【作者】 罗丹；

【导师】 肖水源；

【作者基本信息】 中南大学 ， 社会医学与卫生事业管理， 2009， 博士

【摘要】 背景自从1981年美国发现第一个艾滋病（Acquired ImmuneDeficiency Syndrome,AIDS）病例后,艾滋病疫情蔓延的趋势在世界范围内仍没有得到有效遏制,艾滋病仍然是全球性的重要社会和公共卫生问题。艾滋病的病原体已被明确为是艾滋病病毒（HumanImmunodeficiency Virus,HIV）,其在人群间发生和流行被证实与人类的行为密切相关。尽管世界各国在HIV疫苗研制和抗HIV药物开发上投入了大量的人力和经费且取得了一些进展,但到目前为止,通过行为干预改变HIV感染的危险行为,仍然是是当前艾滋病防治最有效、最主要的策略。在全球所有HIV感染者中,通过性行为传播的约占75%。我国自1998年进入艾滋病流行快速增长期后,性接触已经发展成为HIV传播最重要的途径。因此,危险性性行为,即所有可能导致包括HIV感染在内的性传播感染（Sexually Transmitted Infections,STIs）的性行为,已被公认为是全球艾滋病感染的高危行为。由于性行为的特殊性,对艾滋病经性传播流行的控制比对经血液和母婴传播途径的控制更为困难。在艾滋病的防治领域中,商业性性工作者（Commercialsexual workers,CSWs）、性传播疾病（Sexually transmitted diseases,STD）患者、吸毒人员、流动人口、男男同性恋者等都因其具有危险性性行为和其它高危行为而被列为HIV/AIDS防治工作的重点人群。加强对HIV感染高危人群的危险性性行为研究及干预已成为世界各国AIDS研究及防治工作的重点。但目前已有的大多数研究对危险性性行为缺乏明确的定义,没有包括所有的危险性性行为类型,对影响危险性性行为发生的因素缺乏全面的分析和有力的理论解释。进一步了解重点人群中危险性性行为的发生情况及其影响因素,为通过健康教育、行为干预等手段,改变重点人群的危险性性行为,减少HIV的性行为传播提供科学依据,是HIV/AIDS预防和控制领域的一项重大课题。目的通过对HIV/AIDS预防的四类重点人群（女性性工作者、STD患者、农民工和吸毒人群）危险性性行为的发生情况及其影响因素的研究,为降低HIV经性途径传播提供科学依据。具体研究目标包括:1.描述上述四类人群的主要社会人口学特征、主要的行为特征,性态度状况及HIV/AIDS知识知晓程度。2.研究上述四类人群中危险性性行为（包括商业性性行为,临时性性行为,多性伴性行为,同性性行为和无保护性行为等）的发生率及其分布。3.在国外类似工具的基础上,以前期设计和测试的“性态度开放程度量表（The scale of sexual permissiveness,SSP）”为工具,对上述四类人群的性态度开放程度进行评估。4.对研究对象性行为HIV感染的危险性和性伴HIV感染的危险性进行评估。5.研究上述四类人群危险性性行为发生的主要影响因素,包括社会人口学因素,行为特征因素,性态度开放程度,AIDS/STD相关知识等。6.根据上述研究结果,对四类重点人群的HIV预防策略提出建议。方法采取横断面研究,用非概率抽样（方便抽样）确定研究现场,用概率抽样（分层随机抽样）获得研究样本,包括某省会城市1个城区内21家娱乐场所的210名女性性工作者（Female sexualworkers,FSWs）、某综合性医疗机构性病门诊的300名STD患者、在某社区内流动务工的328名农民工及两所强制戒毒机构的314名吸毒人员。全部以面对面访谈的调查方式,在取得知情同意的条件下,以定式访谈问卷分别对四类人群进行现场调查,具体调查内容包括:1.以自编一般情况问卷分别调查四类重点人群的主要社会人口学特征及每类重点人群的行为特征,如FSWs的开始从业年龄,STD患者的求医行为,农民工的返乡探亲频率,吸毒人员的毒品使用特征等;2.针对不同目标人群,分别设计“性行为调查问卷",调查四类重点人群样本的性行为及危险性性行为发生现况,主要调查和分析指标包括:（1）各类危险性性行为（包括商业性性行为、临时性性行为、多性伴性行为、同性性行为、无保护性行为）的发生情况;（2）各人群样本与不同类型性伴的性行为发生频率、性行为中不使用安全套或不正确使用安全套的情况;（3）不同类型性伴HIV感染危险程度评分及分级:通过四个条目询问被调查对象性伴的感染危险。询问内容为:“在过去一年中,您的性伴:①有无性病/艾滋病感染;②是否与别的异性有过性行为;③是否与同性有过性行为;④是否有注射吸毒行为。根据被调查对象对其性伴上述四个方面情况的知晓程度,设置了“是”、“不清楚”和“否’’三项回答,分别将危险程度计为2分、1分和0分,以总得分表示该性伴的HIV感染危险;（4）对研究对象本人的性行为危险程度进行评分和分级。将性行为中的HIV感染危险分为以下几种情况:①仅有唯一固定性伴,但对其固定性伴的HIV感染危险评估等级不为0;②有多个固定性伴且性行为中非100%使用安全套;③有临时性伴且性行为中非100%使用安全套:④有商业性性伴且性行为中非100%使用安全套;⑤有同性性伴且性行为中非100%使用安全套。上述任何一种情况如有发生,则计“1分”。最后的总分即为该样本的性行为危险评分。危险性性行为分值范围为0～5分,相应的性行为危险程度等级范围为0～V级。3.在参考Reiss等编制的“婚前性行为接受程度量表”的基础上,设计了“性态度开放程度量表”,并已在一个大学生样本中进行了初步的信度和效度检验。用该量表对四类人群样本的性态度开放程度进行评估。4.以中国疾病预防控制中心制定的《艾滋病性病综合监测指南及方案》中AIDS/STD相关知识条目为基础,编订“AIDS/STD相关知识问卷”,分别调查了解四类人群样本的相关知识知晓程度。结果1.在各研究现场,根据各样本入组标准:共有210名FSWs样本入组,172人完成访谈,应答率为81.9%（172/210）,得到有效问卷140份;共有300名门诊STD患者入组,212人完成访谈,应答率为70.7%（212/300）,得到有效问卷200份;共有328名农民工样本入组,268人完成访谈,应答率为81.7%（268/328）,得到有效问卷250份;共有323名吸毒人员样本入组,309人完成访谈,应答率为95.7%（309/323）,得到有效问卷309份。总共完成961名重点人群样本的访谈,得到有效样本899人。2.在调查时点前一年内,四类重点人群样本的各类性行为发生情况为:（1）140名娱乐场所FSWs样本在商业性性行为中,无保护性行为发生率为42.5%（54/127）;122人（87.1%）报告有非商业性性行为,其中临时性性行为发生率为43.4%（53/122）。（2）200名STD患者样本中,192人（96.0%）有固定性伴,22.0%（44/200）有多个固定性伴;临时性性行为发生率为26.5%（53/200）,商业性性行为发生率为40.5%（81/200）,既有商业性又有非商业性性行为的的发生率为39.5（79/200）,多性伴性行为总发生率为62.0%（124/200）;（3）250名农民工样本中,238人（95.2%）有固定性伴;1.6%（4/250）有多个固定性伴,临时性性行为发生率为4.1%（10/250）,商业性性行为发生率为10.0%（25/250）,既有商业性又有非商业性性行为的发生率为8.0（20/250）,多性伴性行为总发生率为13.2%（33/250）;（4）309名吸毒人员样本中,279人（90.3%）有固定性伴;27.5%（85/309）有多个固定性伴;临时性性行为发生率为38.2%（118/309）,商业性性行为发生率为25.9%（80/309）,既有商业性又有非商业性性行为的发生率为23.9（74/309）,多性伴性行为总发生率为62.8%（194/309）。3.对性伴的感染危险程度评估:（1）共有80名（40.0%）STD患者样本认为其性伴无感染危险,46人（23.0%）的性伴感染危险等级为Ⅰ级,74人（37.0%）的性伴危险程度为Ⅱ级;（2）210名（84.0%）农民工样本认为其性伴无感染危险,16人（6.4%）报告的性伴感染危险等级为Ⅰ级,24人（9.6%）的性伴危险程度为Ⅱ级;（3）共有70名（22.7%）吸毒人员样本认为其性伴无感染危险,68人（22.0%）的性伴感染危险等级为Ⅰ级,171人（55.3%）的性伴危险程度为Ⅱ级。因为FSWs样本对其性伴（尤其是其“顾客”）的感染危险程度难以评估,故对该人群没有进行此项调查。4.对STD患者、农民工和吸毒人员三类人群样本的自身性行为危险程度的评分结果为:（1）105名（52.5%）患者样本性行为危险等级最低为0级,其余95人（47.5%）的危险等级在Ⅰ级或以上;危险等级最高为Ⅳ级,有6人（3.0%）。（2）223名（89.2%）农民工样本危险性性行为等级为0级,其余27人（10.8%）的危险等级在Ⅰ级以上;危险等级最高为Ⅱ级,有4人（1.6%）。（3）56名（18.1%）吸毒人员样本性行为危险等级为0级,其余253人（81.9%）的危险等级在Ⅰ级以上;危险等级最高为Ⅲ级,共11人（3.6%）。5.性态度开放程度评定结果显示,72.8%的FSW样本、53.0%的STD患者样本、50.0%的农民工样本和52.1%的吸毒人员样本的性态度开放程度得分在中位数分以上。多因素logistic分析结果发现,FSWs样本的性态度开放程度与其人口学特征、行为特征之间未发现有统计学意义的关联性;性别为男性是STD患者、农民工和吸毒人员三类人群样本性态度开放程度高的主要影响因素,OR值分别为10.150（95%CI=4.776～21.569）、26.214（95%CI=14.322～47.979）、20.311（95%CI=11.081～37.230）。6.相关知识知晓程度调查结果显示,在总分为18分的FSWs知识调查问卷中,FSWs样本的知识中位数得分为13分(P₂₅=11,P₇₅=14);其他三类人群样本在总分为25分的知识调查中:STD患者的中位得分为18分(P₂₅=15,P₇₅=21)、农民工的中位得分为17分(P₂₅=15,P₇₅=19)、吸毒人员的中位得分为18分(P₂₅=15,P₇₅=20)。7.四类人群样本性伴危险等级及自身危险性性行为的单、多因素分析发现:（1）无稳定婚姻关系（OR=3.087,95%CI=1.148～8.301）、农村户籍（OR=2.431,95%CI=1.285,～4.599）是FSWs与商业性性伴发生无保护性性行为的危险因素;（2）单因素Ordinal分析发现,性态度开放程度高（OR=3.857,95%CI=2.239～6.646）是STD患者样本性伴感染危险等级增高的危险因素;多因素Ordinal回归分析结果显示:性别为男性（OR=12.280,95%CI=5.419～27.854）是患者样本性伴感染危险等级增高的危险因素,而初次性行为年龄每增高2岁（OR=0.702,95%CI=0.520～0.947）是其性伴感染危险等级降低的保护因素。多因素logistic回归分析性别为男性（OR=4.764,95%CI=2.027～11.200）、无稳定婚姻（OR=2.533,95%CI=1.057～6.068）及性态度开放程度高（OR=2.563,95%CI=1.177～5.464）是患者样本危险性性行为发生的危险因素;初次性行为年龄等级每增高一个级别,发生危险性性行为的危险性下降,（OR=0.621,95%CI=0.439～0.877）,提示初次性行为开始越晚者危险性性行为发生的危险性越低。知识知晓程度以及其它人口学及人群特征变量对患者样本危险性性行为发生的影响无统计学意义。（3）单因素Ordinal回归分析结果显示性别为男性(OR=5.882,95%CI=3.223～19.163、婚姻状况不稳定（OR=5.517,95%CI=2.918-12.466）、性态度开放程度高（OR=3.397,95%CI=1.609～7.185）是农民工样本性伴危险程度上升的危险因素,多因素Ordinal回归分析结果显示婚姻状况（OR=5.947,95%CI=2.560～13.818）和性态度开放程度(OR=3.455,95%CI=1.197～9.915是农民工性伴危险程度升高的主要影响因素。多因素logistic回归分析结果发现,无稳定婚姻关系（OR=7.059,95%CI=2.892～17.230）、性态度开放程度高（OR=6.985,95%CI=1.481～32.939）是农民工样本危险性性行为发生的危险因素,初次性行为年龄开始晚（OR=0.553,95%CI=0.316-0.968）是农民工样本危险性性行为发生的保护因素。知识知晓程度以及其它人口学及人群特征变量对患者样本危险性性行为发生的影响无统计学意义。（4）多因素Ordinal回归分析结果显示婚姻状况不稳定（OR=2.219,95%CI=1.328～3.629）是吸毒人员样本性伴危险程度上升的危险因素,性别为男性（OR=0.376,95%CI=0.204～0.696）、无注射吸毒史（OR=0.494,95%CI=0.262～0.931）、无STD感染史（OR=0.226,95%CI=0.095～0.540）是吸毒人员样本性伴危险程度上升的保护因素。多因素logistic回归分析结果发现,年龄等级增高（OR=0.774,95%CI=0.616～0.974）、初次性行为年龄每增高2岁（OR=0.690,95%CI=0.499～0.955）是吸毒人员样本危险性性行为发生的保护因素;农村户籍（OR＝2.268,95%CI=1.039～4.953）、无稳定婚姻关系（OR=1.982,95%CI=1.001～3.924）、性态度开放程度高（OR=2.730,95%CI=1.415～5.266）是吸毒人员样本危险性性行为发生的危险因素。知识知晓程度以及其它人口学及行为特征变量对患者样本危险性性行为发生的影响无统计学意义。结论1.四类重点人群均存在较高比例的危险性行为,采取措施改变这四类人群的危险性性行为,对预防和控制HIV向普通人群传播和扩散具有极为重要的意义。2.四类重点人群的混合型性行为（即既有商业性又有非商业性性行为）均有一定的发生率,以FSWs样本最高,其次为STD患者样本,农民工样本最低（P< 0.001）。3.性行为危险等级评分方法可直观地体现HIV感染重点人群整体的危险性性行为发生状况,且能鉴别不同重点人群间的危险性性行为严重程度及分布情况,可作为重点人群危险性性行为监测的参考指标。4.多因素分析发现,有稳定婚姻关系是四类重点人群样本危险性性行为发生的主要保护因素;性态度开放程度高是STD患者、农民工和吸毒人员危险性性行为发生的主要危险因素。5.娱乐场所FSWs样本的无保护商业性性行为和非商业性行为的发生率均较高。6.吸毒人员是除FSWs外的三类人群样本中,性行为危险程度最高的群体（P<0.001）,要将对吸毒人群的危险性性行为干预放在与对其共用针具行为干预同等重要的位置,并在干预中要以该人群中无稳定婚姻关系、有性病感染史和有注射吸毒史的人员为重点。本研究的主要意义与创新点本研究以完整、全面、系统的性行为现况调查指标,以省内四类有代表性的HIV/AIDS预防重点人群为样本,清晰、详实地描述了四类人群的危险性性行为发生现状及其影响因素,主要创新点及其意义有:（1）根据对危险性性行为的操作性定义,全面研究了四类人群危险性性行为（商业性性行为,多性伴性行为,临时性性行为,无保护性性行为等）的发生和分布情况,为估计HIV感染从特定人群向普通人群扩散的危险性提供了基础数据;（2）首次对自身性行为的HIV感染危险程度,性伴的HIV感染危险程度进行了等级评估,可作为与艾滋病防治控制相关的健康教育和行为干预效果的评估指标;（3）从社会人口学变量、行为特征变量、性态度开放程度和相关知识知晓程度几个方面,较全面地分析了四类目标人群危险性性行为发生的主要影响因素,为艾滋病防治控制策略的制定提供了科学依据;（4）在国外类似工具的基础上,以笔者前期的研究为基础,首次在国内运用“性态度开放程度量表（The scale of sexualpermissiveness,SSP）”为工具,对HIV/AIDS预防四类重点人群的性态度开放程度进行了定量评估,并分析四类重点人群性态度开程度的主要影响因素及其与危险性性行为发生的关系。SSP可作为将来长期观察我国人群性态度开放程度及其变化趋势的工具。更多还原

【Abstract】 BackgroundRecent studies offer significant evidence showing that the epidemic of acquired immune deficiency syndrome（AIDS） is still serious all over the world.Human immunodeficiency virus（HIV） infection/AIDS is still a major global public health and social problem since the first AIDS case was identified in 1981.During the past three decades,it has demonstrated that behavioral intervention is the most effective strategy for AIDS/HIV control and prevention.It is reported that almost 75%of the HIV infection and AIDS has spread via sexual contact.Thus,indiscriminate sexual behavior is a serious risk to overall health and well-being.Sexual risk behavior can be described as any sexual practices that result in sexually transmitted infections,including HIV/AIDS.From the available epidemiological data on sexual risk behavior across the globe, some subgroups are to be considered as the main focus of HIV/AIDS prevention and control because of their exposure to high risk sexual behaviors,such as female commercial sexual workers（FSWs）,patients with sexually-transmitted diseases（STDs）,drug users,and migrant rural workers.Greater efforts must be made to conduct research of sexual risk behavior and promote behavioral intervention among these high-risk populations.Many of available studies in China have not clearly defined sexual risk behavior and not all types of sexual risk behaviors were investigated, and there are few studies that linked sexual risk behaviors and multidimensional factors. ObjectivesThis study describes the prevalence of sexual risk behavior among selected groups,which include female commercial sexual workers （FSWs）,patients with sexually-transmitted diseases（STDs）,drug users, and migrant rural workers,and drug users,and its multidimensional determinants.Specific objectives include:1.To describe the demographic characteristics,population-specific behavior characteristics,sexual attitudes and STD/HIV/AIDS-related knowledge among the four groups.2.To understand the prevalence of all types of sexual risk behaviors and to understand the distribution of these behaviors among the four groups.3.To assess the sexual permissiveness among the four main populations by the scale of sexual permissiveness（SSP）,which is composed from overseas quantificational measurements.4.To assess the level of risk of sexual practices,which are based upon the partner’s infection status and the sexual practices which oneself employed.5.To identify risk factors,which include demographic characteristics,population- specific behavior characteristics,sexual attitudes,and related knowledge among the four groups.6.To provide scientific advice for promoting sexual risk behavioral intervention and preventing and controlling HIV/AIDS.MethodsThe design of the study is cross-sectional,in which 21 entertainment establishments in a district of Changsha City,the STD clinic of a polyclinic of Changsha City,a community which had almost 2500 migrant rural workers,and two enforced detoxification institutions in Changsha City are set as primary sampling units.Then,through targeted and facility-based sampling,associate with cluster sampling randomly, face-to-face interviews were carried out among the total of 961 samples were investigated by a self-designed structured questionnaire. The questionnaire was designed to include the items related to the following four domains:1.A questionnaire about background information used by the four groups.The content includes socio-demographic characteristics（e.g.,age, gender,and marital status） and population-specific behavior characteristics（e.g.,the age when FSWs begin their commercial sexual behavior,the health-seeking behavior of patients with STDs,the frequency of returning home for migrant rural workers,and the durg abuse characteristics of drug users.）2.A questionnaire about sexual behaviors was used for four groups, and the recall time that be stetted in the past year.The items related to sexual practices include the prevalence of all kinds of sexual risk behaviors,the number of sexual partners,the frequency of the sexual behavior,the partner’s HIV infection risk status,and the frequency of condom use in sexual intercourse with different partners.3.The scale of sexual permissiveness,which is used to assess the sexual permissiveness of the four groups.4.A questionnaire about STDs and HIV/AIDS knowledge.Results1.With the criterion of each samples,210 FSWs were investigated, 172 finished the interview,response rates were 81.9%（172/210）,and 140 samples of FSWs were successfully investigated.300 STD outpatients were investigated,and 212 finished the interview,response rates were 70.7%（212/300）,and 200 samples of STD outpatients were successfully investigated.328 rural migrant workers were investigated,268 outpatients finished the interview,response rates were 81.7%（268/328）, and 250 samples of rural migrant workers were successfully investigated. 314 drug users in two enforced detoxification institutions were investigated,309 finished the interview,response rates were 98.4% （309/314）,and 309 samples of drug user were successfully investigated. Totally,961 samples that come from four target groups were interviewed, and 899 samples were successfully investigated.2.In the past year,the prevalence of all of types sexual risk behaviors among the four groups occurred as follows:2.1 In 140 FSWs sampled,the prevalence of unprotected sexual behavior with the commercial sexual partners was 42.5%（54/127）.The prevalence of non-commercial sexual behavior was 87.1%（122/140）, including 43.4%（53/122） casual sexual behavior.2.2 96.0%（192/200） of the STD patient sampled had regular sexual relationships,and 22.0%（44/200） of them had more than one regular partner.The prevalences of casual sexual behavior,commercial sexual behavior,and multiple partners sexual behavior were 26.5%（53/200）, 40.5%（81/200）,and 62.0%（124/200）,respectively.And 39.5% （79/200） had sex with both commercial and non-commercial partners.2.3 238（95.2%） of the rural migrant workers sampled had regular sexual relationships,and 1.6%（4/250） of them had more than one regular partner.The prevalences of casual sexual behavior,commercial sexual behavior,and multiple partners sexual behavior were 4.1%（10/250）, 10.0%（25/250）,13.2%（33/250）,) respectively.And 8.0（20/250） had sex with both commercial and non-commercial partners.2.4 279（90.3%） of the drug users sampled had regular sexual relationships,and 27.5%（85/309） of them had more than one regular partners.The prevalence of casual sexual behavior,commercial sexual behavior,and multiple partners sexual behavior were 38.2%（118/309）, 25.9%（80/309）,62.8%（194/309）,respectively.And 8.0%（20/250） had sex with both commercial and non-commercial partners.The prevalence of homosexual behavior was 5.2%（16/309）.3.Excluding FSWs sampled,the remaining third of the groups assessment of their total partner’s HIV infection risk status resulted as below:3.1 22.7%of STD patients sampled reported their partners had no HIV infection risk,and 46 people（23.0%）sampled reported their partners had ’Ⅰ’ hierarchy in the level of risk,74 people（37.0%）sampled reported their partners had ’Ⅱ’hierarchy in the level of risk.3.2 84.0%of migrant rural workers sampled reported their partners had no HIV infection risk,and 16 people（6.9%） sampled reported their partners had ’Ⅰ’ hierarchy in the level of risk,24 people（9.6%） sampled reported their partners had ’Ⅱ’hierarchy in the level of risk.3.3 22.7%of drug users sampled reported their partners had no HIV infection risk,and 68 people（22.0%）sampled reported their partners had ’Ⅰ’ hierarchy in the level of risk,171 people（55.3%） sampled reported their partners had ’Ⅱ’hierarchy in the level of risk.4.Excluding FSWs sampled,the remaining third of the groups assessed their own levels of sexual risk behaviors resulted as below:4.1 52.5%of STD patients sampled reported there was no risk of their own sexual behavior,and the remaining 95（47.5%） sampled were ’Ⅰ’ hierarchy or higher. 4.2 89.2%of migrant rural workers sampled reported there was no risk of their own sexual behavior,and the remaining 27（10.8%） sampled were ’Ⅰ’ hierarchy or higher.4.3 18.1%of drug users sampled reported there was no risk of their own sexual behavior,and the remaining 243（81.9%） sampled were ’Ⅰ’ hierarchy or higher.5.The assessment of the sexual permissiveness of the four sampling groups shows that 72.9%of the FSWs,53.5%of the STD patients,50.0 %of the migrant rural workers,and 52.1%of the drug users sampled had the median scores of the scale of sexual permissiveness.The median scores of knowledge that related with STDs and HIV/AIDS for FSWs, STD patients,migrant rural workers,and drug users sampled were 13(P₂₅ = 11,P₇₅ =14),18(P₂₅ = 15,P₇₅ =21,17(P₂₅ = 15,P₇₅ =19) and 18(P₂₅ = 15,P₇₅ =20),respectively.6.The results of multiple logistic regression analyses are presented as below:6.1 Multiple logistic regression showed that having an unstable marital status（OR=3.087,95%CI=1.148-8.301） and coming from a rural area（OR=2.639,95%CI=1.410-4.941） were risk factors for FSWs sampled having unprotected intercourse with the commercial sexual partners.6.2 Male gender（OR=12.280,95%CI=5.419-27.854） was a risk factor while beginning the sexual debut two years older（OR=0.702, 95%CI=0.520-0.947） was the protective factor for STD patients sampled partners higher in the level of risk hierarchy.Multiple logistic regression showed that male gender（OR=4.764, 95%CI=2.027-11.200）,having unstable marital status（OR=2.533, 95%CI=1.057-6.068）,and high openness of sexual permissiveness （OR=2.563,95%CI=1.177-5.464） continued to be the three most important factors in predicting sexual risk behaviors for STD patients sampled.Beginning their debut later was the most protective factor for STD patients sampled sexual risk behaviors(OR=0.621 （95%CI=0.439-0.877）.There is no significant association between level of knowledge that related STD/AIDS or population-specific characteristics with sexual risk behaviors.6.3 Multiple logistic regression showed that male gender （OR=5.882,95%CI=3.223-19.163）,having unstable marital status （OR=5.517,95%CI=2.918-12.466）,and high openness of sexual permissiveness（OR=3.397,95%CI=1.609-7.185） were the risk factors in predicting sexual risk behaviors for migrant rural workers sampled while beginning their debut two years older was the most protective factor for migrant rural workers’ partners sampled who are higher in the level of risk hierarchy（OR=0.598,95%CI=0.427-0.813）.Multiple logistic regression showed that having unstable marital status（OR=7.059,95%CI=2.892-17.230） and high openness of sexual permissiveness（OR=6.985,95%CI=1.481-32.939） continued to be the two most important factors in predicting sexual risk behaviors for migrant rural workers sampled.Beginning the sexual debut two years older was the protective factor for STD patients sampled sexual risk behaviors （OR=0.553,95%CI=0.316-0.968）.There is no significant association between level of knowledge related to STD/AIDS or population-specific characteristics with sexual risk behaviors.6.4 Multiple logistic regression showed that having unstable marital status(OR=2.219,95%CI=1.328-3.629,male gender （OR=0.376,95%CI=0.204-0.696）,not having a history of STDs （OR=0.226,95%CI=0.095-0.540）,and the history of injection drug use （OR=0.494,95%CI=0.262-0.931） were the protective factors for drug users sampled partners higher level in the risk hierarchy.Multiple logistic regression showed that being from a rural area （OR=2.268,95%CI=1.039-4.953）,having unstable marital status （OR=1.982,95%CI=1.001-3.924）,and high openness of sexual permissiveness（OR=2.730,95%CI=1.415-5.266） continued to be the three most important factors in predicting sexual risk behaviors for drug user sampled.Beginning their sexual debut was the protective factor for STD patients sampled sexual risk behaviors（OR=0.553, 95%CI=0.316-0.968）.There is no significant association between the level of knowledge related to STD/HIV/AIDS or population-specific characteristics with sexual risk behaviors.Conclusions:1.Sexual risk behavior was relatively common in the four main population groups.To modify the sexually risk behavior of these four groups is very important to prevent and control HIV transmission and the spread among general population.2.The dominant pattern of sexual risk behavior was a combination of commercial and non-commercial sex,the most common was in the FSW sampled,next was patients with STDs,and the lowest was the migrant rural population.As a result of these behaviors,these groups form a ’bridging population’ for STD/HIV transmission in the social sexual network.We should intervene to control the transmission and spread of HIV/AIDS in local areas.3.The evaluation method for the risk scale of sexual behavior could directly express the apparent situation of sexual risk behavior in the groups with HIV infection and could differentiate the severe degree and distribution among the different population groups.Also,it could serve as a reference index for the surveillance of the sexual risk behavior in the main groups.4.According to multiple analyses,having stable married status was the main protection factor associated with sexual risk behavior in the four main groups.Sexual attitude with high opening degree was the main factor for the appearance of sexual risk behavior in the patients with STDs,as well as in the groups of migrant rural workers and drug users. There was no significant correlation between the level of knowledge and the appearance of sexual risk behavior in the four main groups.5.Commercial and non-commercial sexual risk behaviors were commonly seen in FSWs working in entertainment establishments.6.In general,among the four main groups,drug users was the group with the highest risk of sexual behavior.We should intervene on the sexual risk behavior of drug users as well as on the sharing of needles, especially paying attention to their unstable married status,history of sexually-transmitted disease infection,and history of injection drug use.更多还原