HIV Prevalence Among Drug Users In Both Rehabilitation Center And Community And Size Estimation In Three Prefectures, Guangdong Province

[Objectives]1. To compare the differences of HIV prevalence between community-based andrehabilitation center-based drug users in three prefectures，Guangdong province in orderto provide basis to correctly assess the HIV prevalence in this population scientifically.2. Estimate the size of drug users in three prefectures to evaluate the preventiveand control strategies.[Methods]Respondent-Driven Sampling (RDS) was applied to recruit community-based drugusers in three cities. After eligibility screening and informed consent, each participantprovided5ml intravenous blood to be tested for HIV, HCV and syphilis antibodies afterface-to-face structured questionnaire interview in a private room. Data ofrehabilitation-based drug users in three cities are downloaded from the system ofNational Center for AIDS/STD Control and Prevention，China CDC. Then analysis wasconducted to compare the data from the two sources.Mulitiplier method was used to estimate the population size of drug users in thestudy. To combine the proportion of drug users themselves and drug users they knowwho had been caught by police angency and the proportions who had been placed intothe detoxification centers during2010,2011and till july2012and the number ofregistered drug users in Local Public Security Bureau detoxification center of each cityin the same given period. Thus, There are six estimated numbers of each city obtained,taking the median as the final, taking the interval value between the minimum and themaximun.Epidata version3.1software was used for data input and management. SPSSversion20.0software was used for analysis which combined the community-based dataweighted by the individualized weights output through the RDSAT with thedownloaded one. Quantitative data was analyzed using univariate and then furthermultivariate logistic(=0.05) regression models. Netdraw version2.118was used to draw recruitment tree. The recruited equilibrium diagram was drawn by EXCEL.[Results]1. The recruited participants of community-based drug users were264,230, and288in three cities respectively.2. The recruited participants of rehabilitation center-based drug users were411,400, and406in three cities respectively.3. A total of1999drug users were recruited inculding39.1%(782/1999) fromcommunity and60.9%(1217/1999)from rehabilitation, with94.5%（1889/1999）beingmales. The maximum age of the participants was69years old and the minimum was16years old. The average age was36.3±7.68（x s）. Most of them were married and fromhigh grade settings which accounted for58.5%（1169/1999）and63.4%（1268/1999）respectively.4. Compared to the rehabilitation center-based drug users, the community-baseddrug users in city1had higher proportion of females, lower proportion of migrants;those in city2had a higher proportion of unmarried persons;thoese in city3had ahigher proportion of immigrants. Most of them had used Heroin, Methamphetamine andKetamine as their main drugs. There were differences in two high-riskbehaviors-injection drug use and needle sharing-between rehabilitation center-basedand community-based drug users(P＜0.05).In city2,the proportion of sharing needlesamong community-based drug users(35.3%) was higher than that among therehabilitation center-based drug users(23.2%);in city1and3the proportion(7.4%,17.8%) was lower than that(29.1%，32.8%). The proportion of using condomeverytime was below70percent in all participants.In city2and city3, the proportion ofusing condom consistently among the community-based drug users was significantlylowere than that among rehabilitation center-based drug users (0.0%vs67.4%, and20.0%vs64.1%,(P＜0.05).5. The pooled prevalence of HIV in three cities was2.8%. The HIV prevalenceamong community-based drug users was4.0%while that among rehabilitationcenter-based drug users was2.0%(P=0.008). In city2,The prevalence of HIV was1.2%among the rehabilitation center-based drug users and6.1%among thecommunity, respectively（P=0.001）. There were no significant difference in city1and3. In city1, city2and city3, the prevalence of HCV among community-based drug userswere72.3%,91.0%and50.9%respectively while those were65.0%,85.5%and61.2%respectively among the rehabilitation center-based drug users correspondingly (P<0.05).Furthermore, The prevalence of syphilis was4.1%among the rehabilitationcenter-based drug users and11.4%among the community, respectively（P=0.001）.6. Weighted multivariate unconditional logistic regression model indicated that thecommunity-based sample（OR=2.54，95%CI：1.44～4.47）,sharing needle（sOR=7.34，95%CI：4.13～13.04）were independent risk factors for HIV infection.Being a male（OR=0.26，95%CI：0.10～0.64）was protective factors for HIV infection.7. The estimated size of drug users in three cities were29268(17485～71154),18950(8473～61998) and5904(3641～7691), respectively.The estimated number ofpeople living with HIV was1248(857～1638),1167(669～1665) and140(96～183),respectively.[Conclusion]1. There were varying degrees of difference in HIV prevalence, demographiccharacteristics and behaviors between the rehabilitation center-based and thecommunity-based drug users in different regions which indicated that the sentinel datamainly from rehabilitation center may not be able to represent the real situation. Forfurther comprehensive assessment, we should combine the community-based drugusers’ data with the rehabilitation-based data.2. Sharing needles and being a female were the risk factors suggested that weshould expand and evaluate the prevention and control strategies inculding MMT，VCTand condom use promotion.3. The estimated number of drug users in three cities were1.8,1.3and1.8timesthe number of registered in Local Public Security Bureau detoxification center of eachcity, respectively.