Methods: We developed a Markov model to simulate mammography screening, breast cancer diagnosis and treatment in a hypothetical cohort of Hong Kong Chinese women. We derived model inputs from local clinical, epidemiological and economic data, and from a literature review. We modeled the benefit of mammography by assuming a stage shift (i.e., cancers in screened women were more likely to be diagnosed at an earlier stage), estimated via model calibrations with trial results. We compared life years saved (LYS) and costs for five strategies: four biennial screening strategies, beginning at age 40 or 50 and ending at age 69 or 79, and a no screening scenario.
Results: Biennial mammography screening resulted in a gain in life expectancy ranging from 4.3 to 9.4 days compared with no screening at an incremental cost of US$1,170 to 2,239 per woman. The least costly non-dominated screening option was screening women from 40 to 69 years of age, with an incremental cost-effectiveness ratio (ICER) of US$64,300 per LYS compared with no screening. In probabilistic sensitivity analyses, the probability of the ICER being below a threshold of US$50,000 per LYS was 15.2%. Extending the screening program to age 79 (the next most effective strategy) would cost an additional US$238,300 per LYS. Biennial screening would be as low as US$50,000 per LYS if the age-adjusted incidence increased by a factor of 1.4.
Conclusions: Our results suggest that mammography for Chinese women in Asia would likely be an inefficient use of scarce public health dollars currently. However, we must remain vigilant and periodically revisit the question of population screening as the breast cancer incidence in Hong Kong is increasing.