Wang Jichuan
Wright State University, School of Medicine, Department of Community Health, Dayton, OH 45435, USA.
AIDS Behav. 2003 Mar;7(1):1-8. doi: 10.1023/a:1022541504943.
Comparative studies on regional HIV seroprevalence or seropositivity rate among injection drug users (IDUs) have focused primarily on assessing the risk factors for HIV infection. This study used a nonparametric analytic approach, known as standardization and decomposition, to to compare HIV seropositivity rates among IDUs between low- and high-HIV-prevalence regions in the United States. The regional difference in HIV seropositivity rate was decomposed into different components: (1) a "rate effect," which was attributed to the differences in factor-specific rates, and (2) "compositional factor effects," which were attributed to the differences in distributions of sociodemographic factors across regions. The analytic results show that the regional difference in HIV seropositivity rate was considerable (21.04%); however, the difference would be adjusted down to 17.65% if sociodemographic factors were proportionally distributed across the regions. Differential distribution of ethnic groups between the two regions accounted for about 15.02% of the regional difference in HIV seropositivity rate. The application of the standardization and decomposition method provides HIV researchers with opportunities to look at familiar data from a different perspective.
针对注射吸毒者（IDU）中区域HIV血清流行率或血清阳性率的比较研究主要集中在评估HIV感染的风险因素。本研究采用一种非参数分析方法，即标准化和分解法，来比较美国HIV高流行区和低流行区IDU中的HIV血清阳性率。HIV血清阳性率的区域差异被分解为不同的组成部分：（1）“率效应”，归因于特定因素率的差异；（2）“构成因素效应”，归因于社会人口学因素在各区域分布的差异。分析结果表明，HIV血清阳性率的区域差异相当大（21.04%）；然而，如果社会人口学因素在各区域按比例分布，该差异将调整至17.65%。两个区域间不同种族的分布差异约占HIV血清阳性率区域差异的15.02%。标准化和分解法的应用为HIV研究人员提供了从不同角度审视常见数据的机会。
