The false-negative fraction for Papanicolaou smears: how often are "abnormal" smears not detected by a "standard" screening cytologist?

Abstract

Screening error rates have been measured in various ways. The false-negative fraction is advantageous because it is valid for comparisons between laboratories, which is crucial if standards for error rates are to be developed. False-negative fractions reported or calculated from data in the literature range from approximately 2% to 28%. Some of the highest published rates have come from prestigious academic laboratories. The lowest rates are from studies that covered terms of a year or less and in which only small parts of each slide were rescreened or the rescreeners were relatively inexperienced. Before standards for false-negative fractions can be set, we must collect more data and encourage more laboratories to measure the false-negative fraction.