 Good afternoon. This is the afternoon session of the House Health Care Committee on February 10th. Our chair is here, but I'm taking over for a little while to coordinate the meeting this afternoon. And we have several members from the Green Mountain Care Board present and on the agenda. Susan, did you have a preference or a plan of who was going to present this afternoon? Yes, I do. Would you like me to share that plan with you, Representative Donahue? Sure. Okay, good afternoon. For the record, this is Susan Barrett. I'm the executive director of the Green Mountain Care Board. I am joined today by Sarah Lindberg, who is our director of data and analytics. Actually her full title, I had to look it up so I got this right. Director of Health Systems, Data and Analytics. I'm also joined by Jean Stetter, who is our budget director, and then I don't see her on the call, but she might be able to join a little bit later. I wanted to introduce you to a doctor from Dartmouth-Hitchcock, who is actually working with us as part of her leadership in preventive medicine residency. She's a fellow in infectious disease, but she is, and I'm looking to see if she's joining us. I don't see her yet. She is doing some government rotation with us at the board. Her focus is on health equity. So hopefully she'll get a chance to pop in. I know she was seeing patients this afternoon, so she was going to try to duck out of clinic to pop on. For the majority of the presentation, Representative Donahue, I will turn it over to Sarah Lindberg, but Jean and I are here to add, but I think she will be the person to be able to set the stage for you on the ask that we have and the work that we're doing around getting claims, insurers to provide race and ethnicity data on their claims. Wonderful. Thank you so much. So if you want to introduce yourself and join us, please. Good afternoon. Hi, this is Sarah Lindberg. As Susan said, I head up a data and analytical team for the Green Mountain Care Board, and I'm here today to present some information. I'll bring that up for you. All right. Everyone able to see the slides? All right. So my objective here today is to introduce you to the healthcare database that is steward by the Green Mountain Care Board. Talk about some of the current limitations we face related to race and ethnicity data, both here in Vermont and more broadly. I want to introduce a few of the strategies we've been pondering to help address those limitations and give you a couple examples of the ways that we would hope to apply better data in this area. So our healthcare database has two main sides. V-Cures or VH-Cures stands for the Vermont Healthcare Uniform Reporting and Evaluation System, rolls off the tongue. That's the All-Payer Claims Database or APCD, which Vermont was an early adopter of these types of databases, so our data actually goes back to 2007. And it's a bit of a misnomer. It really should be the most payer claims database since we aren't getting data for 100% of the insured market here in Vermont, but we estimate that we are getting at least 85% of that information. And we do get claims submitted for Medicare and Medicaid, which are governmental players, as well as commercial insurers, such as BlueShield of Vermont. And it's based on where people live. So whether the person gets care inside Vermont or out of state, we're going to get their medical and pharmaceutical claims in that data set. The other side of it is the hospital discharge data set, or VUDS, the Vermont Uniform Hospital Discharge Data Set. And that's one of the most long-standing data collections. That goes back to the 1980s, and it's a fairly consistent way that states collect information from hospitals. So that is for care that is delivered in Vermont, whether the patient lives in Vermont or comes to us from out of state. There are some limitations associated with that data set. One that's probably most relevant to our conversation here is that it only has the facility side of the care, which can really limit some of our perspective, particularly for outpatient services. So whenever I'm talking about healthcare data, it's a bit of a joke among my team, but I like to think about both the resident perspective and the provider perspective. So resident meaning where does the patient live or the person live? And that's the left-hand portion of this slide. So are they from Vermont or are they from outside Vermont? And then where was that care delivered? Was it delivered in Vermont or outside of Vermont? So outside of essentially benchmarks or comparative information for people who live outside of Vermont and get care outside of Vermont. I'm not too worried most of the time, but for the rest of it. So as you can see, V-Cures, again, with most insured residents, whether they get back care in Vermont or outside of Vermont, whereas the discharge data is largely a provider look. So the facility discharges from our 14 community hospitals. We actually recently were able to also include the Brattle Borough Retreat in that data collection and are in the process of adding ambulatory surgical centers. We also have data sharing agreements with our neighbors, New Hampshire, New York and Massachusetts. Unfortunately, data sharing's gotten surprisingly complex in recent years. So we're a bit delayed in getting those Vermont resident discharges from them, but we're in the process of getting caught up. New Hampshire in particular is a big piece of the puzzle. So we know how important that is. So race and ethnicity data is what I consider a known unknown. As you've been hearing about, I'm sure there are significant limitations in the information available to researchers such as ourselves and policymakers, providers related to good. And by good, I mean reliable and valid information related to race and ethnicity. It's an identified gap that persists despite some attention to it. And I'd say there's other factors that can come into that picture and intersect such as language. This is really important in the healthcare setting. Immigration status can also be a big deal or citizenship status. So these things kind of can all intersect in ways that affect people's interaction with the healthcare delivery system. But this framework I find helpful was introduced by Kilbourne at all back in 2006. And I think it still really stands up. And that is really to reduce or approach any sort of disparity in healthcare. You must understand it. And to really understand it, you really need to be able to detect it. And so we're weak today and being able to detect racial disparities in the healthcare database. And this is something that's much broader than Vermont. Actually, Jean Stetter brought an article to my attention where when the CDC was talking about the first month of rolling out the vaccinations, race and ethnicity data were missing for about half of the vaccination records. And so being able to track that is really vexing them at a national level. And given the seriousness among the people of color and co-morbidities that might be associated with race and ethnicity, that's a big deal in trying to track it. So the current state of race and ethnicity data in the healthcare database, two different stories. So in the discharge data, each facility does collect race and ethnicity data, and those data are submitted to us in a de-identified format, but it does include the reported race and ethnicity. So our vendor does great work in terms of just making sure that the values that are submitted make sense. They're acceptable values of what we would expect, so they're consistent in that way. They check for trends over time to make sure there's a crazy jump that would be unexpected. And they also at a very high level just monitor to see if the rates of discharges they're seeing seem to reflect the underlying population. However, there's been no official kind of audit of that information. And my suspicion is there's probably a lot of differences between the facilities and how they might collect that information. Can we just pause for a minute so we have a question from the chair. I love being on the other end of this being able to raise my hand. Actually, it is a question that comes up anytime we're talking about race data and is this self, is this a self report, or is this a report by someone else? And how are the categories, who's determining what are the appropriate categorizations and there can be significant limitations built in even if you say, oh, we get race data. Thank you. Yeah, and that is actually one of the focuses we currently are doing what we call an enhanced data validation workgroup, and that's one of the very things we're trying to understand is how much variability there is and how that data, those data are collected today and, and, you know, how good we feel about it, how consistently it might be collected, but that's an absolute, like right on the nose point of the type of things that we want to kind of interrogate in that data set. Beacons is a different story. So that's submitted by payers and the payer may or may not even collect race and ethnicity data, at least in the commercial market. And the submission of it is currently optional. So the elements that we collect as part of the all pair claims database are part of a rule. What's great about that is they're really consistent and nobody has to worry about crazy changes. The challenge about that is then if we need to make changes to reflect any old thing that might change. We have to go through the rulemaking process and I'll talk a little bit more about that in the future. But there were, as it was explained to me, there were two main reasons that choice was made back in 2008 when this was set up. One had to do with just the limitations. Payers either felt like they didn't have that data or it was not good data. Unfortunately, some of that still may be true. And the other concern is that, again, we were an early adopter of all pair claims databases. And one of the principal things that people were optimizing for was patient privacy and protection. And, you know, over a decade ago, Vermont had an even wider composition than it does today. And so there were more concerns that including that might increase the risk of re identifying a patient in this data set. So, you know, some of these issues still are at play today, but we think our calculus has changed just in the manner that we think that race and ethnicity data are crucial in trying to understand disparities. And that's why we feel it's appropriate to address. We could take another pause here. Representative Peterson has a question. Yes, thank you. What do you mean when you say ethnicity data? What's that? Sure. So ethnicity is traditionally two options. You're either Hispanic and or Latino or not. So ethnicity is different than race. So race on the, you know, they're big five categories better. I know it's different. But so you're only identifying Spanish or Latino. That's the way ethnicity is currently encoded in the way that we're collecting it. Yeah. Okay. Thank you. No problem. So to talk about some of the strategies that we've been contemplating at the Green Mountain Care Board for addressing these limitations, I'll go over three today. One is developing a standardized approach for collecting race, ethnicity and language information. The second is requiring race and ethnicity data be submitted to the All Pair Claims Database vCures. And the third is thinking about ways to integrate our data with others such as the Vermont Health Information Exchange. So when we talk about standardizing the way that we collect this information, I think that that's a really one of the things that sounds really easy, actually a pretty big undertaking and one that would involve substantial resources. And I would hope that as a state that we work together to get it right. But if we are able to collect race, ethnicity and language information in a uniform way, that would make the data more reliable and valid over time, across different data sources and in different settings that people might be interacting with the healthcare system. And by adopting conventions of how the information is solicited, we also might get more complete data. So just the way that you frame a question might make someone more or less likely to skip that question. So if we really are careful about the way that we frame it, we hopefully would get more complete information. Sometimes these questions confuse people. They don't know what you're really getting at with the way it's framed or if it's not available in the language that they speak best. Then I think if we have it in a consistent format, they'd be more likely to increase familiarity and hopefully reduce some of that confusion. And this is one part that might be getting at part of what Representative, the previous question was about. And that is that CDC does maintain a code set that has over 900 different values for this, these types of variables. And that's really important, especially if you think about a provider. So as an example, I spent a couple of years living in Minneapolis in the early aughts. And at that time, you know, they're, and I'm sure there still is a really large among population. And so they may be Asian and so is, you know, maybe someone who was adopted and is Korean of Korean descent or someone who's been in the United States for generations and is of Indian American descent. So Asian can mean a whole lot of things by having these more refined categories, you can tailor interventions and more, more robustly kind of identify where there might be some potential disparities. For instance, with the Hmong population, there was a, I think a really good impulse to put things in Hmong in written materials. However, it turned out that, you know, most of the Hmong population that we were dealing with in the public education system didn't work literate. The parents didn't read Hmong. So what they really needed were interpreters. And so being able to address that kind of concern really requires this kind of refined categorization that you can also roll up to more meaningful subgroups from at least the Office of Management and Budget perspective, which is what a lot of federal programs use for racial information. And, you know, if we are able to come together and come up with a good approach that we can standardize, then we could kind of roll it out throughout the delivery system and the state of Vermont at large. And, you know, this would be something that would probably require a lot of really careful thought and collaboration to get right and continue to revisit to make sure that we're continuing to use best practices. So if I could ask a question at this point on that last, the prior, yeah. When you're talking about we in terms of, you know, what kind of conventions, for example, that will make people more feel more comfortable with answering questions or what kind of approaches should be extended. Who's are the we who's involved in that conversation and discussion. Yeah, so so far, I've really only been privy to it through the planned work for enhanced data validation work group so this is related to the data assets managed by the Green Mountain Care Board but I would love that to be a much broader conversation and be happy to participate in anything that may already be planned. I guess that yeah that was the indirect question is what whether it involved some of the members of these communities themselves. That would be critical I think to getting it right. So in terms of requiring the race and ethnicity data to be submitted to our all pair claims database. Right now we're really just trying to assess the quality of existing information maintained by the payers and facilities who are submitting data to our respective databases. And our opinion is that we think that there's a lot of work to do to improve that quality before it makes sense to require it submission just because you know if we're going to just collect garbage or or that if we force them to submit it and it's missing. There's not much value to gain there so we have been working on the workgroups since last year and we hope to wrap that up in many of this year. And we also hope soon to launch a rule change for the vcures data so as I said it's I think the rule is from 2008 so it's long overdue for a revisit. And the current plan is we have drafted two separate rules one devoted dedicated to the submission of data, and the other dedicated to the release of that data. And in the submission rule we hope to have a companion guide where we are able to have the elements, a little bit more nimble to update. And that's for reasons, particularly related to interoperability standards at the federal level. So they are making insurers set up ways to exchange information more readily. And so if we could leverage that sort of data exchange, we would reduce burden on the payers and also might be able to get additional information that we're interested in such as the premiums that people pay for health insurance. We have a question from representative page. How long have you been collecting this data and, and could you tell us a little bit about the data that you've already collected? Have you have you had an opportunity to analyze it or. Yeah, so yeah the hospital discharge data we've been collecting since the 1980s and the claims data we've been collecting since 2009 but the data go back to 2007. So we analyze it as a matter of course and in all our business and anytime you see something related to cost that's usually got vcures under the hood. The blueprint for health is one of the power users of the claims database or VDH uses both sides of it but especially the discharge data. But what does it tell us about our race and ethnicity for the state? Does that have any of that information available? So, yeah, the only payer in vcures that really has race and ethnicity data is Medicare and that population I just checked in 2020 is 98% white which makes sense given that we know a lot of our populations of color are younger and therefore less likely to be covered by Medicare. And then for the discharge data set I didn't do a prusal of that but that that's the care delivered at Vermont hospitals and it is anyone who seeks care at that hospital and so that does have race and ethnicity data and has in since inception. Representative Goldman. I have two questions. One is sort of what's the denominator how many people are you talking about when you're talking about these data sets is my first question and when my second question is when you talk about a rule change where does who's responsible for that rule change where does it sit. Oh boy. I'll take the first question first. So, the all payer claims database again we think we're covering about 85% of the insured market in Vermont so we know we don't have the uninsured, or people. You know that the other main missing piece pieces are federal employees, military plans, and then self funded groups since Supreme Court decision. Don't have to submit their data so we're missing self funded groups that choose not to submit their data to the all payer claims database but we think we again have over 85% of the insured market. Yeah what what is that number eight what's the absolute number 85% of what. Oh, so it's about 625,000 Vermont residents. We have just under. Okay, yeah. Yeah, it's a resident base yeah. Okay, so 85% of all Vermonters you have you think you're having data on is what you're saying. So, you're wanting to take about 6605 is probably the insured portion of that yeah yeah yeah so 85% of 600,000 let's say. Okay. Yep. Okay. The rule change. Gosh, I probably won't do this justice. I rule making and changing rules in this state is a established procedure and you have draft rules that you present. And those are, then go through a stakeholder process they go through two different legislative committees i car and l car, not sure what those stands for, but that's a whole. process that takes quite a bit of time so you know we're optimistically hoping we might get through that process by next winter, but that some of that is really out of our hands. Most of our new members don't we've never gone through the explanation of that long process so we'll we'll need to follow up at some point so yeah thank you for that. If I may I might just mention, I'm not sure if it's appropriate in this instance but we also have the authority to change rules by statute. Right. So there are situations in which depending on what's required, things can go faster. So yeah so then once we go through that process whatever path it might take at that point is when we would really take a thorough re look at the data elements like this today and update that submission guide and requiring the submission of race ethnicity is almost a no we would contemplate at that point and hopefully in the interim we've done a good job of improving the quality of that information so that we when we start collecting it hopefully next year, it would be meaningful. So then, probably in my mind the most exciting idea that is on the table over at the Green Mountain Care Board is the idea of data integration so that means combining different data sources at a person level to try and expand the utility of that information. So one example is so, you know, medical and pharmaceutical claims are essentially just the bill that the provider sends to your insurer to say this is how much we want to be reimbursed for that service so it's really good at tracking some things, but not so good at tracking other things. So if we were able to connect the medical claim of a birth with the information on birth certificate, we'd be able to really round out some of the information about that baby so it's it's apt our score, the weight of the baby. So I think what's most exciting to people in trying to work on interventions in this area is connecting it to the parents. So right now, a medical insurance family might not match up that well with the actual family that the baby is born to so that could, you know, with appropriate, you stewardship and making sure that access is appropriate. That could really add some really interesting insights to people working in that area. Similarly, if we're able to connect to death certificates, we would have the cause of death, which can be really intriguing for some research that's going on. Right now, if someone just disappears from the database, we're not always sure if they moved away if they passed away if they lost insurance. So that would at least help us figure out what was going on in the case of death. So one of the prime sources that people talk about in integration is the clinical information contained in the Vermont Health Information Exchange, which is in your, you know, electronic medical record, electronic health record, people call it different things, but that's the system that the provider is looking at in the hospital and connecting that with the information from that shared with the HIE, connecting that with the medical claims so that we can really round out the information that we're able to provide to the people providing care, to researchers working on interventions in this area and public health workers. We know that the V-High already contains race and ethnicity data, so that would just be one of the additional elements that we would hopefully be able to gain from such an integration. So I have a question about that series of bullet points, and it's sort of a broad one so you can maybe answer part of it or the extent you can. How does all of that interplay with patient privacy protections? Yep, it's a great question and something that we wouldn't undertake lightly. So, you know, vital records are one of the first areas that we're interested in pursuing. And so we would do that, you know, with the support of legal representatives on both sides and certainly want to hear from patients. But we would want to know, you know, if the cost outweighs the risk in that case and how we would best protect it. So we have a pretty hefty process for releasing data to folks who are interested in it. And so my hunch is that if any of these integrated data sets would, by necessity, have a more, even more rigorous application process and review process so that we know that the information is being kept safe. So our data is always released in a de-identified format. So, you know, the connections would only be to ID numbers. They wouldn't be by a person's name or anything like that. So we feel very strongly that we want to, you know, maintain privacy. That's a really important consideration in any of these efforts. Representative Page. Yes, I was just curious. What is the cost of collecting all of this data and distributing it, costs and personnel and time, as well as money for the Green Mountain Care Board for all of this work? If that's in a future slide and you want to wait and answer, but if it's not, go for it. Yeah, I guess I don't have those costs at my fingertips. I don't know if Jeanne Stutter might be better. Jeanne might be able to follow up. This is Susan. We can, we can follow up with that. That's carved out specifically in our budget. So we can forward that on to you, Representative Page. You couldn't just tell me, you know, ballpark or. Jeanne, do you have a ballpark number? I don't want to get it wrong. Thank you for the record. I'm Jeanne Stutter, budget director at the Green Mountain Care Board. And there's two things, Representative Page. Sarah talked about first, the rule change work. And that is in the, you know, projected between like 80 and $85,000, which for the Green Mountain Care Board would be 40% general fund and 60% fillback. And then the subsequent work for the data integration is, is estimated to be in that same ballpark for cost. So, so that is, that is what we have is our cost. I think Representative Page you are alluding to the work that would be undertaken by providers as they're asking this question. And I don't know how to quantify that I do know that providers taken a lot of information about individuals so I don't know what the incremental time would be for adding those that question or those questions. I have a couple other questions right now. I'll just add a point of clarification if I'm correct, those budget figures and therefore moving ahead on this are not currently in your budget request. Is that correct. That is correct. Thank you representative Donahue for clarifying that so that's what we identified as the cost. And it's not currently in our budget. Representative China. I'm curious to learn more and it doesn't have to be necessarily right now but I'm curious to learn more about how you extract like what kind of clinical information you're extracting how you're sharing it. How people are consenting or not consenting to that because when most people go to get health care I don't think they're going there with the thought like I'm going to be contributing data about myself to some greater cause that I that is not yet identified. And so, and there might be things that people's data are used for that they don't personally agree with. And if we're just harvesting data and, and providing it to the people at the at a high level even if you're separating identifying information from that data, you're still extracting from the suffering of people like we you know as a system are still extracting from the suffering of people and so even if we're using it for what we might consider to be good. I'm just curious like how much are we respecting people's rights and power and that and, and like so I guess I'm just curious to know more about what that clinical information is like are we going to be, for example in psychotherapy are we going to be as an algorithm going to be scanning all of the notes and looking for patterns of words and extracting something from that, because that's very different than just saying here's the ages of people. So I'm curious if you have any more to share about that. Sure. And so this work is to be determined. So those are the types of questions that we would expect to be grappling with in any planned integration effort. Today's the only clinical integration with vcures of which I'm aware is through the Vermont blueprint for health, and they took information from the Vermont clinical registry while that was still up and running. And that was fed through the v high to the clinical registry and folks had the option to opt out of data sharing. I believe it was even opt in at that point for them for the most of the life of the clinical registry. And I believe the information that was used was pretty limited just test. I mean, not just but largely with test results and things like, you know, blood tests related to diabetes and stuff like that. So yeah, no, I think that all this is really important stuff to be considering and making sure that we're responsible. Representative Goldman. Thank you. I was wondering if you could give an example of how you use the data to inform your policy so that we have a clearer idea of why it's important to collect the data, whatever it may end up being and how it gets used in a forward thinking kind of way. Sure. Yeah, yeah. So this data set the all pair claims database is really the state's only clear lens about the claims based expenditures that Vermonters are incurring. So it's one of the bedrock data sources for our expenditure analysis. It also was the data source that was used for planning the all pair model and is using these to measure the total cost of care while we're monitoring it. It also helped the blueprint for health get set up with the primary care medical homes and was the first chance providers ever got to see some of the outcomes among their patient population. Those are a few examples. We do allow external researchers to apply for the information. A recent example that I thought was pretty interesting was on the RTI international was able to use our data and look at treatment patterns among opioid dependent folks and found some raise some concerns about people who might be tapering too rapidly. So those are just a few examples off the top of my head. That's great. That makes it a little more real, I think. All right. So, and maybe this is also to the point is what will we do if we were to have better race and ethnicity data in these data sets. And I think one of the major things going back to reduce the disparity you first need to be able to detect it. And so our partners over at the Department of Health have had a long standing mission to help, you know, evaluate and understand and reduce disparities based on race and ethnicity. And so this would give them kind of another tool in their toolkit as they're looking at some of these issues in across the state. There are many initiatives that we would hope that this information could help better inform with those variables. And I'll leave you with two examples that are at the top of my mind. So this is data just for convenience. I got out of article that's in pre-publication phase by Fetin et al. Related to the fatality rate for among COVID-19 patients in the United Kingdom. And if you look on the left hand side, you see there's clearly age factor. So those under 65, 2.7 per 100,000 had died during the time period of study versus 147 in the 65 plus. That's the data that we have today in our database. If we were going to expand that and then add the dimension of race and in this case, they just quantified as white or black. We would miss the fact that in both age groups, the fatality rate is twice as high among black people. So 2.5 versus 5.4 among under 65 and 142 versus 321 in the 65 plus. So that's just a really sobering area blind spot of which we're aware, but something that I think we need to do a better job addressing, at least in the data available to us. The other example has to do with substance use treatment. So one of the kind of neat things in my mind about the claims database is that we have eligibility records for people, whether or not they seek care. So if someone has health insurance but isn't going to see a provider, we still know they're there. So if there are concerns about particular subpopulations who are deferring care or otherwise aren't seeking care that we would expect at a, you know, kind of benchmark rate, we're able to get a better insight about that than data sources that rely on someone actually coming to see a provider. We also receive the data in a deidentified format. So the information is transformed into a hashed string of numbers and letters before it comes to us. So that means that submitters don't need to redact claims related to substance use treatment because patient confidentiality is of utmost importance to that population and has been mandated by a federal regulation as 42 CFR part two. We know we have an advantage of seeing some information that a lot of our partners across the health care delivery system might not. And so putting those two things together, we feel like we could really do some interesting things to try and figure out if there might be some racial or ethnic disparities in those populations for this kind of hard to get at information. So that was all I had. I'm happy to take any other questions that people might have questions. I guess they mostly were asked as we went along. I over here we go. Sorry representative vote. Sorry. And I apologize because I missed the first couple minutes. So if this was answered. So if I guess my question would be, if, if we are looking to better understand health care disparities in our state. And we don't have access to this data integration. What will we be able to determine. Sure. So if we are able to improve the quality of the recent ethnicity data maintained by payers and then make it a mandatory variable to submit even if we don't add any other data we've already enriched on the all pair claims database. But it sounds like if we stayed status quo, would it be difficult to get to where we're trying to go. I think it would. And as I alluded to the, you know, there are intersections with other variables such as socioeconomic status, immigration status that aren't, aren't visible either in a clinical record necessarily or a claims record. And so the more as a state we can come together and come up with some more holistic pictures, knowing that there are important tradeoffs to consider related to privacy. So I think, you know, there's some risks in not trying to take advantage of that information. Thank you. Representative Peterson. Yes, could you go back to the last slide. Absolutely. I lost you there and I just want to. Oh, sure. Yeah, my fault. I got. Sure. A lot of detail. Yeah. But they say care. So do we have data on this. We have de identified information. Yes. So I don't know who is who that person is, but I do have records of substance use treatment in a de identified form. So we have data on the subject. Okay. Thank you. No problem. Anyone else. Now, I guess back to Susan Barrett. What is, did you have any other comments from you or any of the other people here or were they just for backup questions? I see we have a Jean Stetter was. Yeah, yeah. Was here for for backup if they're and she backed up beautifully. Thank you, Jean. Thank you so much for budget questions. Oh, are you on mute Jean? Did you want to say something else? Sure. No, I'm not on mute, but I didn't want to talk over you. So, go ahead. No, no, no, go. There was just a couple of things I wanted to, you know, kind of highlight. The report has a very robust data governance process that that covers the release of our data. So I just, you know, that that could be a conversation for another time but there's a there's a robust process in place and I think Sarah, you know, I think it would be appropriate to say that we're our process is a model for other other groups thinking about releasing their data. The other thing is is vCures is used by VDH, you know, it kind of, even though it resides at the Green Mountain Care Board, you know, diva blueprint VDH they all have seats into the database to access it. I just want to, it's something that has longer fingers. And the last thing I wanted to say to, you know, Sarah's point about really getting standards about what you list. That CDC thing that you referenced so if only 50% of the people the federal government said if only 50% they're trying to determine, are we doing a good job or not in terms of getting the vaccine to the BIPOC community. So if 50% doesn't have any information at all, in the, in the information that was reported, the biggest classification was unknown. So, so that just means it's difficult to measure something that they want to measure. So anyway, Sarah, you're better at this. So I'll defer those are the things I wanted to highlight. Thanks, Dean. I'm glad you brought up the data governance council because I don't like to brag, but it's a bragging because we've been working on data governance. It's, oh gosh, since like maybe 2015. And we have, as Jean referenced, other state agencies are emulating our data governance council. I think we have an advantage because we're smaller. We have so much has so much data in HSS so much data. But we have done a ton of work on data governance and it is about oversight of releasing data but it's also being stewards of that data and we're where we're going with that data. We meet every couple of months. I'm the chair of that council. We have membership from blueprint for health, VDH, the other VDH folks have been really tied up lately, as well as by state primary care and others. So thank you, Jean, for bringing that up. You have a web page on your website that that if people wanted to look for more information about the council. I will send that to you. Yeah, and while you're there, there's plenty of examples of what we do with that information. So there's interactive reports and fixed reports and external reports but give you a sense of the application. That would be a great resource to spend some time looking at. Okay, it looks like that's it for our questions. Thank you. I think this was very helpful. I don't know if any of you have had a chance yet to look at the new bill that was introduced just yesterday. So I'm going to look at H 210 on health disparities and there are references in it about, you know, data collection and it might be interesting for, you know, sort of a cross comparison look at what the administration's equity directors recommendations what the bill recommends and what you're looking at and see where the intersectionality is or isn't. But thank you very much. Thank you for having us. So I think probably we're well time for at least a brief. That's only been 45 minutes but but if we don't take a very brief break now, we'd be interrupting our, our next witness midway so maybe we should take a, you know, five minute stretch break and and then come back.