Cytohistological correlation as a measure of quality assurance of a cytology laboratory.

In a hospital-based cytology screening programme for the early detection of preinvasive lesions of the uterine cervix, 166 women with abnormal smears (human papillomavirus (HPV) changes, cervical intraepithelial neoplasia (CIN) and invasive carcinoma) were referred to the central colposcopy clinic between January 1989 and December 1991. The colposcopist (V.S.) was able to take a direct biopsy in 156 cases. In the remaining 10-cases, biopsy could not be taken because of unsatisfactory colposcopy. A cytohistological correlation was obtained in 121/156 (77.5%) cases, and the remaining 35 cases showed a disparity in diagnosis. These were reviewed by one of us (P.S.) and the reasons for underdiagnosis/false negatives and overdiagnosis/false-positive results were analysed. It was found that sampling error was the cause of false negativity and underdiagnosis in most cases while interpretative errors resulted in the overdiagnosis and false-positive smears. The reasons for interpretative errors were studied in detail.