Secondary Analysis of Substance Use In Men

描述（由申请人提供）：最近有充分证据表明，高危人群中新增艾滋病毒感染率的增加与俱乐部毒品（如MDMA，氯胺酮，GHB和甲基苯丙胺）的普及相吻合。本研究旨在通过系统地探索艾滋病毒感染高风险男性药物滥用的流行病学，进一步了解这些联系。该应用程序将对从伊利诺伊州芝加哥获得的三种不同概率样本收集的现有数据集进行广泛的分析：主要的比较分析集中在1996-1997年调查的男性样本和2002-2003年调查的男性样本；作为2001-2002年药物使用一般人口研究的一部分，将采用接受调查的男性样本进行进一步比较。利用平行的工具和措施，在两个芝加哥的样本中，详细的分析将比较和探索物质使用报告的有效性以及高风险和非高风险男性之间物质使用/滥用的相对流行程度。在“认知逃避”概念框架的指导下，该应用程序为了解高风险男性中物质使用流行模式和趋势随时间的变化提供了独特、及时和具有成本效益的机会。多变量模型将系统地检查药物使用与危险行为之间的关系；将比较随时间（跨不同数据集）的关联。为了更好地理解这些联系，我们将探索其他统计方法来组合和调整调查对高风险行为问题的回答，包括单参数项目反应理论（Rasch scaling）和协方差结构测量模型（使用LISREL和MPLUS）。这些缩放方法，反过来，将用于完善和修订统计模型。这一应用将对公共政策产生重大影响，进一步阐明药物滥用在艾滋病毒传播中的因果作用，从而有助于指导今后的艾滋病毒预防和干预工作。

DESCRIPTION (provided by applicant): Recent well-documented increases in the rate of new HIV infections in high risk populations coincide with the spreading popularity of club drugs such as MDMA, ketamine, GHB, & methamphetamine. This research seeks to further understand these linkages by systematically exploring the epidemiology of substance abuse among men at high risk for HIV infection. This application will conduct extensive analyses of existing data sets collected on three different probability samples obtained from Chicago, Illinois: The main comparative analyses focus on a sample of men surveyed in 1996-1997 and an additional sample of men surveyed in 2002-2003; additional comparisons will employ a sample of men surveyed as a part of a general population study of drug use in 2001-2002. Taking advantage of parallel instruments and measures across two Chicago-based samples, detailed analyses will compare and explore substance use reporting validity as well as the relative prevalence of substance use/abuse between high-risk and non-high-risk men. Guided by a "cognitive escape" conceputal framework, this application provides a unique, timely, and cost-effective opportunity for understanding shifts in substance use prevalence patterns and trends over time among high- risk men. Multivariate models will systematically examine associations between drug use and risk behavior; associations over time (across different data sets) will be compared'. In order to better understand these linkages, we will explore alternative statistical approaches for combining and scaling survey responses to questions about high risk behavior/including single parameter item response theory (Rasch Scaling) and covariance structure measurement models (using LISREL and MPLUS). These scaling approaches, in turn, will be used to refine and revise statistical models. This application will have a significant impact on public policy by further articulating the causal role of drug abuse in the spread of HIV, thus helping to guide future HIV prevention and intervention efforts.