Applicability of the Bethesda System 2001 to a public health setting.

Abstract

The degree of applicability of the Bethesda System 2001 (TBS 2001) for cervicovaginal cytology to a public health setting is unknown, and extrapolations from available data are unwarranted. A "before/after" study design was used to evaluate the impact of TBS 2001 on an organized, population-based screening program in northern Italy. Between 2003-2004, 6 cytology laboratories converted from TBS 1991 to TBS 2001. A set of screening indicators based on TBS 2001 (85,012 patients) were compared with those based on TBS 1991 (199,833 patients) by means of their laboratory- and patient age-standardized ratio with a 95% confidence interval (CI). The prevalence of cervical intraepithelial neoplasm (CIN)2-3/carcinoma was stable between the 2 populations. TBS 2001 had no effect on the unsatisfactory rate (1.99% vs. 2.03% for TBS 1991) nor on follow-up compliance rate (93.2% vs. 92.3%). The reporting rate of atypical squamous cells (ASC) decreased from 17.1 to 14.7 per 1000 (ratio, 0.86; 95% CI, 0.81-0.91), the total positivity rate from 31.1 to 29.0 per 1000 (ratio, 0.93; 95% CI, 0.90-0.97), and the ASC:SIL (squamous intraepithelial lesion) ratio from 1.38 to 1.16. Compared with the ASCUS (ASC of undetermined significance) reports of TBS 1991, the predictive value for CIN2-3/carcinoma decreased from 5.2 to 3.5% (ratio, 0.68; 95% CI, 0.48-0.93) among ASCUS reports, but increased from 5.1 to 17.2% (ratio, 3.41; 95% CI, 1.64-6.28) among ASC-cannot exclude high grade lesion (ASC-H) reports. ASC-H had a 5.01-fold (95% CI, 2.23-10.2) greater predictive value than ASCUS. TBS 2001 is applicable to cervical screening in a public health setting.