Objective: To determine the important factors affecting the long-term failure rate of treatment for cervical intraepithelial neoplasia (CIN).
Design: Multivariate analysis of prospectively collected clinical data.
Setting: Colposcopy clinic in a university teaching hospital.
Population: A total of 2455 consecutive women treated for CIN for the first time between 1 January 1989 and 31 December 2004 using excisional techniques.
Methods: The data were obtained from a clinical database into which information had been entered prospectively. Data on community follow-up were obtained from the national Exeter database of cervical screening results. The parameters studied were age, parity, severity of the lesion, extent to which the lesion involved the endocervical canal, seniority of the surgeon, type of treatment, anaesthesia used, dimensions of the treatment biopsy specimen, number of treatment biopsy fragments and completeness of excision.
Main outcome measures: High-grade post-treatment cervical disease detected by biopsy or cytology.
Results: Univariate analysis eliminated several potential factors and identified interactions between some factors. Multivariate analysis identified the grade of disease (P < 0.001) and the extent to which the lesion involved the endocervical canal (P =0.008) as the most important determinants of high-grade post-treatment disease. Of the factors which the surgeon could control, depth of biopsy (P =0.002) and completeness of excision (P = 0.007) were the most important.
Conclusions: This study confirms that clear margins are important to reduce the failure rate of excisional treatment for CIN. An adequate depth of treatment is important when the lesion involves the endocervical canal.
© 2010 The Authors Journal compilation © RCOG 2010 BJOG An International Journal of Obstetrics and Gynaecology.