一个预测新型冠状病毒肺炎流行趋势的简易模型  

A simple model of forecasting COVID-19 epidemic

作　　者：田斌群[1] 陈道远 田卫群[3] 王行环[1] 刘同族[1] 张银高[1] 郑新民[1] 李世文[1] TIAN Binqun;CHEN Daoyuan;TIAN Weiqun;WANG Xinghuan;LIU Tongzu;ZHANG Yingao;ZHENG Xinmin;LI Shiwen(Zhongnan Hospital of Wuhan University,Wuhan 430071,Hubei,China;The Engineering College of Chinese University of Hong Koxg,Hong Kong 999077,China;Wuhan University School of Basic Medical Sciences,Wuhan 430071,Hubei,China)

机构地区：[1]武汉大学中南医院,湖北武汉430071 [2]香港中文大学工程学院,香港999077 [3]武汉大学基础医学院,湖北武汉430071

出　　处：《武汉大学学报（医学版）》2021年第1期28-31,共4页Medical Journal of Wuhan University

摘　　要：目的:分析新型冠状病毒肺炎在武汉市的流行特征,建立预测疫情走势模型。方法:采集官方数据填入Excel表中,生成统计图表,计算每日新增确诊病例平均环比增长率和95%可信区间,据此建立数学模型,预测后续每日新增确诊病例数,并与实际发生每日新增确诊病例数进行比较。结果:平均环比增长率(2020年2月14日—3月8日)为-11.01%,标准差为34.52%,95%可信区间常数为±13.81%。以模型Xn=(1+A±B)n×X0预测8 d(2020年3月8日为起点)的乐观预测新增确诊病例数与实际新增确诊病例数高度接近,经统计学检验,两组数据差异无统计学意义(P>0.05)。结论:以平均环比增长率和95%可信区间为参数的Xn=(1+A±B)n×X0数学模型预测每日新增确诊病例数有一定参考意义。Objective:To analyze the coronavirus disease 2019(COVID-19) epidemic characteristics,and to predict the epidemic trend.Methods:Official data were input into the Excel table to generate statistics charts,and the daily average quarter-on-quarter new growth confirmed cases and 95% confidence interval were calculated on the basis of a simple mathematical model to predict subsequent new daily numbers of confirmed cases.The predicted results were compared to the actual daily numbers of confirmed cases.Results:The average day-on-day growth rate(from February 14 to March 8,2020) was11.01%,and the standard deviation was 34.52% with 95% confidence interval of(±13.81)%.By the model Xn=(1+A±B)n X0 in forecasting the number of new confirmed cases of 8-days(March 8 as the starting point),the predicted results were close to the actual numbers(P>0.05).Conclusion:The mathematical model of Xn=(1+A±B)n X0 is valuable in predicting the daily new confirmed cases during COVID-19 epidemic.

关 键 词：新型冠状病毒肺炎 流行趋势 模型 

分 类 号：R181.2[医药卫生—流行病学]