基于病例-对照设计的中国马流感发生的集聚度探测研究  

Case-control study on cluster detection of equine influenza in China

作　　者：张志诚[1] 王志亮[1] 侯哲生[2] 相文华[3] 王力俭[4] 

机构地区：[1]中国动物卫生与流行病学中心国家外来病研究中心,青岛266032 [2]烟台大学土木工程学院烟台264005 [3]中国农业科学院哈尔滨兽医研究所,哈尔滨150001 [4]新疆畜牧科学院,乌鲁木齐830000

出　　处：《科学通报》2012年第23期2192-2199,共8页Chinese Science Bulletin

基　　金：国家农业公益性行业科研专项(201103008和200903037)资助

摘　　要：基于病例-对照设计、地理风险格局和集聚度探测等方法,开展我国马流行性感冒病例分布特点研究.研究显示,我国是马属动物的养殖大国,马流行性感冒在我国局部地区的马属动物群体中曾在较大范围发生.回顾性分析表明,我国马流感病例主要分布在西北和东北局部地区,2000～2007年病例集聚度探测表明,新疆伊犁州和巴音郭楞蒙古自治州交界地区坐标为(6460039.00,4277548.50)(P<0.001,R=7487976.06M,RR=273.94)病例区集聚度为最高,RR值达到273.94,显示该病例区及附近区域很有可能存在马流感疫情发生所需要的必要风险因素,提示应该加强对该区域马属动物的风险监管和主动监测.此外,在青海和甘肃等省、市和自治区的病例报道预示有沿新-青-甘-蒙一线传播态势.基于病例-对照设计和集聚度探测,初步明确了现阶段我国马流感发生风险主要集中在西北局部地区,应适时加强对我国新、甘、蒙等西北地区马流感的主动监测和马属动物的移动监管.This study investigated the distribution patterns of reported equine influenza （EI） cases in China. Geographical risk pattern analysis of a retrospective case-control study was assessed using geographical information sciences and the likelihood ratio test. The analysis showed that： （1） Equine production in China accounts for 15% of the worldwide total, among which, the highest densities were identified in Yunnan, Gansu, Hebei, Liaoning and Jiling administrative regions. Qinghai, Xizang and most southern administrative regions had relatively low densities. （2） Retrospective studies revealed three major El epidemics across China pre-2000, which were mainly distributed in northwestern and northeastern regions. （3） EI cases were also reported in China after 2000, mainly in the Xingjiang administrative region and neighboring regions. Geographical risk pattern analysis suggested these reported El cases were significantly spatial related in 95% confidence interval （CI）, with an R-value of 0.3415, standard Z-value of 5.7588, a higher critical value compared with the theoretic random distribution. （4） Likelihood ratio testing of the Bernoulli-based case-control study for pre-2000 cases showed that cases positioned at site （6960036, 4432260） （P〈0.001, R=1165880 M, RR=9.11） were most likely clustered, as were cases at site （9768346, 5338305） （P〈0.001, R=1174519 M, RR=6.69）. From a risk management perspective, these two regions should be the focus of active surveillance. （5） Likelihood ratio testing of the Bernoulli-based case-control study for the E1 cases reported after 2000 showed that cases positioned at （6460039.00, 4277548.50） （P〈0.001, R=7487976.06 M, RR=273.94） and nearby locations were most likely clustered. This indicated the existence of risk factors associated with EI infestation in these regions, and active surveillances should therefore be conducted. In this study, case-control cluster detection of EI occurrence indicated that northwestern China is cur

关 键 词：病例-对照 集聚度 马流行性感冒 地理风险格局 主动监测 

分 类 号：S858.21[农业科学—临床兽医学]