Comparing two modalities of management of women with cytologic evidence of squamous or glandular atypia: Early repeat cytology or colposcopy

Early repeat cytology is recommended in most screening programs for cervical cancer in subjects with squamous or glandular abnormalities not amounting to neoplasia (atypical squamous cells of undetermined significance, ASCUS), but immediate colposcopy is also recommended in some countries, especially those where there is easy access to colposcopic facilities. We evaluated the cost-effectiveness of the two procedures in a prospective study of women with cytologic ASCUS, invited to cytocolposcopic assessment after 6 months. Colposcopy-directed biopsy was assumed as the gold standard, and the accuracy of colposcopy at 6 months was assumed to be equal to that of immediate colposcopy. Out of 874 compliers, punch biopsy was performed in 303 cases (34.7%), and 19 CIN2+ lesions were detected (CIN2 = 12, CIN3 = 6, microinvasive carcinoma = 1). Detecting 13 CIN2+ lesions at colposcopy required 874 colposcopies and 303 directed biopsies: the cost per CIN2+ lesion detected with the procedure was 2,749 US$. Detecting 15 CIN2+ lesions at repeat cytology required 874 cytologic examinations, 137 colposcopies, 64 directed biopsies, and 6 diagnostic large-loop resections, the latter being performed in subjects with high-grade squamous intraepithelial lesion and less severe lesions at punch biopsy: the cost per CIN2+ lesion detected with the procedure was 1,961 US$. The policy of repeat smear was more cost-effective than immediate colposcopy. According to such results, the protocol of the Florence screening program has been modified since October 1996.