Mammography Screening and Breast Cancer Mortality in New South Wales, Australia.

摘要

Objective : To investigate the relationship between utilisation of service mammography screening and breast cancer mortality in New South Wales (NSW) women. Setting : Population-based biennial mammography screening was progressively introduced in NSW from 1988, with active recruitment and re-invitation for women aged 50-69 years, and reached full geographic coverage by 1996. Biennial mammography screening participation has varied widely over time and by municipality. Methods : Breast cancer mortality by age, period and municipality was obtained from the NSW Central Cancer Registry. Biennial mammography screening rates for the same strata were obtained from the BreastScreen NSW database. Temporal changes in breast cancer mortality for NSW were summarised as annual average declines using Poisson regression. Breast cancer mortality for 1997-2001 was examined in relation to lagged biennial screening rates by municipality, adjusted for age, area socio-economic and geographic indicators, and breast cancer incidence, also using Poisson regression. Results : For the 50-69 year age group, the mean annual breast cancer mortality decline was 0.8% (not significant) for 1988-1994, and 4.4% ( p < 0.0001) for 1995-2001. Statistically significant negative associations between breast cancer mortality in 1997-2001 and lagged biennial screening rates were found with the highest significance at a four-year lag for women aged 50-69 years ( p = 0.0003) and also for women aged 50-79 years ( p c = 0.0002). From the regression coefficient, a 70% biennial screening rate is associated with 32% lower breast cancer mortality (compared to zero screening). Conclusions : The effect of population-based mammography screening on breast cancer mortality in NSW inferred using this method is consistent with results of trials and other service studies. This suggests that population-based mammography screening programs can achieve significant reductions in breast cancer mortality with adequate participation.