江西省新型冠状病毒肺炎疫情防控效果评估:基于多阶段SEIR模型的研究  

Evaluation on COVID-19 epidemic prevention and control in Jiangxi Province based on modified multi-stage SEIR model

作　　者：张天琛[1] 曾志笠 李中坚 程慧健[1] 潘欢弘[1] 章承锋[1] 丁晟[1] 毛向群[1] 袁辉[1] 李世哲 车达平 ZHANG Tian-chen;ZENG Zhi-li;LI Zhong-jian;CHENG Hui-jian;PAN Huan-hong;ZHANG Cheng-feng;DING Sheng;MAO Xiang-qun;YUAN Hui;LI Shi-zhe;CHE Da-ping(Emergency Research Group of New Coronavirus Pneumonia,Jiangxi Province Center For Disease Control and Prevention,Nanchang,Jiangxi 330029,China;不详)

机构地区：[1]江西省疾病预防控制中心新冠肺炎应急科研攻关课题组,江西南昌330029 [2]华南理工大学数学学院

出　　处：《现代预防医学》2021年第2期339-344,共6页Modern Preventive Medicine

基　　金：江西省重点研发计划(2020YBBGW0001)。

摘　　要：目的回顾江西省COVID-19疫情,评价江西省启动一级响应的及时性和佩戴口罩等防控措施的效果。方法根据江西省防控措施及病例输入情况,构建多阶段新型冠状病毒SEIR模型。将江西省真实病例数与模型预测结果进行比对,验证模型准确度,依据构建的模型评估防控措施的效果。结果江西省COVID-19疫情前期以输入性为主,1月25日后以本地暴发为主。回归分析显示,模型拟合结果与真实病例发生情况比较吻合;截至3月10日模型拟合累计确诊病例946(95%CI:903~995)例。以此模型为“基础模型”评价江西省防控措施的效果,结果显示不佩戴口罩确诊病例数较基础模型上升30.44%;推迟3 d和7 d启动一级响应,确诊病例数较基础模型分别升高91.12%(1808例)和496.72%(5645例),且Rt降至1以下分别推迟5 d和14 d。结论多阶段SEIR模型与真实病例的发病情况吻合度较好,拟合效果较优。及时启动一级响应及提倡佩戴口罩等防控措施的实施对江西省COVID-19疫情的防控起到积极的作用。Objective To review the COVID-19 epidemic in Jiangxi Province and evaluate the timeliness of initiating Grade Ⅰ response and the effectiveness of prevention and control measures such as mask wearing. Methods A multi-stage novel coronavirus SEIR model was constructed based on the prevention and control measures and case import in Jiangxi Province.The model accuracy was determined through comparison between the actual confirmed case number and predicted case number, and the effectiveness of prevention and control measures was evaluated based on the constructed model. Results The COVID-19 epidemic in Jiangxi Province was dominated by imported cases in the early stage and then turned into local outbreak after January 25. The regression analysis showed that the model fit results were in good agreement with the actual case number. The model fit cumulative confirmed cases were 946(95%CI: 903-995) as of March 10. Using this model as the"base model" for effect evaluation, the results showed that the confirmed case number without mask wearing increased by30.44% compared with the base model. The number of confirmed cases would rise by 91.12%(1 808 cases) and 496.72%(5 645 cases) compared with the base model if the initiation of Grade Ⅰ response were to be delayed for 3 and 7 days,respectively, and the time when Rt dropped below 1 would be delayed by 5 and 14 days, respectively. Conclusion The multi-stage SEIR model fits well and the prediction is in good agreement with actual cases. The prompt initiation of Grade Ⅰ response and mask wearing plays a positive role in the prevention and control of COVID-19 epidemic in Jiangxi Province.

关 键 词：新型冠状病毒肺炎 传播动力学 SEIR 模型 

分 类 号：R184[医药卫生—流行病学]