The Effect of Patient and Contextual Characteristics on Racial/Ethnic Disparity in Breast Cancer Mortality

Abstract

Background Racial/ethnic disparity in breast cancer-specific mortality in the U.S. is well documented. We examined whether accounting for racial/ethnic differences in the prevalence of clinical, patient, and lifestyle and contextual factors that are associated with breast cancer-specific mortality can explain this disparity. Methods The California Breast Cancer Survivorship Consortium combined interview data from six California-based breast cancer studies with cancer registry data to create a large racially diverse cohort of women with primary invasive breast cancer. We examined the contribution of variables in a previously reported Cox regression baseline model plus additional contextual, physical activity, body size, and comorbidity variables to the racial/ethnic disparity in breast cancer-specific mortality. Results The cohort comprised 12,098 women. Fifty-four percent were non-Latina Whites, 17% African Americans, 17% Latinas, and 12% Asian Americans. In a model adjusting only for age and study, breast cancer-specific hazard ratios relative to Whites were 1.69 (95% CI 1.46 -1.96), 1.00 (0.84 - 1.19), and 0.52 (0.33 - 0.85) for African Americans, Latinas, and Asian Americans respectively. Adjusting for baseline-model variables decreased disparity primarily by reducing the hazard ratio for African Americans to 1.13 (0.96 - 1.33). The most influential variables were related to disease characteristics, neighborhood socioeconomic status, and smoking status at diagnosis. Other variables had negligible impact on disparity. Conclusions While contextual, physical activity, body size, and comorbidity variables may influence breast cancer-specific mortality, they do not explain racial/ethnic mortality disparity. Impact Other factors besides those investigated here may explain the existing racial/ethnic disparity in mortality.

Relationships between social isolation, neighborhood poverty, and cancer mortality in a population-based study of US adults

“…Furthermore, this study was limited by the unidimensional measure of neighborhood SES that was used. Although neighborhood poverty is a valid measure of area-level SES [ 20 , 47 ], other factors that were not accounted for (such as those described above) may play an important role which was not captured here [ 48 ]. In addition, there is evidence that neighborhood characteristics as perceived by the individual living within the neighborhood may better predict some health outcomes than those captured by census-level variables alone [ 18 , 49 – 51 ].…”

Racial disparities in mortality for patients with prostate cancer after radical prostatectomy

“…Prognostic variables were then added to the baseline model one by one to test their influence on the racial survival disparities, that is, the changes in HRs for race from the baseline model. The significance of the HR changes was evaluated with a previously developed approximate likelihood ratio–based test termed disparity χ 2 32 . The prognostic variables were ranked in order of the significance of the HR changes in all samples.…”

Multilevel social factors and NICU quality of care in California

“…Analyses of associations between zip-code level racial and economic segregation and preterm birth and infant mortality in California showed that women and infants in less privileged zip codes were at increased odds for these adverse outcomes [47]. The contributions of such multilevel resources have also been demonstrated in other areas of health; in particular, this is an emerging area of research in cancer epidemiologic studies [48,49].…”