So my name is Dawn Belt Davis. I'm at the University of Wisconsin-Madison, and I'm actually filling in for the person who was supposed to chair this session, and honestly, I'm sorry, I forgot who that was, but I just wanted to acknowledge that someone else was the official invitee, and I'm just helping out today. So I just wanted to, just brief reminder for everyone, these are gonna be 10-minute talks. We'll have five minutes for questions, and we will just go ahead and jump on in so we can stay on schedule. So our first talk today is from Dr. J. Young Park. He is a resident at the University of Maryland Medical Center, and the title of his talk is Racial and Ethnic Disparities in Diabetes, Technology Use Amongst Persons with Type 1 Diabetes. Dr. Park, hi. So your timer will be on. It's for science, I think it's for science, go here. Okay, thanks. Hi, honored to be here. My name's J. Young Park. I'm one of the internal medicine residents at the University of Maryland Medical Center, and I'm talking on behalf of our department at the Diabetes and Division of Diabetes and Nutrition. And the topic of the talk is Racial and Ethnic Disparities in Diabetes, Technology Use Among Adult Patients with Type 1 Diabetes. So without further ado, I invite you to take a brief mental trip to downtown Baltimore. So this is our center. We're located at the northwest corner of downtown Baltimore, and in front of us is Park Avenue, not named after me. But anyway, so Park Avenue is quite significant in the area because it represents a division among populations. To the east of Park Avenue is the Mount Vernon slash Cathedral Hill area, where the Washington Monument is, the Walters Art Museum, Peabody Music Institute, and the Baltimore Basilica. It's a nice area with lots of good hipster restaurants. On the other hand, to the west of Park Avenue is the Seton Hill area. So it's named after Mother Seton, who's the first American woman to be canonized by the Catholic Church. But this is actually an area that is challenged in terms of socioeconomic disparities. Moreover, the Lexington Market area is the epicenter of drug deals and violent crimes. So to the west and east of Park Avenue, we see two very different worlds on a daily basis. So perhaps because of this, at our center, racial and socioeconomic disparities in the type 1 diabetic patients are quite palpable. And the recent studies by other groups suggested that socioeconomic status is not the only driver of racial disparities. In these studies, the racial disparities persisted even after adjusting for income, insurance level, and health literacy. And in order to truly assess the degree of disparities in our population, and also to explore all the different factors that might exacerbate disparities, we did a comprehensive retrospective core study on insulin pump and CGM use. So we reviewed all of the charts for all type 1 diabetic patients who visited our center between January and December 2019. And patients with a history of type 1 diabetes and LADA were included, but we excluded type 2 diabetes, MODY, and the patients undergoing pancreatic transplant. And then we followed up this 2019 cohort for the next two years, in 2020 and 2021, to check for any changes in diabetes technology use. And we recorded the following parameters. And for income levels, we utilized the 2019 US Census data to estimate the median income for each zip code. And we did multiple logistic regression to account for all of these factors, and also we did some stratified analysis to evaluate for any potential effect modifiers. And among the 731 patients that we screened, 197 patients were excluded, and we ended up with a total of 524 patients, which is the 2019 cohort. Of these patients, 69 patients were lost to follow-up, and nine patients moved to another institution, how dare they, who, so that left us with 446 patients, which gives us the per protocol population. And this, among these patients, only 379 patients had the last fit in 2021. And obviously, if there's a per protocol population, there's going to be a quote-unquote intention to treat population. What we mean by this is the entire 2019 cohort assessed two years later again. So results showed that there were significant differences in terms of the rate of pump and CGM prescriptions among different racial groups. For example, non-Hispanic white patients, their ratio was 54% on insulin pump prescriptions, which is about three times as more as the ratio of non-Hispanic black patients on insulin pump, which is only 18.4%. And these differences were seen again in 2021. And obviously, between the racial groups, there's a significant difference in terms of baseline A1C, insurance type, and the base income. So we did the multiple logics of regression, and here we report the adjusted ratios. So you can see here that the non-Hispanic white patients have a higher percentage pump and CGM use rate compared to any other racial groups. And between 2019 and 2021, the overall rate of prescription for both pump and CGM went up for all races, but the disparity still persisted. And these disparities were statistically significant. So as you can see with these odds ratios, the non-Hispanic white patients are about two times more likely to be on advanced diabetes technology compared to non-Hispanic black patients, even after controlling for age, sex, BMI, A1C insurance type, and ZIP base income. And this was seen in 2019, 2021, in the ITT population, and also the PP population. If we look at the loss-to-follow patients specifically, there were a little bit difference in terms of proportion, 36 versus 39, small difference. And obviously, the per-protocol population was much more likely to have prescribed insulin pumps. But overall, as I discussed in the previous slide, both ITT and PP populations in 2021 had statistically significant ORs. So loss-to-follow didn't make a real dent in the overall population. If we look at insurance types specifically, certain insurance types made the disparities worse, so commercial, Medicare, and other self-pay. And especially, numerically-wise, other self-pay patients were the worst, but this is probably because this is a heterogeneous group of patients consisting of people who are truly uninsured versus young patients who are under Affordable Care Act. So that's why it's a 25. And there may happen a little bit of possible effect modification by insurance type on CGM use, but if you actually add the interaction term to the model, it didn't really change much. So the insurance type is effect modifier, not very significant, and there's disparities regardless of your insurance type. If you look at zip-based income, we had three, we meet three stratas. So patients below 30K annually, which is below the poverty line, patients between 30 and 50K, and patients above 50K. And for all of these strata, there were differences. Numerically-wise, this was the worst in the 30 to 50K. And one might argue that for the patients who are from wealthy areas with an average zip-based income of more than 50K, it starts to matter less because the OR crosses one, but the actual OR is still 2.8. It's probably just due to low ends. And overall, not a significant amount of, regardless of your income, there's disparities. And lastly, we looked at different providers. So this is a graph that represents the range of odds ratios for each provider. Each of these dots represent a provider that at least saw 20 patients at our clinic. And don't worry, all of these provider names are blinded, and I do not know who each of these providers are. I want them as my PI, so I really don't want to know. Anyways. So anyways, there was significant heterogeneity in terms of ORs. Some providers did not prescribe any CGMs denies by black patients, so OR can be calculated, divide by zero, so it looks like it's 100, but it's actually infinity. Only a handful of providers had an odds ratio closer or lower than one in terms of prescribing pumps or unconscious glucose monitors. And we couldn't fit this into the logistic regression model because of the large variety of providers, but we can still nonetheless see the heterogeneity. So in summary, in the year 2019, non-Hispanic white patients were two times more likely to be using advanced diabetes technologies compared to non-Hispanic black patients, even after controlling for age, sex, BMI, A1C, insurance-type ZIP-based income. And two years of follow-up at the clinic did not resolve these differences. Loss to follow-up patients only differed marginally in terms of demographics and disparities were seen in both ITT and PP populations. And insurance and income label may exacerbate racial and ethnic disparities, but not significantly, regardless of what your income bracket is, there's disparities. And there was wide heterogeneity in terms of providers regarding disparities in pump and C-GEM prescriptions. So when it comes to disparities, there's going to be socioeconomic factors that influence it, provider-institutional factors, and also patient-driven factors. And in order to truly really understand all of the factors that go in here, we're going to have to do a quantitative and qualitative analysis. And so far, the studies have focused on socioeconomic factors and provider-institutional factors, a little bit lacking on the patient-driven factors, so what we're doing right now is to conduct a comprehensive survey of all of the patients in the 2019 cohort. And the survey focused on patient-derived and patient-perceived factors. So some patients might be just perfectly happy with multiple daily injections, or some patients might feel that doing some pump is too complicated. So we are also asking open-ended questions regarding what they actually feel in terms of what their barriers are. Recruitment is ongoing, and we aim to use this qualitative assessment to supplement the quantitative approach that we are applying so far. So thank you for listening. I'll open up for questions. Thank you. Okay, thank you very much. If you can just state your name and institution before your question. Hi, Ropa Dei, Denver, Colorado. A wonderful talk. That was really, really interesting, and very well outlines how we, as providers, give unequal treatment to our patients. And I'm wondering if you have any plans to look into that a little bit further, and doing maybe a qualitative study to figure out do we, as providers, think that our minority patients are less likely to be able to handle using a pump, or what are the barriers for us in terms of providing those prescriptions? Yes, I mean, definitely. I mean, we thought about doing a survey to the providers, and I don't think we could have come up with much of a better methodology. The only problem is that then you have the, kind of have an observer effect where the providers now know that you are being observed. And we didn't want to mess up our data yet because we were following for two years. Yeah, so, but now might be the right time to ask them. Thank you. Yehuda from California Clinical. Thank you, that was a great talk. My question is, did you consider patient preference? Because here we're talking about socioeconomic. You talk a lot about economic, which is very good. I wonder in any economic category, patient preference and cultural background was considered, thank you. Right, yeah, so we are going to ask about what the patient perceived factors are during the next phase where we're doing the comprehensive survey for every patient who was in the 2019 cohort. So we hope to address that. We're not going to ask them to self-identify a culture because that might be a little bit, that might be, that might worsen the disparity, like I might be looking at with tinted lenses myself. So, but we will just ask them how they feel, okay. Yes. I found the thing that was most striking was the improvement in adolescent compliance when we were able to get pumped. I had one school nurse who used to hang out in the bushes and watch my patient to see if she was going to the pizza parlor across the street. And kids hate to have grownups telling them everything. And these kids, as soon as they got pumps, they were free, they pushed their own button. And minority kids were every bit as good at the technology as the white kids. It just made a world of difference. Yeah, I completely agree. Regardless of your cultural background, even I don't like being told by the adults in the room what to do. So, then again, the comedy's right here because I'm a 29-year-old talking to a room of very seasoned professionals and pretend that I know anything. Anyways, I'll stop. But yes, I completely agree that there's a lot of implicit biases that we have. And also, during my short experience, I completely agree. People are capable. And no provider's actually more interested in the patient's health than the patient themselves. So, patients generally, and especially if given twice, they're curious and they figure things out. Okay. And yeah, please contact me by email if you have any questions. Okay. All right. Thank you very much. Thank you very much. Yeah. Okay, bye-bye. Bye-bye. All right. So, our next speaker is a somewhat similar topic, I think. And this is Dr. Estelle Everett. She's an assistant professor in the Division of Endocrinology and is an endocrinologist and health services researcher at the University of California, Los Angeles. And the title of her talk is Assessing Longitudinal Disparities in Insulin Pump Use Among Youth with Type 1 Diabetes. All right. Good morning, everyone. So, I'm gonna follow up Dr. Park's talk with the perspective from the youth population. And so, today I'm excited to talk to you about a study we did looking at the longitudinal view of disparities in insulin pump use among youths with type 1 diabetes in the United States. I have no financial disclosures. So, as we all know, insulin pumps have revolutionized the way we manage type 1 diabetes. They have allowed us to provide more physiologic administration of insulin, allowed us to minimize painful injections, and allow our patients to give more discrete insulin administration in public. We now have hybrid closed-loop insulin pumps which combine with CGMs to provide automated insulin administration. And we know from numerous studies that these devices improve glycemic control, reduce hypoglycemia, and improve quality of life in our patients. The problem is that there is a huge divide or huge gap in who's able to access these technologies and gain those benefits of these technologies. The Search for Diabetes in Youth study is one of the first studies that mentioned or described some of these disparities. And they noted that pump users were typically white, had higher income, had a bachelor's degree or higher, and had private insurance. The Type 1 Diabetes Exchange looked a little bit more closely at the racial disparities. And they showed even when looking at patients with an income greater than 100,000, disparities among between black and white patients still persist despite no differences in self-monitoring blood glucose values. So we know over the past 15 to 20 years, there's been a large uptake in insulin pump use in the United States. And so in our study, our objective was to evaluate whether this increase in uptake in technology has resulted in improving or worsening disparities across the United States. So we used data from the Search for Diabetes in Youth study, which is a multi-center observational population-based study of youths with diabetes. They have sites across the United States and have two components in their study. A registry component, which surveils a nationally representative sample of about 6% of the US population less than 20 years old. And they reported incidents of diabetes in youth. And then they had a cohort component where they followed these patients or a subset of these patients to evaluate for diabetes complications, quality of care, barriers to care, et cetera. And so for our study, we did a serial cross-sectional analysis of all search participants with type 1 diabetes who had data on insulin pump use. And we looked over four periods. Period one was 2001 to five, 2006 to 10, 2011 to 15, and then 2016 to 19. And then we evaluated changes in pump use by race, ethnicity, education, household income, as well as health insurance. So we used descriptive statistics to describe the patient population in each time period. We used multivariable generalized estimating equations to estimate the probability of insulin pump use, accounting for repeated measures across time. And then we used multiple imputation to address missing data, assuming a missing at random process. And then we used interaction terms for time period and then our variable interest to assess for temporal trends. So in terms of our patient population, you can see the number of participants increased over time. And that reflect the enrollment of more patients over time in the study. And then similarly, age increased over time as we followed these patients and they aged across the study. We see that about 50% of the patients at all time points were female. This population was predominantly white, non-Hispanic, 75 to 80% at any time point. These patients had a household income mostly greater than $75,000, primarily public insurance and then a more educated population with a bachelor's degree or more. So when we looked at the adjusted odds for insulin pump groups across all years, we found that compared to non-Hispanic whites, those who were Hispanic, black or other races had significantly lower odds for insulin pump use and with those who were black having the lowest odds for use. And then when we look at household income compared to those with a household greater than 75,000, those in the lower income groups also had significantly lower odds for pump use with the lowest income group having the most disparity. And then those with high school degree or less or some college had lower odds than those who had a bachelor's degree or higher and then those with public insurance had lower odds compared to those with private insurance. So when we looked at pump use over time, there was a significant increase in pump use from 2001 to 15, from 32% to almost double in 2016 and 19 with a prevalence of almost 60%. And when we looked at pump use across our subgroups of interest, similar to what we saw in the adult population, there was an increase in pump use across all subgroups. But unfortunately, when we looked at the distribution, there was no difference. So in this chart, the bar graphs, or the column graphs, represent marginal probabilities for pump use across all groups, and then the table on the bottom shows our odd ratio compared to the reference group. So on this table we see, compared to white non-Hispanics, those who are black, Hispanic, or other races, have lower odds ratio for pump use at all time points, and across time. And when we look at the confidence intervals, you'll see that they overlap, and the p-value, which compares the distribution in period one and four, there is no difference. So overall, we see no difference in the distribution of pump use in period one than we do in period four. We see similar findings when we looked at parental education, when we looked at household income, and when we looked at health insurance. So overall, over the past two decades, there have been no changes in the disparities we see in insulin pump use. So some of the limitations include that the data was collected with self-reported data through questionnaires. We were unable to capture changes that occurred in pump use in between study visits. And then this is not a nationally representative population. As you saw in our demographics, most of these patients were white, of higher income, higher education, and had private insurance. And so we think the findings that we see in this study likely underrepresent the disparities that exist in the general population. So briefly, what may be contributing to these disparities? Unfortunately, there hasn't been a lot of studies exploring why these disparities exist. But shout out to Dr. Shirvani Agarwal and Shana Mencher who have started to do some studies looking at why these disparities exist. And so similar to the framework that Dr. Park mentioned, we can think of these disparities at patient barriers, so factors such as concerns about visibility with a pump, provider factors, which can encompass bias or lack of shared decision making, and systemic barriers such as health policy, insurance policy, hospital policies. And so I have several ongoing studies trying to explore some of these barriers. So with the Search for Diabetes and Youth study, some of our next steps we're looking at, the role of financial burden on technology use, the role of social determinants on technology adoption. With some of my collaborators at Hopkins, where we just submitted a paper that looked at differences in prescriber recommendations and prescriber prescribing and use of technology based on race. And then I have a K23 grant to look at technology use in understudied type one populations. And in the study we're doing a RCT using hybrid closed loop insulin pumps in patients with A1C greater than nine, and including a safety net population. And then the last study I'll mention is a provider survey we're doing, we're trying to launch nationally to capture provider perspectives on their prescribing practices. So in summary, diabetes technology, and specifically insulin pumps, are a powerful diabetes tool. And despite we seeing a large increase in overall uptake over the past two decades, the disparities have still persisted. And so we really need additional studies to delve a little deeper into why these disparities exist, understanding the barriers these patients are experiencing so we can test interventions and address these disparities moving forward. And so I just wanted to give a thank you to my collaborators at SEARCH, my UCLA mentors, as well as my funding. And I am happy to take any questions. Thank you. Okay, while we're waiting for someone to be brave and come up in the audience, all right, I will let you start. Go right ahead. Go right ahead. Thanks. Stuart Shalhoub from New Orleans. I was interested in if you have combed down into the data and see whether pump use is actually associated with improvement in glycemic outcomes, specifically time and range, A1C, because an earlier paper from your group seemed to suggest that over these time periods, African Americans still have higher A1Cs compared to Hispanics and white. So there have been several studies in SEARCH as well as from other data sources that do show those who are on insulin pumps do have better A1Cs. But I know there's data from the Type 1 Diabetes Exchange that demonstrated, so for example, although, so a black person who's on the pump versus not the one on the pump may have better glycemic control, but it still may be higher than other racial groups or compared to non-Hispanic whites. I'll go ahead and ask a question as well. So I guess, kind of putting your two talks together, honestly, it was very interesting that we had the similar stories. But I guess one question would be a bit about, and maybe this is something you're looking into as well, but is there decreased awareness, perhaps, of the technology available in these patient populations? And also, I'm wondering about just the simple act of offering this to a patient and whether it's even, if there's a way to track whether that's even been offered versus waiting for the patient to ask for it. That's two excellent questions. So the first one, well, actually, I guess let's start with, I guess, the second question about tracking recommendations. So the study that I mentioned that we'd recently submitted for publication that I did with some collaborators at Hopkins, we looked at all patients, and we did a chart review of over 1,000 patients to look at were there differences in recommendation for insulin pumps? And then for those who were recommended, were there a difference in prescribing? And then acceptance. And so we found that when they compared black and white patients, black patients were less likely to be recommended to get insulin pumps, even after adjusting for several socioeconomic factors. And then although when prescribed the pumps, black patients were less likely to refuse. So if the provider decided to prescribe it, they were less likely to refuse pump therapy. And then your first question. It was about, on the patient side, like just awareness of technology. Oh, awareness of technology. And asking for it. Yeah, that's a good. So Dr. Mencher at Yale recently published a study, a qualitative study that looked at, that surveyed patients and their parents. And one of the barriers, the participants were asked, well, why do you think these barriers exist? And one of the answers included, like many of the patients did not know anyone else with type 1 diabetes. They really didn't have a network or a support system. And so they didn't know anyone with type 1 diabetes. Many reported not even knowing that this technology existed or their providers never even brought it to their attention. And so I think the combination of lack of familiarity with these devices, and then maybe reduced discussion surrounding them, may be contributing as well. Great. All right, thank you very much for a great presentation. Thank you. Okay, and we will move on to our next speaker. This is Dr. Uchana Nwosu. Dr. Nwosu is a chief resident with Kaiser Permanente in Northern California. And the title of his presentation is, Ethnic Differences in Albuminuria Among Adults with Diabetes and Normal Range at GFR. Thank you. Thank you. Good morning, everyone. My name's Uchana Nwosu. Thank you for the lovely introduction. So happy here to present my research. It was kind of quite important findings that we discovered when we were doing this. And it was surprising for all of us. And so we'll get into it. So looking at these ethnic differences. Disclosure. So just starting out, diabetic kidney disease develops in about 40% of adults. And in fact, it's up from about 18% since the 1980s. It's the leading cause of chronic kidney disease worldwide. Current guidelines recommends annual screening yearly with urine albumin creatinine ratio as well as EGFR. In fact, I put this graphic to your right to just remind us that when we think about chronic kidney disease, oftentimes we think of EGFR in particular. But we have to remember that the degree of albuminuria matters as well. And so if you look with people with EGFRs greater than 60, you can also see A2, A3, they also have chronic kidney disease as well. And albuminuria is the first sign of CKD and is also associated with complications like CKD progression, end-stage renal disease, and other cardiovascular events as well. So when we're thinking about that, we know a lot of data has showed that compared to non-Hispanic whites in particular, when we're thinking about lower EGFR, progression to end-stage renal disease, as well as the degree of albuminuria, that a lot of ethnic and racial groups are actually disproportionately affected. And we have a lot of data that has shown us that. But when we think about kind of early kidney disease, when the EGFR is still what we're calling preserved or greater than 60, that we have a lot less data in that area. So what we did for our study is we looked at the prevalence of albuminuria and normal range EGFR in folks with diabetes and a really diverse group of participants. So our method, we did a retrospective cohort study. It was conducted at Kaiser Permanente and multiple centers in the Northern California area. We have about greater than 90% screening rate annually for our patients with diabetes. The inclusion criteria included ages 45 to 74, diabetes prior to 2015, and we want the onset to be before so we could calculate the duration of diabetes actually. And we used multiple ways to figure out the diagnosis of diabetes, specifically looking at ICD codes, we were looking at labs as well as if they had treatment that's consistent with diabetes as well. We also wanted them to have a urine albumin creatinine ratio in 2015. And if they didn't have that, we looked at protein creatinine ratio and then lastly, protein dipstick. And then to calculate the EGFR based off this creatinine as well as the CKD epi equation without race, we thought that was really important to us as well. So our outcomes, really we were looking at any degree of albuminuria. We did classify it as kind of micro and macro albuminuria as well. Our predictors were kind of race and ethnicity as well as Asian ethnic subgroups. And then the covariates that we considered were age and sex, neighborhood deprivation index, which is really a kind of a rough approximation of socioeconomic status for the participants that we had as well. We looked at diabetes duration and then the hemoglobin A1C as well and then a diagnosis of hypertension which was based off ICD codes. And then we used a modified poison regression to really look at the dissociation in terms of ethnicity and albuminuria. So our results, we had about 79,000 participants with diabetes, what we're calling preserved EGFR and urine albumin creatinine ratio. About 96% of our participants had an albumin creatinine ratio. We had about 1% that had a protein creatinine ratio and as you can see, 3% with the urine protein dipstick. Mean age was around 61 plus or minus eight years and then it was predominantly male with about 70% of the participants having the diagnosis of hypertension. And when we looked at our breakdown in terms of the demographics, as you can see, non-Hispanic white make up about 40%, Asian Pacific Islanders, 26% followed by Hispanic Latinx and then lastly, those who identified as black. One important point that I thought was interesting here is that this kind of demographic breakdown mapped extremely well to the actual demographic breakdown in the Northern California area when you compare it to the census at the time as well. So we were pretty happy about that. So just looking at the prevalence prior to adjusting for anything in particular, as you can see, amongst the group, we had about 18.3% of folks who had some degree of albinuria. And then when we look at the race and ethnicity groups, as you can see, Asian Pacific Islanders had the highest burden before adjusting at 21.1%, followed by those who identified as black, then Hispanic and then lastly, non-Hispanic whites. Then we take a look at our force plot. This was after adjusting for kind of all these covariates that I mentioned prior. As we can see, when we're using kind of non-Hispanic whites as the reference group, Asian Pacific Islanders had a 1.3 times higher association with the development of albinuria with all the other racial and ethnic groups having about the same. And then we further go and we break down kind of Asian Pacific Islanders to the different subgroups. We can see that the prevalence is about 21.1% and with the highest burden being amongst Filipinos as well as those who identify as Native Hawaiian and Pacific Islanders at 25.3 and 27.9%, with the lowest burden being amongst South Asians at 14.7%. And again, when we adjust with those same covariates that we adjusted for previously, and this was interesting finding for us that Native Hawaiians as well as Filipinos had a 1.3 and 1.2 times higher risk of developing albinuria, whereas South Asian had about 20% less association with albinuria when compared to Chinese as the reference group. And the study did have some limitations. We didn't account for ACE and ARBs use, and that's something that's really important because it may have masked the degree of protinuria or albinuria that we're seeing in some of our patients, and that's something that we're planning to do kind of on further studies. When it comes to hypertension, our question really wasn't the severity of the duration, we really looked at was the diagnosis of hypertension present or not, and we didn't have that data, it wasn't available. And then we questioned whether it's generalizable given that it's a Kaiser Northern California population, which is a pretty integrated system, and these systems don't necessarily reflect a lot of the systems around the country, so there was that question of whether or not it's generalizable as well. Our strength, this was a really large cohort. We had 79,000 patients. We have a really high rate of screening, greater than 90%, which we were happy, and it allowed us to have such robust data as well. And then like I mentioned before, it was really racially and ethnically diverse, and we were able to kind of use that to kind of come up with the conclusions that we did, so we're really happy with that. So just concluding, you know, just summarizing kind of the findings that we had. The risk in Albanian adults with diabetes and normal range EGFR were higher among Asian Pacific Islanders compared to non-Hispanic whites. Among Asian Pacific Islanders, the risk was highest among Filipinos and Native Hawaiians slash Pacific Islanders, and lower for South Asians when we compare them to Chinese adults. You know, and I think when we think about this research and concluding it, I think more research is needed to understand this observed increase in risk, but I think this data supports more that this move towards disaggregating groups, specifically Asian Pacific Islanders, as we're moving forward, and considering the risk of development of diabetic kidney disease is important. To understand those risks, though, we need to get some more information. So I want to say thank you. I want to say thank you to the kind of funding that I had as well, and thank you to all those that I collaborated with during this project. So I can take any questions. Thank you. Okay, we have some time for some questions. Hi, that was great. That was so interesting. Lots of questions, actually, but I'm curious about a couple of things. One is, what, so your studies clearly show that there's a lot of heterogeneity within these minority groups, and we've known that, you know, you can't lump all Asians together and all Hispanics, et cetera. I wonder about the representation of the different subgroups in your population. Do you have that information? I know it was a large group, but within those subgroups, what was the representation of Filipinos, for instance, in your study? You know, I don't have that data with me exactly, but I remember specifically in the Bay Area and the Northern California area, we have a pretty large population. So, you know, I don't want to approximate or kind of make up a number, but it was pretty high in terms of representation for all of the Asian subgroups because we have such large populations in our area, but I don't remember the numbers off the top of my head. Yeah, it might be worthwhile if you have, you know, a very small representation within the Asian group. If a small representation of Filipinos is there, then, you know, there's more variability. And then the other question for you is, why do you think, what's behind this diversity or these differences? That's a great question. We actually spent a lot of time talking about that. I think in medicine in particular, we like to say things are multifactorial, and I don't think that this is any different. Specifically, thinking about kind of constitutional things, is there a genetic or inheritance pattern that may be related to this? Maybe, but at the same time, I would never underestimate the effects of kind of different social determinants of health, in particular, and what roles those may be playing. I don't think we can necessarily, or we haven't necessarily, just for all of those in our covariates, and that's pretty difficult to do. So, I think a lot of things are playing a role, and it's, like I said, multifactorial. But thank you. Gaurav from California. I wonder if you did correlate that with weight as well, body mass index, because that would indicate the lengths of insulin-resistant, and prior to the detection of their diabetes. Thank you. That is a wonderful point. We know that there's a correlation or association between specifically BMI and the degree of, or the risk of proteinuria slash albinuria as well. And so, as we're moving forward in our studies, not only are we gonna include ACE and ARBs, we're also gonna be including BMI also. Hi, and along that, obviously, you're looking at a time period with rapid adoption of SGLT2 inhibitors, and we know the impact they have on proteinuria, and is that something that you're also going to include in the next round, just to make your life more complicated? Yeah. So, in 2015, SGLT2 inhibitors weren't being used as much or as often, so we honestly didn't include that in this. We can look at that as well. We're definitely looking at ACEs and ARBs, and we thought about looking at SGLT2 inhibitors, but the use at that time was pretty low. But we were gonna revisit those numbers just to make sure we weren't missing those in terms of masking the degree of proteinuria that some groups may have over others. I'll ask one last question, I guess. So, what data is available, or are you able to kind of expand on this study to look at, you're looking at early preserved renal function, albuminuria, and is there data with these different subpopulations of the Asian Americans in progression of kidney disease or chronic kidney disease itself? Yeah, that's a good question as well. When we were doing our kind of literature review, what we found was there was one smaller study. It was a nephrology journal from the Asian Pacific Society of Nephrology, and so they did a study, not in America, but in Singapore. And what they found was that Malaysia, which is kind of Southeast Asian, they actually had an increased risk with kind of early preserved renal function and the progression of proteinuria is what they were looking at, which was different than what we found in terms of our Southeast Asian population. But what was similar is that they found that Southeast Asians also had a reduced association, which was similar to what we found as well. And that was from 2015. Great. Lots more to be done, it sounds like. Oh, one more question, sure, we have a few minutes. No, for years we've sort of known that the various families of hypertensives have different bases of effectiveness varying with the race. And it used to be published and pushed, and now I think since we're not supposed to notice what color our patient's skin is, they aren't pushing it or for whatever reason. But that may well be once the person is on a hypertensive drug, then you get caught up in the blood sugars and all the other things and don't make sure that they're, or may not make sure that they're well controlled. So we maybe should start publishing more about which blood pressure medicines work for which ethnic groups, even if you don't notice what ethnic group your patient belongs to, or racial group. Anyway, I wonder how big that plays in your data. Yeah, you know, I think as I mentioned before, we really just looked at whether the diagnosis was present or not. Didn't look at the severity of hypertension or the duration of hypertension, which may have played a role. You know, whether or not that would have, you know, there would have been differences seen by the different ethnic and racial populations that we had or the subgroup populations, I think would remain to be seen. I don't want to assume that there would have been differences, but thank you. I just want to respond to that. You know, we used to think that ethnic racial groups had different biology, and now that we have so much more genetic information, we realize that these are social constructs. They're not, racial ethnic groups do not define biologically different people, and so we've realized that there are no hypertensive meds that are better for blacks versus other groups, et cetera. Okay, well thank you very much, wonderful presentation. Thank you, Dr. Nwosu. All right, our next speaker is Dr. Juan Frias. He is a Medical Director and Principal Investigator at the National Research Institute for Velocity Clinical Research in Los Angeles. And the title of his presentation is Efficacy and Safety of Terzapatide versus Semaglutide in Hispanic or Latino Population, Prespecified Subgroup Analysis of SURPASS-2. All right, well thank you very much. It's a pleasure to be here. So on behalf of all of the authors, as well as investigators in this important study, pleasure to present this pre-specified analysis. SURPASS-2 was one of the five pivotal clinical trials assessing terzapatide in type 2 diabetes leading to its approval actually last month for the treatment of diabetes. And it's the one trial that has an active comparator which is a selective GLP-1 receptor agonist, semaglutide. So in this analysis, we looked at our patients in this trial that were Latino or Hispanic, and the patients who were non-Latino, non-Hispanic, and looked specifically at safety, as well as efficacy in A1c reduction, as well as body weight. So just by way of background, terzapatide is a dual agonist of GIP and GLP-1. It's a unimolecular peptide. So it's a single molecule. It's actually based on the peptide sequence of GIP, and it's engineered to bind to both receptors. And as I mentioned, it was approved for the use in type 2 diabetes last month. And the phase three clinical trial program included five randomized control trials. And these trials really spanned the spectrum of type 2 diabetes from surpass one, being terzapatide as monotherapy, all the way to surpass five, which was terzapatide added to basal insulin. And there were two, three studies actually that had active comparators. One has surpassed two that we'll be discussing, which compared again to the selective GLP-1 receptor agonist, semaglutide. But surpassed three compared to titrated insulin deglidex, surpassed four to titrated insulin glargine, and then there were two placebo control trials as well. And in all of these trials, we saw significantly greater reduction in A1C and significantly greater reduction in body weight as well, either compared to the active comparator or to placebo. With results I think that were quite surprising when we did the phase two studies. In phase three, we saw up to 50% of patients from an A1C of anywhere between eight and eight and a half achieve normal glycemia, so an A1C of less than 5.7, with very robust weight reduction as well, with up to 60% of patients in some of these studies at the highest dose achieving weight loss of relative weight reduction of greater than 10%. And these were not weight loss studies. So moving on, so the objective of this analysis, as I mentioned, was to look at our Latino-Hispanic group as well as the non-Latinos, non-Hispanics, particularly with respect to A1C and body weight reduction. This looks at the trial design of surpassed two. So these were all patients, adults with type two diabetes, poorly controlled at baseline, all on metformin monotherapy. And in fact, the baseline hemoglobin A1C for these patients was, and the mean was 8.3%. And on your right, you can see the study design. And the design of all the surpassed trials was fairly similar. So there were three doses of trizeptide that were studied, the five, 10, and 15 milligram dose. And in this case, compared to some of glutide at the one milligram dose. At the time of this trial, that was the highest dose for patients with type two diabetes. And all patients were started on 2.5 milligrams of trizeptide. It's a subcutaneous injection administered once a week. And then they were escalated by 2.5 milligrams every four weeks until the randomized dose was reached. So the five milligram dose was reached in four weeks, the 10 milligram in 12 weeks, and the 15 milligram dose in 20 weeks. And then some of glutide was administered per the label. So 0.25 milligrams for four weeks, escalating to 0.5 milligrams for four weeks, and then the randomized dose of one milligram. And this was a, it was an open label trial, but we were, as investigators, we were blinded to the trizeptide dose. If you look at the countries, there were eight countries that were involved, U.S., Argentina, Australia, Brazil, Canada, Israel, Mexico, and UK. I'd like to make the point that about a third of the patients, 34% actually came from Argentina, about 25% from the United States, about 20% from Mexico, and about 10% from Brazil. And that's where most of the Latino patients came from. As you'll see, about 70% of the patients in this trial were actually Latino patients, mostly from Mexico, Argentina, and the U.S., and about 30% were non-Latino. And this, the main study here was published in the New England Journal last year. So with respect to statistical analysis, the efficacy analysis, so our estimate for A1c and body weight reduction, this used the modified intent-to-treat population, so all patients who received at least one dose of either trizeptide or somaglutide. And this was the efficacy estimate, you'll see. So it was A1c or body weight while patients were still on study drug and without rescue therapy. The safety analysis included all of the modified intent-to-treat population throughout the treatment period and also the four-week safety follow-up. And then least square means were calculated using a mixed model with the fixed effect being the pooled country, baseline value, treatment group, visit, and treatment by visit interaction. For body weight, also stratified, the hemoglobin A1c less than or equal to 8.5%. And then the mixed model for the interactions included several terms for ethnicity. And here we see the baseline characteristics, which you see on your left are the Hispanic and Latino, and then on the right, the non-Hispanics. Looking at each of the three doses of trizeptide as well as somaglutide, fairly well matched, but I'll point out that the non-Hispanics tended to have lower baseline hemoglobin A1cs, closer to 8%, and higher in the Hispanics, 8.3%, 8.4%. And the Hispanics tended to be of lower body weight, so 90-ish or so kilogram on average versus 100 or so. Now there were, if you look closely at this, there were more females in the Hispanic group. About 55% of the Hispanic group were female. It was about 49% of the non-Hispanics. So the BMIs are a little closer, but still those patients were, and I think this may have, and we can discuss that, to do with the Argentinian population. But in any case, here we see on your left the change in hemoglobin A1c over time for the overall population. You can see from an A1c of 8.3, very robust reductions in all four of the study groups and significant reductions from baseline, certainly in hemoglobin A1c, with the 15 milligram dose getting to A1c on average of 5.8% and about 45% of those patients achieving an A1c of less than 5.7. And then you see very similar directionally anyway for the Hispanic population and the non-Hispanics as well. And then on your right are the changes in body weight. And again, on average these patients had a BMI for the entire population of about 34 kilograms per meter squared. But you can see again, very nice body weight reductions in all four groups. If we look now though at the estimated treatment difference between terzapatide and semaglutide, and what this is looking at is the two subgroups, Hispanics and non-Hispanics, for the three doses of semaglutide. So on the top, the five, I'm sorry, of terzapatide, the five, the 10, and the 15 milligram. And anything to your left favors terzapatide. So significantly greater reductions in hemoglobin A1c at the three doses and in both subgroups with terzapatide versus the one milligram dose of semaglutide. And you see the interaction p-value there. So there was no interaction between treatment and ethnicity when looking at hemoglobin A1c change. And a similar graphic now for weight change. So again, anything to your left favoring terzapatide. And here actually there was an interaction that was significant between treatment and ethnicity and change in body weight. If we look at safety and tolerability, as with selective GLP-1 receptor agonist, with the dual agonist terzapatide, GI side effects are the most common side effects. And interestingly, actually, we saw higher GI side effects in the non-Hispanics. So this is showing actually the Hispanics. You can see patients with greater than or equal to one treatment emergent adverse event. You can see about 55% of patients, anywhere from 55 to 60% in the Hispanics. In the non-Hispanics, about 80% of the patients having at least, reporting during the course of the trial, at least one treatment emergent adverse event. And you can see differences in nausea. For example, in the 15 milligram group, 16% of the Hispanic patients reporting at least once having had nausea during the trial, compared with 36% of the non-Hispanics. And I don't have it highlighted, but I think certainly in these trials, vomiting is a very objective adverse event that generally gets reported. And you can see anywhere from four to 7% of patients reported vomiting in the Hispanic group. And if you look down, about 10 to maybe 16%. And we haven't really gotten to the bottom of this, but I thought it was a very interesting finding. Very low incidence in both of the subgroups with respect to hypoglycemia, as you would imagine, given the mechanism of action of terzapotide in combination with metformin. And there were only three in the entire study episodes of severe hypoglycemia, which were treated with carbohydrates, and the patients did not. They continued in the study. So to summarize, terzapotide demonstrated greater reductions in hemoglobin A1c, as well as body weight, compared to the selective GLP-1 receptor agonist, semaglutide. There was no interaction effect between the treatment and ethnicity for A1c. And the interaction that was seen in body weight with ethnicity could be due to the lower body weight and lower BMI in the Hispanics. And also there was a difference, particularly in the 15 milligram dose, with less efficacy in the Hispanic group with respect to body weight than the non-Hispanics. And in both groups, gastrointestinal side effects were the most common. I would say that as with the selective GLP-1 receptor agonist when we saw these, most were mild to moderate in severity and tended to occur during the dose escalation period and then dissipate over time. And relatively few patients discontinued study drug due to GI side effects. So consistent with the primary results of SURPASS-2, what we showed that terzapotide is effective in reducing glycemic control or improving glycemic control, reducing body weight compared to the selective GLP-1 receptor agonist semaglutide in type 2 diabetes. Thank you very much. Okay. We're open for questions. It looks like Dr. Park is going to kick us off. Go ahead. Yes. So, maybe you mentioned this briefly, but I was just curious what prompted the question to look at the Latino population in the first place, if there are any known, like, a reason why, like, it doesn't work as much, and the other question is a minor point, but if you're comparing the Latino or Hispanic, shouldn't the comparator be not Latino and not Hispanic? Yeah, that was what was done. I mean, we really weren't necessarily comparing the two. We were looking at the effect in the Hispanic Latino population in this trial and looking at the effect of the non-Latino, non-Hispanic in this trial and comparing terzapatite in those subpopulations to semaglutide, so we're not doing here a direct comparison between the two. Now, with respect, there isn't really a reason to believe that it would have been more or less effective in one population or another. I mean, I think that's a really good question, but since type 2 diabetes is so common in this population, I think it is important at least to explore it and to make sure that these differences do hold up in the Latino as well and different populations as well, but there was nothing specific leading to this. It's been seen certainly with the selective GLP-1 receptor agonists or many other medications in there. Okay. All right. Thank you. Mm-hmm. Hi. Allison Alvear, University of Minnesota, and I had the same first question, but my second question was, was the method to, is it, in all, throughout all of the population centers in all of the countries, was the method to identify who was considered Hispanic and Latino the same? Yeah, it was the same throughout, but I think, you know, someone mentioned before about heterogeneity in the Asian population, and there's certainly a lot of heterogeneity also in the Hispanic population. So, I mean, particularly, and I think, I mean, as we look at the sort of lower body weight, for example, in Hispanics versus the non-Hispanics, I think had this study been done solely in the U.S., we probably would have seen higher body weight in the Hispanics, potentially. We haven't looked specifically at just the U.S., the 25 percent of patients who were in the U.S., but I think having 33 percent of patients in Argentina, where most of these are very European, Hispanics may have made a difference in the body weight. I'm just speculating here, but we're going to be looking at this further to see. In the UK PDS and some other studies, they found that different agents have differential effects on A1C and glucose. Did you check and see where the impact was here? Are you having more of an effect on A1C, you know, a different mechanism, and it's not all through glucose change? Yeah. I've not looked at that in these two subgroups. We certainly looked. We did not look at continuous glucose monitoring. For example, in this study, there was a study, surpassed three, that had a subgroup looking at continuous glucose monitoring. We did do seven-point glucose profiles, and what you see there is lowering both the fasting and postprandial, but whether there's a difference between these two subgroups, don't know at this point. With your agent in general, compared to, say, insulin or other agents? I don't understand that question. Does your agent have a bigger effect on A1C than it does on glucose? You know what I mean? Right, right, right. You could have an effect on A1C and not really affect glucose very much. I don't think so. This will be based on the surpassed three, and again, the CGM sub-study. When you look at sort of the estimated A1C based on the mean glucose of CGM, from my understanding, it matches up very nicely. The mean glucose is going down. Exactly. Right. Yeah. Okay. Okay. I think in the interest of time, we'll probably have to leave it there. Thank you. Thank you so much, Dr. Frias. I know you have a flight to catch, so I'm going to get you out of here. All right. Our next speaker is Dr. Hafiz Shaka. Dr. Shaka is the Chief Resident of Internal Medicine at the John H. Stroger Junior Hospital of Cook County in Chicago, Illinois. And the title of his talk is Widening Economic Disparity in Rates of Readmissions Following Hospitalization for Hyperglycemic Emergencies Among U.S. Adults Over a Decade. Okay. Dr. Shaka, come on up. Yeah. Okay. Here we go. This is your pointer here. And just in case, for your background, if you need a laser pointer, it's not going to get there. Okay. Hi. Good morning, everyone. So, this screen is a bit awkward for me. Yesterday was the International Day Celebrating People with Albinism, of which I'm one of them. We literally have very low vision. So, apologies if I have to squint a little bit to view this. So, my topic today is going to be on widening economic disparities in rates and readmissions, basically in the U.S., among patients who present with diabetic emergencies. Of note, we did this from the National or Nationwide Readmission Database of the United States. I'm the presenter. My co-authors are... Sorry, I didn't list them here. So, I have no disclosures nor conflict of interest during this particular study. So, as an introduction, hyperglycemic emergencies, including diabetic ketoacidosis and hyperglycemic hyperosmolar state, they are very common amongst adults with diabetes. Some may be the initial presentation, especially among patients with type 1 diabetes. They may initially present with diabetic ketoacidosis. The hyperglycemic hyperosmolar state is way more common amongst patients with type 2 diabetes, and you could have variable presentation, but it's usually a marker of severe insulin resistance. Now, the household income is a well-known determinant of health, especially when you map it on different zip codes in the U.S. There have been studies that have shown that you could have significant shifts in life expectancy, even within the same county, just because of the mean income related to that zip code. So, based on this index, we wanted to assess to see if this factor had an impact on patients who were hospitalized and then readmitted. Readmission is one of the main metrics by which the Center for Medicare and Medicaid assesses a lot of hospitals, and there have been concerted efforts recently to sort of decrease readmissions across board, to the extent of penalizing centers if your readmission rate is high. Interestingly, Medicaid currently tracks about five or so diagnoses, including heart failure, but diabetes and its complications isn't one of the main metrics which is being tracked. So our study here wanted to find out the trends in readmissions based on the mean household income across the past decade. This decade actually coincided with when the Center for Medicare and Medicaid started the Hospitals Readmissions Reductions Program, the HRRP. So we used the longitudinal trend analysis using the Nationwide Readmission Database. It's the largest all-payer database in the United States that helps to track both index admissions and readmissions for patients within the US. Over a year, I think it tracks over 30 million plus inpatient visits. The study was from 2010 to 2019 and involved adults with DKA, mainly inpatients with type 1 diabetes, and adults with HHS, mainly inpatients with type 2 diabetes, as the principal discharge diagnosis. We excluded patients who were admitted electively for any reason. And then we had to exclude December hospitalizations because during the period, if you want to track 30-day readmissions, the December period might cause a swing. And the database doesn't actually translate year on year as a way of protecting the identification of this large database. There's a variable in the NRD that helps you to link patients and link visits. So that enables you to know if a patient gets readmitted within the hospital. And that was one of the metrics we used. And there's also a variable that tracks the median household income for the patient's zip code. Now, outcomes that were assessed were the trends in 30-day all-course readmissions and the emergency-specific readmission rate. In this case, either 30-day readmission, again, for an episode of diabetic ketoacidosis or for another episode of hyperglycemic, hyperosmolar states. We also did a trend in the mortality and length of stay over the 10-year period. We calculated the annual percentage change adjusted for basically the patient's sex. So this graph shows the results of our study. From 2010 to 2019, there's been a considerable and increased trend towards increasing readmissions in all populations. But this was more significant amongst patients in the low-income quarter. As you can see, amongst patients with DKA hospitalizations, patients in the lower quarter had a higher 30-day all-course readmissions compared to patients in the low-income quarter. This is shown in the highest graph present that is colored in red. And you can see that there's a sustained increased trend over the study period. Now, we also assessed the DKA-specific readmission rate. And again, there was a significant increase over the years. But also, this metric was found to be significantly higher amongst patients in the low-income quarter relative to patients in the high-income quarter. Now, the second figure here shows a similar, well, not quite similar, but it still shows a trend towards increasing readmission rates for patients who were admitted for hyperglycemic hyperosmolar states. Initially in the decade, there was a lower rate amongst patients in the high-income quarter. However, you can appreciate that that is sort of flattened out while patients in the low-income quarter have a sustained increase in their all-course readmission rates. The HHS-specific readmission rates were kind of flat with no trend over the decade. However, the high-income patients sustainably had a lower readmission rate compared to patients with the low-income quarter. Now, in conclusion, there seemed to be a rise in readmission rates following hyperglycemic emergencies. And this is a particularly worrying trend. It's like a marker towards failed management, especially inpatient. But we know there's a significant overlap between the inpatient and the outpatient management of diabetes and its complication. And majority of these admissions could also be attributed to failure of proper outpatient management. And with the differences that are occurring due to the income, it calls into question about health disparities in terms of access to health care, in terms of access to medications. Like, you could have the newest medications that can significantly lower your HbA1c or control your glucose level. But if these are not accessible to the poorest people, we're still going to be incurring significant health care costs in the management of this patient. Thank you. Thank you very much. We have time for a few questions, if anyone would like to come on up. Hi, I'm Amanda Sheehan. I'm a nurse practitioner at Joslin in Boston. And I was thinking about when I wrote Tate through the hospital and we covered the ED. And as you know, often our underserved population most often seeks care through the ED. And I'm wondering if there was any differentiation between the readmission rate. Like, was there any stratification for, were they discharged after OBs in the ED? Or was it just they were admitted and then discharged? Did that account as a readmission? Because sometimes these patients come into the ED, they can't get admitted. There's not enough beds. We'll monitor them in OBs. And then they're sort of discharged from there. And I find that the discharge process is much more rushed, even if they're just sort of there for a 24-hour period. And I often wonder about those patients because if they're admitted, they've had more time for discharge teaching, maybe we provided more recommendations for additional medications for follow-up. But the ED is just a very different animal. So I'm wondering if you could speak to that. All right, thank you for the question. The advantage here is those patients who are likely not included in this study because this study was mainly for patients who have attained inpatient status, which is typically the case in patients who have these emergencies, like diabetic ketoacidosis. Most of them will be managed in the intensive care setting. So it's highly unlikely that a significant proportion of the included patients here were discharged at the emergency department. I'll just say, I feel like since things have been so busy since COVID, sometimes we were having to manage people in the ED. If it was like a mild DKA, it was for a while, and then especially with some atypical forms of diabetes. But anyway, thank you so much. Thank you. Hi, yeah, I personally found it sad for one of my patients with who keeps coming in with DKA, had to prescribe insulin as an indigent medication. And he sold the insulin for cocaine, and he came back. So that was interesting. So have you looked at how many of the patients actually got a new insulin prescription at discharge? So unfortunately, the database, that was one of the limitations. We cannot assess the medications these patients were discharged on. But it still boils down to the question of, you have that significantly different population, the high versus the low income. So low income patients have also been associated with lower health care literacy, and also low access to their medication. So those might be contributing factors towards readmissions down the line. While we're waiting, I had one question too, which is, do you have any data about the length of hospital stay, and whether that potentially impacted the readmission rate, given maybe a push to quicker discharge may lead to more readmission down the road? Yes, we had that metric. And across board, a lot of studies have shown significant reductions in length of stay, actually, over the past decade. So we don't know if this is directly correlated to readmissions rate. But then it's something that across multiple conditions, and not necessarily just diabetic emergencies, including conditions like heart failure and stuff. Not most of these conditions haven't really shown a rise in the readmission rates over this period. Great. Thank you. Go ahead, and we can have time for one more question. I have a general question and a comment. I'm seeing that most of the studies boil down to economic level. And economic income, or the income of anybody, is dependent on many social factors, like level of education, like many other socials. And if we do these studies to find the results, and based on those, to solve the social problem of our patients that will return again and again. I have so many of them, they do the same thing. And the question is, maybe instead of the economic level, we should subdivide them in, for example, what about the education level? How many of them had high school education? How many of them were employed? I think those many, many other factors we can add, because that would help us to solve the problems of readmission better, because based on your tremendous, fantastic study, when we have the results, we can apply it to the society. Thank you. Thank you very much, and I agree with you. It just raises a question of, basically, primary prevention. So, if we can prevent or educate these patients to better control, take ownership of their management of their condition, that would definitely improve their control, and then their likelihood of getting admitted for this. There are different studies that still talk about health literacy, and we know that there's no perfect correlation with education, and then the adherence to medication, per se. So, you always run into different substratification. So, I think we can only do the best we can. Thank you. That's a great way to end that. All right, thank you very much, Dr. Charko. Thank you very much, everyone. Thank you. Okay, and our last speaker for the session is Tracy Zhu. Ms. Zhu is a public health researcher at the University of Nevada, involved in the ECHO study, and the title of her presentation is Combating Therapeutic Inertia. Project ECHO for Diabetes Improves Primary Care Providers' Comfort and Use of Diabetes Medication and Technology. Do we have her screen in presenter mode, or is that not possible? No, that's not possible. Okay. We're good. Hi, everyone. Thank you. My name is Tracy. I'm presenting on behalf of the co-authors. Again, this topic is about combating therapeutic inertia in Project ECHO for diabetes to improve primary care providers' comfort and use of diabetes medication and technology. So a little background about Project ECHO. It stands for Extension for Community Health Care Outcomes. It's a hub-spoken model in which specialists tele-mentoring primary care providers. Project ECHO also combines short diabetes didactics and patient case discussions, of course, the identified, to connect providers with specialists, especially in the medical and deserved communities. Project ECHO was established in 2003 at the University of New Mexico, and George Washington University also launched its first ECHO back in 2018. And then University of Washington Diabetes Institute replicated this model using this curriculum in 2020. Again, the University of New Mexico was the first ECHO program that indicated knowledge and confidence improvement, and our data also indicated the same results. The 547 intervention patients showing their recent studies with data that decreased from 10.5% to 9.3% comparing the pre- and post-intervention periods. And also providers' ability increased from average amount peers competent to very competent, and the confidence level increased, and their knowledge of diabetes medication increased as well, showing by our survey using a seven-point Likert scale. And changes in prescribing practices. To our knowledge, we are the first program that investigated into the changes in providers' prescribing practices, clinical inertia, especially in the settings to use newer medications and diabetes technologies continues to be a barrier in terms of optimizing glycemic control. In a recent retro-respective study that showed GLP-1 use increased but remained low, especially in those with coronary artery disease and type 2 diabetes. So the use was even lower in Asian Black and Hispanic patients and those with low income. The method we used here involves three unique diabetes ECHO programs, which evaluated comfort or perception for providers in terms of prescribing practices changes among the local community primary care providers. The total amount of PCPs were 74 in Illinois, District of Columbia, and New Mexico and Washington. So we used the REDCap survey evaluating in a cohort comparing the pre- and post-surveys followed by post-session surveys using four-point and seven-point Likert scales. So in total, 45 reported pre- and post-surveys and another site which used post-session surveys, only 29 providers reported those kind of results. So here is our data in terms of GLP-1 and SGL-2 uses among primary care providers. Our data were significant, showing that both pre-prevention and the increase after the intervention, providers' confidence level definitely increased from sometimes to always. In terms of the CGM and basolin uses, we also see an increase in the provider outcomes, showing by the pre- and post-survey differences. And using the post-session evaluation, we see the GLP-1 result increased with amount as well as the CGL-2 amount and then followed by CGM and multiple dose insulin, basal insulin. Although the insulin pump data remained the same. So in terms of conclusion, the Project ECHO definitely highlights the mechanism to combat clinical inertia and increase providers' confidence in prescribing medications and using newer technologies. Of course, more research will be needed for patient-level outcomes and cross-evaluation at other ECHO sites using different cohorts. And then there is an ongoing analysis for patient-level outcomes for the National Capital Area ECHO Program. Our goal is to increase the endocrinologist's accessibility to increase patient-level outcome. And then participation in ECHO definitely associated with improvement in PCP likelihood of prescribing newer medication and using newer technologies. And with that, I'm open to questions. Thank you. I previously was internal medicine director of a small community clinic in Minneapolis, actually kind of large one with university, and we as providers in like from 2003 to 2008, so a while back, wanted to see how we did as far as were we meeting all the recommended cures, just getting regular A1Cs and things regardless of insure, and we found we did really poor even though we felt like we did good. So I gave us surveys, you know, before and every year of what we thought we were doing and what we actually were doing based on the data we could get from Medicare and from billing, and we always think we're doing better than we are, and we think we're ordering tests in a more socioeconomic and regardless of insure and status and regardless of race, et cetera, and we don't. We did really poorly and even with, you know, that feedback, we continued to do rather poorly. So I'd say you can get the prescribing data from, and that would be a more reliable method to see, you know, if people's prescribing practices actually did change. Thank you for that recommendation. Yeah, in a recent cohort that we did, we are monitoring the PA student and nurse practitioner students in terms of using the prescribing behaviors early on as early learners, and I do think you brought up a great point. If we can look into the actual prescribing data compared to just the perception or the self-efficacy increase, that would be another way to look at the changes of prescribing behaviors. Okay. Any other questions from the audience? All right. Well, thank you so much for a wonderful session. Everyone, thanks to all of our speakers.

albuminuria

kidney disease

diabetes

prevalence

ethnic differences

Asian Pacific Islanders

Black individuals

Hispanic/Latinx individuals

therapeutic inertia

diabetes management

primary care settings

SGLT2 inhibitors