A nationwide mammographys screening program including women aged 50-59 year
s at the rime of the first invitation and involving more than 100 radiologi
sts was started in Finland in January. 1987. From 1987 through 1997, a tota
l of 1690496 invitations to biennial two-view mammography screening was sen
t out. The compliance For screening was 88.5% with 495744 screening examina
tions performed during this 11-year period. There were 49020 recalls (3.28%
of those attending) for further work-up studies and 9689 women (0.65% of t
hose attending) were referred for surgery. The total number of screening-de
tected breast cancers was 5595, giving a detection rate of 3.7 cancers per
1000 screening studies. More than half (57.7%) of all surgical biopsies rev
ealed breast cancer and 67.8% of the invasive cancers were at Stage I. The
positive predictive value of referral to surgical biopsy increased from 33.
2% in 1987 to 65.5% in 1997, and the ratio of malignant to benign biopsies
more than tripled From the first to the fifth year of screening. The observ
ed/expected ratio of invasive cancel detection was 2.44. Only 0.27% (1 out
of every 372) of ail screening mammograms were followed by a benign biopsy,
and 2.90% (1 out of every 34) of all screening mammograms were followed by
the women bring recalled for further studies and /o not found to have brea
st cancer. This gave a specificity of recall after screening mammography gr
eater than 97.0% and a specificity of referral to surgical biopsy greater t
han 99.7%. Measures of specificity improved considerably during the first t
hree years of the screening program. The high specificity of screening mamm
ography can be attributed to the nature of the screening process as well as
to the opportunity for individual radiologists to attain a greater level o
f experience and competence. The decision to recall appears to have been cr
ucial in determining the specificity.