目的本研究以血吸虫调查的2送6检Kato法粪检方式为例,建立2次送检粪便样品的阳性Kato片数的联合分布概率模型,得到血吸虫实际感染率的最大似然估计。方法在血吸虫感染率概率模型中引入参数:人群血吸虫感染率π,血吸虫病患者的粪便中有虫卵的概率πD和有虫卵粪便样品的每张Kato片检出阳性的概率π1,由此建立概率模型,通过对参数π,πD和π1取不同值,构造了20种情况,进行随机模拟,计算感染率的最大似然估计与真实值的绝对误差和相对误差,并做综合评价。结果在π、πD、π1取不同值的各种情况下所得到的π、πD、π1最大似然估计值都非常接近其真实值(设定值),绝对误差的均值控制在0.04以下,相对误差的均值控制在9%以下,即便进行多次模拟所得结果也很稳定。结论本研究给出的概率模型可以应用于各种血吸虫流行病学调查现场的感染率估计。
Objective Based on the 2 feces samples with 6 smears Kato Katz technique of Schistosomiasis surveys,this research constructed the joint probability model of the positive smear number of the 2 feces samples so as to obtain the maximum likelihood estimation of the schistosomiasis prevalence.Methods Three parameters were inducted into the probability model of schistosomiasis prevalence. They were π(schistosomiasis prevalence of the population),πD(the probability that schistosomiasis patients have eggs in there...