SSM - Population Health, Год журнала: 2024, Номер 29, С. 101736 - 101736

Опубликована: Дек. 9, 2024

Malignant neoplasm of the breast was fifth leading cause death among women in Germany 2020. To improve early detection, nationwide cancer screening (BCS) programmes for 50-69 have been implemented since 2005. However, has not reached European benchmark 70% participation, and socio-demographic inequalities persist. At same time, challenges exist to identify groups at high risk non-participation, it is likely that this due disadvantages on multiple social dimensions. This study, therefore, aimed intersectional higher attending BCS by comparing two analytical strategies: a) evidence-informed regression b) decision tree-based regression. Participants were drawn from German 2019 Health Interview Survey (N = 23,001; 21.6% response rate). Two logistic regressions using cross-classification based relevant PROGRESS-Plus characteristics adjusted age built. The approach selected variables literature tree best-performing tree. first identified low-income born outside Germany, living rural areas cohabiting with their partner never (OR 9.48, p 0.002), whereas second, a Classification Regression Tree (61.91% balanced accuracy), determined widowed alone, children, or other arrangements, residing specific federal states (i.e. Bavaria, Brandenburg, Bremen, Hamburg, Saarland) 3.43, < 0.001). Compared regression, yielded discriminatory accuracy (AUC 0.6726 vs AUC 0.6618) added nuances identification at-risk groups, going beyond known inequality dimensions and, helping inclusion under-studied populations screening.

Язык: Английский