海南省疟疾疫情时空分析及影响因素研究

发布时间：2019-07-04 12:25

【摘要】：研究目的通过对1990年-2010年海南省疟疾疫情时空分析及影响因素的研究，为该省及类似地区疟疾防控提供科学依据。通过尝试时间空间分析技术与统计分析方法在疟疾研究领域的综合应用，为类似研究提供方法学参考。研究方法收集整理1990-2010年海南省疟疾疫情监测数据，将全国1:100万电子地图加工处理为1:100万海南省市县边界图。提取海南省各县市的气温、降水量、湿度等气象与环境数据，人口、GDP、卫生、人民生活指标等数据，建立海南省市县综合信息数据库。应用时空扫描统计量、时间序列分析方法、主成分分析以及多元线性回归分析方法对资料进行分析。所采用的软件包括Excel2003、SatScan9.0、SPSS13.0、Mapinfo7.0等统计软件。研究结果 1.1990年-2010年海南省疟疾呈整体下降的趋势，总体来说，海南省疟疾发病最高峰为7、8月份，2月份为全年发病最低峰。海南省西南部地区是疟疾疾病的高发区。人群分布，30-50岁男性为高发人群，农民、工人及民工占总报告人数的85.11%。 2.运用时空重排扫描统计量对2005-2010年疟疾发病数据进行分析发现，2005-2010年海南省可能存在7个疟疾聚集区域（P0.05）。对2010年疟疾发病数据进行时空扫描发现有3个聚集区域（P0.05）。运用2005年1月-2009年12月建立的ARIMA模型对海南省2010年1月份的疾病发病率进行预测，预测值为0.15/10万，95%的可信区间为[-1.04,1.33]，实际检测发病率为0.17/10万，疟疾月发病率实际值落入了预测值的95%可信区间内，预测相对误差为11.8%。将2010年1月份疟疾实际发病率纳入时间序列模型，重新拟合，，并对2010年2-12月份疟疾发病率进行预测，实际疟疾发病率比理论疟疾发病率平均下降了85.75%。 3.将15个温度指标和13个降水量指标进行主成分分析得其主成分，并以2010年市县疟疾发病率为因变量，男性人口、农业人口、少数名族人口、GDP、第一产业、第二产业、第三产业、人均GDP、卫生机构总数、执业医师人数、农村居民家庭人均纯收入、城市居民家庭人均纯收入、温度主成分、降雨量主成分、湿度等为解释变量构建结局变量为连续型变量的多元线性回归模型。其中，疟疾发病的危险因素的包括“城市居民家庭人均纯收入”、“降水量主成分2”、“湿度”这三个因素，疟疾发病的保护因素包括“男性人口”、“农业人口”、“人均GDP”、“医疗机构个数”、“执业医师个数”、“农村居民家庭人均纯收入”这6个因素。研究结论本课题重点探讨了1990-2010年海南省疟疾流行病学特征和时空聚集性，还重点研究了海南省不同市县疟疾疫情存在差异的主要影响因素。在海南，疟疾的流行有一定的周期性、地域性，并在某些特定人群中有较高的发病率。时间序列模型可以很好的拟合疟疾发病率在时间序列上的变化趋势。在防治措施、人口免疫状态及人口流动没有发生大幅度变化的时候可以用来预测疟疾的发病率变化。海南省2010年起开始实施的消除疟疾计划作用明显，时间序列模型从侧面也证明了这一点。另外，降水量和湿度将影响海南省疟疾在各市县的发病，而经济、卫生和人民生活水平也是影响疟疾发病的重要因素。这一研究结果将有助于当地进一步在多方面对疟疾疫情进行防控。创新点本课题在整个研究过程中，将多个领域的信息与知识进行了整合。将疟疾疫情数据进行时间、空间上的整合，打破了时间与空间的限制，不仅仅研究了有网络直报系统以来的数据，而且还研究了从1990年来的疟疾发病数据。在空间上，打破了县市的界限，对海南省全省疫情进行空间扫描，为疟疾疾病的预警提供新思路。将人口学、经济学及卫生学知识相融合，使得疟疾疫情及影响因素的分析更加的立体，更加的贴合实际。将自然因素与社会因素相结合的分析方法在国内外的疟疾研究领域较为罕见。
[Abstract]:Purpose of the study To provide scientific basis for malaria prevention and control in the province and similar areas through the study of the time-space analysis and the influencing factors of the malaria epidemic in Hainan Province from 1990 to 2010. According to the comprehensive application of the method of time-space analysis and statistical analysis in the field of malaria research, it provides methodological reference for similar research. An examination. The research method collects the monitoring data of the malaria epidemic in Hainan Province from 1990 to 2010, and the national 1:1 million electronic map processing and processing is 1:1 million Hainan. The boundary map of the city and county. The data such as air temperature, precipitation, humidity and other meteorological and environmental data, population, GDP, health and people's living standards in various counties and counties of Hainan are extracted to establish a comprehensive city and county of Hainan Province. Information database. Application of space-time scanning statistics, time series analysis method, principal component analysis and multivariate linear regression analysis method The materials used include Excel2003, SatScan9.0, SPSS13.0, Mapinfo 7.0. equal statistical soft According to the results of the study, the total incidence of malaria in Hainan Province from 1990 to 2010 is the trend of overall decline. In general, the peak of malaria in Hainan Province is 7, August and February. The lowest peak in the whole year of the whole year. Is a high-risk area of malaria. Population,30-50-year-old male is a high-risk group, and farmers, workers and migrant workers account for the total report. The number of malaria cases was 85.11%.2. The analysis of the data of malaria in 2005-2010 was carried out by using the time-space rearrangement scanning statistic, and there may be 7 malaria cases in Hainan province from 2005 to 2010. The area of aggregation (P0.05). The time-and-space scan of the malaria incidence data in 2010 was found to be 3 The disease incidence of Hainan Province in January 2010 was predicted with the ARIMA model established in January-December,2009. The predicted value was 0.15/ 100,000, and the confidence interval of 95% was[-1.04, 1.33], and the actual detection was made. In the case of 0.17/ 100,000, the actual value of the monthly incidence of malaria falls within the 95% confidence interval of the predicted value, pre- The relative error is 11.8%. The actual incidence of malaria in January 2010 is included in the time series model, re-fitted, and the incidence of malaria in the period from February to December 2010 is predicted, and the actual incidence of malaria is higher than the theoretical malaria incidence The average decrease of 85.75%.3. The main components of 15 temperature indexes and 13 precipitation indexes are analyzed to obtain the main components, and the incidence of malaria in the city of the city in 2010 is the dependent variable, the male population, the agricultural population, the minority population, the GDP, the first industry, the second industry, Industrial, tertiary industry, per capita GDP, total number of health facilities, number of medical practitioners, net income per person in rural households, net income per person of urban resident, main component of temperature, main component of rainfall, humidity, etc., and the construction outcome variable of the interpretation variable is continuous The multivariate linear regression model of the continuous variable is the three factors including "per capita net income of urban residents", "Precipitation principal component 2" and "humidity", including "male population", "agricultural population", "per capita GDP", "Number of medical institutions", "Number of medical practitioners",

"per capita net income of rural housa. s" 【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：R531.3

【参考文献】