Expert Addresses Ways of Mitigating Racial Disparities in Oncology Practices

Chasity M. Washington, MPH, CHES

While clinical trials often collect data regarding racial and ethnic groups when evaluating treatments in different patient populations, these are not usually large enough groups to represent a whole racial or ethnic community in 1 study. In addition, many groups are often lumped together in the data, such as Cubans, Puerto Ricans, and Mexicans being grouped together as Hispanics. These are only a few of the challenges that can be addressed with oncologists and study investigators to mitigate the racial disparities that exist today in oncology practice.

As an oncologist, an important step to overcoming these disparities would be to better understand your community to see whether your hospital or clinic is serving these populations. At the Ohio State University Comprehensive Cancer Center—James, their team is trying to overcome some of these barriers by sending out a van into the community with healthcare workers to conduct cancer screening and tests. This allows the community to be tested without having to come into the healthcare system right away, and it encourages them to come in person for follow-ups.

Clinical trial investigators could also collaborate together; while many racial and ethnic groups are often represented in small numbers for a single clinical trial, collaboration with other clinical trials could improve representation of these groups. Hospital systems could provide biases tests and resources for overcoming potential biases for staff because often times biases present due to a lack of experience or knowledge.

In an interview with Targeted Oncology, Chasity M. Washington, MPH, CHES, director, Center for Cancer Health Equity, The Ohio State University Comprehensive Cancer Center–James, shared her understandings of the racial disparities in oncology and how investigators can help mitigate these challenges within their communities.

What racial disparities are commonly seen in the oncology world?

Most of the largest disparities are usually between African Americans and their white counterparts, but we're starting to see disparities in other racial and ethnic groups as well, especially if you do what we call a disaggregation of the data. A lot of times we lump some of these smaller groups together because the statistics and data folks need larger and significant numbers, so if you disaggregationthat data and start looking at subgroups of what we call Latin or Hispanics, there are certain disparities from different disease sites that are starting to be seen. In particular, I was at a conference where they shared that there's actually a disparity in breast cancer for Puerto Rican women, but again, if you roll that up, you lose that data.

In just cancer incidence and mortality overall, there's always a disparity. Again, the African American men and women are more often going to die from cancer than their white counterparts and other racial and ethnic groups. In particular with African American women, what we're seeing is now they're getting cancer just as often as the white females, but they're dying from it more often. We used to be able to say that for breast cancer, African American women were diagnosed less often and then were dying more often, so because of screening and outreach, I think that's why the incidence rate has finally caught up. Now they're getting screened, so we're finding more cancers, but there's still a disparity in the outcomes for these women.

Do we know why there is this disparity first of all, and then second, what could oncologists and other clinicians do to mitigate this disparity?

What I suggest is that there are things that you can do internally; you can look at your data, your institutions, policies, and procedures to make sure that those are friendly for those populations and make sure that you're serving those populations. I think a lot of places do a good job of looking at your service area and are the patients that are coming to our hospital or clinic or facility representative of those that are in our community. If not, that's the first place that you should start. Why are they not coming here because they're clearly getting cancer? Where are they going? Are they not going anywhere? You just basically should find out where they are in those communities and how you can outreach to them. A lot of things that I suggest are going outside of the walls of the cancer hospital or the cancer center or the clinic, and doing community outreach and education and engagement. I know a lot of nurses already do some of that with their clinical ladder and magnet status and things of that nature, but just be mindful about those initiatives and things that they do. Make sure that they have interpreter services and that they're using it properly, and they're not using family members. Simple things could be even doing an environmental scan of their facility. Would I feel welcome walking into your institution? When I see someone who looks like me when I see art on the walls that reflects my community? Would I see a magazine that I could relate to?

You should also consider an LGBTQ community. If they're going into the breast center, which we like to make everything pink and is very gender-specific and very feminine, how would that make them feel as a LGBTQ person coming into that facility? Some of those things can be done within the institution, and then I think a lot of it needs to be done outside. We need to know who we're serving, and we need to know what the resources for them are, we need to get them in for screening sooner, address the barriers to screening, get them back from their diagnostic testing into treatment sooner, etc. A lot of times there's a delay from diagnosis to initiation of treatment in these communities, so we need to make sure that we're addressing any barriers that may cause those situations.

Knowing who you're serving and knowing what those barriers are using the data that exists instead of just always using it to highlight the disparities in the issues, we should use it to address those instead. We've done a lot at our institution; my department is our community outreach and engagement arm of the hospital, so we get to go out into these communities and address those barriers and identify those barriers. We have community health workers that get them in from mobile mammography because they don't always want to come to the hospital, so we pull up to their community, and they can just do 15 minutes on the mobile van, and then if there's anything abnormal, we call them back, and there's someone from their community that speaks their language give them those results. This encourages them to come back and explains the importance of that.

How can clinicians identify these biases and approach them to ensure the best care?

There's a lot of implicit bias training that goes on. There is an implicit association test where you can take these assessments, and it will tell you if you have an implicit bias. There's 1 for gender, there's 1 for race, there's 1 for LGBTQ religion, politics, etc, so you can take those and see where you might have that bias. However, a lot of times people take these or they go through one-time class, and what I would say is that you need to then do things to mitigate that bias. If you realize you have a gender bias, then what do you do to deal with that?

I would encourage institutions, if you're going to have implicit bias training or encourage your physicians and staff to take these sessions, then you also follow up with sessions on different communities and populations and how to use interpreters and having even patient advisory boards or patient advisory councils that you engage and share that information, those sort of things so that they can have the skills to address that. Telling them that they have the bias and then not giving them the tools to do anything about it is probably more harmful than good. A lot of times, it's just simply, if you have a bias towards a group, then learning about the group, engaging in the group, and talking to people from the group, experiencing things like a festival or so to just submerge yourself in that culture because a lot of times it's just getting familiar with something.

A lot of times, biases are just fear or unknown or just things that, again, you're not exposed to, so the more experiences you have, the broader your horizons and the less likely you are to have some of these biases.

Why is it that these groups are lumped together in these larger groups in clinical trials? Is there anything that can be done in academia moving forward to address this challenge?

I think a lot of times it's because statisticians have to have a certain sample size or number for it to be what's called statistically significant, and sometimes, depending on the community and the population you're collecting in, then we can't truly, from a research standpoint, say anything about it but if you look at it individually, then add it collectively across the country in places where there are other larger minority populations, then I think that's when you can start to see these differences and parcel them out. I think we are moving towards that.

I've been to a few conferences where people are encouraging that. I was on a call earlier where a researcher had done that with the Asian population and started looking at specific Asian groups versus lumping everyone together, but we also have to start collecting that. Sometimes when we do data collection, we don't collect the subgroups, so if you're asking someone for their race, and then their ethnicity, you'll say Hispanic or non-Hispanic, and their race choices, we don't say Cuban, or Mexican, etc, so we don't give people the options. There needs to be a movement towards that. I think in some places there are movements towards that, like more than 1 race has been added, so we're starting to capture that for folks. I think a lot of times Africans are also lumped in with African Americans, so Somalis and other groups either are lumped in with African Americans or they're lumped together as Africans, so we then wouldn't be able to tell if there were disparities in those groups either.

I think it's important, and it's something that the data folks will need to help figure out, and then people who do research should think about how we can collectively collect that information better.