基于宁波、舒兰和郑州三地新冠肺炎潜伏期的研究

Research on the COVID-19 incubation periods of Ningbo,Shulan and Zhengzhou

通过对宁波、舒兰和郑州共3组新型冠状病毒肺炎潜伏期数据分别进行正态分布、对数正态分布、Weibull分布以及BS分布拟合检验,发现这3组数据都服从两参数BS疲劳寿命分布。针对多个两参数BS分布总体,给出了参数齐性的近似检验方法,就讨论的3组潜伏期数据可以认为其参数是相同的,并得到了刻度参数β与形状参数α的点估计,进而求得了确诊病例在其暴露开始后的14 d和21 d内出现临床症状或检测试剂呈阳性的概率分别达到0.968412和0.997324.与文献相应结果比较后可知,用两参数BS分布来拟合新冠肺炎潜伏期比用Weibull分布与对数正态分布来拟合的优势明显,所分析的结果与当前的“14+7”的防控策略相吻合。

Three sets of COVID-19 incubation period data from Ningbo,Shulan and Zhengzhou,are respectively fitted with Normal distribution,Lognormal distribution,Weibull distribution and BS distribution.It can be concluded that these three sets of data all follow two-parameter BS distribution.The approximate test method for the homogeneity of parameters is proposed for multiple two-parameter BS fatigue life distribution populations.It can be considered that the parameters of these three sets of incubation period data are the same,and then the point estimates of scale parameter β and shape parameter a are obtained.Thus,the probability of a confirmed case showing clinical symptoms or positive test reagents within 14 d and 21 d of the onset of exposure reaches respectively 0.968 412 and 0.997 324.Compared with the corresponding results in the references,it can be found that two-parameter BS distribution has obvious advantages over Weibull distribution and Lognormal distribution to fit the COVID-19 incubation period.The analysis results are also consistent with the current "14+7" prevention and control strategy.