Understanding the effects of socioeconomic status along the breast cancer continuum in Australian women: a systematic review of evidence

BACKGROUND: Globally, the provision of equitable outcomes for women with breast cancer is a priority for governments. However, there is growing evidence that a socioeconomic status (SES) gradient exists in outcomes across the breast cancer continuum – namely incidence, diagnosis, treatment, survival and mortality. This systematic review describes this evidence and, because of the importance of place in defining SES, findings are limited to the Australian experience. METHODS: An on-line search of PubMed and the Web of Science identified 44 studies published since 1995 which examined the influence of SES along the continuum. The critique of studies included the study design, the types and scales of SES variable measured, and the results in terms of direction and significance of the relationships found. To aid in the interpretation of results, the findings were discussed in the context of a systems dynamic feedback diagram. RESULTS: We found 67 findings which reported 107 relationships between SES within outcomes along the continuum. Results suggest no differences in the participation in screening by SES. Higher incidence was reported in women with higher SES whereas a negative association was reported between SES and diagnosis. Associations with treatment choice were specific to the treatment choice undertaken. Some evidence was found towards greater survival for women with higher SES, however, the evidence for a SES relationship with mortality was less conclusive. CONCLUSIONS: In a universal health system such as that in Australia, evidence of an SES gradient exists, however, the strength and direction of this relationship varies along the continuum. This is a complex relationship and the heterogeneity in study design, the SES indicator selected and its representative scale further complicates our understanding of its influence. More complex multilevel studies are needed to better understand these relationships, the interactions between predictors and to reduce biases introduced by methodological issues.