Good morning, everyone. Welcome to our presentation, Implementation and Capture of Occupational Data for Health, or ODH Elements, from Federally Qualified Health Centers, or FQHCs. Well, hi, everybody. Really happy to be here with you guys today, and especially happy to be here with these two wonderful guest speakers. I've been working with these two people for the past almost two years now, mostly on Zoom. So, actually, this is the first time I've actually met Catherine. I met Ray at a previous conference, but really thrilled to be able to tell you about this Occupational Data for Health, what we're doing with it, how we've implemented it into federally qualified health care centers. And I knew I liked these people when I first met them, because Catherine is in charge of research at Health Choice Network, family medicine background. And the first thing she said is how important work is to health. So I was like, yes, I was with my people, with somebody who understands. And Ray is informatics. And so, as you know, the way to the future is informatics, is to basically get this information into the medical record. And so Ray knows how to do this and has experience, and it's been really just exciting working with them. So we're excited to tell you guys about what we've done. So without further ado, we have nothing to disclose. What we're going to do, I'm going to tell you a little bit of the overview of Occupational Data for Health and how we got here. Ray is going to tell you more about USCDI and how data standards, terminology, exemplars, how we get the ODH into the record. And then Catherine is going to tell you more about the real world experience in a federally qualified health care center. And hopefully sometime at the end for discussion, because what we want to do in this talk is basically get you guys as excited as we are about this and implement it in your own centers to advocate for it. I love our specialty of occupational medicine, but we're a very small portion of all of medical care. So the gist of Occupational Data for Health is that by getting job and industry into the primary medical record, we can really leverage what we do. We can see where people are, help the individual patient, help populations, help public health. So I'm going to tell you, again, a history of ODH, the uses that NIOSH envisioned in patient care, population health, public health, a touch on USCDI, and then the overview of the pilot projects that I'm on. And before we go any further, how many of you have already heard of Occupational Data for Health? Okay, great. So, yeah, basically what it was was NIOSH's idea that how we would get standardized interoperable job information into the primary record. You know, a lot of times it's maybe in free text or different areas within the record, but never in a standardized interoperable way. So it's really, again, for the primary record. As occupational medicine physicians, we use the primary record, but it's this Occupational Data for Health is meant for that record, not just for our records. I always love this quote, health starts where we live, learn, work, and play. Robert Wood Johnson Foundation, Social Determinants of Health. And as you all know, work is one of the hugest social determinants of health. I always liked family medicine. I started in family medicine. I like seeing my patients. But what frustrated me is that you only have that short little time with them, and then they go into these environments where you really didn't have much say. So that's where I really was drawn to occupational medicine, is that through the work environment, we can affect their health. So Occupational Data for Health is a way that we can allow others to do this. Work is one of the biggest social determinants of health, but it's very multifaceted. As you know, there are very many good things with work, you know, access to income, insurance, a social network. But there's also risks, you know, both mental, psychological, physical, and work, and these risks are not distributed evenly across race, gender, or ethnicity. So certain groups are more prone to pandemics, recessions, climate change. How do we reach these groups? Through work. As I've gone on in this project, it comes up often, why don't we just use income as a social determinant of health? Certainly, that's an important one. You know, what you make, whether you have a job or not, is huge. But I thought this was interesting. This was just from a simple Google search Indeed.com, 25 jobs that pay $50,000 a year without a degree. So you're already taking education out of there, it's just income. But look at how different they are. I don't need to tell this group that the health risks of a firefighter EMT is very different than those of a hotel manager. So again, you can't just use income, you need to know where they work. We know that EHR can provide tremendous benefit. We can, public health, we can identify risks, track patterns of disease, track trends of illness and injury. EHRs provide opportunities for intervention on clinical care, sentinel cases, co-worker protection. Really, again, it's the wave of the future. I know we all have mixed feelings about EHR, but it is here to stay. And again, I think we need to embrace the informatics potential. However, again, this often comes up where all this EHR information is taken away from patient care. And I have also heard, you know, as I talk about this project, well, isn't that just more information that's going to burden the clinician to have to collect? Isn't it going to slow the patient down, slow the record down? Well, the great thing about occupational data for health is that NIOS researched this for a very long time and really made it very patient-centric. They were aware of this problem from the start and really made it so that the patient enters the data. So they developed this vocabulary so that people would understand it and could put their own data in, and that is, again, for the betterment of their own care. So we don't want that. We do want this. We want work to be in there with all the great things that we're going to have in the future for electronic health records. I always loved in sci-fi movies how they have these little transponders and they go around and scan people and just all their health information comes up with the sci-fi doctors of the future. Well, we're going to get there someday. We're going to be able to say, you know, okay, this person, what are their determinants? What's their past history? Why not? Where do they work? And certain groups of workers have certain risks. So again, the computer will be able to tell us that. So with this in mind, NIOSH worked with the Institute of Medicine back in 2011. The Institute of Medicine is now the National Academies of Science, Engineering, and Medicine. And they came up with this letter, Benefits of Incorporating Occupational Information into the Electronic Health Record. Improve quality, safety, and efficiency, reduce health disparities, engage patients and families, improve care coordination, and improve population and public health. And there were 10 recommendations to advance the likelihood of incorporating occupational data for health into the general EMR while protecting privacy. So again, that last part is very key, while protecting privacy. This comes up often too, you know, where are you getting this information? Is it protected? Yes. It's just like any other HPI. It's something the patient supplies and that we tell them, we guarantee that it is going to be kept with all their other sensitive medical information. So the goals in developing occupational data for health were threefold, patient care, population health, and public health. So I'll go through them one by one. And first, clinical care. And this is what really excited me from the start, you know, it's basically that person sitting in front of you. If you knew their work information, how would that help you? And so these three things here, asthma, diabetes, and the return to work after a low back pain, have been worked on and researched and developed into clinical decision support tools. These articles are on the ACOM website for those of you that might be interested to look them up. But in any case, the idea is that once we have the work information, these decision support tools will help the clinician with that patient sitting in front of them. So for example, asthma, what about these guys that come in or girls that come in with respiratory complaints time after time to urgent care or express care? No one's asking them about work. The record, again, in the future will trigger. Where do these people work? Why are they continuing to come to their urgent care? As you all know, a lot of the working population sometimes doesn't have a primary care doc, they go to express cares or urgent cares for their care. Diabetes, healthcare workers are a good example. You know, we concentrate a lot on the risk of healthcare workers, but what if they're also diabetic and working long shifts? Will their doctor be triggered to ask, you know, hey, what's your shift work like? Is that what's causing your hemoglobin A1C to go up? Return to work after low back pain. I know this is always a hard thing for non-occupational medicine docs. They tend to just put people off. But as we all know, that's not the way to get better from a low back pain injury. You need to keep moving. So again, if we knew more about their work, then we could guide the activity prescription moving forward. So really exciting opportunities for that patient sitting in front of you. Other ones, Lyme disease in agricultural workers, maybe consider that. Truck drivers' medication use. Many of you might know Epic already does have a module that they worked on with the National Transportation Safety Board, whereby if a person is a CDL driver or a pilot, it can be put into the record. And then, say, their family doctor tries to order a sedating medicine, that can pop up. So that's the gist of getting this information in the record moving forward, is things like that. The need for accommodations during pregnancy. Assist patients' successful return to work after illness or injury. As I've gone in this project, I've asked my colleagues in other specialties, too, do you ask about work? What do you put into your records? And one of my orthopedic colleagues pointed out, he said, yeah, I try to ask. A lot of times my partners don't, or I get these consults where I have no idea what they do. But he said, if I'm going to do a shoulder surgery, then I want to know, are they going back to a jackhammer or a desk? So things like that, we're going to need to know. We do need to know. Oftentimes, stuff like what my friend brought up, he has to dig for or spend extra time asking the patient. If it's in the record already, that's going to improve care from the start. Patient populations. The first three examples were from a pilot where they collected occupational information in free text, and then they back-coded it into industry and occupation so that it made sense to the EHR. They then worked with public health to see more about what these patients, where they lived, who they were. And what they found was that a lot of their population was Portuguese-speaking, and that a lot of the women were housekeepers. A lot of the men were house painters. So what they did is tailor the information for prevention in their native language. They intervened on a culturally appropriate level to these people with that information they found. So this pilot was great, but unfortunately, coding on the backend is not scalable or sustainable. Other population health possibilities are to provide public-facing workers with vaccinations for COVID, conduct appropriate screening for silicosis. That's a big initiative right now. I know WOMA is working on that. I think it also went to the House of Delegates. And you're going to hear more about this one, identifying common ICD codes within populations to target interventions. So again, that's a huge possibility if we have the information. In public health, we can provide industry and occupation to cancer registries, provide current industry occupation employer address with electronic case reports. We can identify patients with reportable conditions like pesticide poisoning. So many, many possibilities for this. With that, we have advocated for occupational information to be in U.S. CDI. And how many know what you see U.S. CDI is? So basically, what it is is the United States Core Data for Interoperability. And I'm going to very crudely explain it here. Ray's going to go into much more detail, but basically moving forward, it's what we're going to have in all records throughout the country. It sort of replaces meaningful use. So basically, there's data elements, data classes, and groups advocate, okay, this is what we need for data elements. This is what should be in the class. People weigh in on that. And then it goes through a process whereby it gets adopted into U.S. Core Data for Interoperability. So that's, again, formatted in a way that if you're a nurse in Philadelphia and you go to San Diego, the record is going to have the same language. It's going to be able to pick that up. It's not going to be siloed. It's going to be interoperable. And this is a big win for us. Education and industry are in final version three of U.S. CDI, so yes, ACOM, and NAC, and we all advocated for this to happen, and it has happened. So the government has recognized how important this is. So there's a process with this, though. We're still not there all the way, because there's the standards versions advancement process. So when we're in the final version, we need to show that it works. How is it working? You know, what do people need to do? So it goes through the standard version advancement process in the next step, and that's also where we are right now. And that's where our pilot comes in. So as I told you, I've had the opportunity to work on this pilot with Ray and Catherine. Several different sites throughout the country, National Association of Community Health Centers has implemented pilot testing of this. Basically, what they do is respect what the clinic is already doing. They have a human-centered design framework from the start. You know, when I first started on the project, I said, okay, what are we going to do? How are we going to lay it out? They said, we need to go to the individual clinics and see what they're already doing, see what's feasible for them as they ask occupation and industry information, what questions they have, how comfortable they feel with it, whether it impacts their clinic flow, whether it impacts their EHR. So we're working at three different sites to do that. And the goal is that by connecting the implementation of occupational data for health with the large-scale data for COVID-19 public health surveillance and expertise, this project can identify essential workers. So we can identify the essential workers to protect them, keep our economy running. We can then scale and spread what works. So this pilot is moving along in that direction. The main questions we ask are the occupation, the industry, the employment status, and their employer name and address. One of our star clinics has been at Health Choice Network. Catherine is going to tell you more about that. And Ray is first going to tell you how we get that information in the record. Thanks so much, Catherine. And very nice to meet everyone. My name is Raymond Uy and call me Ray. I'm the current physician informaticist with the National Association of Community Health Centers. My background is in standards and clinical informatics from the National Institutes of Health, specifically the National Library of Medicine. So policy to practice, I saw two, three folks raise their hands when Catherine asked who is aware of what's USCDI or U.S. Coordinator for Interoperability is. I can see why you guys don't know it, but we're here on this side working on it and I want you guys to know what it is, why it matters, and how does it apply to your everyday work. So U.S. Core, as we call it, USCDI, it's urgently called Meaningful Use or the Common Clinical Dataset. It started in 2010. 2010 was when they started publishing the Common Clinical Dataset and they published it in 2011. It's currently supported now by what's called the 21st Century Cures Act, which requires specialty content to be certified. So electronic health record systems, EPICS or whatever you're using in your clinics or in your institutes have to be certified under this 21st Century Cures Act or the 2015 edition of this. So USCDI creates this framework for annually updating different data elements and data classes. What do we mean by data elements and data classes? Think about it this way. So folks who hate their EHRs, who are not really tech savvy, computers speak their own language. They don't understand English. They don't understand your narrative or your soap note. They understand structured information and that's how things are recorded in the database, in the relational database. So again, paragraphs and narratives, your soap notes. You have to use what's called natural language processing to actually extract the concepts you put there. But there are not a lot of people and it's not automatic. It doesn't work immediately. You have to have someone who's doing that and specializing in that to actually extract it. So when you're recording work information in a narrative, in a paragraph, just like what Catherine said earlier, someone had to go back into that, find the industry and occupation there and recode it. And that's like a lot of overhead. There's a lot of time involved. It doesn't make sense, especially now in 2023. You have AI everywhere, AI this, AI that. So we need to speak the same language. We need to understand the language computers speak. And computers speak ones and zeros and codes. So that's what this is all about. So who cares about ICD? We do. Who cares about SNOMED, LOINC codes, those things? This is what USCDI is all about. So the core principles are, again, interoperability. Why is this so hard? It's so hard because there's not a lot of money to be made from interoperability, let's be honest. So we're trying to approach this from a policy perspective, like let's make it a requirement. As a business, why do I want my data to be interoperable with Epic and Cerner? I want to lock them in in my system so that they'll remain in my health system and I can make money from them. So I'm just really speaking out there. This is the core idea, the core principle that all health information, if you're supporting public health and population health, they need to be interoperable across the United States. And we're not doing a good job with this. So that's what USCDI is all about, creating a core set of data needed to support and facilitate proper care. Think about it this way, OBGYN, you're pregnant, you go to an OB clinic, what are the things that they need? We all went through Bates' guide to physical examination and history taking. The first data elements you think about is how far along are you? What's your age at gestation? What's your last menstrual period, blah, blah, blah. In the same way, occupational medicine, the first thing I can think of, I'm not an O.M. practitioner, what is your work? And what industry are you working on? You spend eight hours a day, every weekday, doing this work. The other eight hours is sleeping. What's the other eight hours? It's up to you. But that's why this is so important and that's why we're doing this work. So, again, USCDI matters because it's included as a standard in the final rule in 2020. Standards for other people, that's kind of boring, right? But remember, the iPhone in the EU is now USB type C. It doesn't use the lightning connector anymore, not like here. So in the same way of you buying a converter for your plug or finding the right thing to charge your phone, we're applying the same thing to health and medical information, right? Using the same standard set of codes that machines and EHRs, whatever your software vendor is, using the same exact one code to represent occupation and occupation industry, for example, and employment status. So USCDI version one replaces the CCD and it has transitions of care documents, other clinical information, right? And the important thing here to see is the last bullet on the third bullet is access to data via APIs. What's an API? It's an application programming interface. What it does, it helps facilitate or translate your software with other software, right? If you're not using the same language, you won't be able to understand each other. Just like if I speak Japanese right now, I know maybe one or two people understand me, but not everyone. There. So why it matters, this is kind of a blown up picture of this. In 2020, I believe, U.S. CDI Version 1 is now standard, which means if you are an HR vendor and you are certified under the 2015 or the 21st Century Cures Act, all of the data elements and data classes in U.S. Core Data for Interability Version 1, you are required to report and adhere to those data elements and the standard codes that represent those concepts. Those codes, let's see, ICD-10 code for X, Y, and Z, or Z code, for example, for Occupational Exposure. I wonder how many folks here code for Z codes? Ah-ha, thank God. We need more folks to do that. So that's what Z codes are for, to represent those concepts so we can do research on it. U.S. CDI Version 2, I believe, is under the 2022 SVAP process. What's important to note is starting December 31st, 2022, that all HR vendors across the United States are now required to adhere to what's called the FHIR standard. FHIR, what's that? F-H-I-R, as we call FHIR, Fast Healthcare Interoperability Resources. It makes health records or health information for patients human-readable. You don't have to know programming or C or C-Objective or Python to understand FHIR. You can actually look at the code and see patient name, patient diagnosis, lab notes, blah, blah, blah. That's what FHIR is all about. So starting, again, just a few months ago, all of them are required to open their FHIR API. Maybe I'll skip through this. You can just check this out online. Basically every year, this is open for public comment. So if you are in a specialized field and you want your data element or specific information, such as maybe an angle of something, you can submit it. It is open to public and your voice will be heard and that will be put into consideration so that that information can be interoperable across the United States, across EHR systems. So there. Ah, interestingly, this is U.S. version 1 on the left. You see it's just a little bit, you know, just to give software vendors, you know, a little bit of time to address all of these, which they should be supporting already. You'll see U.S. CDI version 2 on the right, and I've highlighted a couple of things here that's related to what we've done with NAC and our folks and our partners, is social drivers of health assessment goals and interventions as a machine code so that your social workers and your community-based organizations can actually do something about the Z code and the diagnosis that you saw. For example, they're homeless. What do you do about it? If we've coded homelessness as a code in your EHR, then other systems can identify that same code and use a code for what intervention they actually did for that homelessness or food insecurity. Food insecurity, SNAP, food stamps, or homelessness, housing support, occupational medicine, occupation problem, or unemployment, employment recruitment places. So those things. And part of that is SOGI also, Sexual Origination and Gender Identity. Pregnancy status, and again, what Michelle showed, occupation and occupation industry. Thanks to our partners with NIOSH and CDC for getting this through. And we are commenting here as well, NAC and our partners are supporting inclusion of occupation and occupation. This is a big win. There. There you can see it on the bottom right, occupation, occupation, industry, with the exact definition of that concept of occupation and industry, and the actual standards that use them. For example, for occupation uses, what's called SOC codes, for example, and occupation industry using NAICS, N-A-I-C-S codes, or S-I-C codes. So beyond this, there's a lot of players, but there are data sets that are needed beyond U.S. CDI, and that's U.S. CDI Plus. So if there's an immediate need for your organization to support something that's not part of U.S. CDI, then there's the program by the Office of the National Coordinator to help push that through and help recommend and set the standards for the vendors on how to represent those information. So I'll skip through that. You can check that yourself. So again, occupational data for health. There's the guide for health IT developers, probably not the right audience for this, but talk to your clinical informaticist. Talk to your data nerds. Talk to your informatics folks about this, because all of the codes and all the things that are recommended by CDC NIOSH are in that guide. So it's freely available online, a quick Google search, you will get to that. There's also HL7. What in the world is HL7? So HL7 is the standard for transmitting health information across different systems in a secure way. All right? That's what HL7 is. And HL7 represents also the FHIR that I talked to you about. And Work and Health has already, there's a functional profile that has been published on how to implement and consolidate clinical document architecture templates of clinical notes containing occupation and work information. So here are what's called value sets. Value sets are just codes that contains the codes that represent the concept. So you're seeing on the top here, top right, representing industry, past or present industry, useful industry. It's published in a website called FinVADS. And you see those codes there, the 2.16, the blah, blah, blah. That's a very unique code specific to that set of codes containing all of the NAICS codes available across the United States that represent industry. In the same way, there is a value set for occupation. You see it down here representing occupation. There's FinVADS occupation CDC census 2010. That's the OID or the OID or the object ID. And that's specific to that. So that when you're coding, you can always refer to that set of codes, basically. Last but not the least, you'll see LOINC, L-O-I-N-C. LOINC code is another standard. It's free. It's made by Regenstrief Institute in Indianapolis. And those are the exact one, that's the one code that should represent past or present and useful industry in your EHR. So when you're capturing this data element, ask your EHR vendor, are you representing useful industry or past or present industry using that LOINC code? And you guys know what LOINC code is. It's for ordering labs, right? But not just labs. It's for establishing and recording observations. LO in the LOINC is logical observations. So I just wanted to add that. So employment status is also here. There is one in one single code that represents employment status on LOINC, that's 74165-2. And there is also a value set for different types of employment status, fully employed, part-time, not employed. That's all there. So what we want is for EHR vendors to use those exact same codes that's also being used internationally. This is not limited to the states. All across the United States, Japan, Korea, Philippines, UK, are using these exact same codes. So what we want, not just here in the United States, if you go travel, go on vacation, you get sick there, you get injured, they can get that information using that specific code because that's the language that we are using in the EHR field. So aha, here it is. So on the top left, employment status with exact specific code that's supposed to be used without spaces, of course. On the bottom left would be work classification. The middle bottom here would be, sorry it's small, but check out the slides if you have the chance. And on the very right would be the occupation SOC reference terms and codes that represent those occupations. So again, we don't want work and occupation to be recorded as a narrative. We want them to be recorded at the back end using these codes. So ask your data nerd, ask your informaticist if they know about this because they have to, especially if they're supporting your EHR in your practice. So what did we want to do with getting industry and occupation codes as funded by CDC? We can use industry and occupation code information to identify essential workers. So the CISA, or the Cybersecurity Infrastructure Security Agency, published a mapping of different SOC and NAICS codes and which ones are deemed to be essential or an essential worker. And not just that, this was also used for vaccine phases. You remember that year when there were, these are the priority folks who need to get the vaccine first? They have a mapping for that and it uses SOC and NAICS codes, standardized codes to represent industry and occupation. So again, if you type the work in industry in your soap note, we won't be able to do any of that work. It goes nowhere unless you do a research. So without further ado, Dr. Chung Bridges will show you how they used ODH and what they were able to do with it. Thank you so much. Good morning, everybody. What a pleasure it is to be here speaking with you this morning. It's been quite a journey with these colleagues of mine implementing the collection of occupation and industry in the electronic health record. We have a specific use case that we were a part of. And so my presentation, I want to talk to you a little bit about the why, although we've already talked quite a bit about why it's important to collect occupational data. But I still want to give you from my perspective, first of all, let me back up and say, so I'm the director of research at Health Choice Network. Health Choice Network is a health center controlled network that works with federally qualified health centers. So I'll tell you a little bit more about HCN, Health Choice Network. In my role as director of research, I'm charged with helping our federally qualified health centers to conduct research that's meaningful to their patient populations, that moves the needle on health disparities, and that has a health equity lens. We know that research studies generally do not include enough diverse populations. So FQHCs or federally qualified health centers, it's very critical from a health equity standpoint for those health centers to be involved in research. And of course, all that we're talking about in collecting occupational and industry data allows for research to be done. Because when we have that unified coding for occupation, even internationally it sounds like we have it, imagine the types of research studies that we can do on that variable of occupation that's unified among health centers. So occupation, I'm speaking to the choir when I tell you occupation is a key social determinant or driver of health. There's a Weber classification that I really like that classifies socioeconomic status based on three categories of class, status, and power. So you can imagine a school teacher, for example, who might have a lower income, which class corresponds to the income, but their status in society is pretty good. We all esteem our teachers. And then power can differ depending on whether a teacher is a part of a union, if they're organized. So there are three, just want to draw your attention to the fact that socioeconomic status itself is very complex, but occupation relates to each of these components in a different way. It determines income, the neighborhood one lives in, access to insurance status and quality of their health insurance. We've already talked about physical and psychological exposures to environment over time. And it impacts the health of the entire family. So there's a generational component to occupation. Conceptual model, it's everything that you think of in the way of how occupation might impact on somebody's health. We've already talked about health insurance, the environment they live in, whether or not they're able to take time off to take care of their health issues. So all the different ways that work impacts on health and by extension on the existence of health disparities. You'll see, as I talk about the use case in the particular health center, that the population was largely Latinx in the health center where we implemented this occupational data for health model. So wanted to draw your attention to why work is a very critical and important health determinant, especially for our Latinx populations. So for example, we know that while unemployment skyrocketed during the pandemic, that our Latinx workers were affected to a greater extent than other workers. We know that Latinx workers are underrepresented in managerial and professional jobs relative to other populations. We know that Latinx men or Latino men are overrepresented in construction, maintenance occupations, but again, underrepresented in managerial and professional roles. And Latino women are more likely to be employed in service occupations than other women workers in the United States. You know, thinking about some of the mechanisms through which work can impact on our health, I thought that this was interesting, Latinx workers tend to have less advanced notice of their work schedules. So you can imagine that that lack of control over your work schedule could have an impact on your health. And so it shows that this particular population is less likely to know from day to day what their work schedule is going to be like. And we can imagine how that many different ways that would impact on somebody's health. This one, of course, just brings it, you know, really to our visualization to say that our Latin workers have the highest rates of occupational fatalities in the United States. So that says a lot about the jobs that our Latinx workers have. And are overrepresented in some of the occupations with the highest rates of workplace illnesses and injuries. So you can, at your leisure, take a look at some of these slides and some of the statistics, but really just wanted to set the stage as to why it's so important for us to collect this data, especially to focus on certain populations that are impacted by health disparities. So getting into our federally qualified health centers. Who are they? Who do they serve? Federally qualified health centers serve one in seven racial ethnic minorities in the U.S., one in five Medicaid beneficiaries, one in five uninsured, and one in three living in poverty. So clearly, this is a very underserved population that's being served by our federally qualified health centers. Federally qualified health center patients tend to suffer from chronic conditions at higher rates than the general population. Hypertension, high cholesterol, asthma, diabetes, poor health. So you can see that this is a very challenged population that our health centers, our federally qualified health centers are serving. Who is Health Choice Network? I quickly mentioned that we are a health center controlled network, meaning that we work for the federally qualified health centers that are part of our network. We have 60 safety net organizations in 20 states and territories of the United States, and nearly 3 million patients. What's neat about our network is that we are responsible for the electronic record for a proportion of the federally qualified health centers that are a part of our network. And what that means is that we have access to the data of those health centers. And then we are also able to implement research projects, whether secondary data analysis, but also interventional studies in each of those health centers. So the focus of this use case of this pilot was La Casa Family Health Center in Portales, New Mexico. Just to let you know about their population, in 2021, 73% of their population identified as a member of a racial or ethnic minority group. They were 69% Hispanic or Latinx, 27% non-Hispanic white, 3.5% black or African-American, and 23% preferring a language other than English. 65% had incomes below 100% of the FPL, federal poverty level, and 91% had incomes less than two times the federal poverty level, 17% uninsured, and 51% Medicaid and CHIP beneficiary. So this goes along with what I talked about with who FQHCs serve. This is a very underserved population. Insurance is an issue, quality of insurance is an issue, income, all that we would expect. So we were excited to have this opportunity by NAC and CDC, gave us the funding that we were able to implement this project and implement the collection of occupational data for health at La Casa Family Health Center. We recruited La Casa. Our chief medical director has been amazing, Dr. Gidel Tom. He's worked with us on this project. And again, just amazing that this data has not been collected in a structured way up to this point. And the fact that we were able to implement this, and I'll show you one of the data analyses that we did to show a pattern of health outcome among a particular occupation or industry group. So the data is very valuable, and it's incredible that we're not collecting it to this point. So what did we do? So we reviewed previous methods of capturing occupation-related data, which were minimal. There might have been a question on employment status. Are you employed? Are you unemployed? We developed a plan for EHR optimization and implementation. So the optimization was we have a team who works on our electronic health record, and they were able to incorporate the occupational categories and industries into the EHR. So we worked internally in HCN on the optimization. We trained the La Casa staff on the new workflow. We went live. We reviewed and analyzed the data, did feedback and improvement, always having conversations with the patients. Patients often came and asked staff at La Casa, why do you want to know my occupation? So there was a lot of education that had to happen to help patients to understand the importance of that piece of data and why it was being collected. Data analysis and review was always done with the La Casa staff to make sure that everything that we were doing was in line with their normal workflows, that it wasn't going to be disruptive. FQHCs do a lot of wonderful work with very little as far as resources of time, staff, finances. So we don't want to overtax them or do anything that's going to cause them to be unable to meet the other challenges that they need to meet. So we were very conscious of that in working with La Casa. And then, of course, just figuring out once we looked at the data and saw the patterns, what could the health center do to address what was seen? So the data elements that were established were, are you currently working? What is your employment company name or industry? And what is your job? And as mentioned, we did some of this gathering of information, worked with stakeholders. A train-the-trainer approach was used, where we at HCN trained the staff of La Casa in collecting the data, and then the La Casa staff trained their staff on how to collect the data. So there were several challenges that we confronted as we implemented the project. There were limitations on customization on the number of categories that we could include. Because when you get deep down into occupation, I mean, there are, is it thousands of categories? You could really have thousands of categories. But from a logistical feasibility standpoint in EHR, to have somebody searching a list of a thousand categories can be challenging. The limitation in the EHR system that we had previously was that it was not a searchable list. So we had to give the health center a separate Excel data file that they searched for those categories because it was not possible to do it within the EHR. Now they've switched to a different EHR where the list is searchable. So that's a very important component. It has to be a searchable list where they're able to put in an industry and occupation and a list comes up and they can choose from those. We had to add not in the labor force to distinguish from unemployed as a category. And I mentioned the fact that we created this Excel data file so that there was a searchable list for the health center. So just the different roles and the fact that there had to be constant education about why this was important. The patients had to be educated about why it was important. The staff had to be educated about why it was important and even the providers to some extent. I mean, providers know, they collect this information every day when they're seeing the patient. They ask questions about occupation but to have it as a field, I think it was a light bulb for many to realize how critical it was to have that as a data field. And just the cycle we went through with figuring out the feasibility, doing the optimization, implementation, data collection, quality check and insights as we went along and then improvement opportunities. So I want to get to the insights that we found where we focused on this group of dairy and cattle farm workers. So essentially we asked the LA CASA staff, we said, okay, so now that we are collecting occupation, what can we do with this data? Is there a group that you're interested in knowing a little bit about their health status? And they said, yes, we have a lot of, I think cheese is very big in New Mexico, the production of cheese and these cattle farms, dairy and cattle farms. And so there was a group that they were interested in looking at. So we said, okay, let's look at the data and see what we can find. So they felt that many of the patients at the health center were part of these dairy farms and they were exposed to poor environmental air quality due to chemicals at neighboring sites, allergens, fumes from the cows. So there was a question as to whether or not being a part of these types of jobs or industries created a higher rate of asthma or respiratory tract issues among these essential workers. So this is a work category that we were able to track. And so there were 63 patients that were identified as employees of these dairy or cattle farms and 49% had a diagnosis of respiratory tract and or asthma issues. I mean, that's just astounding. We were shocked. And just to tell you about who these patients were, you can see kind of more middle age, 40 to 49, definitely more male, Spanish speaking, Hispanic or Latinx. And yeah, it's very striking just to see who these patients were. So what did La Casa do? They developed an onsite COVID-19 vaccine clinic for these dairy farm workers. They offered vaccines between work shifts and invited family members to clinic for vaccines. But the moral of the story is that knowing the risk of COVID for people with asthma and respiratory issues based on an occupation or industry category allowed La Casa to do true public health for that particular population of workers that made an impact that we are yet to see. We'll do some analysis to see what the result was of some of that work, but just very important from a standpoint of public health, population level health, but research as well to capture and understand these categories. And I'd like to acknowledge the wonderful work of our La Casa Family Health Center team. I mentioned Dr. Tom, the Chief Medical Director, Ms. Armijo, our Director of Operations and Information Systems Analyst. Our HCN research team has been amazing. And of course, our colleagues at NAC and CDC without whom we would not have been able to do this critical work. Thank you so much. Thank you. Thank you. Thank you. Thank you. Thank you. Thank you. Thank you. Thank you. Thank you. Thank you. Thank you. There are no stupid questions. So please, we would like to have a discussion. How are you guys doing? Are you interested? Are you collecting this type of data or kind of ideas? Well, you guys, this is awesome. This is really awesome. Great work. So I had a couple of little questions. One is, rather than have a patient input the data, you had the staff. Yeah, can you talk a little about that? So I think that's the ideal. The ideal would be for the patient to self-enter that data. And the current EHR system that we're working with now allows for that more. They have a very good platform for the patients, very friendly. You know, there's a lot about the educational level of the patient, the ability of the patient. There are some patients who it's harder to do. There's a technology gap, a technology disparity, so to speak. So there are some things that need to be overcome, especially for an underserved patient population. Not to say that it can't be done. I think that that's the direction that we're going in for sure. That's the ideal. Okay, yeah. And the guide does mention that it would be patient-provided entry is the recommendation in the guide from NIOSH. But we saw that. We were looking in Cleveland. We thought that was gonna be a problem. So interesting. Two little quick other questions. What's not in the labor force versus unemployed? Did you have both? And what's the? So at first we only had unemployed. We didn't have not in the labor force. So unemployed implies that the person wants to be in the labor force. They currently are not employed, but they would like to be in the labor force. Versus somebody who's not in the labor force and is happy not to be in the labor force. So we thought that was a very important distinction. Yeah, that's me. Okay. And last question. Sorry, one more thing on the status too. That's been a point of contention. That's actually why that status is not in one of the elements of US CDI. Because different groups have different terminology for this. As Catherine brought up, there's retired or part-time. So we really are trying to narrow it down to just three categories. So we're continuing to advocate for that. Because again, we just want interoperable data. We don't want to confuse the picture. Last quick question. Was NIOSH staff involved in this? Okay. Yes. So congratulations on successful implementation of this project. So that you know that I'm not making this stuff up. It serves the discussion to let you know that I sit on the board of directors for one of the largest federally qualified health centers in Northern California, La Clinica de la Raza. And knowing that the front desk clerks who are tasked with doing multiple, multiple things just to get the patient registered, all right. Having to search through multiple fields just to select the correct insurance, all right. Even though they are on Medicaid, all right. From a user perspective, having to comb something like job classification is potentially a fire pit, okay. And put it on the patient to do it. Okay, well, let's say the patient finds that, oh yeah, I work in construction, okay. But what kind of construction does he or she do, all right? Are they working for a large construction company that does commercial type construction? Or do they work for a small contractor, okay, who just does indoor renovations for residential units where they may be working hands-on with engineered stone products and they don't wear even a dust mask? Yeah, that's all great points. And you mentioned about our NIOSH colleagues. We would be remiss if we didn't acknowledge all the work that went into this and is continuing to happen. So again, they have looked into just what you described. I know Jenny Lundsman, one of her favorite examples is just cowboy, you put down cowboy, okay. Well, do you work on a ranch or are you in the entertainment industry? So again, the prototype does drill down into those things. The vast amount of data is a problem. And I'm gonna let Ray just speak to that for a minute. Yes, thanks. Whew, that's a big informatics problem as you just described and a lot of folks from HRQ, the Agency for Health Research Quality. There are a lot of informatics people trying to do usability studies on drilling down and getting those nuanced information that you were mentioning. So some of the folks, some of our partners, which is Alliance Chicago and OCHIN, which are other two health choice networks, who are doing a network-wide implementation of capturing employment and occupation information. We're trying really hard to talk to the EPIC folks to really digest the whole terminology and to make it easy to search. I think it's a big problem with standards. For example, if you look at ICD and SNOMED, there's always lack of granularity. You can't really get into that specific information. So some of our partners added another field next to occupation industry, which is for them to, they don't have to choose the code. They don't have to choose it from the structured list. They can type it also. Just so if it's really nuanced, they can do it. But that is, I think, a very important limitation that you just described, and we haven't really solved it yet. But we will try our best to really work on it. I just want to mention, too, that the creativity of the health centers are always amazing. So they developed these workarounds for some of these issues, and for example, they had a cheat sheet internally for some of these common occupation and industries that they were seeing at the health center, and there was almost like a learning curve among the staff who were collecting that data so that they would get to know better how to classify certain of these industries and occupations that they were seeing quite a lot of. There was the issue of the front desk getting the bottleneck with this new data that needed to be captured. There was another workaround of actually having the patient write down what their job was, what their industry was, and then going back retrospectively to enter that into the EHR, calling the patient to check to make sure that they were along the right lines once that information was entered into the EHR. So lots of creative ways to work around the challenges because yes, the health center's busy, can't have that bottleneck, gotta keep moving, but then also that learning curve that I think is as we work more with occupation and industry, I think the staff members become more familiarized with the categories, and it's not so daunting. That nuance is why you guys are here because you guys get that nuance. So thank you, Occupational Medicine. Hi, there's so many great points today. So I'm Wendy Tenassi. I've met with you all on the phone a few times. Really happy to talk to you. I want to challenge your premise that there are no stupid questions. So I'm sure I've got a thousand of them. Your point about the stoneworkers is fascinating because it really worries me. And if we had this in place before, we would have picked up pulmonary fibrosis and death in a population. But how do you get all the way down to the fact that there's stone fabricators I think is hard, except when you put it all together, so I've got a lot of questions and a million thoughts. A couple of things. One is probably regionally, there are regional jobs. So the clinic that you were working in had a lot of a large group of farm workers at Stanford Hospital, we don't. So probably it's daunting when it's an entire list, but it may be that there is a cohort of stoneworkers down near me, it would be down in the San Jose area. But they're not gonna be in Yreka as much. So I just wonder if it gets a little bit easier if you can regionalize the job types. I also wanted to know, so I think the EPIC, you know the EPIC thing is important to me too. I'm on the EPIC steering committee for AHRQ Health. Stanford is EPIC for patients and we're gonna go to AHRQ Health and then there's gonna be this blending. And that also for places that are doing that, like UC Davis is doing it and some others, they're trying to get the whole UC system, you could potentially use occupational health as that portal into the record rather than the primary care physicians. You know, at least you're gonna capture a lot of the, you're gonna capture the employees, you're gonna capture a portion because everybody is self-insured at Stanford. They all go through Stanford. So you can capture them maybe in different clinics was one of my parts. It doesn't all have to bottleneck in primary care. And then a question I had to get there was, sort of can we, if we talk to Stanford, roll it out as a test pilot in just one small segment of the population? So I've just learned that a bunch of our surgeons are getting cataracts. Hi, Amir. And did you know this? You're nodding. So a bunch of our surgeons are getting cataracts. Turns out that their eyes are a couple of inches from like the fluoroscope and there's no leaded loops. So I'm just learning this and we're just gonna put together research. So I'm gonna talk with any of you, but I was thinking, gosh, if I knew there were a surgeon, could I go back and just search surgeon and cataracts and this would be so easy? And then, or could we do it proactively that there's some sort of alert? So anyway, I love it. Maybe I wanna know, maybe can we, with Epic test trial, like a certain subpopulation of the whole unit? So then the question is, with the data you've already collected, do we have access? Is it stored in a national database somewhere where we could access? Because for the farm workers, I wanna know about tuberculosis because of bovine tuberculosis and farm workers and state of origin. And that would be something that I would be interested in looking up and maybe Amir would not. Okay, thank you. I guess I can begin. So I'll try to work backwards if you don't mind. Farm worker status is a social driver of health data element and there's only one machine code in the standards that represent migrant or agricultural worker, which is a SNOMED code. And if you ask EHR vendors, I have this list of SNOMED codes. Can you give me the data to show me all of these people? It's literally, it's on my phone. It's one code and the terminology and the concept represents agricultural and agricultural worker or farm worker. And they don't, because we've published lists of value sets for these and I've never received in multiple CDC and data projects of a list of SNOMED codes. So that's one big limitation is that they do a test to support SNOMED CT, but if you query the database, it doesn't exist. Number two, I want to address your idea. I think it's really the end user experience. I think a human centered design approach to creating these solutions to these problems is to ask, like you said, there is that surgeons in cataract problem that is a very niche, not really very niche, but that is a very specific problem that you have to ask and work your way up to the solution. And I think this is only a piece of the puzzle. Like I said, you start with the pilot and really show them that it works. And that's why we're working through USCDI to make these small incremental steps. I think in the future, we could do something similar to what the oncologists made, which is called MCODE, M-C-O-D-E, which is the minimal clinical oncology data elements, which is ask any oncologist out there what is their cognitive framework for how they approach patients and what are the common data elements and create that framework and represent them through all of these standardized codes. I think something similar should be created for occupational medicine and specific to your use case for surgeons and doing fluoroscopy. Sorry, Michelle. No, you have all great points. And yeah, so many things you brought up, great ideas. So yeah, back to the surgeons and the cataracts, very important, but you have to, again, also this ties into occupational medicine clinics getting the data. You have to be careful where you get the data because as occupational medicine docs, oftentimes we're seeing the patient on behalf of the employer. So the employer, we know what that person does. The employer has told us what that person does. So we got the information from the employer. We've talked about the challenges of collecting it from the patient. Sometimes the patients aren't tech savvy. They don't know how to enter the data. So you have to just, again, be very clear that the patient is supplying this data as PHI, especially if you're going to move forward and then use that data to connect it with other variables, such as a cataract or whatever else. So that's just a consideration to keep in mind. Before we, just one more thing I remembered you mentioned is about availability of this data across the United States. So we have what's called HIEs or health information exchanges. I think you're aware of that already, but it is not mandatory to join or report health data to these health information exchanges, which is incredibly strange to me. And I think I began that part of my conversation earlier is that is there money to be made or is there a law that they will be punished if they don't follow it? So now we're doing the law approach because is there money to be made in interoperability? Sorry, maybe somebody has a business model on that, so. And I was just going to say about this specific data that was presented is, so it's owned by the FQHC as well as Health Choice Network. We have access to that data. The unfortunate piece is it's just the one health center that we're collecting the occupational data. It's such a missed opportunity. I'm looking forward to the day when we are collecting this throughout our network because we mentioned 3 million patients. Imagine if we had occupational data on 3 million patients nationally, what we could do. It'd be amazing. We'll talk some more. A quick question for you all. You gave percentages for these occupations, but it'd be nice to have some denominators. Have you thought about that in terms of getting rates? I'd just be interested in that. Absolutely. I think there's some refinement that can be done in the analysis that we did. This was very high level, I think, some of these analyses that we did, but absolutely. Yeah, I can speak to that data as well. That's pretty early on during the year one of our projects. We are currently on year two, and we are getting monthly data from each of our three health center network partners. That was preliminary data, but thank you for bringing that up. Yeah, I just want to bring out one other point that was touched on earlier. Different groups have different occupations within their patient population, so that has been useful, I think, to Health Choice Networks and to others that are starting this. Within your population, what do most people do? If you can get those codes, and then, again, have those at the ready, you won't have the full denominator of all your patients, but you'll have a big chunk, so that's been a way in. I'm just going to say, I'm an epidemiologist at heart, so I get that question. Thank you for that. Thank you. We'll refine it. So I think we are- I understand. It's a pilot study, and it's the rate that you're paying. Thank you. Awesome. And I think we are at time. Thank you for all of those questions, and my last takeaway here, please use those Z codes, specifically Z57, occupational exposure, yeah, those things. And last but not least, for those working with farm workers, there's a US CDI data element that is currently in level one. There's three levels, level one, level two, and then acceptance into the draft, so we submitted farm worker status. If you are a farm worker or not, and it's stuck in level one, so if you have some time, get your informatics to hopefully comment and support the inclusion of farm worker status in US core data for interoperability. Thank you so much. Thank you. Thank you. Thank you. Thank you. Thank you. Thank you.

occupational data

Federally Qualified Health Centers

work and health

electronic health records

occupational data collection

challenges in collecting occupational data

La Casa Family Health Center

respiratory tract issues

occupational data standardization

United States Core Data for Interoperability