Abstract

We have observed that there is some resemblance between the problems of sampling in industrial quality control and the process of diagnosis of mixed cell populations in cytogenetics. This resemblance enabled us to draw from the methodology of the former science to solve some diagnostic problems in the latter. We considered which of the several sampling procedures available for quality control would be more efficient and more suitable in clinical cytogenetics, concluding that 'sequential sampling' combines both features. We also studied the effect that some abnormalities due to technical factors had on sample size required to reach a diagnosis with a given level of confidence. We give a set of tables for sequential sampling in diagnostic laboratories, illustrating their use in discriminating between mosaicism and pseudomosaicism in prenatal diagnosis and in the diagnosis of the fragile X syndrome. Finally, we discuss the merits of sequential sampling as shown here, comparing it with the current method used in cytogenetics to exclude chromosomal mosaicism.