Assessing Patterns of HIV Risk Among IDUs in Neighborhood Environments over Time

描述（由申请人提供）：本研究响应PAR 07-143“环境、行为和艾滋病毒/艾滋病途径的研究”，采用创新的方法研究了1987 - 2005年旧金山湾区注射吸毒者（IDUs）艾滋病毒风险行为的时空模式和艾滋病毒预防服务。我们建议建立社区特征、社区艾滋病预防服务和注射吸毒者个体艾滋病风险行为结果之间的关系模型。关于艾滋病毒和非法药物使用的研究传统上只侧重于产生风险的个人层面特征。艾滋病毒和非法药物使用领域的新研究表明，在研究个人健康行为时，必须考虑到社会环境因素，包括获得保健服务的机会。拟议的研究涉及将新出现的空间分析方法应用于次要数据集，以完善关于社区特征如何与街头注射吸毒者获得艾滋病毒预防服务和艾滋病毒风险相关的知识。我们的工作将改编著名的弱势群体行为模型，将环境变量与个体行为之间的空间相互作用的成熟概念化纳入其中。我们从1987年至2005年的城市健康研究中对街头注射吸毒者进行了35000次连续横断面访谈，获得了个人水平和艾滋病毒检测数据。其他使用的数据包括来自美国人口普查的档案数据、人口普查邻里变化数据库、艾滋病毒/艾滋病报告、犯罪报告和艾滋病毒预防服务的地理编码位置（包括艾滋病毒检测站点、药物治疗项目和注射器交换项目）。具体目标包括：描述19年来旧金山湾区人口普查区街头注射吸毒者社区特征、艾滋病毒预防服务和艾滋病毒风险行为的地理分布；具体目标2：研究旧金山湾区社区特征（如社区劣势或剥夺）、艾滋病毒预防服务（如药物治疗、注射器交换、艾滋病毒检测）和注射吸毒者艾滋病毒风险行为（如共用注射器、无保护的性行为）之间的统计关联。目标1将涉及制作普查区域一级的地图，以说明关键变量随时间的地理分布。目标2将研究社区特征、艾滋病毒预防服务和艾滋病毒风险行为之间的横断面和时间关联。分析方法将包括多层次选择偏差模型和空间回归模型。该项目是首批利用丰富的地理空间背景数据集以及19年个人层面数据的独特连续横截面的项目之一，利用先进的空间分析方法，更好地了解背景变量与街头注射吸毒者中艾滋病毒风险行为之间的关系。公共卫生相关性：该项目是第一批利用丰富的地理空间上下文数据集的项目之一

DESCRIPTION (provided by applicant): The proposed study responds to PAR 07-143 "Research on Pathways Linking Environments, Behaviors and HIV/AIDS" by using innovative methods to examine the space-time patterns of HIV risk behaviors among injection drug users (IDUs) and HIV prevention services in the San Francisco Bay Area from 1987 to 2005. We propose to model the relationship among neighborhood characteristics, neighborhood HIV prevention services, and individual HIV risk behavior outcomes among IDUs. Research on HIV and illicit substance use has traditionally focused solely on individual-level characteristics that confer risk. Emerging research in the areas of HIV and illicit substance use has identified the importance of considering social environmental factors, including access to health care services, when studying individual health behaviors. The proposed study involves applying newly emergent spatial analytical methods to secondary datasets in order to refine the knowledge regarding how neighborhood characteristics might be associated with access to HIV prevention services and HIV risk among street-based IDUs. Our work will adapt the well-known Behavioral Model of Vulnerable Populations to incorporate a well-developed conceptualization of the spatial interaction between contextual variables and individual behavior. We have individual-level and HIV test data from 35,000 serial cross-sectional interviews among street-based IDUs from the Urban Health Study from 1987 to 2005. Other data to be used include archival data derived from the U.S. Census, the Census Neighborhood Change Database, HIV/AIDS reports, crime reports, and geocoded locations of HIV prevention services (including HIV testing sites, drug treatment programs, and syringe exchange programs). The specific aims include Specific Aim 1: To describe the geographic distribution of neighborhood characteristics, HIV prevention services, and HIV risk behaviors among street-based IDUs in San Francisco Bay Area census tracts over a 19-year period; and Specific Aim 2: To examine the statistical associations among neighborhood characteristics (e.g., neighborhood disadvantage or deprivation), HIV prevention services (e.g., drug treatment, syringe exchange, HIV testing) and HIV risk behaviors of IDUs (e.g., syringe sharing, unprotected sex) in the San Francisco Bay Area. Aim 1 will involve the creation of census-tract-level maps to illustrate the geographic distributions of key variables over time. Aim 2 will examine the cross-sectional and temporal associations among neighborhood characteristics, HIV prevention services, and HIV risk behaviors. Analytical methods will include multilevel selection bias models and spatial regression models. This project is among the first to utilize a rich set of geospatial contextual data in conjunction with a unique serial cross-section of 19 years of individual-level data, using advanced spatial analytic methods to better understand the relationships between contextual variables and HIV risk behaviors among street-based IDUs. PUBLIC HEALTH RELEVANCE: This project is among the first to utilize a rich set of geospatial contextual data in conjunction with a unique serial cross-section of 19 years of individual-level data, using advanced spatial analytic methods to better understand the relationships between contextual variables and HIV risk behaviors among street-based injection drug users (IDUs). This work will advance our methodological understanding of how best to study this population using cutting-edge methods that examine the spatial patterns of HIV risk behaviors and the relationship between neighborhood context and availability of HIV prevention services in metropolitan areas. The findings will benefit policy makers, service providers, and researchers.