Sizes Estimation Of High Risk Populations For HIV/AIDS Using Network Scale-up Method In Chongqqing Municipality, Taizhou City In Jiangsu Province And Harbin City In Heilongjiang Province

Objective:To estimate the social network sizes in general population and the sizes of high risk populations for HIV/AIDS in Chongqing municipality, Taizhou city in Jiangsu province and Harbin city in Heilongjiang province, using the network scale-up method(NSUM) and explore the feasibility of application of network scale-up method for HIV/AIDS in China.Methods:Participants were recruited through a stratified multistage sampling method and a household survey using network scale-up method (NSUM) was conducted in Chongqing municipality, Taizhou city and Harbin city from August to October in2011. The average social network sizes (c value) were estimated, and the c value was adjusted by proportion of demographic characteristics, back estimation and outlier elimination. Using the adjusted c value, the number of acquaintance of high risk populations and the respect level toward high risk populations were used to estimate the sizes of high risk populations.Results:9,024questionnaires were collected, among which8,318were valid including2,957from Chongqing municipality,2,783from Taizhou city and2,578from Harbin city.1,173respondents among2,957from Chongqing municipality were male (39.67%) and1,784(60.33%) respondents were female and mean age was39.21±11.51years.1,334(47.93%) respondents among2,783from Taizhou city were male and1,449(50.27%) were female and mean age was39.39±10.71years.995(38.60%) respondents among2,578from Harbin city were male and1,583(61.40%) were female and mean age was41.03±10.81years.The average personal social network sizes for Chongqing residents were310,424for Taizhou and267for Harbin, which differed from place, sex, age, educational level and marriage status.The estimated population sizes from Chongqing municipality were28,418(95%CI:26,636-30,201) for female sex workers(FSWs),163,199(95%CI:156,490-169,908) for clients,37,959(95%CI:34,888-41,030) for drug users(DUs), 14,975(95%CI:13,047-16,904) for injecting drug users(IDUs),16,767(95%CI:14,602～18,932) for MSM,189,736(95%CI:182,502-196,970) for people who had multiple sex partners. The ratio of clients to FSWs was5.74:1. The proportion of IDUs among DUs was38.59%. The estimated size for FSWs accounted for0.37%of the total number of the female aged15to49and the clients accounted for2.09%of the total number of the male aged15to49. The proportion of MSM among the total number of the male aged15to49was0.21%and1.23%for people who had multiple sex partners among the total number of persons aged15to49.The estimated population sizes from Taizhou city were6,370(95%CI:5,886～6,853) for FSWs,15,202(95%CI:14,560-15,847) for clients,6,607(95%CI:6,064～7,151) for DUs,1,488(95%CI:1,230～1,746) for IDUs,1,203(95%CI:1,034-1,372) for MSM,22,689(95%CI:21,903～23,476) for people who had multiple sex partners. The ratio of clients to FSWs was2.39:1. The proportion of IDUs among DUs was22.52%. The estimated size for FSWs accounted for0.52%of the total number of the female aged15to49and the clients accounted for1.28%of the total number of the male aged15to49. The proportion of MSM among the total number of the male aged15to49was0.10%and0.94%for people who had multiple sex partners among the total number of persons aged15to49.The estimated population sizes from Harbin city were6,353(95%CI:5,709-6,996) for FSWs,14,645(95%CI:13,641～15,649) for clients,1,826(95%CI:1,591～2,061) for DUs,849(95%CI:689～1,009) for IDUs,2,984(95%CI:2,464-3,503) for MSM,17,783(95%CI:16,677～,890) for people who had multiple sex partners. The ratio of clients to FSWs was2.31:1. The proportion of IDUs among DUs was46.50%. The estimated size for FSWs accounted for0.20%of the total number of the female aged15to49and the clients accounted for0.44%of the total number of the male aged15to49. The proportion of MSM among the total number of the male aged15to49was0.09%and0.27%for people who had multiple sex partners among the total number of persons aged15to49. Conclusions:Multiple population sizes estimation can be conducted in one survey using network scale-up method, and the process was cheaper and faster compared to the traditional methods, but the estimated sizes were affected by the selection of known populations and transmission effect, barrier effect and estimation effect.The estimated sizes for high risk populations using network scale-up method were lower than traditional methods, and the estimated sizes for MSM and DUs were dramatically different compared with the results according to the traditional methods.This study was the pilot survey using network scale-up method in China. Based on experience from other countries, this study was adjusted according to the realistic situation in the survey districts, which provided some suggestion for the further research.