性和退伍军人管理局劳动力的不断变化的模式 容量这些目标的完成将使退伍军人事务部能够确定个人、社区和机构因素 与糖尿病预后不良相关，并针对社区和系统层面的努力， 低资源地区。

Anticipated Impacts on Veterans Health Care: This project will put forth a comprehensive geospatial framework to address the VA Blueprint for Excellence Strategy 3: Leverage information technologies, analytics, and models of healthcare to optimize individual well-being and population health outcomes. By creating a spatially referenced dataset incorporating health information, workforce productivity, neighborhood deprivation, we will develop a comprehensive database to examine multiple dimensions of diabetes care. Through the use of advanced GIS and spatiotemporal statistics, we will identify hotspots of high disease risk, poor neighborhood resources, and low VA workforce capacity. This information will improve access to care by helping VA policy makers better match resources to areas with poor outcomes. Finally, by pinpointing areas with excessive health expenditures, the VA can develop cost-reduction measures to improve Veterans’ health while containing costs. Background: Diabetes is the seventh leading cause of death in the United States, can lead to serious complications, and is associated with increased healthcare costs. Prevalence estimates for Veterans show a disproportionate burden of disease, with estimates close to 25%, as compared to 8% of the general US population. Evidence consistently shows racial minorities have a higher prevalence of diabetes, worse outcomes, higher risk of complications, and higher mortality rate compared to non-Hispanic whites. This disparity persists after controlling for patient-level factors such as education, income, knowledge, health literacy, and self-efficacy; provider-level factors, such as bias, communication, and trust; and system-level factors, such as access to care. Little attention has been given to differences that may be explained by regional variation in patient-level resources, community-level resources, and health workforce resources. Objectives: This study seeks to identify and explain spatial and temporal variation in health outcomes, community resources, VA workforce capacity, and health disparities among patients with type 2 diabetes. Aim 1 will examine spatiotemporal trends in diabetes outcomes, including metabolic control, cost, and mortality. Aim 2 will develop a new spatiotemporal neighborhood deprivation index and examine its association with diabetes outcomes and racial disparities. Aim 3 will develop and validate a novel geographic workforce deprivation index to examine its association with diabetes outcomes and racial disparities. Methods: We will construct a cohort of veterans with type 2 diabetes receiving either inpatient or outpatient care at the VA during the years 2000 through 2015 by linking multiple patient and administrative files from the VHA National Patient Care and Pharmacy Benefits Management databases, using a previously validated VA algorithm. Using advanced GIS and spatial statistical methods, we will examine spatiotemporal trends in diabetes outcomes among Veterans with type 2 diabetes. In Aim 1, we will develop a flexible Bayesian spatiotemporal model to identify hotspots of high prevalence of diabetes-related outcomes. In Aims 2 and 3, we will use spatiotemporal latent factor models to develop novel neighborhood and workforce deprivation indices, allowing us to investigate evolving patterns in community resource availability and VA workforce capacity. Completion of these aims will enable the VA to identify individual, community, and institutional factors associated with poor diabetes outcomes and to target community and system-level efforts to improve health in low-resource areas.