Comparison of methods of counting skin window coverslips A statistical analysis

Abstract

Using the skin window technique, many investigators have expressed their results in terms of an increase or decrease of one cell type versus another on the skin window coverslips. These percentages of cell types are often then discussed as to how they represent the cellular aspect of inflammation and-lor hypersenrtsitvuity. The usual method of arriving (it these percentages of cell types is to count approximately 200 cells representative of the relative proportion of cell types seen on scanning a skin window coverslip. This method of counting (referred to as the “representative” method) was drought to be too -subjective, and a new method of counting was elaborated (referred to as the “fixed” method). Using the “fixed” method, one counts all the cells which fall under a given unit (e.g., a poiniter tip) as the, entire coverslip is viewed in a fixed pattern.: To compare' the two methods, the scatter of the samples obtained by the two methods was calculated by finding their standard deviations about, the true percentages of cell, types on each coverslip sampled. The true percentages of cell types were obtained by counting call the cells on the coverslips: Eight to 10 samples were -taken from each of coverslips by both methods (126 samples from a total cell population of .58,611). Samples obtained by the “representative” method had standard deviations twice those of the “fixed” method. The “fixed” method was found to come closer to the binomial standard deviations of the true percentages of cell types on the coverslips than the “representative” method. Hence the “faxed” method of counting is proposed as a new means-of sampling skin window coverslips.