Page 43 - Screening for Cervical Cancer: Systematic Evidence Review

Chapter III. Results

confirmation of CIN 3 and carcinoma and achieved an 82% verification rate, whereas Formso

and colleagues relied on cytology alone. The differences between their findings are certain to

37
reflect the inclusion of unscreened women in the prevalence study but not in the incidence
35
study. The differences may also reflect discrepancies between cytology and histology among

older women and unstable estimates based on small numbers.

The remaining publications reflect falling risk with age and low prevalence among older

41
women. These include the Cecchini et al. study of previously unscreened women in Italy,
39
Cruickshank’s study among British women who are actively advised to be screened every 3

years, and the work by Lawson and colleagues based on opportunistic screening in the US

40
National Breast and Cervical Cancer Screening Program. Among American women being
40
screened for the first time in the study by Lawson et al., rates of CIN 3 or cancer were 2.3 per

1,000 Pap tests among women ages 50 to 64 and 1.7 per 1,000 among those ages 65 and over,

followed by a reduction to 1.3 per 1,000 and 0.7 per 1,000 at the second screening opportunity.

Screening History and Interval

The significance of prior negative smears is directly addressed by both age-adjusted and

age-specific analyses in the examination by Sawaya and colleagues of the probability of

abnormal Pap after 1, 2, or 3 prior normal Paps. The probability of HSIL decreases with each

subsequent normal smear in women ages 30 and older. Reduction within age brackets is detailed

in Evidence Table 1. Investigation of the influence of interval suggests that longer intervals are

associated with similar or increased detection of high-grade lesions; 26,35,43 other work not

reviewed here (but considered by the US Preventive Services Task Force [USPSTF] in the mid-

43