Breast Cancer Mortality After Implementation of Organized Population-Based Breast Cancer Screening in Norway

We estimated breast cancer (BC) mortality reduction associated with invitations to a nationwide population-based screening program and with changes in treatment.

BreastScreen Norway started in 1996 and became nationwide in 2005. It invites women aged 50-69 years to biennial mammographic screening. We retrieved individual-level data for 1 340 333 women from national registries. During 1996-2014 (screening window), women contributed person-years in noninvited and invited periods. We created comparable periods for 1977-1995 (prescreening window) by dividing the follow-up time for each woman into pseudo-noninvited and pseudo-invited periods. We estimated BC mortality for the four periods, using the so-called evaluation model: counting BC deaths in each period for all women diagnosed within the period and counting BC deaths and person-years after screening-age for those diagnosed within screening age. We used a multivariable flexible parametric survival model to estimate hazard ratio (HR) for the effect of invitation and improved treatment.

Using the regression approach, we found 5818 BC deaths across 16 533 281 person-years. Invitations to screening reduced BC mortality by 20% (HR = 0.80, 95% confidence interval [CI] = 0.70 to 0.91) among women 50 years and older and by 25% (HR = 0.75, 95% CI = 0.65 to 0.86) among screening-aged women. The treatment effect was 23% (HR = 0.77, 95% CI = 0.65 to 0.92) for women 50 years and older and 17% (HR = 0.83, 95% CI = 0.74 to 0.94) for screening-aged women.

We observed a similar reduction in BC mortality associated with invitations to screening and improvements in treatment during 1977-2014, among women 50 years and older.