Consistency of care across US may have ‘biggest impact’ on breast cancer outcomes
Click Here to Manage Email Alerts
Initial breast cancer care across the United States is inconsistent, which is largely attributed to regional factors rather than patient factors, according to data published in JAMA Oncology.
However, researchers said most of the variance they saw in breast cancer care in the country was random or unexplained.
“We know that breast cancer care varies and that some patients don’t receive recommended care, but we don’t know whether that variation is driven more by region or by patient factors (eg, race/ethnicity, socioeconomic status, age),” Michael J. Hassett, MD, MPH, attending physician in the department of medical oncology at Dana-Farber Cancer Institute, told Healio. “Being able to identify the regions or the patient groups that are most likely to experience suboptimal care will help to design strategies that improve quality of care.”
Establishing metrics
Hassett and colleagues conducted a retrospective population-based cohort study using data from the Surveillance, Epidemiology and End Results (SEER)-Medicare database. They included 31,571 patients (83.7% non-Hispanic white; median age, 71 years) who were diagnosed with stage I to III breast cancer between 2007 and 2013.
Five metrics of care were established:
- stage I at diagnosis (proxy for screening mammography and timely surgery);
- chemotherapy receipt;
- radiation therapy receipt;
- endocrine therapy initiation within 1 year of diagnosis; and
- continuation of endocrine therapy 3 to 5 years after diagnosis.
A simulation was used to determine treatment variance across the five metrics, and whether it could be attributed to randomness, patient factors, health service area (HSA)/region or something outside of these three factors, which the researchers deemed “unexplained variability.”
Assessing treatment variation
Among patients eligible for various treatments, 17,297 of 21,190 (81.6%) received radiation therapy, 7,204 of 9,903 (72.8%) received chemotherapy, 13,115 of 26,855 (48.8%) started endocrine therapy and 13,944 of 26,855 (52.1%) continued endocrine therapy, according to the study.
Variance attributed to HSA/region was greatest for endocrine therapy (39% for continuation, 28% for initiation), and 12% of variance in radiation therapy, 4% of variance in stage I at diagnosis and 3% of variance in chemotherapy were attributed to HSA/region.
“For each of the metrics, the variance attributed to the region (HSA) was multifold larger than that explained by patient factors,” which only accounted for 1% to 4% of the variation, the researchers said.
They also found that “most of the observed geospatial variation is random and cannot be explained, and there was significant geospatial variation in the use of radiation and endocrine therapy, but not for the use of other cancer treatments,” Hassett told Healio.
“To improve breast cancer care and outcomes, reducing unwarranted geospatial variation (ie, ensuring that care is consistent wherever it is provided in the U.S.) may have the biggest impact on outcomes irrespective a person’s age, race/ethnicity, socioeconomic status, etc.,” said Hassett, who is also an assistant professor of medicine at Harvard Medical School. “Notably, quality improvement efforts should target a small number of lower performing regions for endocrine therapy and radiation therapy.”
Moving forward, Hassett told Healio that research should examine variance in breast cancer outcomes and determine the characteristics of relatively low-performing regions.
Perspective
Back to Top
Breast cancer disparities in the U.S. are a long-standing problem that have consistently worsened over time. For years, extensive research has been published describing the perpetual differences in breast cancer outcomes by race: women who are Black have a 40% higher risk of dying from breast cancer compared with women who are white.
Years of research have identified several potential and complex factors contributing to this persistent and stubborn injustice. Renewed awareness of our nation’s insidious racism — from which medicine and health care are not exempt — has generated genuine enthusiasm to uncover and address health inequities. These efforts across multiple sectors of our society have resulted in expanded attempts to understand and correct these patterns. However, we must be cautious about the risk of oversimplifying this complex issue.
In a brief report in JAMA Oncology by Hassett et al, the authors examine variation in breast cancer care by “region vs. patient factors.” The choice of categories is problematic given that it is impossible to dissociate socioeconomic “patient factors” from where people live. As we know, race and poverty are linked: legacies of structural racism influencing the Great Migration, redlining, predatory lending and racial wealth gaps have a distinct effect on geographic segregation by race. How “patient factors” or regions determine health access, quality and treatment decisions is complex. Perhaps this is a major reason for the substantial unexplained geographic variance in this brief report. As the authors correctly conclude, efforts to identify the unexplained variance are warranted.
Overall, analyzing how patient factors and HSAs/regions contribute to breast cancer disparities are critical and certainly to be encouraged. Unfortunately, how we define categories and address the inherent complexity of treatment decisions remains a significant challenge. Further, large datasets such as the SEER database are blunt instruments at best. We must remain cautious in drawing conclusions about causation and reflectively note the assumptions in our findings. Otherwise, research may influence clinical interventions that could be unwarranted, expensive and inadvertently assign undue responsibility to the patient.
Reference:
Wilkerson I. The Long-Lasting Legacy of the Great Migration. Smithsonian Magazine. Published September 2016. https://www.smithsonianmag.com/history/long-lasting-legacy-great-migration-180960118/. Accessed Feb. 6, 2022.
Collapse
Hassett MJ, et al. JAMA Oncol. 2022;doi:10.1001/jamaoncol.2021.7337.
Read more about
Click Here to Manage Email Alerts