OBJECTIVE: To evaluate the effectiveness of various screening tests for detecting genital human papillomavirus (HPV) in a community-based population and to determine the prevalence of cervical lesions on colposcopically directed biopsies in patients found to have HPV by any screening test. DESIGN: Cross-sectional analysis of 208 female patients screened for HPV by clinical examination, Papanicolaou test, dot blot hybridization test (ViraPap, Digene Inc, Silver Spring, Md), and polymerase chain reaction (PCR) analysis. All persons with abnormal or positive results by any method were offered colposcopic evaluation. SETTING: Two community-based family practice offices in southeastern Michigan. MAIN OUTCOME MEASURES: The presence of HPV as determined by each test and results of colposcopic biopsies (gold standard) in patients who had HPV identified by any test. RESULTS: The prevalence of HPV infection was 20.3% by PCR analysis, 3.1% by ViraPap, 3.0% by the Papanicolaou test, and 0% by clinical examination. Symptomatic patients (those with complaints of vaginal odor, swelling, or itching) were more likely to harbor HPV as determined by PCR analysis than were asymptomatic women (P = .03, odds ratio = 2.65). Human papillomavirus type 16 or 18 was found in 41% of patients with positive PCR analyses and in all patients with positive ViraPap tests that were typed. Colposcopy with biopsy (gold standard for the presence of HPV disease) was performed on 34 of the 41 patients who tested positive for HPV by PCR analysis. Histologic results revealed that 79.4% of these patients had cervical disease: 14.6% had cervical intraepithelial neoplasia, 38.3% had condyloma, and 26.5% had cervicitis. CONCLUSIONS: Human papillomavirus as detected by PCR analysis was present in 20.3% of women in our population and was often one of the higher-risk types (16 or 18). A positive PCR analysis was predictive of cervical disease on colposcopic biopsy. The ViraPap test, Papanicolaou test, and clinical examination were insensitive measures for detecting HPV-related lesions in this population.
All Science Journal Classification (ASJC) codes