Health Equity Actions: Understanding Data Disaggregation
Listen to “Health Equity Actions: Understanding Data Disaggragation” on Spreaker.
August 5, 2022Join us today as our guest host, Adriene Thornton, Children’s Minnesota Health Equity Manager, discusses the importance of understanding data disaggregation and the impact on health equity with guest, Dr. Nathan Chomilo, MD. Dr. Chomilo is Medical Director for the State of Minnesota’s Medicaid and MinnesotaCare programs, as well as a general pediatrician.
Transcript
Dr. Angela Kade Goepferd: This is Talking Pediatrics, a clinical podcast by Children’s Minnesota, home to the Kid Experts where the complex is our every day. Each week we bring you intriguing stories and relevant pediatric health care information as we partner with you in the care of your patients. Our guests, data, ideas and practical tips will surprise, challenge and perhaps change how you care for kids.
Welcome to Talking Pediatrics. I’m your host, Dr. Angela Kade Goepferd. On today’s episode, we have another equity action segment with Adriene Thornton. This episode today will look at the issue of data disaggregation. I think that many of us, when we see healthcare data segregated by African American, Asian American, and other ethnicities, assume that that data has been disaggregated. But on today’s episode, you’ll learn that truthfully, we are still lumping in several different ethnicities within those categories and why that’s problematic and what we need to do about it.
Adriene Thornton: Welcome to Equity Actions, a part of Talking Pediatrics Podcast at Children’s Minnesota. I am Adriene Thornton, your host, and the manager of health equity at Children’s Minnesota. I am very excited to welcome our guest today, Dr. Nathan Chomilo. Dr. Nathan T. Chomilo is a medical director for the state of Minnesota’s Medicaid and Minnesota Care programs, and he also practices as a general pediatrician with Park Nicollet Health Services Health Partners. He also served as the state of Minnesota’s COVID-19 vaccine equity director, and is currently a senior advisor on equity to the Minnesota commissioner of health. He is an executive committee member of the American Academy of Pediatrics section on minority health equity and inclusion, and serves on the boards of Reach Out and Read in the Minnesota chapter of the AAP.
He also is an adjunct assistant professor of pediatrics at the University of Minnesota Medical School, and is a co-founder of the organization Minnesota Doctors for Health Equity, for which I have been on that website many, many times. So I appreciate you helping to create that organization because it’s been very helpful in my journey as the manager of health equity.
So today, we’re going to be discussing data disaggregation. Most people, when we’re looking at health equity, health equity data, we’re looking at aggregated data, and there have been several studies published that talk about data disaggregation, why it’s important, what is the impact of it, and why we should be doing it. So we will be talking about that with Dr. Chomilo today. Dr. Chomilo, welcome.
Dr. Nathan Chomilo: Good morning, Adriene. Thanks for having me.
Adriene Thornton: Yeah, thank you for being here. Let’s get right into it. A lot of people don’t realize that when they’re looking at data, that it’s aggregated. For instance, when you’re looking at Black, African American, that includes anybody who identifies as Black or African American. It doesn’t account for anybody who may be African and West Indies, or someone who may be actually a native of one of the countries in Africa. When you’re looking at the Asian population, there are different ethnicities within the Asian population and they don’t all function, live, or have the same cultural values. So, can you talk about how we got to the point where we started lumping everybody into five or six different race and ethnicity categories and not really looking at those nuances?
Dr. Nathan Chomilo: It starts with an interrogation of what race is? Race is a social construct. It’s not a biological one. So it’s been defined really through a social, political, economic lens over the course of history, really used initially to justify institutions like slavery, land theft, to say that these people are different than us and therefore we can oppress them; we can benefit from their labor or from taking what they own. Then as we progressed and our country grew, we started setting up these categories to really structure access to resources, who has rights and who doesn’t have rights, and who is treated as a regular class citizen and a second class citizen. That’s been the legacy of a racial categorization in our country and in other countries around the world.
When you look at census definitions of who is white, for example, there’s periods of our time when folks with Italian or Irish background weren’t considered white. If you look at the most recent census categories, folks who are Egyptian or Lebanese are considered white, and there’s been a gradual shift to trying to capture more Hispanic folks as white folks. So really, when you look at race ethnicity data and disaggregation, that is, I think, some of the underlying issues we have with using it to define first, let alone start to address health inequities and health disparities that arise from those more structural inequities that we see
Adriene Thornton: Interesting, because I’ve been doing a lot of work with equitable distribution of the COVID-19 vaccine. Children’s Minnesota has very intentionally gone out into the community and offered vaccine in areas and zip codes that are more socially vulnerable to make sure that everyone has access. I attended a grand rounds and I also read a report that was published by the Coalition of Asian American Leaders. It was really interesting to me because even though I’m in this work, I’m looking at the data, I had not realized the nuances of the data related to the Asian American community in Minnesota because I knew that, oh, 4% of the deaths related to COVID in Minnesota are in the Asian community. Okay. Well, then we need to focus on making sure that we get education and resources to that community and make sure we offer vaccines to that community.
But which Asian community? Because just looking at the overall COVID data, it doesn’t tell you which Asian community needs all the resources. It doesn’t tell you which zip codes you should focus on. It doesn’t show you that 49% of those Asian deaths were in the Hmong community, which is very different from the Karen community, which is very different from the Karenni community. Seeing those nuances, even though it gives you a very small number in terms… When we’re looking at data, you always want big numbers. You want lots and lots and lots of data, and the data numbers that get smaller in terms of your end, but that doesn’t mean it’s not important. So can you talk about the importance of disaggregating data in terms of looking at health disparities and how we deploy resources and how we actually provide care in a culturally sensitive manner?
Dr. Nathan Chomilo: When you talk about disaggregation of data for health equity, race ethnicity is one component. But we could also talk about how we define areas as rural or what’s been termed rurality. We certainly would talk about definitions of sexual orientation and gender identity, who is living with a disability. Then pediatrics in particular, we talk about children and youth with special healthcare needs as well, and those kind of medically complex children and wanting to be able to disaggregate some of the data around how they interact with our systems. But if you go back up, talking about your point with CAAL, the Coalition of Asian American Leaders and the work they did, it clearly showed that, yeah, while you look across these five broad groups that Asian Pacific Islanders in Minnesota seem to be doing relatively well or better when it came to the impacts of COVID, so the number of cases, hospitalizations, and deaths seemed to be doing relatively well at getting access to the vaccine.
When you actually were able to use zip code data and look at even things like death certificate data to get at that more disaggregated picture, you saw that there was a real disproportionate impact happening in specific communities with COVID being the leading cause of death, as you noted in the Karen, Hmong, and Karenni communities in 2020. That type of work is intensive and really difficult because to get accurate data, you do need to have a good estimate; you need a good denominator to know who’s all in the community and are you capturing everyone. So that’s one of the big challenges that we face when we’re trying to do population health, and as you’re noticing it getting lower and lower. What is our data source and how do we know that we’re getting the right picture from the data, because it can be a really helpful signal?
That’s how I look at it in the context of how do we advance health equity, how do we provide culturally relevant and responsive care and solutions, is that data is really the initial signal. It shows us that there is a gap in either an outcome or access to a resource. Then from there, we should really start to ask more questions, and we should start to go to the communities that we are identifying as experiencing that gap or experiencing that inequity and ask them what’s the lived reality, and does this data match up with what they’re seeing? Then if so, what are some possible answers and solutions that we can go to from there? So I think it can be hard for folks who like to really just sit and crunch the numbers to take that next step and say, “All right, how do we go and have a community-led conversation about what these numbers mean and where to go from there?”
But I think when you look at COVID-19 as an example, that’s certainly what we’ve tried to do as far as really having community at the table in interaction with the department of health, through several different venues, whether it’s the COVID community coordinators, there’s [inaudible 00:09:57] community liaisons; there’s the diverse media vendors and community engagement contractors that have all been trying to really follow what we were seeing in that MDH data that relatively early in 2020 got released, showing the disparate impacts of the pandemic, with what are the solutions. So I think that’s the first piece, when you come to disaggregation, is using it as a signal, then, and not just stopping there, going to the communities and asking.
But I also think it’s real important at the beginning to think of what are your definitions. Are you really capturing everyone within a group? If you think, like you were mentioning about the category of Black, the category of even American Indian and all the different tribal nations that fall under that, Indigenous peoples that fall under that, Asia Pacific Islander, Hispanic, and even white. There’s a lot of different, as we talked about, that because of these shifting definitions, there’s a lot of different groups and communities that get captured in those. So understanding what your definitions really are. How are you asking this? How are you collecting the data? Is it asked when they first see a health provider? Is it asked as they enroll for their insurance? Is it asked as they fill out some other form and what completion rates?
So that’s something that we’ve looked at in Medicaid. When it got to Medicaid, initially, our race ethnicity data, we were missing it for about 27% of enrollees in Medicaid. We’re able to work and use some different ways to look at other public programs that folks have signed up for, and other ways that we collect data to decrease that gap so that it’s roughly around only 10% missing right now. But I think that’s the other piece, is if you’re looking at a population, and you only have the race ethnicity data, for example, for 50% of the population, and you’re going to come and say that, “Oh, this group’s experiencing a gap.” Well, you really don’t know because you don’t have insight into half of your community.
So I think understanding from the get how this information is gathered and what it’s based in, is really important. Then having a clear conversation with community about what those definitions are, how do they want to show up in the data? What do they want to be listed as? What would they feel comfortable marking on a sheet when they’re asked their race and ethnicity? And seeing where that can take you.
Adriene Thornton: Children’s Minnesota participates in a national pediatric collaborative that is looking at race, ethnicity, and language, how we collect it, and what we’re finding is there is an unknown category. Right now, there are about 30 hospitals that are initially participating in this cohort, and across the board, about 20% of the patients have unknown listed as their race or ethnicity. I’m like, “Well, you’re living and you’re breathing.” So that creates a little caveat because you’re having to guess. Are these unknowns a part of a specific race or ethnicity group? Are they a part of multiple? It really changes what the data looks like.
So, we’re working at Children’s Minnesota to try and capture as much as we can in our computer system. We have 82 different ethnicities listed thus far, and we’re not done. I look at that and I just think it’ll be wonderful to capture all of that nuance. But when I pull the data, it’s going to be interesting to see what it looks like. Then it’ll be interesting to see how many people will come to one of our facilities and not fit into one of those 82. Where do we put them? It definitely does speak to how do you focus where you put your resources if you don’t really understand who you’re serving?
Dr. Nathan Chomilo: The other piece of that is reconciling, again, with how race has been used historically in our country and in the medical system around the issue of trust and the violation of trust that medicine and other institutions have taken in our Black, Indigenous, and other communities. They’re asking you to identify your race. A lot of folks rightly will ask why? And do I want some institution that has wronged me or my family before to have that information, to potentially use it in a way that could discriminate or be biased? That’s something that we’ve seen, again, on the Medicaid enrollment side and heard from other states that they’re also dealing with this concern, particularly around with some of the federal actions around immigration in the last five, six years, that folks just aren’t real comfortable always volunteering that information, particularly if you don’t take the time to talk about what the information’s being used for.
We have seen some improvements when you normalize and say 85% of folks fill out this information when they’re filling out this application or that this information is used to ensure quality in our programs for everyone. So this is how we’re going to use the information. Then as you are touching on, having more specific categories so folks feel seen. As a biracial Black man, myself, I’ve always struggled with some of these forms, especially the ones where I have to pick one or the other. I can’t pick multiple, like the biracial category itself and how that’s used. So I do think that those are considerations as we are thinking about how we build systems that collect the data. Then what you’re getting to is then what do you do on the back end once you have that data, if you haven’t been intentional to be getting, you’re going to be stuck with what to do with data and how good is it at getting at what you’re trying to do.
Adriene Thornton: I was reading a report by the Robert Wood Johnson foundation. They were saying that the importance of data disaggregation is that you take health equity from focusing on healthcare systems and social determinants of health as being the catalyst for improving health equity, and you go to an individual level. It’s the person. What does this person need for us to improve their health? In my mind, I’m thinking, “Well, it really isn’t a huge difference between the two.” It’s a play on words. But I suppose if we collect 82 categories of ethnicities, we are truly getting down to the personal level at that point, which is wonderful. Do you think it’s more important against the person level or is it okay for us to focus at a higher systems level to see what we can do to improve health disparities?
Dr. Nathan Chomilo: At least when we’re talking about race, I say that race ethnicity aren’t really super helpful at determining decisions at the bedside. But they are good markers of the impacts of structural racism at the population level. As a policy maker, it really is hard to think of making a policy that gets down to the individual side. We really are aiming more to improve systems so that most folks from a group experience less barriers, experience better outcomes. As a provider, I find it helpful to know the cultural experiences and differences with my patients. But I wouldn’t ever want to assume just because I have a Somali patient or a Hmong patient or Vietnamese or Liberian or African American in front of me that I know exactly everything I need to know about them and their lived experience. So I think the data is helpful in us capturing these experiences and having a broad understanding. But if we’re talking about getting down to the person level, we always want to start with asking that person what their experience is and being open to it being different from what we see in the data per se.
Adriene Thornton: Well, and I just want to tell you that at Children’s Minnesota, we appreciate the work the health department has done to get the zip code level information. Because what we did for our community COVID clinics is we took the core town one zip codes that the department of health created. And we overlaid our patient population over it. We created a heat map that showed us where our patients lived, how many of those lived in the Q1 zip codes, and then it helped us determine where we should go based on the COVID-19 vaccine rates for the state, for our community clinic.
My final question for you, when we’re looking at data disaggregation, we know that it’s important. We know that it will help us funnel our resources. We know that it can have a positive impact on community health. Go through this a little bit, but I just want to circle back to it because a lot of us already have a lot of data that we’re working with, and even though we’re continuing to collect data, we are still using our historical data. So, in terms of impact to the community, does it matter if we disaggregate the data when we’re collecting it versus when we’re analyzing it versus when we’re reporting it? Is there any point that it’s more important to disaggregate the data or does it matter where you do it as long as you do it?
Dr. Nathan Chomilo: I think it’s most important to have a clear definition of what is the data that we’re trying to capture and that is reflective of the communities that we have talked to, that we know have experienced inequities. Again, race and ethnicity is one, but there’s multiple different communities that we can look at the data and disaggregate in a way that we can better see any gaps or barriers that they experience to care. So I think at the beginning is the most important part because that’s some of the struggle I’ve seen, is that if we try to work just with the data we have, we have real limits on, well, what are our denominators? How do we know? There’s some data sources that show that in certain age groups, we’re vaccinating over 100% percent of Asian Pacific Islanders in Minnesota. But there’s not extra people just floating in the air from that community. But that’s because the data source itself isn’t as accurate as we would like to get.
I think determining which data source to use really can start to be defined by what are your definitions you’re using and what communities you’re trying to capture.
Adriene Thornton: Thank you so much for joining our podcast today. That was a wonderful conversation. As a data geek, I love talking about data. So that was amazing. Thank you to everyone who joined our podcast today. This is Adriene Thornton signing off for Equity Actions, a part of our Talking Pediatrics Podcast.
Dr. Angela Kade Goepferd: Thank you for joining us for Talking Pediatrics. Come back each week for a new episode with our caregivers and experts in pediatric health. Our executive producer and showrunner is Ilze Vogel. Episodes are engineered, produced, and edited by Jake Beaver. Amy Juba is our marketing representative. For more information and additional episodes, visit us at childrensmn.org/talkingpediatrics, and to rate and review our show, please go to childrensmn.org/survey.