摘要
目的分析某省2005~2008年流行性腮腺炎(腮腺炎)监测资料,建立指数平滑模型对数据趋势进行预测和预警。方法对某省2005~2008年腮腺炎日报告发病数进行7d移动求和消除数据中"周末效应",用2005~2007年数据进行指数模型拟合,对2008年腮腺炎发病情况进行预测和预警。结果霍尔特-温特斯(Holt-Winters)指数平滑模型,对2008年数据预测趋势与实际发病情况接近,预测效果较好,对突发疫情警示的灵敏度76.92%,特异度83.33%,及时率80%。结论指数平滑法用于传染病趋势预警实际可行,对传染病预防控制有实用意义。
Objective To analyze the daily data of epidemic Mumps in a province from 2004 to 2008 and set up exponential smoothing model for the prediction.MethodsTo predict and warn the epidemic Mumps in 2008 through calculating 7-day moving summation and removing the effect of weekends to the data of daily reported mumps cases during 2005~2008 and exponential summation to the data from 2005 to 2007.ResultsThe performance of Holt-Winters exponential smoothing is good.The result of warning sensitivity was 76.92%,specificity was 83.33%,and timely rate was 80%.ConclusionsIt is practicable to use exponential smoothing method to warn against epidemic Mumps.
出处
《中国疫苗和免疫》
CAS
2010年第3期233-237,共5页
Chinese Journal of Vaccines and Immunization
关键词
流行性腮腺炎
指数平滑法
预测
预警
Epidemic Mumps
Exponential smoothing method
Prediction
Warning