A multicentre study comparing cervicography and cytology in the detection of cervical intraepithelial neoplasia

Objective To evaluate in a multicentre setting the performance of cervicography compared with cytology for the detection of cervical intraepithelial neoplasia. Design Prospective comparative multicentre study. Setting Three hospitals with outpatient gynaecology clinics and three cancer screening clinics. Participants and methods Cervical cytology and cervicography were performed on 5724 women. If one or both tests showed an abnormality suggestive of at least a low grade squamous intraepithelial lesion, a colposcopy with directed biopsy was carried out. Cervicograms were evaluated by four experienced 'senior' assessors and by ten new 'junior' assessors. Results Results were fully analysed for 5192 women (91%). A cervical biopsy was carried out on 228 women and this confirmed a true positive lesion in 116 cases (incidence rate: 2.2%). Of these, 72 cases (62.1%) were detected by cervicography and 64 (55.2%) by cytology. This difference was not statistically significant (McNemar: P = 0.475). Only 20 cases of CIN (17%) were concordantly detected by both tests. Senior assessors performed significantly better with a detection capacity of 80.6% compared to a detection capacity of 56.6% for the junior assessors (chi(2) test: P = 0.034). Conclusions Cervicography must be considered as a complementary test to cytology. Overall detection of CIN is improved, but this is mainly due to the detection of more low grade lesions. The lower sensitivity and specificity in high grade lesions compared with cervical cytology is the main limitation of cervicography in screening for GIN. An important finding was that the performance of cervicography was highly dependent on the assessors' experience.