An incremental analysis was performed and the total costs and QALYs for all ten strategies were reported. The dominated and extendedly dominated strategies, which were more costly or less cost-effective and less effective than other strategies, were excluded. A trade-off plot was presented to demonstrate the cost-effectiveness or domination of the strategies.
Mammography and clinical breast examination every two years and given in alternating years for those aged 40 to 79 years, had an incremental cost-effectiveness ratio (ICER) of $35,500 per QALY gained compared with no screening.
Biennial mammography and annual examination for those aged 40 to 79 years, had an ICER of $90,100 per QALY gained compared with the previous strategy (mammography and examination every two years from age 40 to 79 years).
Annual mammography and examination for those aged 40 to 59 years plus biennial mammography and annual examination for those aged 60 to 79 years, had an ICER of $169,500 per QALY gained, compared with the previous strategy.
Annual mammography and examination for those aged 40 to 79 years (a guideline strategy), had an ICER of $367,100 per QALY gained, compared with the previous strategy.
Annual mammography for those aged 40 to 79 years, plus triennial examination for those aged 20 to 39 years and annual examination for those aged 40 to 79 years (another guideline strategy), had an ICER of $3,939,000 per QALY gained, compared with the previous strategy.
Two other guideline strategies were dominated. At the commonly accepted willingness-to-pay threshold of $50,000 per QALY, only the first strategy (mammography and clinical breast examination every two years for those aged 40 to 79 years) could be considered to be cost-effective.
The results of the one-way sensitivity analysis showed that varying the sensitivity and specificity altered the relative cost-effectiveness of the strategies. Compared with the base-case results, two strategies became dominated and two ceased to be dominated. Sensitivity analysis assuming the alternative survival model has the same impact on the results.