Impact of the more restrictive definition of atypical squamous cells introduced by the 2001 Bethesda system on the sensitivity and specificity of the papanicolaou test: A 5-year follow-up study of papanicolaou tests originally interpreted as ASCUS, reclassified according to Bethesda 2001 criteria

BACKGROUND. The Bethesda 2001 (B2001) classification system raised the threshold for reporting atypical squamous cells. The category of Atypical Squamous Cells of Undetermined Significance (ASCUS)-favor reactive has been eliminated with a recommendation for cytologists to judiciously downgrade those Papanicolaou (Pap) tests that would formerly have been in this category to Negative for Intraepithelial Lesion or Malignancy (NILM). The effect of this change on sensitivity and specificity of the Pap test is not yet known. METHODS. A total of 535 consecutive SurePath Pap tests interpreted as ASCUS during a period from March 1 through December 31, 2001, with follow-up polymerase chain reaction(PCR)-based human papilloma virus (HPV) testing, were independently reviewed by 3 pathologists and 1 cytotechnologist who reinterpreted these tests according to Bethesda 2001 criteria. Follow-up HPV and biopsy results were compiled for a 5-year period. RESULTS. By consensus of the 4 observers, 169 (32%) of the ASCUS cases were downgraded to NILM. These cases showed a lower rate of underlying high-risk HPV infection (11% vs 30%) and cervical intraepithelial neoplasia of grades 2 to 3 (CIN 2/3) (5% vs 10%) on follow-up than those tests that were reinterpreted as ASCUS or higher. Nine women with follow-up CIN 2/3 would have had the Pap test interpreted as NILM instead of ASCUS under Bethesda 2001 (20% of all CIN 2/3 found). Individual reviewers downgraded 29% to 42% Pap tests to NILM including those of 5 to 10 women with follow-up CIN 2/3. CONCLUSIONS. The ASCUS threshold established by B2001 prevents a sizeable subset of women from having follow-up for ASC. However, as a consequence, a few women with CIN 2/3 are downgraded to NILM. The cost savings thus achieved must be weighed against the lost opportunities to detect CIN 2/3.