Display Settings:


Send to:

Choose Destination
See comment in PubMed Commons below

Screening for Cervical Cancer [Internet].


Rockville (MD): Agency for Healthcare Research and Quality (US); 2002 Jan.
U.S. Preventive Services Task Force Evidence Syntheses, formerly Systematic Evidence Reviews.



Methods that improve detection of serious cervical lesions while minimizing excess screening are the key to advancing cervical cancer prevention.


To examine the evidence about benefits and harms of screening among older women (ages 65 and older) and those who have had hysterectomies, and to examine the diagnostic performance of new technologies and human papilloma virus (HPV) testing for detecting cervical lesions.


We identified English-language articles on cervical neoplasia, cervical dysplasia, and screening from a comprehensive search of the MEDLINE database from 1995 through June 2000. In addition, we used published systematic reviews, the second Guide to Clinical Preventive Services, and peer review to assure a complete update of specific topics.


We included articles that reported on screening for squamous cell carcinoma of the cervix if they included the age distribution of the study population and presented analyses stratified by age or if they included hysterectomy status as a covariate. For diagnostic tools, we required that the test be used as part of a screening strategy, that the method be compared with a reference standard, and that all cells of a 2x2 table can be completed.


We extracted the following data from articles addressing screening among older women and those who have had a hysterectomy: study design, objectives, location and timeframe, source of the data (e.g., population-based registry), participants, screening program used, outcomes and measures, and results relevant to age and screening interval. For articles about diagnostic tests, we extracted study design, test methods, location, patient population, outcome measures (emphasizing documentation of the reference standard), prevalence of lesions, and test characteristics including sensitivity, specificity, and predictive values. We used scoring checklists to summarize strengths of the publications; we also evaluated the validity of each article and the overall quality of the evidence.


The evidence about age and hysterectomy is observational, predominantly from population- or care-based data. The findings are consistent: risk of cervical cancer or abnormalities falls with age; high-grade and more severe lesions are detected in fewer than 1 per 1,000 Pap tests among women older than 60 who have had prior screening; and longer histories of prior normal Pap tests further reduces risk. After hysterectomy, high-grade vaginal lesions are rare, fewer than 2 to 4 per 10,000 tests. The literature about new diagnostic tools is limited by lack of histologically validated performance. Using tools such as liquid cytology, neural-net rescreening, and computer-based review algorithms improves sensitivity; however, this improvement is predominantly for detection of low-grade lesions. The impact on specificity is poorly documented. Sensitivity of HPV testing for screening detection of high-grade lesions is competitive with conventional cytology (roughly 82%); specificity is lower (78%); and negative predictive value is good (99%). For triage of women with abnormal Pap tests, sensitivity for detecting high-grade lesions is 85%, specificity is 60%, and negative predictive value is 97%.


The yield of screening among older women who have been previously screened decreases with age; if recommendations are not modified, older women are disproportionately likely to have evaluations for false-positive findings. The prior recommendation of the US Preventive Services Task Force to discontinue Pap testing after hysterectomy for benign disease is supported. For making decisions about screening modality in US populations, evidence about these new technologies for cytology screening and HPV testing is currently limited. Controlled trials and prospective cost evaluation of new screening strategies in each of these areas are required. Important trials will be completed in 2001 that may clarify our conclusions.

Books & DocumentsFree full text
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Write to the Help Desk