摘要
目的：在理论上探索应用混合样本方法估测媒介感染率的可行性。方法：用二项分布确定混合样本中感染者的概率，根据预估的媒介感染率的最高限，确定最大混合样本数，使阳性混合样本中只有一个感染者的概率大于预定的显著性水平，据此用阳性的混合样本数估测样本感染率，并导出了判断预估感染率是否合适的指标。结果：本文建立的方法虽然需要预估当地感染率来确定最大混合样本数，但估值的方差和精度不受混合样本数的影响，在样本数较少和感染率高时，估测效率较高。结论：应用混合样本估测媒介感染率在理论上是可行的，用混合样本阳性比估测媒介感染率适用于对当地流行背景不清和精度要求不高的情况；而最大混合样本法需预估当地的感染率，但具有较高的精度。
AIM:To investigate theoretically the feasibility of the pool sampling method in estimating the vector prevalence.METHODS:The probability distribution of infected individuals in a given pool was analyzed using binomial distribution,based upon the upper limit of predicted vector infectivity.The maximal pool size was determined in case that the probability of only one infected individual within this positive pool is larger than a predetermined significance level,in this way infection rate is estimated by the number of positive pools and the judgment of predicted prevalence was addressed by the threshold of positive pool ratio.RESULTS:In comparison with the pool sampling based on the positive pool ratio the method suggested in the present paper poses a variance and precision of estimates independent on pool sizes although it requires a predicted prevalence to determine the maximal pool size. In situations of limited samples and a high prevalence this pooling method has higher estimation efficiency.CONCLUSION:Theoretically the estimation of vector prevalence using pooling methods is feasible.The estimation based on positive pool ratio is suitable for situations of lacking epidemiological data and rough precision estimation,whereas the method based on the maximum pool size provides more accurate estimates and needs a prior predicted infection rate.
出处
《中国寄生虫学与寄生虫病杂志》
CAS
CSCD
北大核心
1998年第1期29-33,共5页
Chinese Journal of Parasitology and Parasitic Diseases
