Table 3.

Key articles and data use

Article reference	Processing discussion
Raftery and Chorozoglou12	Conducted in the UK, this article calculated QALY data collected from a variety of UK-based sources and inputting them into their Southampton Model to calculate QALY data for breast cancer screening. QALY data were extracted from this article. In order to evaluate these data, we created ratios between the papers which took into account the potential harms of breast cancer screening and those that did not. It was then possible to apply these ratios to our QALY data which did not take into account potential harms to calculate the true QALYs gained. This article produced markedly different QALY data to other articles produced at the same time in the same region, as it accounted for risks associated with mammography screening, such as increased risk of breast cancer, overdiagnosis, and overtreatment.
Madan et al28	This article conducted a CUA on the breast cancer screening, 6 years prior to this analysis based on one UK-based randomized control trial and data from the screening program itself. QALY data, probability data, and costing data were obtained from this article for analysis.
Pharoah et al25	This article used its own primary data from cohorts of 364,500 women over a 15-year period to calculate key outcome data and QALY measures in the UK, these data along with costings were extracted for analysis in this report. Probability measures were also obtained for this report.
Robertson et al29	This article used some primary survey data and combined this with systematic review data to calculate QALY and costing data, these were extracted for us in the analysis. Probability measurers were also obtained from this report.
Groot et al30	This article was an exception to the data collected from the other four articles, and it contained data obtained internationally. These data were obtained as part of a project by the WHO and were converted into UK health care system data to ensure robustness of the four UK-based sources. This study used a large sample of data and will ensure estimates are as accurate as possible. In order to convert the data from a US sample to a UK sample, relative health care spends were taken into account, a ratio was created and the figures for the US divided by this ratio. The remaining values were then converted from USD to GBP. Probability measurers were also obtained from this report.

Abbreviations: CUA, cost–utility analysis; GBP, British pound; QALY, quality-adjusted life year; USD, US dollar; WHO, World Health Organization.