For Your Informatics: Episode 47 - Social Determinants of Health at AMIA 2023 Annual Symposium

- Welcome to another edition of For Your Informatics, a podcast where we explore the limitless world of medical informatics. Created and led by the women in AMIA, we offer insights into career paths, leadership, and education. Thanks for joining us as we highlight lives to inspire greatness, inclusion, and diversity in the field of informatics. - Hello and welcome to For Your Informatics, a podcast exploring the limitless world of medical informatics. My name is Dr. Leila Wursami, and I'll be your host for this exciting episode. Thank you so much for joining us. Today we have the pleasure of introducing and interviewing two guests. On my left, I have Dr. Joshua Vest, and on my right, I have Tiffany Harmon. Dr. Joshua Vest is a health services researcher with an interest in adoption and effective application of health information technology to organizational, clinical, and population health outcomes. Much of his work is focused on adoption, utilization, impact, and policy issues of technologies that facilitate the sharing of patient information between different organizations. As a former local public health practitioner, Dr. Vest has a particular interest in effective public health information systems and the use of social determinants of health data. He's a professor of health policy and management at the Indiana University, Richard M. Fairbank School of Public Health, and a scientist at the Regan Striff Institute Center of Biomedical Informatics. Welcome. Tiffany Harmon is a dedicated nurse informaticist who I have the pleasure of knowing and working on in the FYI Informatics podcast. She's been a part of 3M HIS since 2010. Presently, Tiffany serves as the global coding and clinical content manager and is also leading efforts in the social drivers of health program for 3M. Her passions include population health, clinical document improvement, computer assisted coding, and nurse terminologies. Tiffany has authored several articles and presented her findings at numerous society conferences. Tiffany is also a member for the FOIA Informatics podcast working group. Welcome, Tiffany. Hello. So I would like the guests to introduce themselves and tell us about their journey into informatics. Josh. Yeah, so howdy. My real interest and kind of jump into informatics goes back to about 2005. I was the local epidemiologist for Austin, Texas, and I was faced with a task of doing a community health needs assessment, particularly around asthma and pediatric asthma. And so we're out collecting data, trying to get all this information together. And if you know anything about pediatric asthma, a really important indicator is gonna be emergency department usage. And so we called up the hospitals, there were several of them in Austin, and said we'd like to know information about admissions for pediatric asthma, and they told us, no, we're not gonna tell you that because it's not reportable by law, so you don't get to have it. And I was complaining about that, stonewalling to a colleague, and she said, oh, you gotta go talk to these guys. They've got everybody's data. And I went and met this, what was termed at the time, the Energy Care Collaborative. It was one of the first operational health information exchanges in the nation. And I was captivated by this whole sharing of data and the power of using it back and forth for community purposes. And as a health service research student at the time, I was also really fascinated in the role of technology among hospitals, when hospitals, like they can really hate each other. And the ones that were here in Texas, they hated each other a lot, but they were managing to share data and work in that way. And I found that to be a fascinating combination, both from the informatics side, but also on the people and organizational side. I've been hooked ever since. - Wonderful. Tiffany, would you like to share your journey into informatics? - Sure, love to. In 2010, I was working as a graveyard nurse at a local hospital. And there was a position available for a data analyst at 3M. And I thought, well, gosh, what is this job about? I applied for the position. It was for a group of nurses to research and development project for a group of nurses to go through and look for a semantic shift and drift between data as it was dispersed out between numerous military treatment facilities. So when I very first started my job, I had no idea what semantic shift and drift meant. When we laughed a little, when we read the objective of the project overview, but soon we fell in love with informatics. And here I am almost 14 years later. My first conference I ever presented at was at AMIA in 2012 where I was an associate nurse with my presented my findings on this research and development project. So kind of grown from there. - Thank you for sharing your journey with us, Tiffany. So let's just jump right in. SDOH is a buzzword that has been in the healthcare field recently. And I would like to know, Josh, how do you define SDOH and how do you use it in your work? - Yeah, so I think there's been a lot of great new conceptual work in defining and about how we think about social determinants of health. And I kind of approach it the, I approach it kind of following the concept of levels. At the highest level, we have what we would often call the social determinants of health, like the actual term. These are the structures, the institutions, the policies, the environments in which people live that kind of either shape or constrain or enable their life. Below that, we have more pressing individual level characteristics, those are health related social needs. And I tend to do both in our work, mostly when I work with health systems and we talk about how we understand social determinants, how we measure them, how we respond to them and how we help patients around them. That's usually talking about the social needs component, that individual level focus. When we're moving towards the policy space and thinking about how we make broader systemic changes, it's kind of in that SDOH level. But we use both data, kind of using my public health hat and my informatics hat. - Wonderful. Tiffany, how do you define SDOH and how do you use it in your work? So I define SDOH on the patient level since my background is in nursing and social determinants of health really play a vital role in shaping the wellbeing and quality of life for individuals and communities. Social risk is clearly linked to outcomes and where a person lives does provide a lot of information about their health. And what I find fascinating is only 20% of outcomes are addressed in the healthcare setting and 80% of a person's health is really influenced by non-medical factors. So when 3M brought this project forth and said, "Hey, can you lead the social determinants of health "at 3M using our computer-assisted coding?" I jumped for joy. So right now we're working with different hospital systems that are utilizing our computer-assisted coding and we're looking for ancillary documents that contain SDOH data, narrative about your patients of what their health risk is. And so we're turning that data on within our software and auto-suggesting it for the coders and for the hospital initiatives. So Tiffany's told us why it's important and how she uses it. Josh, why do you feel SDOH is important and how do you use it in your work? Yeah, so I completely echo what Tiffany said, it's just dramatically important to health and well-being. We, and I guess not a clinician, don't even pretend to be one or play one. But when I go out and I talk to a lot of doctors and nurses and when I go out and talk and do either just visits to clinics or the hospital or do focus groups and we say something, a very basic question like tell us about the most complex patients you had this week or the most difficult patient. Everybody who's a clinician will smile and they've all got a story ready to go. And almost universally that story isn't some kind of like house-esque deduction of some rare condition. It is a, Mrs. Smith keeps coming back and we can't get her diabetes under control because she can't afford a medication. Or we have a kid continually returning for uncontrolled asthma. Or we keep making appointments for somebody and we just can't get them to the homeless shelter or the food pantry or to come back. It's always social, right? That's almost universally the underlying problem and challenge that folks have. And that's where I think is really kind of interesting where as informatics is spreading out and moving towards more, being driven by policy, being driven by the fact that so much of it is of health and well-being is driven by social factors and social conditions to including more of that data. I think it's really interesting and an opportunity for informatics. Wonderful. So let's talk about policy. Health data and now SDOH as what we're talking about today, is it regulated by the government and how do healthcare organizations utilize that data? Josh, do you wanna take a stab at it? Well, so I'll start with the first 'cause I think the most recent and very important change in the social determinants landscape is CMS's new rules about requiring screening as a quality metric for inpatient visits. That is a big shift, right? I think, as Tiffany said, people have known it's important and been measuring it and attempting to measure for quite some time, but until you actually have that policy or regulatory framework behind it, it gets really hard for health organizations to jump to it and really get behind it. Yeah, and right now in 2023, the SDOH one and two measures are voluntary. In 2024, they're gonna be required. So as a hospital, you are going to have to report all the patients that were admitted over the age of 18 that were screened for one of the five drivers of health and that also was positive for one of the five drivers of health. And those drivers are food insecurity, housing, transportation, utility, and interpersonal safety. So all of a sudden, the CMS one's the biggest, right? But recently, the joint commissions come out as well being in support of it and tying it to disparities measurements. And I believe HEDIS has also come up with quality metrics too. Yeah, that's right. Hospitals are being rated based on health disparities. The US News and Report ranks its annual best hospitals on the Health Equity Index, so that's available. And then NCQA made a HEDIS measure for the social needs screening and intervention. Yeah, so they're planning on using that. So the landscape has changed with this regulatory changes. So can we talk about what impacts it will have from an individual level to a population level? There is one other huge impact that actually was just mandated and came out in October of 2024. Three homelessness codes in the ICD-10Z Code range for SDOH. And they've moved those from a non-CC to a CC. So these homeless codes require more resources on the hospitals, you know, bigger healthcare teams, so higher reimbursement rate. Sorry, can we go back? No, we can go back, but first, for our listeners that do not know CC means what? Oh gosh, it's complication and comorbidity. I got that right? Yeah. I'm looking at you, you're nodding, okay. All right, so let's talk about the impact from an individual level to a population level. What are the outcomes that we are kind of striving for using this data? I think, you know, for me as a clinician, if hospitals can start to aggregate this data, right, you're gonna be able to see, you know, in a city, here's where our food deserts are, here's where our homelessness population is, because we're collecting all this data on inpatient and outpatient, and now we can start to identify where high-risk locations are and where we are gonna need more resources, more money allocated, those type of things. Josh? I would totally agree. I also think, in many ways I still think it comes back to just the triple aim or the quadruple aim as we know that these factors are associated with increased costs. They're associated with poor outcomes, and so these are opportunities for us to improve population health in that perspective. For healthcare organizations, there also is obviously a financial component, either, you know, improving the bottom line or potentially reducing exposure to penalties through like the readmission reduction program or other kind of quality metrics, and so I think we are talking about overall quality and, you know, improving cost outcomes as well. Is there anything else that you wanna add? I think we're gonna start to run into some big challenges as we start to aggregate this data, 'cause we're gonna be able to see it on inpatient and outpatient, but do we have the hospital resources? So now that we're collecting it, do we have enough social workers? Do we have enough case managers to manage these workloads? And then on the other side is once we start to identify that these patients are high risk and they need resources, do the states have the funding to allocate resources for these patients? And I think that's gonna be, I think that's another topic. - No, I think that's a critically important one because there's a couple pieces. One is both from the patient and the provider side. When these questions get asked or people get invited to share their social determinants or social needs, patients expect response, right? Like they don't say, patients consistently in surveys and focus groups will say, most actually are fairly accepting about being screened. That acceptance of screening is actually very high, but if you're doing it, you have to do something with the information. Nobody just wants to be kind of a answer bank or being gleaned for data. There has to be something. At the same time, the clinical staff is also on board the same thing of saying, if we're asking this and finding out that somebody doesn't gotta ride home, but we don't have anything to do about it, like that's terrible, maybe unethical. It makes us feel like we're not doing our job. So the resourcing is really important. And that's like, I'm excited by this area. I think it's interesting, but I'm also a little bit maybe cynical in some way. Because I'm a public health guy, right? Public health guy and public health always kind of has a little bit of an axe to grind about people don't do prevention well and things like that. The healthcare system doesn't screen well for depression. It doesn't screen well for tobacco use. It doesn't do a lot of routine screenings well that are preventative. And those have very clear medical clinical pathways. Social needs and social risks have a completely different pathway to addressing them. And they have different ideologies and they're very, very complicated. I'm not exactly sure unless people like CMS come in and say, "Thout shall do this." Most healthcare organizations were not really equipped to do this probably very well. And then I think you come with the other part is the resourcing is a really problem. The large accountable health communities demonstration project said, "It takes a lot to respond." Tiffany mentioned hiring social workers. Yeah, social workers, care managers, navigators. The caseloads are very, very high, mostly because the US does not have a very good social safety net or social system. And so therefore identifying these problems doesn't necessarily mean there's the resources to deal with them. 'Cause almost all the research says the percent of patients who have an unmet social need is going to be very, very high. It's going to be high. I'll give a couple of anecdotes. One of the health systems we work very closely with, I'm on their advisory boards. They came and said, "We're expecting 20% "to respond positive." And they were horrified by 20%. And I said, "You probably should double that." And I was even underestimating. I think they're running about 45%. And if you're in a safety net system, it's possibly more like 90% of people have at least one need. It's very high. So that brings the question. And I'll put you both on notice on here. I do want you to take a stance on it. Who do you think should brunt? Who should take the brunt of the cost of this? Providing these resources. Do you think it's a federal level? Do you think it's a state level? Do you think the healthcare organizations or payers need to do that? Tiffany. I think it's a mix between state level funding and payers. That's how I, I mean, that's, you would think, and I don't know if they're doing this because these new initiatives and mandates to now capture and report SDOH are fairly new within the last couple of years, but you would almost think like, as I'm putting out funding for domestic violence for homelessness, and this is what we have budgeted by state, I would take this data that the hospitals are gathering and say, oh, well, I don't think we've budgeted enough for domestic violence because we've got, there's 60% of people within a community that are experienced this or homelessness or whatever it may be. So I would hope that the states would use this data to say, this is how much money we need to allocate for advocates and for resources for these high-risk patients. - So I think there's a couple of different ways and I'm not trying to dodge your question. One is, I think health systems are going to be stuck in a make versus buy choice going forward, right? And I think the CMS is going to push them into that, having it measured in them saying we're gonna respond. And their health systems are gonna be forced to saying, are we either going to partner in the community to try to figure out how to do this, or are we gonna start taking those services in-house? And there are many organizations that have been trying to do it in-house, and there's been many organizations that have been very effective in creating partnerships. And then there's gonna be a set of organizations where neither one of them is a possibility, right? Like if you're in some lower resource to area, you just may not have the places to refer people to, right? You may just not have those community partners. It may also be at the same time, hard for you to hire and recruit those folks who would handle in-house. I kind of tend to believe, so again, I'm not trying to dodge your question, there's two things, who's ultimately gonna be the one? I think we're talking about societal and structural things. If you back up to transportation, housing, right? Availability of fresh foods, employment, that's very big state government, federal government, institutional level interventions. Now there are, again, there are innovative forward-thinking health systems that are trying to do some of this stuff on their own. They're either working on advocacy to make the changes, or they're doing things like building housing, which is really cool, I think it's great. Scalable, probably not, sustainable, questionable. I think on the other side, I would also add to your other question about also paying for it, putting on my health services researcher hat of to whom is accruing the benefits of these interventions? And probably broadly, it's society and the payers, right? And payers being very broad of all payers, right? If we're assuming that we're gonna eventually reduce utilization, we're going to improve care, health and well-being, that's who actually is accruing the benefits of these interventions. - And the more I think about it also, it may be a trifecta, right? It may be state, it may be the payer and also the hospital, because they also benefit resorting utilization on the hospital side. - Sure, yeah, well, and I think it's the places where you have the most interest so far in the health sector from this is the places where they are the payer, right? They have a large attributed health population or, so where they have or they're the predominant payer in an area where their community and their population have a high degree of overlap. So they're the ones who are bearing the risk. - Right, vertical integrated organizations are the ones who are gonna be the vanguards, you think, when it comes to this? - They're the ones who I think have the initial, the most pressure. They're also the ones who we've now had 20 years of health policy pushing towards integration. And so they're also the ones who are better equipped either inherently or by design or by policy to be equipped for more population health level interventions and management. - So any last words before I close out? Usually I ask for resources that if anybody wanted to get more information when it comes to SDOH or policy regarding SDOH, where would you point them to? - So I'll start with the couple things 'cause I think a lot of our conversations which have been great, it's been focused more kind of on the policy and the broader environmental stuff. But in the informatics space, we haven't even talked about things like, well, Tiffany did mention to start about coding, some of the things, but like how you use those codes, how you build them in referral systems, the whole thing about closed loop referrals and the gravity, so one thing I would point people is to like the gravity project, there's a lot of interest and informatics development around, and people like Loink have done a lot around coding and standards and how we share this information back and forth as well from the informatics side, but that's kind of like even the, we've been kind of on the, for lack of a better term, upstream broader space, but in the internal informatics space, there's that. - Yeah, and I think what I do see with some of our clients that are capturing this data is it's in multiple areas within the electronic health record. So some of our sites are like, we are only going to capture this information on a health related social needs screening form. So that's the only way we're capturing it in our system, but we have clients where it's being captured in the social worker notes, the case management notes, and then also the nurses are capturing this information when they are seeing their patients, think of a labor and delivery nurse, they come in, they wanna make sure that the mom has a safe place to go home, there's no domestic violence. So it's getting the information and understanding where it exists so that you don't have just data that we're not capturing or that's not available or lost data. - So in summary, I will try to summarize the points that were brought up. SUH has levels from structure of healthcare organizations all the way down to the individual level. SUH definitely impacts health outcomes, the landscape is being changed right now with the regulatory changes, but how are healthcare organizations going to react to it? How is it gonna impact budgets of states and federal resources that they have assigned to social determinants of health? And if you are collecting this data, is it actionable? Are you making it actionable and are you using informatics strategy in making it actionable? - And as a clinician, I will say that I do ask my patients their social determinants of health, but most of the time unless I send them off to the social worker, I have no idea how to help them. So definitely creating those resources at the point of care where I can actually give them a resource when I'm talking to them versus pointing them to someone else would be helpful. - To go back to what you were talking about, resources. There are resources that are out there to connect people to the different type of community resources that are available. So Feeding America has a website. AAFP has a website called the Everyone Project. It's a neighborhood navigator. So in your neighborhood navigator, all you need to put in is your zip code. And then from there, you can find all the different resources that are available within your zip code. - And if I can plug two more things, I'll plug two more. One, Tiffany mentioned about all the different documentation. I think it was very recently. Jamie, I had a very nice article on some very foundational, good, solid informatics work on how documentation of SDOH exists in the EHR. And the other resource that's out there, the SIRON Network out of University of California, San Francisco has done a phenomenal job of putting together tools, information, a lot of good resources to really think about social determinants and social needs. - Okay, before we go, any other last words? This is Dr. Layla Rosami with Dr. Josh. - Best. - And Tiffany Harmon. And we would like to thank them for joining us at Amia Symposium in New Orleans at our live episode. Thank you. - Thank you. - Oh, thank you for having us. (upbeat music) - Thank you for joining us for this edition of For Your Informatics, a podcast where we explore the limitless world of medical informatics. Follow us on Twitter, Instagram, and LinkedIn at FYInformatics and never miss an episode. We would love to hear from you. Let us know what you think about the show, ideas for future topics or guests, and other suggestions. Until next time.