OBJECTIVES: We estimated stage-specific loss of quality adjusted expectancy (loss-of-QALE) to evaluate the cost-utility of breast cancer (BC) screening using real world data considering false positive (FP) costs.

METHODS: We interlinked four national databases. A total of 106,199 females with BC were identified from 2002-2015 and followed until 2017. We collected the utility values measured by EQ-5D questionnaire from 1,181 women who visited the National Cheng Kung University Hospital with 2,247 repeated measurements from 2011- 2021. Using iSQoL2 R package, we estimated QALE and lifetime medical expenditures weighted by survival probability. We estimated the loss-of-QALE by subtracting the QALE of BC from that of the sex-, age-, and calendar year-matched referents with the utility values collected from the National Health Interview Survey. Women with FP were defined as those who were not diagnosed with BC within one year after screening. Meanwhile, FP costs were sonograms, mammography, biopsies, and surgeries reimbursed by the National Health Insurance. We calculated the incremental cost-effectiveness ratio (ICER) by comparing the loss-of-QALE of those detected by screening versus non-screening during 2004-2013.

RESULTS: The QALEs of stages I, II, III, and IV were 30.0, 25.4, 17.9, and 4.3 QALYs, respectively, while the loss-of-QALEs were 0, 3.8, 11.0, and 23.0 QALYs, respectively. About 55% of women with FP only received one sonogram in outpatients and 9% received twice, and the FP costs ranged from 6 to 8 million US dollars (USD). The ICER of mammography would be lower than $ 1,231 USD per QALY gained if the coverage rate exceeded 30%.

CONCLUSIONS: Mammography screening program in Taiwan seemed less than 1 GDP per QALY after accounting for FP related costs under a universal health coverage system. Future studies are warranted to evaluate from both medical and societal perspectives.