基于移动流行区间法的甘肃省手足口病流行阈值估计及强度分级研究

Threshold estimation and intensity classification of hand, foot and mouth disease in Gansu province based on moving epidemic method

  • 摘要:
    目的 使用移动流行区间法(MEM)估计甘肃省手足口病流行阈值,建立流行强度分级预警标准,为防控措施提供科学依据。
    方法 从中国疾病预防控制信息系统收集2015—2023年甘肃省手足口病分周报告发病数,每年划定两个流行季节,分别以2015—2023年和2015—2019年发病数为历史数据集,通过MEM计算甘肃省手足口病流行阈值和强度阈值,并通过交叉验证法评价预警效果。
    结果 以2015—2019年历史数据建立模型的预警效果优于以2015—2023年历史数据建立的模型,夏季主高峰的流行阈值为160例,中流行强度阈值为550例,高流行强度阈值为850例,超高流行强度阈值为1 031例,流行期预警的灵敏度为92.77%,特异度为83.64%,约登指数为75.41%;秋季次高峰的流行阈值为63例,中流行强度阈值为158例,高流行强度阈值为390例,超高流行强度阈值为581例,预警的灵敏度、特异度和约登指数低于夏季主高峰。
    结论 剔除历史基线中报告发病数异常的年份可以优化模型效果。 利用MEM建立手足口病预警模型,可实现甘肃省手足口病流行强度分级预警,能够为手足口病的早期预警和防控工作提供科学依据。

     

    Abstract:
    Objective Moving epidemic method (MEM) was used to estimate the epidemic threshold of hand, foot and mouth disease (HFMD) in Gansu province for the establishment of the early warning standard of epidemic intensity and provide scientific basis for the prevention and control of HFMD.
    Methods The incidence data of HFMD weekly reported in Gansu from 2015 to 2023 were collected from Chinese Disease Prevention and Control Information System, and two epidemic seasons were set in a year. The incidence data of HFMD during 2015-2023 and 2015-2019 were used as historical data sets, the epidemic threshold and intensity threshold of HFMD in Gansu were calculated by MEM, and the early warning effect was evaluated by cross validation method.
    Results The early warning effect of the model established based on the historical data from 2015 to 2019 was better than that of the model established based on the historical data from 2015 to 2023. The epidemic threshold of the main peak in summer was 160 cases, the medium epidemic intensity threshold was 550 cases, the high epidemic intensity threshold was 850 cases, and the superior high epidemic intensity threshold was 1 031 cases. The sensitivity, specificity and Youden’s index of early warning in epidemic periods were 92.77%, 83.64% and 75.41%, respectively. The epidemic threshold of sub-peak in autumn was 63 cases, the medium threshold was 158 cases, the high epidemic intensity threshold was 390 cases, and the superior high epidemic intensity threshold was 581 cases. The sensitivity, specificity and Youden’s index of early warning were lower than those of main peak in summer.
    Conclusion The model effect could be improved by excluding the years in which the number of reported cases was abnormal in the historical baseline. The establishment of early warning models of HFMD by MEM could realize the graded early warning of epidemic intensity of HFMD in Gansu and provide scientific basis for the early warning, prevention and control of HFMD.

     

/

返回文章
返回