Proficiency testing assessments for nematode worm egg counting based on Poisson variation

Abstract The Poisson distribution provides an appropriate model for the variation within laboratories in worm egg counting. This is demonstrated by the results from annual quality assessment exercises in which laboratories in Australia tested multiple samples from the same mixtures prepared using different worm egg densities. Confidence intervals based on simulations using the Poisson distribution are recommended in the analysis of the results to identify laboratories showing significant bias or overdispersion, hence indicating possible procedural errors.