温特斯乘法模型在呼吸道症候群监测中的应用  

Application of the prediction model of winters method-multiplicative in respiratory syndrome surveillance

作　　者：孙乔[1] 袁政安[2] 陶芳芳[2] 谢维[3] 白云 吴红岩[1] 朱渭萍[1] 储强[1] 

机构地区：[1]上海市浦东新区疾病预防控制中心,上海200136 [2]上海市疾病预防控制中心,上海200336 [3]上海市卫生局,上海200040 [4]上海市浦东新区卫生局,上海200125

出　　处：《中华疾病控制杂志》2011年第10期905-908,共4页Chinese Journal of Disease Control & Prevention

基　　金：上海市卫生局优秀学科带头人培养计划项目(08GWD26);"十一五"国家科技重大专项(2009ZX10004-201)

摘　　要：目的探讨时间序列预测模型在症状监测预测预警方面的应用。方法对症状监测信息系统中采集到的上海市浦东新区三林和高东社区卫生服务中心2007年7月～2008年6月门诊就诊中呼吸道症候群数进行建模拟合，2008年7月的数据用来进行模型验证。结果温特斯乘法模型能很好地拟合呼吸道症状在时间序列上的变动趋势，与2008年7月的预测值平均相对误差为4．17％。结论温特斯乘法模型能够较好应用于症状监测预测预警，为早期识别异常信号提供科学依据。Objective To explore the application of time series forecasting model in forecasting and early warning in symdromic surveillance. Methods The number of outpatient respiratory syndrome from July 2007 to June 2008 was collected from Symdromic Surveillance Information System in Sanlin and Gaodong Community Health Center in Shanghai Pudong New Area. The model was established and the data in July 2008 was used for model validation. Results Winters method can fit on the changes in time series trend of respiratory symptoms. The average relative error of predictive value was 4.17%. Conclusions Winters method could be applied in forecasting and early warning in symdromic surveillance and it provide a scientific basis for earlv identification of abnormal signals.

关 键 词：人群监测 季节律 预测 

分 类 号：R181.8[医药卫生—流行病学]