Résumé
A large waterborne outbreak of cryptosporidiosis in Milwaukee, Wisconsin, USA in 1993 prompted a search for ways to prevent large-scale waterborne outbreaks of protozoan parasitoses. Methods for detecting Cryptosporidium parvum play an integral role in strategies that lead to appropriate treatment of surface water, but are criticized because they produce results that are highly variable. The US Environmental Protection Agency developed a set of criteria to evaluate detection methods for protozoan parasites in water. As a consequence, the Agency has had to develop approaches to reducing uncertainty of evaluations. The variability and accuracy of various methods of producing small numbers of Cryptosporidium spp oocysts were tested. The least variable and most accurate method was used to spike seven surface water, and one tap water sample to compare 4 detection methods that had been reported in the literature. The least variable and most accurate method for spiking specified numbers of oocysts into samples was found to be flow cytometry. The most effective of the methods tested for detection in surface, tap and reagent water was solid phase cytometry.