Experiments were performed to determine whether ingested particles (beads, seeds) and helminth eggs are randomly dispersed in the feces. In human volunteers, it was found that a high proportion of ingested particles remained in the intestine for an average of about 3 days, and were so well mixed that meal- or day-grouping was essentially lost as the elements became randomly distributed in the fecal mass. Thus, it was inferred that helminth eggs, like beads and seeds, are randomly scattered in the stools, even if introduced into the fecal stream (above the lower colon) at irregular intervals as much as several hours apart. Variability of counts was not affected by stirring of the stool prior to sampling.
Based on the Poisson distribution, which was shown by statistical tests to be appropriate, the probabilities of detecting helminth eggs at different levels of infection were determined, and it was found that the size of the sample employed in either the dilution or the direct smear methods is adequate for routine estimation of infection intensity. When compared by the index of dispersion and by the percentages of counts deviating beyond arbitrary boundaries, egg counts by direct smear were found to be, on the average, less variable from day to day than were those by the dilution method.
Present address: Department of the Army, Walter Reed Army Institute of Research, Walter Reed Army Medical Center, Washington, D.C. 20012.