I'm Priya Valenki, I'm a faculty at the Emory University School of Medicine and welcome to the guideline session on the new updated inpatient treatment of hospitalized patients with type 2 diabetes in the non-critical care setting. So it's been about 10 years since the last update and it's published online on the Endocrine Society website and it'll be published in Journal of Clinical Endocrinology and Metabolism in August. And on the panel with me are the panelists are part of the guideline team and with me are co-chairs Dr. Mary Kortkowski and Dr. Ranganath Muniappa who are sitting up here and also with us are Dr. Marie McDonald and Dr. Amy Donohue and the session today will initially be an introduction to the methodology that was used to come up with the guidelines and the second part of the session will be a summary of the new guidelines, what is new, what is updated and then we'll go through a bunch of cases during which we ask the audience to participate and after two cases, the first two cases we'll have a question and answer session and then we'll have three to five cases, cases three to five after which we will have question and answers. So I think again the moderators are myself and Dr. Freeby who's sitting in the audience and also the panel members and this is the full guideline development panel. Some of you who are in the room, sitting some of you in the front seat and these are all our disclosures so I'll give you a minute. And again today's the agenda, we'll go through the great methodology and an overview of the guidelines and the cases after which you will have time to ask questions. And this is the QR code so if you put your phone up to the QR code, you'll be taken to a website where you can put in questions or you can also type in the specific website up there. So I'll leave this on for about 30 seconds. Okay, I think everyone got a chance to scan the QR code. All right, so at first I'll introduce Dr. Marie McDonald. She's an associate professor at the Brigham and Women's Hospital at the Harvard Medical School, and she'll go through the grade methodology with you. Dr. McDonald. Thank you, Priya, and good morning, everybody. So it is really a great pleasure to be presenting to you today representing the Clinical Guidelines Committee for the Endocrine Society. I am the chair of that committee, and my task today is to provide you a brief overview, a very brief overview of the grade process. So many of you think of the grade process perhaps as a scoring system for the strength of recommendations, strength of evidence, and quality of data, but it actually is a very comprehensive process that allows the most unbiased approach to achieving a recommendation through a panel of experts. So walking you through that, essentially at its core, despite its complexity, at its core it's a very simple concept, which is that you start with a question, a question that is commonly asked in clinical practice, you then, that leads you to the evidence discovery, the discovery phase of the guideline process, which yields different levels of evidence, randomized trials, non-randomized trials, non-randomized clinical studies. Finally, leading to statements that include recommendations. Now this might be reminiscent of the scientific method, which is part of the concept, and the goal is to minimize bias. Now compare that to the alternative method, which has been used sort of traditionally in the past, and we're trying to move away from that idea of starting with recommendations. If you have experts in a room who've been treating collectively thousands of patients, there's merit to the concept of starting off by saying, I think we need to tell the readers this, and let's find the evidence to prove what we're trying to say. This method actually does introduce quite a bit of bias, and therefore we're switching it around, and starting instead with a clinical question. I think this will advance eventually. There we go. All right, so you'll notice when you open the guideline, I hope that you do, it is published online today, yesterday, you'll notice that there's a difference in the grade, this inaugural grade guideline compared to prior Endocrine Society guidelines. First you'll notice it's more concise, it's more transparent, it'll be clear how the experts arrived at a conclusion, and as a result of the methodology, it'll be easier to update over time, which is a priority that the CGC does take seriously, and we're working on all the time. Okay, so as I mentioned, grade starts with questions, and these questions are the clinical questions asked by both patients and providers that the experts bring to the table. So they're pertinent for decision making, they're not necessarily questions that have been commonly asked or answered by available or well-known research, especially when basic science or translational science is incorporated. The idea is that you compare management options around the question being asked. The questions, we call them PICO questions, this is an acronym, and it stands for Population Intervention Comparator and Outcome. So for example, the population has to be specifically, as specifically as possible defined in order to answer the question reasonably. So sometimes it's broad, like adults with hyperglycemia in the hospital. That sounds specific, but it's actually quite broad if you think of all the different adults in the hospital for different reasons in receiving care. The intervention, we describe with, again, as much detail as possible, which we'll go over an example, but it could be a medication or an approach, as long as the approach is well-defined and tested, testable, and then a comparator is the alternative. So we have the intervention and a comparator, similar to how you would design a prospective trial. And finally, the outcomes. This is perhaps the most challenging part of the PICO question for the panel because it cannot be exhaustive. It has to be a list of prioritized outcomes that are most important for clinical practice. And sometimes these outcomes are not the outcomes that were chosen by those who designed the best randomized trial for a specific question, but the panel has decided is actually the most important clinical question. So I would say, and maybe Mary would agree, that I think the outcomes was the biggest challenge. Okay, let me give you an example. So this is PICO-2 in our guideline, management of inpatient hyperglycemia, specifically adult patients who are hospitalized for non-critical illness and experience hyperglycemia while receiving glucocorticoids. So the intervention for the PICO is the NPH-based insulin regimen. The comparison was a basal bolus insulin regimen using a long-acting insulin, for example, Glargine. Outcomes that the guideline panel just chose included mean hospital glucose, number of hypoglycemia events, and length of hospital stay. Now there were other outcomes that were uncovered along the way, but these were prioritized for the systematic review. Now at its core also, GRADE requires a systematic review for every PICO question, and this is very important. And I'll show you, you'll hear many examples of how that plays out in real time here. Now beyond the systematic review, we also consider other factors that are visible in the evidence documents that we receive, that again could be non-randomized clinical data, including randomized, and in addition to randomized clinical trials. And on balance, the panel is asked to decide, are there more desirable consequences or undesirable consequences for a specific question? So desirable consequences, if there are more on that side, that comes in favor of the intervention. On the flip side, if there are more undesirable consequences, it is in favor of alternative. Now most of the evidence, of course, for this balance comes from randomized trial data. This is obviously prioritized as the highest quality data. But the panel will consider all available evidence related to the outcomes, and will also consider factors such as feasibility, acceptability, and equity. So as a result, and we call this evidence to decision, this whole process. As a result, we end up with a recommendation that has taken us really on a journey. We've arrived at the recommendation after considering all of the factors on balance. So in the end, the recommendation will be either strong for or against the intervention versus the comparator. And strong recommendations in general is that it results from the determination that the evidence is irrefutable that the intervention, for example, the desirable effects or consequences outweigh the undesirable. And just to put it into perspective, the methodologists remind me that this means that in general, over 90% of the time, this recommendation should pertain to an individual patient. So you really need a lot of strong evidence to support that. Most of the time, in general, in endocrinology, we see conditional recommendations, which basically says most of the time, probably rather, which is the same concept, the desirable consequences, for example, probably outweigh the undesirable. And hence, that would form the recommendation. Now beyond the strength of the recommendation, we also look at quality of data. And I'll play this out through this recommendation 2.1. Our final REC is in adult patients who are hospitalized for non-critical illness and experience hyperglycemia while receiving glucocorticoids. We suggest glycemic management with either NPH-based insulin or basal bolus insulin regimens. Suggest is the verbage we use for conditional recommendations. And as you can see, the coding that follows, the two indicates the strength of the recommendation, two indicates conditional, one would be the highest quality of evidence. And then following that, or sorry, the strength of the recommendation number one would be the strongest recommendation, two is a conditional, rather. So in this case, you see a two. And then following the four icons there, you see two circles with pluses that indicates low quality of evidence on balance. If you saw one, it would be very low. Two is low, three is moderate, and four is high. So in this case, we have a conditional recommendation with low quality of evidence. And for that reason, we're suggesting either. Okay, I think that concludes my introduction to GRADE. I hope that helps. But what I'm gonna actually take a moment to do is to highlight what you've probably heard, which is the work and leadership involved in taking this panel through the GRADE process. This is the first time we did a complete GRADE process. And I think the two chairs that we have here, Mary and Raghunath, were already highly regarded in their field. But they have earned an enormous amount of new respect from their colleagues through this process. And we just want to take a moment to recognize the two of you from the Endocrine Society. Thank you. This is exciting for you. We want to thank you both. We wouldn't have been able to do this without you, so thank you very much. And Brea Hickman wants you to take it out of the box. And as we're doing that, since we're getting a photo op, I just want to acknowledge Andrea Hickman and Maureen Corrigan, who are Endocrine Society staff. Those of you who've taken part in guidelines before have learned how amazing they are and how critical they are to the success of the society in this endeavor. In addition, Rob Lash, Rob's around here somewhere, I think, who's our incredible CEO and just making, putting his stamp on everything in such an important way. Thank you, Mary and Ranganath. Okay, I next want to introduce Mary Korikowski. She's a professor at University of Pittsburgh School of Medicine in Pittsburgh. She is our illustrious chair. Marie, thank you. I have to say, that gift-giving was not part of the slide set I looked at this morning that said final slide set. So thank you. Thank you very much for the recognition. I'm actually very moved by it, and I want to second your statement about I was honored to be invited to chair this group in 2019. And if I didn't know Ranganath, we really didn't know each other before, but if I ever have to do another guideline, you're my co-chair, or maybe you're the chair this time. I think we can switch places. And also, though, we got to choose the people who worked on this panel. And it was truly a group effort with lots of discussion and lots of asides when we sort of found that, were we all working with the same set of definitions for each of the issues that we were addressing? And so the whole panel just worked together in a great way. So anyway, but now I'll go to my scripted part of the presentation this morning. So let's see. So you already know the name of this guideline and there were 11 of us that were involved. And it represented primarily adult endocrinology, but we also had a general internist who was a representative from the American College of Physicians. And we had a patient representative on this panel, which was really an excellent idea. I think that's something that the Endocrine Society is adding now to all their guidelines. And I have to say that our patient representative was really amazing at keeping it real and letting us know how patients looked at any of the data. We also had a group of methodologists. We had Dr. Hassan Murad from the Mayo with his group who actually performed all the systematic reviews and the systematic reviews are being published concurrently with the guideline itself in the JCE and MNR available for you to review to see how they selected the references and how they did the forest plots. And we also had Dr. Nancy Santeso who is at McMaster University and she really guided us through the grade process. This was new to many of us following this particular guideline and she really kept asking us the questions that sort of kept us on track when we wanted to sort of go into expert opinion, which we are more comfortable with than really the rigor of this process. But we learned, we all learned and we learned a lot from them. So just to review briefly with you some of the areas that we covered. We had to select 10 questions. We had more than 10 when we started but we had to narrow them down to what we thought were the most important. And so the 10 areas that we covered were the use of continuous glucose monitors in the hospital in conjunction with point-of-care glucose monitoring. We looked at glucocorticoid and enteral nutrition associated hyperglycemia. We looked at the ability of patients who use insulin pumps in the outpatient setting and we know there's a lot of them. Could they continue to safely use these while they're in the hospital setting? We looked at the importance of diabetes self-management education in the hospital. We know that most people consider this an outpatient initiative but most people are not getting it in the outpatient area and so we looked at the studies that have looked at inpatient diabetes self-management education and the effect on outcomes. We also looked at the whole issue about what the glycemic goal should be before someone undergoes elective surgical procedures. Let's see, I'll move this on. We investigated the practice of administering carbohydrate-containing oral fluids to patient before having surgical procedures. We looked at, we also sort of took on the question about what many of us in this room know as sliding-scale insulin and we keep trying to substitute the term correctional insulin but sliding-scale seems to stay with us. But is this ever okay to use? And if it is okay to use, where are the circumstances where it is okay to use this? And when scheduled insulin therapy is always recommended and when correctional insulin therapy could be recommended. And the whole issue which many of us have questions in our mind about the use of non-insulin therapies in the hospital and certainly performing these reviews let us know where we need more data. But the two areas that we're focused on here were the dipeptidyl peptidase inhibitors and the GLP-1 receptor agonist where there is data for us available to review. So with that review, what we'll do now is we're going to go through five cases for you. There's audience responses with some of them that you'll be able to vote on your answers but to just guide you through where we were with some of the recommendations that were made in this guideline. So just take a minute for those of you who didn't have a chance to scan the QR code. We'll just give everyone a chance to get your cameras ready and hopefully everyone's ready. Okay, so case one. This is a 59-year-old woman with type 2 diabetes, also with COPD and CKD who was admitted to the hospital with pneumonia for treatment with intravenous antibiotics. Her home diabetes medications include Metformin, 1,000 milligrams twice a day and Glymyperide, four milligrams once a day. In A1c, two months prior to admission was 7.6% so she did not need another one at time of admission. Her weight is 216 pounds. Her oxygen saturation at time of admission was 88%. Her blood glucose at time of admission was 185. A repeat blood sugar done four hours later was 193. She had a mild degree of CKD. So the question for you is, okay, this is a woman with pneumonia being admitted. She has type 2 diabetes treated with oral agents. So what would you recommend for diabetes management for this patient in the hospital? Would you stop her home medications and start correctional insulin dosing? Would you stop her home diabetes medications and start scheduled insulin dosing with the basal bolus insulin regimen? Would you continue her home diabetes medications and add correctional insulin to this? Or would you add sitagliptin with correctional insulin to her regimen? So take a few seconds to answer this question. I think I have to move this forward. There, it should be appearing on your phones now. Okay. So the majority of you are choosing stopping her home diabetes medications and starting her on basal bolus insulin, which is, or it could be basal insulin with a correctional insulin protocol depending on how much she's eating. The other questions, stopping her home diabetes medications and starting correctional insulin. She already has a blood glucose of greater than 180, which is what we really identified as the threshold for hyperglycemia requiring insulin therapy. Choice C would be to continue her home diabetes medications. We're really not sure if that's going to be safe. She has a mild degree of CKD. We don't know what her hospital course would be. And so we're not really sure that it's safe for her to be taking metformin. There's really no data one way or the other. And the use of a sulfonylurea in the hospital is also subject to a contributor to hypoglycemia that can actually be prolonged in some patients who are not eating regularly. And adding sitagliptin is certainly a consideration, but that is really reserved for patients with milder degrees of hyperglycemia who have persistent blood sugars less than 180. And just to guide you through where we were with this process now, I think, is this a pointer? Oh, okay, sorry. Okay, so these are the forest plots that were associated with this particular question. That here at the top we have hypoglycemia. And you can see here that if we look at hypoglycemia alone, it favors the use of correctional insulin alone compared to basal bolus insulin. But if we look at mean daily glucoses, it favors the use of basal bolus insulin where the risk of the mean daily glucose level was about 26 milligrams per deciliter less is sort of the mean data from all the studies that we looked at. When we read through the individual studies for this particular question, we found that some of the patients who were put on correction insulin alone actually had more severe degrees of hyperglycemia if correction insulin was used alone. So that was the underpinning for this recommendation. So when we looked at the question about correctional insulin versus scheduled insulin for someone with type 2 diabetes on non-insulin therapy prior to admission, our recommendation here was that initial therapy could be with correctional insulin alone or with scheduled insulin therapy. Either one would be correct. What's not on this slide that was part of this is once a blood sugar is confirmed as being above 180 milligrams per deciliter and admission blood glucose confirming it several hours later, which she was on 193 several hours later, that this was an indication to change from correctional insulin alone to scheduled insulin therapy. So, let me just advance this here. There we go. So, staying with the same patient, on day two of her hospitalization she remains on supplemental oxygen with antibiotics. She feels weak, she has a low appetite. Attempts to wean her from oxygen are unsuccessful. Her bedside glucoses on the regimen that's been prescribed are ranging between 132 and 194 milligrams per deciliter. And a decision is made to start her on glucocorticoids with prednisone 40 milligrams once a day. So, she has received prednisone in the past when she's had difficulty with her COPD and has experienced home blood sugars greater than 300 when she's received these before. So, what do you recommend at this point? Do you increase the intensity of her correctional insulin dosing from a low to a moderate scale? Do you increase all current doses of scheduled insulin therapy by 20%? Do you add NPH to her current insulin regimen timed to the administration of prednisone? Or do you add a glucagon-like peptide one receptor agonist? And this does not have, this particular question does not have an audience response. Oh, it does? It does? Let's see. Oh, okay, no, it does not. So, we're telling you the answer here. I could have asked for a show of hands. Here we felt that either was okay. I suspect that she would need more than a 20% increase in her dose of scheduled insulin therapy along with an increase in her correction insulin scale. But she could also have a dose of NPH added to whatever insulin regimen she's also on timed to the administration of the prednisone dosing because the pharmacokinetics of prednisone are very similar to those associated with NPH insulin. Here, when we looked at the data, what we found is that the data were sort of on balance. You'll see one of our forest plots here. But we said that either was okay, but the important thing would be to be proactive in adjusting her insulin doses to prevent her from getting severe hyperglycemia, which could interfere with her ability to recover from the pneumonia. What we found with looking at this data is that there was wide variability in how and when NPH was dosed. There was wide variability in the doses of glucocorticoid used. So even though on this forest plot you're seeing that the mean daily blood glucose was about 40 milligrams per deciliter in the NPH-based group compared to the basal bolus insulin group, the quality of the evidence was in fact low. And so we decided, made this a conditional recommendation for either the intervention or the comparator. So that's question one. So at this point, I will turn this over to the co-chair for this committee, Dr. Muniyappa, who will take us through the second question. Okay. Question, case two. So this case is one of our PICOs. We spent a lot of time discussing this particular case. So the case goes as such. A 75-year-old man with type two diabetes is seen by an orthopedic surgeon for severe OA, and he has limited ability to exercise. So for this reason, he's been recommended a knee replacement. And the surgeon tells him, which is a common occurrence, right? They'll, you know, you need to get your A1c below 7%. Anyway, his current diabetes medications include glargine, metformin, and cetagliptin. You can see the physical exam there. And his most recent A1c was 7.7. In the past, it has ranged from 7.3 to 8.1, and his home blood glucose monitoring reveals a range between 108 to 192, with occasional episodes of mild hypoglycemia. So that's the case scenario. And here is the question. The patient is referred to us to improve his glycemic control, right? The surgeon tells us, get the A1c to below 7. So what do you advise? So there you see the options. Have a discussion with the surgeon and the patient that an A1c of eight is acceptable for moving forward with the surgical procedure. Add pre-meal insulin, increase the dose of basal insulin by 20%, or add a sulfonylurea. So those are the choices. Give a few seconds to Malo over it. You all were on the guideline panel, I'm sure it's less than eight, right? So that helped. There won't be too many questions on that, so, okay. Thank you. We will address the 16 percent, the 22 percent, I think that's important too, in the question and answer session. Okay. So here is the recommendation. Thank you for all recommending this, right? So the key point is when feasible, right? That's the key point. And as Dr. McDonald was mentioning in terms of the grade process, when we looked at the data, I mean, we all know preoperative glucose, whether it's immediate or past, in a month or just on the day of surgery, or postoperative glucose concentrations have a graded relationship with postoperative complications, right, be it wound infection, wound dehiscence, length of stay, which is actually a surrogate for all the adverse effects of hyperglycemia. What we found was that there was a dose effect. That means if you categorize the A1Cs to less than 6.5, less than seven, greater than eight, you always saw sort of a relationship suggesting there is some underlying causal mechanisms, although it's correlational, but nonetheless, I mean, clinically we can apply the first principles and say, hey, there is causal mechanisms. So the question always becomes a trouble in terms of the threshold. Which threshold do we pick, 8 percent or 7 percent? So as Dr. McDonnell mentioned, in the grade approach, especially when the panel discussed this issue, the key points were equity, cost-effectiveness, feasibility, and applicability, right? We all know to bring A1C to less than 7 percent can delay surgery, especially if your patients are from socioeconomically lower status, or they don't have immediate access to care, like surgery. So in that situation, we considered all of these factors and then put a threshold of less than, you know, less than 8 percent, greater than 8 percent. That's the threshold that we chose. So when feasible, that's what we want to achieve. So I want to make sure that that's how we came to this recommendation. So just to follow up on this particular case, the patient was then scheduled for a total knee replacement and arrives on the same day surgery unit, the morning of the surgery, and his blood glucoses are found to be 275, okay? The patient reports that his home blood glucose, the week preceding surgery, has been, you know, 120 to 156, well within the recommendations that we set. I love the surgeon. He keeps consulting us again and again. The surgeon wants to cancel the surgery, but the patient's request calls you for consultation, right? So they call us again, call you all again, and this is the question, which of the following facts do you consider when making a recommendation to the surgeon? The patient leaves 130 miles away. You can see that. His daughter took the day off from work to drive him to the hospital for this procedure. These are all very important considerations, though, mind you. I think that's where the great approach differs from just looking at benefit and harms. The preoperative glucose, 275, can be treated with insulin, as you can see there. The patient drank a carbohydrate-containing beverage. This also comes from the surgeons, by the way, before coming to the hospital, according to the preoperative instructions that were given to him, all of the above, okay? For the fellows here, any time you see all of the above, you don't be on the boards this way, but pick all of the above, no, I'm just kidding, okay? So that's, in fact, the right answer, right? So this is why it's absolutely important for us as guidelines, when we develop guidelines, to consider all the other issues that is evidence to decision-making issues. We can discuss this further. So there is a recommendation by the surgeons, not by us endocrinologists, that recommend using caloric-containing carbohydrate drinks before surgery, okay? Don't ask me why. At least the hypothesis is that during stress hyperglycemia, say during surgery, you have increased insulin resistance, and apparently if you take a carbohydrate-containing drink, somehow it'll improve the insulin resistance. That is the hypothesis, although the evidence supporting that is rather slim. Nevertheless, this is an important problem for us when we are consulted of a case like this, you know, the surgeon giving carbohydrate drinks. What happens in a patient with diabetes was an important question that the panel thought was important to address, and so this is one of the PICO, and if you see the language that we use, I mean, it's fine, it just says, we suggest not administering a carbohydrate-containing oral fluids preoperatively, especially in people with diabetes, because there is absolutely no data at all, and the data that we have is a couple of studies, and it means nothing. So considering the harms that you get with it, it's better, and there's no benefit, demonstrated benefits, so we recommended not to give carbohydrate-containing drinks. Anyway, to summarize this recommendation, the previous one, where we have a threshold is if even feasible, try to get it less than eight, or if the glucose is between, you know, 100 to 180 milligram per deciliter. When you can't, I think we said, right, when it isn't feasible, when you consider all the other factors for the patient, we suggest targeting the preoperative blood glucose between 100 to 180, even the day of the surgery. In fact, the panel remarked that anywhere from one to four hours, if you can bring the sugars down there, and maintain it postoperatively, that's not in the guideline, but nevertheless, you'll have better outcomes than before. So with that, I'll stop here. All right, great. Okay. So we'll do a third case. No, we're not doing a third case. We're doing a Q&A. I'll be back. Thank you, Marie, Mary, and Raghunath for kicking off this session. I want to thank the Guidelines Committee for putting together these very important questions and answering questions that we have in the hospital on a daily basis. So as mentioned, I am going to lead the Q&A for the first session here. We have another Q&A afterwards towards the end, but any questions from the audience, and I also have questions virtually as well. So step up to the microphone. Good morning, Alison Myers from Montefiore in the Bronx. My question is for the previous, the last speaker. You were talking about ERIS, and you found that there was no utility of using it. So we did a small, I was working with the colorectal surgeons, and we did a small, unfortunately, it was retrospective study, and we actually found that most of our patients with diabetes, because if you look at the literature, they really never really tease out the diabetes patients versus non-diabetes, which probably made your life miserable. But we did find that there was a less length of stay in our patients with diabetes. Granted, if you look at the sample we had, most of our patients were not coming in with A1Cs of 10, 12. They were more around the 7 to 9 range. So I think that it's not so much that we can't do it with our patients, but we just have to do it in the right setting, and what I realize is that if you don't have an endocrinologist who understands this concept, or a primary care doctor who doesn't understand this concept, it can be a real disaster, because they assume the patient's having colorectal surgery, they'll be NPO after midnight, not understanding that, no, they're drinking about 60 to 100 grams of carbohydrates the night before and the morning of, and especially in a type 1, that could be a disaster. So I think we can do it if you have good consultation in the preoperative time, and like I said, we did find that there was decreased length of stay in our patients who did do ERP, opposed to age, sex match controls that did not use ERP. Thank you. Actually, that's extremely well said. When you make a guideline, we have to, like was discussed, you have to balance benefits and harms, and you all said, you said the right reasons, right? You have to have somebody consulting you. Somebody has to come and manage the blood sugars very operatively and preoperatively, and there wasn't much evidence to suggest at all. The only evidence we saw was there was a better, from the patient's standpoint of it, they liked it, meaning, I mean, they like having a oral carbohydrate ring. Other than that, the evidence was rather slim. Mind you, that's a very low, very low certainty of evidence, and our suggestions is basically, it's a suggestion. So you have to apply that individually, and especially type 1s and type 2s uncontrolled. I think the benefits is rather slim compared to the harms. If I may, I'm just going to comment on this too, because that's exactly right. As Dr. Munopia mentioned, that it's really a balance of harms and benefits, and one of the harms is the action that we have to take, including insulin injection for some of these patients. With, by the way, no evidence in people with diabetes, isolating patients with diabetes, and looking just at carbohydrate-containing drinks. That was really our issue, that there was no study looking just at diabetes alone. So the EROS bundle, as our, the person who asked the question has wisely stated, actually has a lot of good things, including early use of antibiotics, early mobilization. This is, this is called enhanced recovery after surgery. We do not want to exclude people with diabetes from these things that really make a lot of good sense. The bundling with carbohydrate drinks is the problem, and that's exactly what we address. So fully agree that EROS does reduce length of stay. It's very well established, I would say at this point, but no studies in diabetes with carbohydrate drinks, which I think a lot of us here who study physiology would agree it doesn't make a lot of sense. Two more questions from the audience. Thank you. Good morning. I'm Pasquale Passarella from Albany Medical Center. My question was back about the patient starting glucocorticoids in the hospital. How would you determine or go about determining the initial dose of the NPH based on steroid dose and, of course, the individualized patient? Thank you for that question. It's a very good question, and there are several protocols published in the literature for addressing glucocorticoid-associated hyperglycemia. The easiest one is probably the oldest one that was published in endocrine practice some years ago that we've adopted at our own institution of using a weight-based algorithm based on the dose of glucocorticoid being used. It was Dr. Klor at Medical College of Virginia who published this a number of years ago, and he started with 0.1 unit per kilo for 10 milligrams of prednisone or prednisone equivalent up to a max of 0.4 units per kilo of insulin for any dose of 40 milligrams per day or above. That's only one of several publications, and although there's not strict guidance, that wasn't the purpose of the guideline itself, there are references to other methods of calculating doses with people receiving glucocorticoids. So I'm going to follow up on that. So we've had a number of questions on virtually in terms of glucocorticoids, and one of the questions we have is whether or not to replace NPH for basal bolus insulin or to add on. That's an excellent question. And in the studies that... I mean there will be some people who wouldn't have been on insulin before getting glucocorticoids who could have glucocorticoid-associated hyperglycemia, but in those already on a basal bolus insulin regimen, in the majority of the studies, the NPH was added on to a pre-existing basal bolus insulin regimen. And then the dose could be tapered as the dose of glucocorticoid was tapered, or if it was abruptly stopped, then they would have only gotten their usual doses of insulin and not receive the dose of NPH. So it could obviate the risk for hypoglycemia in these patients. Thank you. One more question. Julia Vargas, Lincoln Hospital in the Bronx. The question is, a patient that presented incidental finding of adrenal incidentaloma, woke up consistent with ACTH independent cushions, A1C of 14%. What will be your time range in this situation? Like how much would you wait to do the surgery in this patient? Because A1C is high. We know it is the hypercortisolemia who's doing this. In this scenario, what would you do? Okay, I would, I'd say there's two questions there, right? There's the person who has Cushing's syndrome, so we don't even know. You're saying it's from an adrenal adenoma. Okay, so they are going to be scheduled for surgery, but in the meantime, they have a very high HbA1c, but you want to take them to surgery. So I guess going to the Raghunath's point is that I don't think you want to delay surgery indefinitely to bring them from an A1C of 14 down to 8, but you could treat their hyperglycemia, so they could have a blood glucose of less than 180 prior to undergoing the surgical procedure throughout the perioperative period and the post-operative period. I think that's another point that we made is that post-operative glycemic control is as important, if not more important, after surgical procedures. All right, thank you. So we'll move on to case number three. Okay, here we go. Did everybody get the QR code? I know that we had a request to check it out again, so for folks who need to, please do. We'll have a couple of more questions for you. Okay, so this is... we're going to go right to the case. So which of the following patients would be a candidate for initiating continuous glucose monitoring in the hospital where policies guiding the use of these devices are in place and nursing personnel have been educated on the use of the devices? A. A 48-year-old patient with type 2 well-controlled on metformin and a DPP-4 inhibitor admitted for observation after presenting with chest pain. B. A 6-year-old patient with type 1 diabetes with hypoglycemia unawareness admitted with hyperkalemia and acute kidney injury. C. A 72-year-old patient without a prior history of diabetes found to have random blood glucose of 162 and 175 in the first 12 hours of hospital admission following an MVA. Or lastly, a 22-year-old patient with one-year history of type 1 treated with basal bolus insulin at home admitted for emergency appendectomy. Appendectomy. So lots of words there. I'm just going to give you a few seconds. Just... I'm going to take the opportunity to also remind folks that this guideline did address... It was all encompassing for hyperglycemia. It wasn't specific to type 2 diabetes. Just a quick note on that. Okay, let's go. So... 10 seconds takes a long time. Aha! I knew this was going to be a softball for you guys. But I think that the point of these questions as you can see is really to illustrate our point and the issues that we're trying to address. So first of all, as we will discuss shortly, the use of CGM is complex. There are some issues around feasibility and acceptability and what we found is that those at highest risk of hypoglycemia probably have the most to gain. If a hospital system were to decide to support this through policy, it would probably be targeted toward those individuals. Now walking through these options, the individual on oral agents at home clearly not only probably is at very, very low risk of hypoglycemia, might not require insulin and also might be out of the hospital soon. We should consider that as we know the warm-up period for CGM devices is still... is still relatively long compared to what we need often in the hospital as far as needing data quickly. That warm-up time is shortening as we know with new devices. But we also know that the mean absolute relative difference, which is how we look at accuracy with CGM improves after day two, certainly after day one. So patients who are going to be in the hospital a very short time are less likely to benefit. Of course those who are already on CGM, which we'll talk about briefly, those patients will probably stand to benefit the most and they'll be the most comfort with in that case. So looking at item C, this is another patient who may or may not need insulin during their hospital stay and hence would have probably the lowest risk. A patient presenting after a motor vehicle accident with glucose below 180 during the highest stress time during their hospitalization, assuming they're not going to get glucocorticoids, it's a good predictor that they're probably not going to need insulin. Usually stress hyperglycemia is more indicative on the first day. And then lastly our patient with type 1 diabetes, hopefully is already on CGM, but putting that aside, this patient is going to the operating room and we still do not have enough data to support safe use of CGM in the setting of electrocautery, otherwise known as diathermy. Companies are working on that problem, but at this point we still have to not... we have to safely protect patients from problems around accuracy related to that. So we remove devices, generally speaking, before surgery. Okay. So as far as the data that we looked at for CGM, the data is interesting because it was derived from both randomized trials, as you see in the top panel, as well as cohort studies that were mostly ancillary studies tacked on to ongoing randomized trials, and for the most part that data was blinded to both the investigator and the patient. So that becomes relevant, especially when we talk about hypoglycemia detection. But first let me just point out, we didn't see a lot as far as benefit around mean daily blood glucose in the randomized trials, and that's where you really would see that benefit if it existed. And remember some of these trials go back to 2015 and 17 when the devices were not yet at the generation level we were at now for accuracy. But we did see, if you look just at the RCTs, a mean difference in the mean daily blood glucose of about 15. We decided that was not very impressive and more data is needed, I think, for us to understand that. But overall favorable. The hypoglycemia is much more interesting to us. We have placed a high value as a panel on this. And as you can see, and just to highlight for you, the randomized trial data, which is on the bottom here, I did a little switch for you, actually is telling us more about prevention of hypoglycemia. That's why you see the forest plot showing favorable on the left here. This is favorable towards CGM. Whereas the cohort studies, which are, again, blinded, they were more interested in understanding if CGM was better than capillary blood glucose in detecting hypoglycemia. And in fact, not surprisingly, CGM did better than capillary blood glucose in identifying or detecting hypoglycemia. In those trials, the investigators could not act on it because that was not the design of the study. They were more about accuracy. So on balance, it appeared to us as a panel that you could, that CGM was good at detecting hypoglycemia and allowing prevention when you looked at the different trial designs. Okay. So the case three, our recommendation here is in our population, which is non-critically ill hospitalized adult patients with diabetes, mellitus, when we compare CGM plus the use of the scheduled use of capillary blood glucose per routine versus just the routine capillary blood glucose alone, in adults with insulin-treated diabetes who are at high risk of hypoglycemia, we suggested the use of real-time CGM with confirmatory bedside capillary blood glucose monitoring. This was a weak recommendation basically because of the quality of the data in addition to what we could derive from the point statistics that you saw as far as how much the benefit was, the quantity of the benefit essentially. Now there are a couple of remarks that we did add to the guideline that are really important in interpreting this. First of all, there are CGM accuracy concerns that we did list in the guideline. For example, extensive skin disorder, hypoperfusion, the use of pressor therapy. The list here basically represents what the exclusion criteria have been in general for the limited number of randomized trials that we have seen. So consider that very important. And then finally, who are these people who are high risk for hypoglycemia? And as you'll see here, these are all derived from specific studies that have studied the risk of hypoglycemia in the hospital. At the time of publication, this is the data we had available to us, but it's pretty broad. So older adults, lower BMI, total daily dose of insulin more than 0.6 units per kilo per day, and then you can see classical risk factors including CKD, liver failure, stroke is known to be a risk factor, active malignancy, pancreatic disease, CHF, and a history of hypoglycemia at home or unawareness, so a large group. One other comment is that this guideline statement, this recommendation, I think has a lot of applicability in the near future. Many of us through during COVID were aware of the fact that the FDA did grant the emergency use of authorization of CGM in the hospital during the pandemic. We also recently, the FDA granted breakthrough status for CGM use in the hospital, which only means that when the data becomes available that is ongoing around accuracy and safety, when it becomes available, the FDA will allow a rapid pathway to approval for the hospital setting. So we saw the writing on the wall and we saw this as something that is probably very near future state for our hospitals, and we wanted to put that guidance out there now rather than wait another five years. So... All right. So I think we are up for our next case. Amy Donahy. Hello everyone. Thank you so much for attending. I'm so proud to be part of this panel. I have case number four for us and I will read it to you. A 42-year-old woman with type 1 diabetes is admitted to the hospital with pylonephritis and she started on IV antibiotics. She states that she's not been taking her insulin regularly for the past few days because she's felt sick. She reports infrequent and random home blood glucose testing with all of her glucoses greater than 200 at home. She denies ever experiencing a low blood glucose. Her admission blood glucose is 415, her A1C 9%. At home she's supposed to take insulin glargine 12 units at bedtime and insulin lispro 4 units prior to meals. On admission the admitting intern orders her home doses plus a low dose correction scale, which would be typical of what I would expect. Over the next two days her clinical status improves. She's getting ready to be discharged. She's switched to oral antibiotics and she's going to go home tomorrow. Her glucoses on the regimen that she takes at home in addition to that correction are there for you to look at. I'll give you a second to take a look at those glucoses, which given her infection really don't look terrible. So here is our audience poll question. What do you recommend for this patient before she is discharged? Is it A, increase her basal and bolus doses because she is having some residual hyperglycemia? Would you place a personal CGM on the patient? Perhaps in many hospitals we're giving out free samples of these CGMs now. Do we put one on those and have the patient follow up in clinic in four weeks? C, should we advise the patient to make an appointment with a diabetes educator after discharge for diabetes self-management education and support? Or D, should we consult the inpatient diabetes educator or the health provider specifically tasked at your hospital to provide inpatient diabetes education to see this patient prior to discharge? So you'll have a few seconds to enter your choice. Okay. So many of you do agree with what the panel is recommending. We would definitely want to consult the inpatient diabetes educator or in the absence of a dedicated educator at a facility because the panel also acknowledged that many hospitals do not have diabetes educators. It would be the person who is specifically tasked to provide inpatient diabetes survival skill education and support to the patient. Option A, increasing the insulin doses. Not the most ideal answer. She already admitted she doesn't really check her glucoses too frequently. She likely doesn't take her insulin at home like she's supposed to. So that might not be the best plan, especially if we encourage her during hospitalization that she should be taking her insulin when she goes home. Option B, place a CGM on the patient. Although a good option, because we know that CGMs are wonderful, especially in the outpatient setting and teaching patients, placing a CGM that only lasts 10 or 14 days and then following up in four weeks, definitely not a good choice. We know that when we place CGMs on patients, we need to provide that education on what to do with those numbers. We know that even though she is infrequently checking at home, that CGM that's just going to last 10 or 14 days is not a long-term solution to her glucose monitoring. And because all of them require prior authorizations, without that early follow-up in the outpatient clinic to facilitate the continuation of the CGM, not the best choice. And option C, to advise the patient to make an appointment with a diabetes educator afterwards. Of course, that would be an important thing to do, but we know that patients need to leave the hospital knowing how to take care of themselves, how to survive to that first outpatient appointment. And so that's why option D is the correct answer. So the guideline panel recommends, the recommendation is, in adult patients with diabetes who are hospitalized for non-critical illness, we suggest providing inpatient diabetes education as part of a comprehensive diabetes discharge planning process. I'll point out the level is two, and there's three pluses there. So this is moderate certainty of evidence. And just like with all of our recommendations, there are some significant remarks that really go along with this recommendation. On your cards, they all have the remarks. If you go online and look at the actual guideline, you can look at our decision to evidence tables if you go to the links within the guideline itself, and you can see where we came up with these remarks and how we came up with the decisions for our recommendations. So the first remark here is that that inpatient education has to be accompanied by comprehensive diabetes discharge planning. It's not enough just to explain to the patient, you need to take your insulin, here's how to do it, and then hopefully they know how to do it. We really need to make sure that we are validating that they have the skills needed to survive, those diabetes survival skills. We need to make sure we're referring for outpatient follow-up for diabetes education and support. We need to schedule the diabetes follow-up appointments for patients and we need to ensure patients have access to their medications and supplies. We shouldn't be sending patients home with prescriptions for insulin and not sending them home with the pen needle prescriptions that they could actually take the insulin. So all these things are very important in ensuring that patients don't come back to the hospital. The other remark that we made was that inpatient diabetes education is best provided by a diabetes care and education specialist. Now again, our panel acknowledged that there are limited availability of these specialists. However, all the studies that we reviewed to come up with this recommendation either had the education provided by the specialist or had the education provided by a healthcare provider who had the support and oversight by one of these specialists. The diabetes care and education specialist can serve as a resource to healthcare providers and that can include nurses, pharmacists, dieticians, other healthcare providers within the hospital and they can provide training and support so that all patients will be able to go home with standardized and consistent education. Here are our data that helped us support this. When we looked at A1C reduction, and this is inpatient education versus no, but it's really not no education because all the studies had standard care. So the patient was either randomized to the formalized education or they were randomized to whatever the standard education was that all patients would leave with. There was an A1C reduction at three months of about 1.25% and there was also a reduction in hospital readmissions, especially in a couple of the studies, early readmissions, so seven day readmission rates. Okay, so we'll move on to case number five. Case number five is a 24 year old woman with type 1 diabetes treated with insulin pump therapy for the past six years. She was admitted with community acquired pneumonia and she was actually looking really sick when we saw her. She reports generalized weakness, decreased appetite, she has a productive cough, shortness of breath, she's kind of fallen asleep as she's talking to us. Her A1C levels have ranged between 6.5 and 7.2% since she started insulin pump therapy. But she has reported that within the last few days while she's been sick her glucoses have been running pretty high. Again, she looks very ill when we evaluate her. Her vital signs are there on the slide. Her lung exam showed decreased breath sounds. Her neurologic exam shows that she's alert and oriented but she's very drowsy. Her admission glucose is 256. She's able to tell us what her pump settings are. Her target glucose is between 80 to 140. She has two basal rate settings at midnight .4 units per hour and then it changes at 7 a.m. to .6 units an hour. She uses an insulin to carb ratio, one unit for every 15 grams of carbohydrates and her insulin sensitivity factor is 50. Come on. This is not an audience response question so you don't have to break out your phones. What's the next best step for this patient? A, should we continue the insulin pump with an increase in her basal rates of 20%? Should we discontinue the insulin pump and transition the patient to correctional insulin administered every four hours? Should we discontinue the insulin pump and transition the patient to basal bolus insulin? Or should we continue the insulin pump with the addition of correctional insulin scale administered every four hours? I'll give you one second to think in your head of what that answer is and I'll show you the answer now. So the best answer in this case is that we would discontinue the insulin pump and transition the patient to basal bolus insulin therapy. Many patients who wear insulin pumps are able and should be allowed to use their pump while in the hospital. The caveat is they have to be alert enough to have the mental and the physical ability to manage their pump. They have to be able to self-manage their diabetes using this pump because we know that the healthcare providers around them are not knowledgeable about the pump. So in this particular case, our lady was drowsy. She was acutely ill, she was not able to continue her pump and so that's why A and D are not the best answers in this case. And B, we would never want to stop an insulin pump and just institute correction insulin because then of course we don't have any long-acting basal insulin on board. The patient would be at risk of course for hyperglycemia and perhaps even DKA. So the panel, the recommendation is in adult patients using insulin pump therapy for diabetes management prior to admission for a non-critical illness, we suggest these patients continue insulin pump therapy rather than changing to subcutaneous basal bolus insulin therapy. And the key here is in hospitals with access to personnel with expertise in insulin pump therapy, in insulin pump therapy. So even though these patients will be self-managing with their pump, they need to have access to somebody in the hospital that has some knowledge of pump use, whether that be an endocrinologist or a diabetes care and education specialist or another person who has expertise in this area. And again, the remarks, very important to accompany these recommendations. And these are the remarks that apply to the patient case that I just provided. Inpatients who have impaired level of consciousness or inability to appropriately adjust their pump settings or if they're critically ill or if they're being admitted with DKA or hyperosmolar hyperglycemic state, obviously they would not be good candidates for inpatient use. And it's important to not just make that assessment at the time of admission but also throughout the hospitalization. If there's ever a change in the status of a patient who's being put on a pump, then obviously that would be a decision to, a change in status would make a decision to perhaps convert them over to a basal bolus regimen. The Forest Plot that accompanies this recommendation, the studies, limited number of studies in the inpatient setting for pump use. There's a small decrease in mean glucose that is associated with patients being able to continue their pump versus not. However, we found that in many of the studies, patients were much more satisfied being able to continue their pump and having the control over their diabetes, which might be something that they feel they can control as opposed to the acute illness which might be out of their control while they're in the hospital. In addition, there was little harm with continuing the pump. There was no difference in rates of DKA or hypoglycemia associated with patients continuing their pump in the hospital setting. So with that, I will turn this back over to our chair to close out our session. So just briefly to review some of the other topics that were covered in the guideline that we didn't include questions for, which was the use of non-insulin therapies for glucose management, which I know is a big question for everyone, the use of what type of regimens to use for people receiving enteral nutrition, should we use carbohydrate counting or fixed meal, pre-meal prandial insulin dosing, and again recommendations for a scheduled versus correctional insulin in people with newly recognized hyperglycemia versus those who are on insulin prior to admission. There are several resources available to you that are accompanied with the publication of this guideline and these include an app that you can download on your phones that's already available. There's an online CME program that's also accessible to you. There are patient resources and there is a pocket guide, which I'll use a visual card here, that I believe is available for you to take with you when you leave. And at this point, thank you for staying with us through all these presentations, but I think I'll turn this over to the chairs. Moderators. Moderators. All right, we've had a number of questions online. I just want to ask one just to start off with because there's a very common theme amongst some of these questions. In particular related, and it was just addressed, but commenting on the use of non-insulin therapies other than DPP-4 inhibitors, what might you consider the data, trials, and in particular also more recently just with the SGL-2 inhibitors and the use from our cardiologists in the hospital. Any comments? Thanks, Matt. I'll start with this. I think we all have a lot of feelings about this actually, but what we don't have is a lot of data. So we recognize that probably among the questions, this is probably the one that we wished we had better data. For example, a lot of us do potentially employ non-insulin agents in certain scenarios in the hospital. All of us practice hospital-based medicine to some extent in a consultative capacity. And so we do have a remark that does state that as part of a transition plan to home, a use of supervised and risk-assessed use of non-insulin therapies could be decided you could decide to do that. It does actually make a lot of sense at the end of a lot of hospital stays. As far as the question about the data we did have, on balance there was more harm with the use of GLP-1 receptor agonists and we didn't have a lot of data for, we didn't really have any data for SGLT-2s, but we commented on the perceived increased harm from some very small reports. And also, more importantly in that regard, a lack of evidence of benefit in the hospital. So I think what we're saying is the DPP-4 inhibitors are the only drugs that had very robust randomized trial data despite the fact that the trials were done with different outcomes and methodologies. So we've concluded that that's for sure a safe approach and in patients who have very well-controlled diabetes, not on a significant amount of insulin, and I think we define that as less than .6 units per kilo per day, right? And who came into the hospital well-controlled, those are the ones that benefit from a DPP-4 inhibitor to minimize hypoglycemia with the very judicious use of supplemental insulin. Yeah, the audience, please state your name and where you're from. Suchitra Nincharla from Northern Virginia. So with that last case, I just have a suggestion. The patient was so ill, a type 1 diabetic, before you give the choices, DK was excluded because there was just a glucose and she was sick, she was sleeping on you, et cetera, et cetera. So and then my question is, I see that there's no reference to pregnant women. Pregnant women, GDMs, type 2s, type 1s, coming in for labor or a C-section. It would have been so helpful to OBs to have some guidelines. Where do you maintain the sugar? Do you give them insulin, et cetera? Thank you, that's an excellent suggestion. I have to say, our minds didn't go there, but they probably should have. That's a very important area of glycemic management that we did not address, but I agree with you. I think we all agree with you. Sure, thank you. I think one of the frustrating things for the panelists was we were limited to the number of questions we were able to address because of the grade process and the systematic reviews that would accompany them. So there was a lot of questions we wanted to include and we wish we had more, so maybe we'll have to have another guideline soon, I think. That's right. Thank you, thank you for all the work you did. Vidya Kadambi from Medical College of Wisconsin. In patients who are on scheduled insulin, I see a lot of variability between providers as to some choose basal, some choose the short-acting insulin. In an outpatient world, you start with basal insulin and I always get into this trouble. So what is the inpatient data suggest when you are thinking about starting scheduled insulin? I want to just be sure I under... What criteria do we use for scheduled insulin? No, whether it should be start with basal insulin or short-acting insulin. Because you could do both, I mean... So under the recommendation for correctional versus scheduled insulin therapy, we addressed three populations. The patient population who has no prior history of diabetes but has elevated glucose levels on admission. The ones that we talked about in the case, the ones with type 2 diabetes treated with diet or non-insulin therapies prior to admission. And then the people who are already on scheduled insulin before coming into the hospital. And those people definitely need, that latter group definitely needs a scheduled insulin. For the people that you transition or start scheduled insulin therapy in the hospital is for the first time. I think I speak for the panel. I think the majority of us, Dr. Umpiras is here who's done a lot of these studies with the weight-based insulin dosing is to start with a basal insulin. I know some of the studies use just a basal insulin alone based anywhere from .1 to .4 units per kilo based on the assessed insulin resistance for the patient, age, BMI, et cetera. With correctional insulin or with bolus insulin divided before meals. And some of these eating regular meals. Thank you. All right, so I'm gonna go back to an online question. And one that I think was addressed in the guidelines. But regarding pre-op and when an A1C is elevated because of the lag in A1C, can we use blood glucose monitoring as a tool for preparation? As well as an add-on to the question with use of CGM and can we use GMI? Good question. But we didn't address that. And I don't know if we had enough evidence even to address that particular question. So I'll just stick with the GRADE process and what we recommended. We don't have any data on it unless the panel has some thoughts about it. Which are not recommendations, by the way. The recommendations are published. This is just fun. But I think that we all do different things depending on what the patient is able to access, et cetera. But I think that those who are already on CGM, we've heard from Dr. Roosh and others that the GMI is a good alternative to the A1C. Especially if you need the last two weeks, which you may or may not be able to get a good GMI from that. But yeah, I think the mean blood glucose is probably reasonable pre-op to assess a patient's risk. But as Dr. Korekowski mentioned, it really is about the glucose prior to surgery. And then the glucose right prior to surgery predicts the post-operative glucose, which is the strongest predictor for post-op complications. So I'm a fan of using CGM if you can. But more importantly, guiding a patient to achieving glucose control. If the surgical team is comfortable with the higher A1C, with your commitment to helping the patient achieve better control pre-op to that below 180, which is what we stated in the guideline. However you do that is up to you. That's what we do. Thank you. Hi, Toure Zahedi from New York. I noticed that the panel is considering using newer antidiabetic agents. And I want to bring a word of caution here. I saw the enthusiasm in this year's meeting about SGLT2 inhibitors and GLP-1s and their cardiorenal protective effects. And I want to bring some issues that we have to take into consideration to take all of these in with a grain of salt. After Eugene Braunwald wrote the article in New England Journal reviewing SGLT2s, I went back and reread all of the supporting articles. And first of all, all of them are sponsored by the drug companies that make them. Second, the one that was sponsored by more reputable drug companies like Merck, and those have good track records as far as scientific integrity, didn't show any benefit. On those that did show the benefit, the benefits are minuscule. But they were hidden under the auspices of relative risk reduction and also number of events saved per 1,000 patient year, which are very, could be very deceiving and misleading. So I took it upon myself to go and see actually how much would it cost if there is integrity to the data and we believe that they are correct. How much would it have costed to save one event in the placebo group as opposed to those who actually got the medications? And the number is in the millions of dollars per event saved during the duration. In the Empire Reg, it was about $1.5 million per one event saved over a period of, I guess it was three and a half or four years in about 4,000 or 5,000 people. And what is your question regarding the use of those agents? That we have to be very, very careful and take the data with a grain of salt and also consider the huge burden that the cost of these medications, the average wholesale price for these medications is over $600. Yeah, I just wanna say we agree that we, I think the studies you're mentioning are outpatient studies. But we agree that the data's not there so that's why we didn't recommend the use, the general use of non-insulin agents. But I'm seeing the enthusiasm, they are ascending on the charts as the second line treatment for outpatients. Cardiac and nephrology groups are very enthusiastic. They're bringing it into four and I'm just worried about the cost that this would cost on the healthcare and take away the public health and other things. Thank you very much. That's an important point. Dr. Walia, do you wanna come up? Hi everyone, thank you so much for your very hard work. I'm sure it took a year plus. I'm Amisha Walia, I'm from Northwestern. My question is actually a methodology question and looking to the future. So it's very difficult to randomize patients at a patient level. And it's become much harder as well in the peri-COVID era, being in this space for a long time. In addition, when we're looking at glycemic control, when you're randomizing by the patient level, the effects in some place like inpatient is really systematic, right? So the nursing has a huge thing to do with it as well as upper level barriers. So my question really is about sort of like methodologies in terms of randomization at the level of the units or the level of hospitals, randomized clusters, interrupted time series. How do you forecast? Because I think that, I mean everyone's like, we need more data, the level of data. But I think for the next five years in this era, we might still be in this era where it's very difficult to randomize at that level. So do you have any comments as well as in terms of like the grade criteria as well about how we can utilize some of these other methods and improve like our level of evidence? You raise a really good point. And I just, I think we were having this conversation yesterday among our group about how difficult, like even when you talk about patients with diabetes, you're already talking about so many different types of patients with diabetes admitted to so many different parts of the hospital with so many different reasons for admission. And can you, Angela, you're shaking your head. You may have been the one who brought up about people being admitted for an AMI. But is that the same person, same as the person who's been admitted for pneumonia? And can you extrapolate the person? So you raise really important questions. I'm not really sure we have the answers. I know for some of the diabetes education studies, it was randomized based on nursing unit. You know, it would be a unit where the education was provided just to prevent cross-contamination for a patient who might be in the next bed who was randomized to the no education arm. So I think you will help guide us through that in the future as we look at these alternative methods. Just to add one point. I think the guideline, if you look at the document, for every PICO, there's a whole bunch of research questions that the panel members wanted to ask. And you're absolutely right. How do we do it in the era of funding related to inpatient management for diabetes and what not? So those are all valid questions, which we really need if we continue to use the great approach, right? Otherwise, how are we going to make a recommendation that's much little bit more than a weak recommendation that we currently do? So those are all excellent questions that we have to think over. Can I make a follow up comment? I think that's exactly right. I think we're gonna have to stretch ourselves beyond these criteria in order to really understand as we move in the post-COVID era. Thank you. Thank you. I think we have time for one more question. Okay. I'm Beatrice Chung from Brazil. And I would like to ask you, first of all, congratulations for the guidelines. Very interesting and a very good job. I would like to ask you, considering that most physicians that will treat these patients are known specialists, they are known endocrinologists, they are not comfortable with basal bolus insulin, even with all this stuff to read and to guide them. What do you think about basal only, considering also a tertiary care unit hospital when patients have many other diseases and a high risk of hypoglycemia? Is this clear in the guideline to use basal only? So we defined terms and we found we had to define terms for ourselves. We had to sort of back up and come up with some decisions. But scheduled insulin therapy was defined as a basal insulin used in combination with correctional insulin for people who are eating the correctional insulin administered before meals. Or as a basal bolus insulin regimen. And I know Dr. Umperas, who is here, actually did a study in people with type two diabetes that were hospitalized comparing what he called a basal plus regimen to a basal bolus regimen. The basal plus being the basal plus correction. And found similar outcomes in this selected group of patients with type two diabetes. So there will be some patients who will do okay with basal plus a correction insulin scale. But people with type one diabetes in the hospital who are eating meals, they're, they. Not type one, just type two, of course. No, or in hospital hyperglycemia, for example. Okay, thank you. All right, thank you everyone. I think, I'd like to thank the panelists. And if you have any questions, please feel free to come up. Thank the moderators.

inpatient glycemic management

glycemic control

adult patients

hospital guidelines

hyperglycemia

surgery

blood glucose levels

continuous glucose monitoring

diabetes education

discharge planning

insulin pump therapy

non-insulin therapies