Hu Pei, Zhong Fei, Cheng Wei-Bin, Xu Hui-Fang, Ling Li
School of Public Health, Sun Yat-Sen University, Guangzhou 510080, China.
Zhonghua Liu Xing Bing Xue Za Zhi. 2012 Jul;33(7):667-71.
To develop a human immune deficiency virus (HIV) infection risk appraisal model suitable for men who has sex with men (MSM) in Guangzhou, and to provide tools for follow-up the outcomes on health education and behavior intervention.
A cros-sectional study was conducted in Guangzhou from 2008 to 2010. Based on the HIV surveillance data, the main risk factors of HIV infection among MSM were screened by means of logistic regression. Degree on relative risk was transformed into risk scores by adopting the statistics models. Individual risk scores, group risk scores and individual infection risk in comparison with usual MSM groups could then be calculated according to the rate of exposure on those risk factors appeared in data from the surveillance programs.
Risk factors related to HIV infection among MSM and the quantitative assessment standard (risk scores and risk scores table of population groups) for those factors were set up by multiple logistic regression, including age, location of registered residence, monthly income, major location for finding their sexual partners, HIV testing in the past year, age when having the first sexual intercourse, rate of condom use in the past six months, symptoms related to sexually transmitted diseases (STDs) and syphilis in particular. The average risk score of population was 6.06, with risk scores for HIV positive and negative as 3.10 and 18.08 respectively (P < 0.001). The rates of HIV infection for different score groups were 0.9%, 2.0%, 7.0%, 14.4% and 33.3%, respectively. The sensitivity and specificity on the prediction of scores were 54.4% and 75.4% respectively, with the accuracy rate as 74.2%.
HIV infection risk model could be used to quantify and classify the individual's infectious status and related factors among MSM more directly and effectively, so as to help the individuals to identify their high-risk behaviors as well as lifestyles. We felt that it could also serve as an important tool used for personalized HIV health education and behavior intervention programs.
建立适合广州男男性行为者(MSM)的人类免疫缺陷病毒(HIV)感染风险评估模型,为健康教育和行为干预效果的跟踪提供工具。
于2008年至2010年在广州开展横断面研究。基于HIV监测数据,通过逻辑回归筛选MSM中HIV感染的主要危险因素。采用统计模型将相对危险度转化为风险评分。然后根据监测项目数据中出现的这些危险因素的暴露率,计算个体风险评分、群体风险评分以及与普通MSM群体相比的个体感染风险。
通过多元逻辑回归建立了MSM中与HIV感染相关的危险因素及其定量评估标准(人群组风险评分及风险评分表),包括年龄、户籍所在地、月收入、寻找性伴的主要场所、过去一年的HIV检测情况、首次性行为年龄、过去六个月的避孕套使用比例、性传播疾病(STD)尤其是梅毒相关症状。人群平均风险评分为6.06,HIV阳性和阴性者的风险评分分别为3.10和18.08(P<0.001)。不同评分组的HIV感染率分别为0.9%、2.0%、7.0%、14.4%和33.3%。评分预测的敏感度和特异度分别为54.4%和75.4%,准确率为74.2%。
HIV感染风险模型可更直接有效地对MSM个体的感染状况及相关因素进行量化和分类,有助于个体识别自身的高危行为及生活方式。我们认为它也可作为个性化HIV健康教育和行为干预项目的重要工具。