Why that’s a big data problem for population health.
Fewer black women get breast cancer proportionally, but more die from it, according to data from the American Cancer Society. And yet, very few breast cancer studies factor race or other socioeconomic indicators into their analysis, according to a report published by researchers at the Johns Hopkins University Bloomberg School of Public Health.
In a commentary titled, “Social factors matter in cancer risk and survivorship,” the authors argue that in addition to race and ethnicity, social factors like income, access to healthcare, and socioeconomic position should be regularly incorporated in cancer research because they influence medical outcomes.
In 2015, the main risk-assessment tool for breast cancer was adjusted to accommodate race. Known as the Gail model, the tool is used to target at-risk groups and determine eligibility for prevention trials. Previously, it had been validated only on white women over 35 and had greatly underestimated the risk levels of black women. Once adjusted, the tool increased the percentage of black women eligible for chemoprevention trials to 17.1 percent, up from 5.7 percent.
But despite the recognition socioeconomic factors play a role in cancer risk and outcomes, the vast majority of research continues to ignore these factors, according to the authors. They found that of 57 articles published in 2016 on breast cancer, fewer than 5 percent—of those that were not explicitly designed to study such disparities—reported findings stratified by race and other socioeconomic factors.
Lorraine Dean, ScD, a social epidemiologist at Johns Hopkins and one of the paper’s authors, said that given the evidence from the Gail model, race shouldn’t be the only factor considered.
Dean began to focus on this issue after working with women who suffer from lymphedema, a chronic condition common among women who undergo breast-cancer surgery. Dean found that while the raw data show that there is a higher incidence of lymphedema among black women, some peer-reviewed papers found that race was not a statistically relevant factor.
A 2009 paper that looked at risk factors for arm lymphedema by race concluded that, “While black women had higher prevalence of arm lymphedema than white women (28 percent vs. 21 percent), race was not associated with lymphedema risk in models adjusted for multiple factors.”
Dean says controlling for race while assuming “all else being equal” misses the point because in real life, race can’t be isolated from other social factors. “Saying that when we adjust for it statistically, that [disparity] goes away isn't that meaningful, because they still walk into the clinic living a different experience,” she notes. “It could be things like greater levels of stress. It could be all types of other social factors that essentially race is coding for.”
Not all researchers agree that social factors should be included in medical research. They argue against their inclusion because race and ethnicity are not modifiable factors, and poverty or socioeconomic status are beyond the scope of healthcare practitioners to treat.
Dean acknowledged that in comparison with factor like income, race is extensively reported. But even when disparities across race are documented, they are not often meaningfully addressed. The papers “don't seek to unpack why actually there might be a reason, what social differences are happening across racial groups that make this difference.”
The authors argue that as the evidence in breast cancer research shows, race and socioeconomic factors affect risk and survivorship, and should therefore should be collected, documented and analyzed much more extensively. And not only in breast cancer research, but across cancer research more generally.
Get the best insights in healthcare analytics directly to your inbox.