[Estimation of vector prevalence by using pool sampling].

Abstract

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.