Assessing new technologies for cervical cancer screening: beyond sensitivity

New technologies have been proposed to replace cervical cytology as the medium for cervical cancer screening, especially testing for human papillomavirus (HPV). Many of these alternatives have been endorsed because their single-test sensitivity is superior to that of cytology. The sensitivity of a single Pap test may be as low as 50%, but lifetime testing has significantly greater sensitivity, and other parameters are also important in defining the optimal screening test. Test accuracy depends not only on sensitivity but also on specificity and disease prevalence; both of these are problematic when considering HPV as a screen. A screening test also must be acceptable to both clinicians and patients. HPV testing is regarded as test for a sexually transmitted disease by many women. Clinicians often fail to appreciate the transient nature of HPV infection, prompting overly aggressive treatment that risks future preterm birth, and there is no consensus about strategies for follow-up of abnormal HPV screening results when obtained without concomitant cytology. An emphasis on sensitivity also may be inappropriate when most cervical cancers occur in unscreened women, for whom the sensitivity of testing is irrelevant. HPV testing and other technologies have substantial promise as tests that may replace cytology in cervical cancer screening, but obstacles are significant, and premature adoption of these methods may result in patient harm.