The unreliability of the Kato-Katz technique limits its usefulness for evaluating S. mansoni infections.

The Kato-Katz technique, a (semi) quantitative stool examination technique, is generally recommended for diagnosis and evaluation of Schistosoma mansoni infection by schistosome experts. However, egg counts are subject to important variability. In order to quantify the reproducibility of egg counts using the Kato-Katz technique, field data of 1255 observations on 299 subjects infected with Schistosoma mansoni were analysed. Agreement between repeated observations was assessed both categorically (kappa statistic) and continuously (analysis of variance). The day-to-day variation of egg counts was much greater than the variation due to different observers or different slides. The quantitative reproducibility was low: the weighted kappa statistic was 0.39 between specimens of different days, 0.62 between slides of the same specimen and 0.81 between observers of the same slide. Therefore the classification of individual patients into groups based on egg counts, used as a measure of morbidity, must be interpreted with great care, especially in longitudinal studies. Usefulness of the Kato-Katz technique appears limited. Its reproducibility is low. It cannot be recommended as a routine test in a primary health care setting or in a hospital laboratory because safety and detection of other parasites are better assured by other techniques. It can be used in epidemiological studies and evaluation of schistosomiasis control programmes, but here too, other techniques might be preferred.