The role of colposcopy in the follow up of women treated for cervical intraepithelial neoplasia

Objective To determine the effect of colposcopy on the sensitivity and specificity of follow-up procedures in the detection of residual or recurrent disease after treatment for cervical intraepithelial neoplasia (CIN). Design A retrospective study of information held in a colposcopy database. Setting A teaching hospital colposcopy clinic. Population A total of 2244 women treated in Hammersmith Hospital for histologically confirmed CIN between 1 January 1988 and 31 December 2002. Methods Data from the records of women treated with some form of local conservative therapy for CIN1-CIN3 between January 1988 and December 2002 were extracted from the colposcopy database. Women with histological confirmation of post-treatment disease were identified. Main outcome measures The sensitivity and specificity of cytology alone was compared with the sensitivity and specificity of the combination of colposcopy and cytology. Results Colposcopy improved the sensitivity of cytology for the detection of high-grade disease from 64 to 91% but reduced the specificity from 95 to 88%. With a 3% rate of post-treatment high-grade disease, colposcopy detected 8 extra cases per 1000 women but resulted in 88 more false alarms per 1000 women. Among women in whom the treatment margins were involved or uncertain, colposcopy detected 13 extra cases per 1000 women but resulted in 12 fewer false alarms per 1000 women because the prevalence of post-treatment disease was higher. Conclusions Colposcopy does improve the detection rate of post-treatment disease but at a cost of additional false alarms. The benefit of colposcopy will be greater in high-risk groups of women with higher rates of treatment failure.