Gynaecological cytology screening in South Australia: a 23-year experience

The Cytology Laboratory at The Queen Elizabeth Hospital, Adelaide, processed 630,131 gynaecological smears from 347,389 patients between September 1, 1959 and July 31, 1982. The data suggested a lack of compliance with the recommendations of screening frequency, with an average of only 1.8 smear-tests per woman being performed during the 23-year study period. There was a two-fold increase in the incidence of histologically-confirmed cervical dysplasia and carcinoma-in-situ from 1977, but no increase in the incidence of invasive carcinoma. The age distributions of the incidence of dysplasia and carcinoma-in-situ moved towards younger age-groups during the study period. By the end of the study period (1977-1981), the incidence peaks of dysplasia and carcinoma-in-situ were both at ages 25 to 34 years compared with 30 to 39 years of age and 35 to 44 years of age, respectively, in 1962-1966. Invasive carcinoma had a much broader age distribution, with two incidence peaks, and the distribution shifted towards older age groups. The first peak had moved from 40 to 49 years of age (1962-1966) to 50 to 59 years of age (1977-1981) and the second peak moved from 60 to 69 years of age to 75 to 84 years of age. Error rates in cytological assessment were determined by comparing the histological diagnosis with the result of the most recent smear-test. By means of this method the sensitivity was 84.6%. The sensitivity was increased to 91.4% if the estimate were based on the most abnormal result of a smear-test before the histological diagnosis. A poor correlation was found between a mildly-atypical (inflammatory) cytological result (class 2) and a benign histological diagnosis: 48% of class-2 smears were diagnosed histologically as dysplasia or worse. In very few of these cases had an elapsed time of more than five years occurred between the class-2 result of a smear-test and the subsequent diagnosis of neoplasia. Thus, the majority of these cases were either underdiagnosed by cytology or had developed within five years, which indicates the importance of the careful follow-up of women with class-2 smears. Finally, the data illustrate the ability of gynaecological cytology screening to assist in the detection of extracervical gynaecological neoplasms of the uterus, ovary, vagina and vulva, as 132 such cases were diagnosed during the study period after positive results of a smear-test led to further investigations.