NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Hartmann KE, Hall SA, Nanda K, et al. Screening for Cervical Cancer [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2002 Jan. (Systematic Evidence Reviews, No. 25.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Screening for Cervical Cancer

Screening for Cervical Cancer [Internet].

Show details



To place choices about cervical screening in context, we need to highlight 4 factors not explicitly investigated in this systematic evidence review (SER).

First, the successes of cervical cancer screening to date, using conventional cytology, have been achieved using a diagnostic test that in other settings would be considered weak. Synthesis of the highest quality literature evaluating traditional Pap testing produces estimates that its sensitivity is 51 percent and specificity is 98 percent.32,50

Second, these test characteristics have proven adequate because cervical dysplasia is a slowly progressing and often self-resolving condition. The estimated progression time from severe dysplasia or carcinoma in situ to invasive cancer is 10 to 15 years.21,32 Thus, repeated screening builds in redundant opportunities to detect abnormalities.

Third, cervical dysplasia itself does not cause morbidity or mortality. Low-grade squamous intraepithelial lesions are the most common changes detected by cytology in regularly screened populations, but they are also the most likely to resolve spontaneously. These low-grade changes are most common among young women, as are intermittent, recurrent, and resolving infection with human papilloma virus (HPV).

Fourth, an ideal screening system will optimize detection of high-grade lesions and minimize evaluation of low-grade and false-positive test results. However, this ideal can be achieved only with improved use of screening among the currently unscreened and with reliable systems for assuring follow-up of abnormal screening findings.

Major Findings and Limitations of the Literature

Who Should be Screened and How Often?

We focused this question on screening among women age 65 and older and for those who have had a hysterectomy and on identifying research that directly compared methods of selecting screening interval. In summary, all the available evidence is observational, predominantly from large population-based data sources and from a small number of prospective cohort studies. Given these sources of information, the findings of these studies are highly coherent and support the following conclusions:

  • The risks of high-grade cervical lesions and cancer fall with age.
  • A history of prior normal Pap tests further reduces risk.
  • If screening recommendations are not modified with age, older women are disproportionately likely to have evaluations for false-positive findings.

Among previously screened women with a history of normal Pap tests, fewer than 1 individual per 1,000 screened (in some scenarios as few as 1 per 10,000) screened will have a high-grade cytologic abnormality. As an example, if the sensitivity of cytology is 60 percent and the specificity is 98 percent for detection of high-grade abnormalities, then 34 women will be evaluated for high-grade squamous intraepithelial lesion for each true high-grade cervical lesion identified; moreover, two high-grade lesions will have been missed by cytology for every three cases identified. The ratio of true positives to false positives is much higher if low-grade cytologic changes are considered. In unpublished work, Sawaya and colleagues' report that 231 additional procedures - 112 extra Pap tests, 33 colposcopies, 30 biopsies, 35 endocervical curettages, 8 endometrial biopsies, 4 dilation and curettages, 7 loop electrosurgical excision procedures, and 2 cone biopsies - were done in response to 110 Pap tests reported as atypical squamous cells of unknown significance (ASCUS) or greater on cytology among a group of menopausal women from a prospective cohort. 44 At the conclusion of these evaluations, a case of cervical intraepithelial neoplasia 2 and a case of vaginal intraepithelial neoplasia III (not the target of cervical cancer screening) had been identified.

The literature provides fairly reliable estimates of the number of women who need to be screened to detect serious lesions. Recommendations can be made to discontinue or substantially lengthen the interval, beyond 3 years, for screening among women age 65 and older who have a history of prior normal Pap tests, depending on tolerance for missing rare cases that would have been detectable under different screening systems. The difficult trade-off between overscreening and missing rare but potentially preventable cases is a challenge for policy in this area. It suggests at minimum that women and their providers should be fully informed about the relatively larger risk of overintervention compared to the much smaller risk of failing to detect a high-grade lesion that would lead to morbidity or mortality.

Prior recommendations of the US Preventive Services Task Force to discontinue Pap testing after hysterectomy for benign disease are clearly supported and should be re-emphasized. Lastly, no direct comparisons between outcomes of proposed screening systems or interval for screening were identified. These findings confirm that this literature is substantially less well developed than that of other areas of screening tests, such as colorectal cancer screening, in which trials comparing use of screening methods over time are available to inform decision-making.

New Methods for Preparing or Evaluating Cervical Cytology

This portion of our SER evaluated the performance of new technologies for preparing or interpreting the cytology specimens; these included liquid-based cytology collection systems, neural-net screening and rescreening tools, and computer algorithms for selecting slides for screening or rescreening. This literature is fundamentally limited by lack of histologically confirmed performance measures -- no gold standard is used for comparisons. Of the very few studies using colposcopy and histology to verify the diagnostic test characteristics of new tools, the most common shortcoming was failure to apply the gold standard to test negative population or a subset to allow estimation of specificity.

To receive approval from the Food and Drug Administration, each of the new technologies currently in use (ThinPrep®, PapNet®, and AutoPap®) had to demonstrate improved sensitivity over conventional Pap testing. As noted in the introduction of this chapter, improving sensitivity, especially if it amplifies detection of low-grade lesions, may not be a benefit at a population level in terms of reducing the burden of morbidity and mortality associated with cervical cancer. This is especially true if increased sensitivity is accompanied by decreased specificity, requiring evaluation of a greater number of false positives and increasing costs. It is precisely these parameters that are most poorly measured in this literature.

Overall, the quality of this literature is poor for the purposes of making decisions about choice of screening systems in US populations. No randomized trials or prospective cohort studies relate use of a screening modality over time to outcomes for individual women. The cost-effectiveness of use of new technologies has only been estimated, not measured directly. The most sophisticated and thorough cost model to date is significantly hindered by the limitations of the literature. Nonetheless, it demonstrates that, at present, new technologies are more costly than conventional cytology; only if used in screening intervals of 3 years or longer will new technologies fall within the traditional range considered to be cost-effective ($50,000 per life-year). 32

The Role of HPV Testing in Cervical Cancer Screening and Triage

HPV testing will have the greatest utility if it can aid in identification of those with high-grade cervical lesions that require prompt treatment and confirm low-risk status among those with comparatively minor Pap test abnormalities such as ASCUS or cervical intraepithelial neoplasm (CIN) 1. The performance characteristics of tests for high-risk HPV types suggest greatest benefit will emerge from the latter role or from combinations with other screening modalities in the former role to improve detection. The current literature on diagnostic test performance is of fair quality with good use of histologic tools to verify HPV test results; its primary limitation is lack of prospective and experimental evidence for its role in screening or triage.

At least 8 studies evaluating HPV testing in large populations are under way or recently completed but not yet in the published literature. 21 The Atypical Squamous Cells of Undetermined Significance/Low-Grade Squamous Intraepithelial Lesions Triage Study (ALTS) in the United States has completed enrollment of 5,060 women with ASCUS diagnosed by conventional cytology and then randomized at enrollment to immediate colposcopy, HPV testing, or repeat Pap with ThinPrep®, the latter two arms are used to triage patients to colposcopy or less intensive follow-up. Participants are followed for clinical outcomes for two years from enrollment. The ThinPrep® arm was closed before completion of the trial, suggesting that HPV is at least more effective than repeat cytology using a more sensitive cytology tool. Definitive evidence on which to base recommendations and further model performance of HPV will be available from US and European trials within the year.

Benefits and Harms

The tangible benefit of cervical cancer screening is reduction of cervical cancer morbidity and mortality by identification of treatable precursors or early stage disease. Intangible benefits (such as reassurance for those with normal Pap tests) or concurrent preventive care at the time of Pap testing (such as screening for sexually transmitted disease or contraceptive counseling) are not documented. These intangible benefits are of particular interest because presumably they play a role in the persistence of annual screening among low-risk patients. Although some observers postulate that such overscreening occurs as a result of provider habit and the tradition of the annual pelvic examination, published evidence does not address either patient or provider knowledge, attitudes, or desire for annual testing.

Harms of screening are poorly understood. The majority of relevant research documents the psychological distress associated with the having an abnormal Pap test. Qualitative studies suggest: (1) women have unmet information needs about the meaning of abnormal results and lack factual information about what to expect during subsequent evaluation; (2) they experience distress and anxiety before and after evaluation while the diagnosis is unknown; and (3) patient education interventions can successfully reduce the fear and uncertainty associated with follow-up care for abnormal Pap testing.74-80 Long-term concerns of women with and without definitive diagnosis of cervical abnormality and the influence of previous evaluation and outcomes on future screening behavior are unclear.

Future Research Needs

Future research must address outcomes of specific screening strategies, not only descriptive statistics and evaluation of test characteristics. On-going projects in the area of HPV testing may be the only areas of experimental research testing explicit hypotheses about the best approach to Pap testing. Until large randomized trials or well-designed prospective studies (capable of measuring and adjusting for anticipated biases) are done, especially ones that address topics such as age to discontinue screening and actual costs of new technologies, the relative benefits of one screening strategy over another are unproven and difficult to promote. The notable exception to indecision in the face of insufficient evidence is consumer demand for use of new technologies, at intervals unlikely to be cost-effective.

The community of cervical cancer researchers oriented towards preventive interventions and screening have consistently called for research of these types:

  • Study of the factors that determine uptake and continuance of screening, provider and patient preferences, and adherence to appropriate screening, follow-up, and evaluation;
  • Investigation of the potential for automated screening processes, such as computerized re-screening and screening, to reduce between laboratory variations in quality;
  • Comparison of cytology technologies and HPV testing methods with a histologic reference standard including verification of the status of individuals, or a sample of individuals, with normal test results;
  • Direct, prospective comparisons of screening strategies that include assessment of health outcomes and cost;
  • Continued study of the natural history of cervical dysplasia and HPV infection;
  • Investigation of the consequences of diagnosis, evaluation, and treatment of low-grade cervical intraepithelial abnormalities or detection of HPV with respect to potential psychological distress, sexual behavior, future reproductive and sexual function, and demands on the health care system;
  • Trials of preventive interventions such as HPV vaccination, use of topical retinoids, and smoking cessation intervention among women with CIN;
  • Outcomes studies evaluating use of HPV testing to guide cessation of Pap testing among older women, and
  • Evaluation of health care systems methods for documenting the prior Pap test history of individual women, to promote appropriate screening and use of screening resources.

Advances in detecting high-grade lesions and minimizing overscreening must go hand-in-hand with active research on promoting uptake and continuance of appropriate Pap testing. The majority of cervical cancers still occur among women who are unscreened, who are inadequately screened, and who do not receive appropriate follow-up for abnormal screening test results. As Table 5 summarized in Chapter 3, a robust literature, including randomized intervention trials, provides insight into methods to promote uptake and continuance and to improve follow-up. New approaches to screening and new tools for screening need to be deployed within a care system that improves documentation of women's screening status, enhances the knowledge of both patients and providers, and optimizes reaching the unreached.


  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...