Abstract

A pilot study of an external quality assessment scheme was run between February 1985 and September 1986 to (i) assess the feasibility of running one from a district general hospital; (ii) to estimate the time required to organise and run it with a computer; (iii) to provide sound statistical results with which future schemes could be compared. Seven laboratories participated, and the 20 smears selected from each laboratory were circulated in three rounds in batches of seven, seven, and six according to a prearranged order. Results analysed using the kappa statistic showed moderate levels of interlaboratory agreement, with complete agreement emerging only on a small proportion of cases.