Should population-based screening for breast cancer be extended to Canadian women aged 70 to 79? A model-based approach.

Statement of problem. This thesis applied established criteria for evaluation of a screening program to assess the population health impacts of extending population-based breast cancer screening to Canadian women aged 70 to 79. Methods. Experimental evidence was systematically reviewed, complemented by analyses of Canadian data for incidence, mortality, screening participation and treatment patterns for women aged 70 to 79. The MISCAN micro-simulation model was adapted to the Canadian population to assess benefits and harms. Results. Compared to stopping at age 69 and depending upon the level of screening participation and sojourn time, continuing to screen women aged 70 to 79 resulted in an estimated 530 to 547 false positive mammograms per 10,000 screens, 20 to 40 additional biopsies, 10 to 25 "extra" cancers, and 12.2 to 13.7 prevented cancer deaths, while 30 fewer women will be diagnosed with stage T2+ cancer. About 110 life-years will be gained per 10,000 screens, with 210 to 440 life-years lived in lead-time. Quality adjustment of life-years gained reduced the benefit by up to 31% to 48% with a 5% discount factor for shorter vs longer sojoum times respectively. Between 733 and 821 screens will be needed to avert one breast cancer death, with an overall gain in life expectancy of about 4 days per screen. Overall, an estimated 459 breast cancer deaths can be prevented per year in Canada if the national target level of 70% screening participation is reached. Conclusion. Overall, results indicate favourable population health impacts of extending breast cancer screening to screen women aged 70 to 79. More work is needed to discern the sojourn time and the costs associated with the benefits