To address disparities in cancer care and, more specifically, to design interventions to address disparities that actually work, it's important to drill down on the details. "We can't fix what we can't measure," noted Katherine Reeder-Hayes, MD, MSc, MBA, Associate Professor of Medicine-Oncology and Section Chief of Breast Oncology at UNC-Chapel Hill Lineberger Comprehensive Cancer Center. That's a big part of the reason Reeder-Hayes and colleagues collected data to investigate patients in North Carolina who are more likely to experience breast cancer treatment delays. The researchers looked at data collected from 2004 to 2017, which included 32,626 patients with breast cancer, of whom 19 percent were Black (Cancer 2023; https://doi.org/10.1002/cncr.34573).
Researchers shared the following findings. The chance of having delayed care is significantly higher if you are a Black woman in the state overall and within the same geographic region. Some regions had bigger gaps between Black and non-Black women than other regions. Accounting for individual factors like cancer stage, treatment type, insurance type, and the distance the patient lived from cancer treatment facilities didn't explain why Black women had more frequent delays than non-Black women in the same area. "That doesn't mean these factors weren't important or didn't matter to treatment delays. [It was] just that those factors weren't the reasons why Black patients did worse," Reeder-Hayes said. She shared her thoughts on why the data are so important in designing and implementing interventions to address these disparities.
1 How is this different from other data investigating this question?
"Historically, much of the research around race and cancer care has been descriptive, meaning research that describes pattern differences between racial groups in different aspects of care without testing interventions to close those gaps. Awareness is important, but our research team strongly believes that awareness alone will not change the patterns we continue to see over the past 30 years of Black people getting less appropriate cancer care and benefitting less from cancer detection and treatment advances.
"The research you see in this publication is intended to identify places in our region where cancer disparities are particularly acute, so we know where we might be able to have the biggest impact from interventions. It is tightly connected to other projects that share data with people in communities and ask for their input on needed solutions and also feeds data into simulation models that look at which interventions might lead to the biggest improvements in racial equity. Ultimately, we hope it will also inform interventions in these problematic areas that we see on the map."
2 What were the key findings reported and how generalizable are they to other geographic areas?
"We found that, while racial disparities in the timeliness of breast cancer care around the state of North Carolina were widespread, the size of those racial gaps and the timeliness of care overall varied quite a lot. Interestingly, we also found that these geographic differences were not explained by the characteristics of the patients living in the regions or the characteristics of their cancer. So, if cancer care teams in certain areas say: 'Oh, it's particularly hard to treat breast cancer in our area because people are poor or have really advanced stages of cancer when they come in,' then our research does not bear out that explanation.
"We were not surprised to find that disparities existed or that they varied by geographic region. But I think we have been surprised, as residents of North Carolina, by which regions performed better and not so well. Each geography will likely have its own patterns and that is the point of the research. To effect change, you will need to know the patterns of care in your own area of interest for your own disease. However, the overall lesson that there will be variation, and that this variation is not explained by the characteristics of the patients living in each location, is I think very generalizable.
"North Carolina is also a great place to conduct such research because it is very diverse. About 21 percent of our population identifies as Black, and the state includes large urban areas, very rural areas, and areas with large Black populations, as well as some with heavily White populations. These characteristics mean that we can learn a lot from this one state about how similar areas across the rest of the country might function."
3 What types of interventions do you plan to test to address these disparities?
"It's hard to change the complexity of the United States' health system with any one intervention. However, if the main driver of delays in care-particularly for groups of people who have been historically marginalized-turns out to be a complex health care system, then interventions that offer better navigation (particularly more carefully targeted to the patients who most need it) may turn out to be effective.
"The data in this particular study say more about what doesn't explain the delays, not the characteristics of the person such as their level of social deprivation, distance to care, insurance type, or the particular features of their cancer. Our hypothesis, therefore, is that treatment timeliness may be a function of how the local health system is built, rather than the particular people using it. Our team continues to work with these data in what we call multi-level modeling to measure these area-level characteristics and test this hypothesis.
"Whenever navigation has been tested in the cancer space, it has generally been beneficial to patients and particularly to patients at risk of poor outcomes. How to make it maximally efficient and to pay for it outside the research context are the challenges we're now facing."