摘要
目的了解1997—2020年中国大陆登革热病例数量变化情况,并使用时间序列模型进行拟合研究。方法从国家卫生健康委员、国家统计局网站收集年度法定传染病疫情及年末人口数据,分析我国大陆登革热病例数量变化。构建各类自回归积分滑动平均模型和指数平滑模型,比较不同模型效果,对登革热发病趋势进行拟合。结果2014年后我国大陆的登革热输入性病例数量呈明显增加趋势。单指数平滑、双指数平滑、Holt-Winters无季节性模型拟合2018年我国登革热病例数的相对误差均小于5%;双指数平滑模型获得的拟合值与实际值最接近,相对误差小于1%。结论在不发生大规模暴发流行的年度,指数平滑模型能够较好地拟合登革热病例数量变化,将有助于建立疫情监测预警机制。
Objective To investigate the quantitaty changes in China's Mainland from 1997 to 2020,to predict dengue fever cases by time series fitting.Methods The bulletins about dengue fever cases and population at year-end from National Health Commission and National Bureau of Statistics were retrieved.The quantitative changes of dengue fever cases were analyzed in China's Mainland.The ARIMA models and the exponential smoothing models were established respectively,the effects of these models were compared.The tendency of dengue fever cases was fitted.Results Since 2014,there was a significant increasing tendency in imported cases of dengue fever.The relative errors of predicted dengue fever cases in 2018 from single exponential smoothing model,double exponential smoothing model and Holt-Winters(no seasonal)exponential smoothing model were below 5%,ubiquitously.The fitting result from double exponential smoothing model was the nearest approximation of actual dengue fever cases,and the relative error was below 1%.Conclusion The exponential smoothing model should be used to fit quantitative changes of dengue fever cases during years without massive outbreaks,which is conducive to establish mechanism for surveillance and early-warning.
作者
胡学锋
吴霜
吴媛媛
邱文毅
符丽媛
刘洋
吴海磊
HU Xue-feng;WU Shuang;WU Yuan-yuan;QIU Wen-yi;FU Li-yuan;LIU Yang;WU Hai-lei(Jiangsu International Travel Healthcare Center(Nanjing Customs District Port Clinic),Nanjing,Jiangsu 210019,China;不详)
出处
《中国国境卫生检疫杂志》
CAS
2022年第3期216-219,共4页
Chinese Journal of Frontier Health and Quarantine
基金
南京海关科研项目(2021KJ25,2021KJ29)。
关键词
登革热
时间序列
卫生检疫
早期预警
Dengue fever
Time series
Health inspection and quarantine
Early-warning
