【作者】 王敏；

【导师】 聂绍发；

【作者基本信息】 华中科技大学 ， 流行病与卫生统计学， 2012， 博士

【摘要】 研究目的：1.构建与我国结核病流行特征相适应的监测预警指标体系,为结核病疫情的预警和控制筛选出敏感性高、时效性好、可操作性强的重要指标,进而为结核病的防控工作提供有力的科学依据。2.探讨我国不同地区结核病患病率和发病率,了解农村、城市和流动人口几类人群的发病特征,为今后不同特征人群的结核病防控工作提供参考。3.揭示研究地区总人群的发病影响因素,以及不同暴露因素对结核病发病的贡献程度,根据研究结果来提出合理可行的防治措施。研究方法：1.采用定性和定量相结合的研究方法进行结核病监测预警指标体系构建,主要包括文献研究、德尔菲法与专家会议法。2.采用横断面调查和随访研究的方法,获得调查现场人群基本资料和患病率,以及人群年发病率和发病特征资料。3.采用卡方检验来对影响总人群发病的因素进行单变量分析,采用Cox回归模型对影响总人群发病的因素进行多元分析。将四个现场人群的特征属性按照能否被干预措施改变分为不可调节的背景变量和可调节的危险因素。运用SAS程序计算各危险因素的全人群归因危险度百分比(PARF)和偏人群归因危险度百分比(PARP),以确定各因素以及因素间的不同组合对结核病发病的贡献。研究结果：1.本次研究,共咨询专家18位,进行了两轮专家咨询,构建了一套包括4个一级指标、9个二级指标,48个三级指标在内的结核病预警指标体系。第一轮咨询专家的熟悉程度范围为0.508-0.967,均数为0.801；权威系数范围为0.704-0.933,均数为0.850。第二轮专家的熟悉程度范围为0.840-0.967,均数为0.922；权威系数范围为0.893-0.957,均数为0.917。经归一法计算权重以后,权重排名前5位的指标为：新涂阳病例登记率、治愈率、完成治疗率、总体到位率、结核病的地区分布。2.分别选取代表流动人口(广东)、中部农村人口(湖南)、城市人口(上海)和东部农村人口(江苏)的样本做现场研究。基线调查总人群活动性肺结核患病率为63.55/10万。广东、湖南、上海、江苏的活动性患病率分别为：103.23/10万、122.89/10万、16.52/10万、46.21/10万。总人口活动性肺结核标化发病密度为41.75/10万、涂阳肺结核标化发病密度为18.59/10万；各省活动性肺结核标化发病密度不完全一样。3.对总人群发病影响因素分析,多因素分析结果显示,男性(RR=2.67,95%CI: 1.80-3.97)、年龄每增加5岁(RR=1.18,95%CI:1.03-1.36)、有结核病史(RR=3.83, 95%CI:2.00-7.31)、已婚(RR=2.1,95% CI:1.09-8.30)可能会增加发病的危险性；而职业为学生和儿童(RR=0.12,95%CI:0.02-0.57)、文化程度在初中及以上(RR=0.81,95%CI:0.67-0.99)的人群发病风险较小。4.对四个现场的人群进行单独分析,结核病史是四个地区共同的危险因素,其他因素在不同地区对发病贡献不完全一致。吸烟对发病的贡献在广东和江苏最大(广东：PARF=8.02%, PARP=6.26%,上海：PARF=23.99%, PARp =20.61%),结核病史对发病的贡献在湖南和江苏最大(湖南：PARF=9.24%, PARP=7.85%,江苏：PARF=21.90%, PARP=17.32%)。如果能同时降低结核病史、结核病接触史、吸烟史和糖尿病史在人群中的暴露水平,则广东、湖南、上海和江苏至少能减少11.50%、13.83%、33.37%和47.73%的发病。研究结论：1.所建立的监测预警指标体系与我国结核病流行特征相适应,可为结核病的防控工作提供有力的科学依据。2.我国目前结核病高发区集中在经济不发达地区和中部农村,其次是在流动人口中。提示防控工作依旧要以农村地区和经济欠发达地区为重点,在老年人和男性中要做好结核病人的发现和治疗工作。3.通过病因分值研究,提示在开展结核病防控工作的过程中,重点控制传染源,提高结核病人发现率和治愈率。提倡禁烟和控制血糖可有效降低结核病发病。创新点：1.本研究首次构建了我国结核病监测预警指标体系,为结核病疫情早期预警提供有用工具。同时,还能为今后呼吸道传染病单病种监测预警指标体系构建提供方法学基础。2本研究首次在国内建立大规模结核病人群观察研究现场,并获得发病率资料。3本研究首次利用全人群归因危险度百分比和偏人群归因危险度百分比的概念来区分人群中背景变量和环境暴露因素引起的发病。更多还原

【Abstract】 Objectives:1.To construct a monitoring and early warning index system which adapt to tuberculosis popularity characteristics in our country, supporting the warning and controlling of tuberculosis with important index of high sensitivity, good timeliness and strongly maneuverability, thus providing tuberculosis prevention and control of with powerfully scientific basis.2. To obtain tuberculosis prevalence and incidence in different areas, and to understand the incidence characteristics of rural, urban and floating populations, so as to provid further tuberculosis control and prevention with reference of different characteristics of the groups.3. To discuss the influence factors of incidence of the whole population and contribution of expose conditions to the incidence in four different fields, according to the results of this study to propose prevention and control measures reasonably.Methods:1.Both the qualitative and quantitative research methods were used for tuberculosis monitoring and early warning index system construction, literature study, the Delphi method and expert meeting method were included. 2.The Cross-sectional survey method was used to get demographic data and prevalence of the population, then the observation object were followed-up to obtain the annual incidence and its features.3.Using a chi-square test for single variables analysis of factors which affect the total population’s incidence, with the Cox model for multivariate analysis. According to the characteristics whether can be changed or not, they were divided to background variables which can not be adjusted and risk factors which were adjustable. SAS program was used to calculate full population attributable risk percentage (PARF) and partial population attributable risk percentage (PARP) of risk factors, so as to determine the contribution of factors and the different combination of them to tuberculosis incidence.Results:1. In this study, a total of 18 experts attended two rounds of expert consultation. A monitoring and early warning index system of tuberculosis including 4 primary indexes,9 secondary indexes and 48 third indexes was build. The first round of the expert familiarity degree ranged 0.508-0.967, and the mean score was 0.801. Authority coefficient ranged 0.704-0.933, and the mean score was 0.850. The second round of the expert familiarity degree ranged 0.840-0.967, and the mean score was 0.922. Authority coefficient ranged 0.893-0.957, and the mean score was 0.917. The weight was calculated by normalization method and the top five indexes were: notification of new smear positive case, cure rate, complete treatment rates, the overall rate of patient in place and the regional distribution of tuberculosis.2. Guangdong, Hunan, Shanghai and Jiangsu were chosen to represent the floating population, the mid rural population, the eastern rural population and the urban population. The prevalence of total population was 63.55/100,000. The prevalence of active tuberculosis of Guangdong, Hunan, Shanghai and Jiangsu was 103.23/100,000, 122.89/100,000,16.52/100,000,46.21/100,000 respectively. When standardized with merged population, the incidence density of the active tuberculosis was 41.75/100,000, and the incidence density of smear positive case was 18.59/100,000. The incidence density of the active tuberculosis after standardization was different in four fields.3. Multivariate analysis shown male (RR= 2.67,95%CI:1.80-3.97), age increase pre five years (RR=1.18,95% CI:1.03-1.36), tuberculosis history (RR= 3.83,95% CI: 2.00-7.31) and married (RR= 2.1,95% CI:1.09-8.30) may increase the risk of disease. For students and children (RR= 0.12,95% CI:0.02-0.57) or population with education level above junior school (RR= 0.81,95% CI:0.67-0.99), the risk was lower.4 The population in four fields shared the same risk factors of tuberculosis history. Other factors contributed differently in four study sites. In Guangdong and Shanghai population, smoking contributed most to the disease (Guangdong:PARF= 8.02%, PARP= 6.26%, Shanghai:PARF= 23.99%, PARP= 20.61%), in Hunan and Jiangsu, the history of tuberculosis was the greatest contribution to incidence (Hunan:PARF= 9.24%, PARP= 7.85%, Jiangsu:PARF= 21.90%, PARP= 17.32%). If measures can be taken to reduce the history of tuberculosis, contact history with patient, smoking and diabetes in the crowd, the incidence of Guangdong, Hunan, Shanghai and Jiangsu can at least drop by 11.50%,13.83%,33.37% and 47.73%.Conclusions:1. The score of coordination and concentration degree of the warning indicator system were high, that shown the index system adapt to our country’s characteristics of tuberculosis, and could provide our prevention and control measure of powerfully scientific basis.2. Tuberculosis was highly concentrated in the less developed areas and the mid rural area in China, then also among the floating population. It was showed that rural areas and undeveloped areas were the keys of prevention and control work, and measures should point to elderly and male on case detection and treatment. 3. In process of prevention and control of tuberculosis, the key point was to control the source of infection and improve the case detection rate and the cure rate. In carrying out health education, advocating quit smoking and control of blood sugar were also effective measures to reduce incidence of tuberculosis.Innovations:1. This is the first time to establish tuberculosis monitoring and early warning index system in China, the system is useful for tuberculosis forecast. At the same time, it can work as methodology in index system construction of respiratory infections single disease monitoring and warning.2. At the domestic, this is the first time to establish such a large-scale observation field, and, the incidence data is obtained directly.3 It is the first time to introduce the concept of full population attributable risk and partial population attributable risk to distinguish the contribution of disease incidence between background variables and exposure conditions.更多还原