Background: Population-based cancer registries are essential for assessing breast cancer screening at the population level. When linked with screening programme data, they enable early evaluation through standardised indicators. However, their use remains heterogeneous and inconsistently reported across Europe. This review examined how registry data have been used to evaluate European population-based breast cancer screening programmes by comparing cancers according to detection mode (screen-detected, interval, and non-screened) and reporting outcomes related to methodological variability, temporal trends, and evidence gaps.
Methods: Following PRISMA-ScR, we included studies evaluating organised European screening programmes that linked their data with a population-based cancer registry. An extraction grid captured study characteristics, linkage approaches, detection mode definitions, outcomes (mortality, survival, tumour characteristics, and screening quality indicators), and bias adjustments. The Reach, Effectiveness, Adoption, Implementation, and Maintenance (RE-AIM) Framework guided the methodological analysis.
Results: Twenty-six studies were included, mostly from Western and Northern Europe. Over 70% were published before 2010, indicating limited recent registry-based evaluations. Screen-detected cancers consistently showed lower mortality, better survival, and more favourable tumour characteristics than interval and non-screened cancers. Reporting of screening quality indicators was inconsistent. Definitions of detection modes, especially interval cancers, and approaches to bias adjustment (e.g., lead time, selection) varied widely, reducing comparability. Few studies used comprehensive bias correction or clearly reported linkage methods.
Conclusions: Registry-based evaluations provide valuable evidence on breast cancer screening impact, yet methodological variability and inconsistent outcome definitions restrict comparability. Renewed and harmonised approaches, including standardised definitions, core indicators, and systematic bias correction, are needed.
Methods: Following PRISMA-ScR, we included studies evaluating organised European screening programmes that linked their data with a population-based cancer registry. An extraction grid captured study characteristics, linkage approaches, detection mode definitions, outcomes (mortality, survival, tumour characteristics, and screening quality indicators), and bias adjustments. The Reach, Effectiveness, Adoption, Implementation, and Maintenance (RE-AIM) Framework guided the methodological analysis.
Results: Twenty-six studies were included, mostly from Western and Northern Europe. Over 70% were published before 2010, indicating limited recent registry-based evaluations. Screen-detected cancers consistently showed lower mortality, better survival, and more favourable tumour characteristics than interval and non-screened cancers. Reporting of screening quality indicators was inconsistent. Definitions of detection modes, especially interval cancers, and approaches to bias adjustment (e.g., lead time, selection) varied widely, reducing comparability. Few studies used comprehensive bias correction or clearly reported linkage methods.
Conclusions: Registry-based evaluations provide valuable evidence on breast cancer screening impact, yet methodological variability and inconsistent outcome definitions restrict comparability. Renewed and harmonised approaches, including standardised definitions, core indicators, and systematic bias correction, are needed.
Evaluating Breast Cancer Screening in Europe Using Population-Based Cancer Registries