Detection of high-grade cervical dysplasia: impact of age and Bethesda system terminology.

Pap smear, colposcopy, and biopsy results were collected from 1988-1993 at a group of family planning clinics. Positive predictive values and likelihood ratios were calculated for diagnosis of high-grade lesions based on age and Pap smear results. One thousand and forty-seven colposcopies were logged; 771 had a biopsy or endocervical curettage. Seventy-nine (10%) were high-grade lesions. If only human papillomavirus (HPV) was reported on the Pap smear, the likelihood of a high-grade biopsy was lowest (positive predictive value, 4.5%; likelihood ratio, 0.4). Women under age 25 were less likely to have high-grade biopsies (positive predictive value, 7.3%; likelihood ratio 0.7). Repeat Pap smears for atypical cells of undetermined significance (ASCUS) and low grade squamous intra-epithelial lesion (LGSIL) showing only HPV in women under age 30 would have reduced the immediate colposcopy rate by 60% and delayed diagnosis by 23% of high-grade lesions. Consideration of patient age and whether HPV is the only Pap smear finding may reduce referral for immediate colposcopy.