My name is Bryn Marks. I'm a pediatric endocrinologist currently based in D.C., but about to move over to the Children's Hospital of Philadelphia, and I'm here with my co-chair. Hi, I'm Amisha Walia from Northwestern University in the Division of Endocrinology, Metabolism, and Molecular Medicine. Welcome today. We have a wonderful lineup for you this morning, really focused on diabetes technology, covering a vast array of that. So our first presenter this morning is Dr. Greg Forlenza. Greg is an associate professor at the Barbara Davis Center and the director of the Type 1 Diabetes Pediatric Technology Research Team, and he'll be talking to us today about the future of automated insulin delivery. Hello, good morning. Can you guys hear me? There he goes. All right, so thank you guys for coming today. My name is Greg Forlenza. I'm a pediatric endocrinologist at Barbara Davis Center, and I'm actually an Atlanta native and went to Georgia Tech. Go Jackets. So hopefully there's some Georgia Tech fans in the crowd here, too. Where's the advancer? Here, this guy's not working. So these are my disclosures. I'm always proud to say I work with every company in the United States that makes technology for kids with type 1 diabetes, and so I definitely have a very strong bias towards technology, which will come out as I talk today, but I don't have any bias towards any particular product. And so my first slide here is basically a definition of some of the terms that I'm going to use. And so we use a lot of different terms over the years, and most of them have very bad to kind of non-existent definitions. And so the term that I tend to use is hybrid closed loop. The bigger category would be AID, or automated insulin delivery. Automated insulin delivery is basically any system where the CGM is talking to the pump, the CGM is telling the pump how much insulin to give, and then insulin is being modified based on that. Hybrid closed loop would be a subcategory of that. Hybrid closed loop is any time where the system's automating part of it, and you're also having to jump into the loop and to perform an action yourself. And then fully closed loop is any system that is fully automated where you won't have to interact with the system at all. Artificial pancreas is a term we definitely should have stopped using a long time ago. We still use it because it's kind of sexy. When you put artificial pancreas in a grant, people are like, ooh, artificial pancreas. But I've also had families come in to our outpatient diabetes center with me, a pediatric endocrinologist, and ask when I'm going to implant the artificial pancreas. And I'm like, I didn't consent you for any of that. This is an outpatient center. We don't do any of that. So it's clearly a very misleading term. We should stop using it, except we're asking for money. And so the topic that my friend Jen Shear asked me to speak about is artificial pancreas, the future is now. And so in order to talk about the future, I think we have to go back a little bit and talk about the past. And so for people that have been involved in the field for a little while, this is Aaron Kowalski's artificial pancreas development roadmap. 2009 is a time we're still using that term. I'm going to go through this in more detail in a second. But this was Aaron's original roadmap. And it's pretty surprising how now, almost 12 years later, or 13 years later, this is pretty accurate for what we saw. There's really only one design on here that we never did, which is step three. But we had threshold suspend, predictive low glucose suspend, hybrid closed loop, fully closed loop, and multi-hormone systems are the systems that we have somewhat developed and are on a pathway to still developing. This was Aaron's revised roadmap from 2015. And what we see here is sort of a bifurcation that you could have a single hormone system or a multi-hormone system. So as we talk about what's come so far and what's coming, I think it's good to kind of review how Aaron's roadmap did in terms of what we've seen. And so it's not a perfect graph. I kind of jury rigged Aaron's a little bit and made this myself. But here's kind of the basic idea. So we had threshold suspend systems. We had the Medtronic 530G and 630G, which were threshold suspend systems. We've had predictive low glucose suspend systems. We had the Medtronic 640G in Europe, never came out in the United States. And then Tandem Basal IQ was a predictive low glucose suspend system. And then since there, we've had multiple hybrid closed loop systems. So we had the Medtronic 670G, 770G, and 780G would still be considered a hybrid closed loop system. As well as Tandem Control IQ, Insulet Omnipod 5. And then hopefully later this year, we'll have the Beta Bionics Islet system also coming out as a hybrid closed loop design. And then there's also several systems in Europe that I don't have personal experience with. So I'm not gonna talk about them as much, but Diaboloop and CAM-APSFX. We will also have hybrid closed loop multi-hormone systems, the Beta Bionics Islet. We're gonna be starting trials later this year with that system as a multi-hormone system. And then no big surprise, in the future, every company will basically be going to some form of fully closed loop system. Whether they've started their trials or not so far, I can present a little bit of data on that. But that's kind of the big picture design is that we will have fully closed loop systems that are both single and multi-hormone. So in order to talk about any of these systems, the big thing we have to talk about first is CGM. I think most people are gonna be familiar with CGM, where it is, what it does. But this is a graphic that I like to update. It's based on an original slide from University of Padua that I've updated several times. And what we see here is the subsequent generational improvement in the accuracy of CGM systems. And one of the things that I like about keeping this kind of living slide within my deck is that I need to update it every three to six months. And so the Abbott Freestyle Libre 3 data actually came out last week at ADA, and so I've had to add it since then. And so even though Dexcom G7's not on market yet, they've published their accuracy data. And so what we see here is that really every two to four years, we're coming out with a new generation of CGM that is significantly more accurate than the next generation. That there's multiple companies bringing these systems online, and obviously with a little bit of an outlier there for some of the Abbott Navigator, which had way better accuracy than I think we appreciated at the time, they're all generally within the same box, which I think is pretty interesting. You know, they compete like crazy, they definitely fight with each other, but they're kind of doing very similar things when you look at it kind of from an agnostic field picture. And so the other thing that we see is that we're kind of now hopefully hitting a plateau in the it's accurate enough that it doesn't matter except for outlier range, which is essentially where we're getting to now. And so this is my summary for what exists right now and what we expect to see coming. And so Dexcom G7, we hope to see come to market later this year. It's been with FDA kind of like everything during COVID, a little longer than we would like to see, but it will be a 10-day sensor with a 30-minute warmup period and a published MART of about 8%. The Medtronic Guardian 4 is also under FDA review right now. They haven't published their accuracy data for that, but I imagine it will be somewhere at or below 10%, kind of in this similar range. The Guardian 4 will be factory calibrated. We have over 100 kids using it right now in a variety of studies and it works. It doesn't have the same limitations that we've seen with Guardian 3. And then we'll be starting studies later this year with Medtronic Synergy, which is the thing that I have an image for there. They're kind of one piece disposable sensor. And then Abbott Freestyle Libre 3 was actually FDA approved just last week. If you were at ADA, it was kind of funny. They had like buses with Abbott logos for Freestyle Libre 2 on them. And then Freestyle Libre 3 came out kind of at the start of the visit. And you're like, oh, you guys wasted some money there. But it's good to see the technology move forward. And then hopefully later this year, we'll see it integrate with hybrid closed loop designs so that patients can actually have some variety. You know, you could use product A with sensor A, you could use product B with sensor A or product A with sensor B and mix things up and result in some competition, which I think is really good for patients. So the last kind of background I'm gonna present is a study that I published with Todd Alonzo and an excellent med student and resident at our center looking at kind of the four categories of device use. And this is something that Kelly Miller and Nicole Foster have published variations on this with the Type 1 Diabetes Exchange. But obviously due to issues with the exchange, there was kind of a lull in them publishing new data. And we came out with sixth generation sensors and hybrid closed loop systems in that interim. So we wanted to do an analysis on our 4,000 patients. And so that's what this is. What I show here is that the black circle is the least technology using group. It is the MDI injection and finger stick group. The black square is the pump and finger stick group, a group that used to be very popular and is now decreasing with size over time. The gray triangle is the CGM with injection group, something that five, six years ago we didn't think would be a group. And now is actually a pretty large group. And then the last group, the gray triangle, I mean, the gray diamond is the fully technology hybrid closed loop and sensor. And hybrid closed loop pump and then sensor and pump combined, I'll break that on the next slide. And what I think is really interesting about this is that it shows pretty clearly to me that CGM is what matters. That CGM group is much lower for both cohorts than the non-CGM group. And to some of us, this was surprising when we started to see this trend three years ago because we thought that pump was what mattered and CGM was adjuvant to pump. But what we're seeing is that CGM is what matters and pump is adjuvant to CGM. And so I'm gonna talk about automation, which is my thing, but the takeaway, I think, for someone who's not as engaged in technology is we have pretty clear data that CGM is what matters. And I actually draw this out on the back of the check-in sheet for families sometimes. When they say, I'm only gonna wear one thing, what should I wear? I say, we've answered that question. I don't have an opinion on it. We have data and the data's pretty clear. These are correlational data, they're not RCT. I know that, but I still think they're pretty powerful data. Breaking out now the full technology using group to non-automated and automated. And so our hybrid closed-loop group, our AID group here, is a combination of Medtronic 670G people who continued it and TANF Control IQ users. And what we see here is that among the almost 700 hybrid closed-loop users, the A1C is significantly lower than the A1C in the other groups, excluding the two to six where it's still off-label, so our numbers are a lot smaller. And so again, we see that automation is doing a really good job. And again, I'll point out that the 12 to 18-year-old group, that's the peak of the lifetime A1C curve. The average A1C in that population in the recent exchange studies has been somewhere between nine and 10%. So again, there's some selection bias here. I know that, but still to have an A1C in that age group that's right around seven and a half, I think speaks to the power of these systems. So this is what's on the market right now in the United States. People are probably familiar with this. I just kind of include it for completeness. We have the Medtronic 770G from Medtronic. 780G is under FDA review right now. I keep hoping it's gonna be approved between the time I make the slides and the time I give the talk. Hasn't happened yet, so we'll have to keep waiting on FDA for that. And then we have TANF Control IQ, which has been on the market since 2020. We have Insulet Omnipod 5, which is just starting to roll out now. And we're finally starting to see some commercial approvals for that. If people have questions after the meeting, we can talk about that. But basically, our center has had a 14-person task force meeting with two to six people from Insulet once a week to try and figure out how to get kids the system. So if you feel like that's a similar experience that you've been having at your practices, and I have a lot of friends at Insulet. I've run the studies since the first in human. That's the experience we've been having, and they're working on it. And so here we have a slide that I published with my friend, Reyhan Lal, earlier this year. And this is a summary of the pivotal trial data for all the systems that are currently commercial somewhere in the world. This was published in Diabetes Technology and Therapeutics. I see Sue Brown here from UVA, but I don't see Boris, and that's good, because Boris gave a talk about how no one should ever make this slide. And so the challenge with this is that we're using different sensors, we're using different study designs. This is not something where you could say system A is better than system B, system B is better than system A. And I've told all my friends in industry, please don't pick apart a 1.2% difference between your product and your competitor's product. That's not the point of these studies. In my opinion, the point of these studies is this. Excluding 670, which we've published a lot of data about the challenges that people had with that. Adults using these systems are achieving a time and range of somewhere in the mid-70s. Children using these systems are achieving a time and range of somewhere in the high 60s. They're doing that with a hemoglobin A1C value that for adults is right at or below 7%. For kids is right at 7%. And they're doing that with a time below range that's generally at 2% or less. And this is something else that I'll kind of draw out for families when they say, pick the system for me, which is the best system. And I'll say, there isn't a best system, there's the system that's the best fit for you. The analogy I tend to use for this is, and it's one of the challenges we have in starting these things at diagnosis, it's like you're trying to buy a car. And when I talk about this at diagnosis, what I basically say to families is the challenge is, it's like trying to buy a car if two weeks ago you didn't know there were cars. How do you know which car is the best fit for you? How do you know if you need a minivan or a sports car? If your goal is to drive around and look cool and they give you a minivan, that's not a good fit. If you've got five kids, you've got to schlep around to soccer practice and they give you a sports car, that's also not a good fit. And so it's about finding the system that's right for you, your lifestyle and your priorities. It's not about one company winning. And so that's my main advice working on all these systems. So the summary of where we are today is that we have three commercially approved systems. There's at least two more under FDA review and the hybrid closed loop designs have the trends that I said. And we're basically hitting right at ADA goal for the first time since the DCCT in terms of where hemoglobin A1Cs are with these systems. And so the next thing I wanna talk about is kind of the future where we're going. And so what I'm gonna talk about is a system that we just recently started presenting pivotal trial data for, which is the beta bionics islet. This is the system some people may have heard about developed by a guy named Ed Damiano from Boston University. The famous story about that is Ed's a control systems engineer. His wife is a physician. Their son David was diagnosed with diabetes when he was 18 months old and Ed started seeing what was going on. And he said, we can do better than this. I can build a system that talks to itself. I can get that and develop it. I can have that out within the next few years. David was diagnosed when he was 18 months old. David's currently a college sophomore. And so they were off a little bit on the scale of time that it takes to develop a commercial product, but they were right on in terms of what we're able to do and what technology is able to do. And so the cool things about the islet and I put it some on my slide from ADA here is that it's initialized only with patient weight. You don't put any other parameters in it. You put in patient weight and you say go bionic. I think the go bionic's a little dorky. It kind of fits with the mood of the company, but it's still, that's all it does. And there's no basal rates. There's no carb ratios. There's no correction factors. You set a target that's conceptual. You say about normal, a little higher blood sugar, a little lower blood sugar, and that's it. This is the first system that we've developed that's actually using machine learning in real time to adapt those parameters. When 670 came out, I remember hearing people talk about all the ways 670 was gonna learn and I would sit with some of the people from Medtronic and they'd be like, oh my God, we didn't develop any of that. Where are people getting these ideas? And so this system is actually the first one that is doing that in real time, is actually using machine learning. And so what the designs of the system are doing is enabling us to kind of expand this technology to a greater group of patients. And so what we see here is that this study actually enrolled a wide array of adults who had a much more diverse racial and ethnic background than in our previous hybrid closed loop studies and I think equally important, had a much more diverse hemoglobin A1C profile than in our previous studies, with 23% having hemoglobin A1C of eight to nine and 11% having hemoglobin A1C of greater than nine. And what we saw with the results of this system, this is one of my two favorite slides. The other one is on the pediatric data, is basically a squinching of the A1C distribution. And so bear in mind, people were doing this without carb entry, without correction factor entry. All they're doing is telling the system, I'm eating breakfast, it's about a regular breakfast, I'm eating breakfast, it's a small breakfast, I'm eating breakfast, it's a large breakfast. And they were able to do that over three months of system use. You can see here in the control group that the A1Cs basically stayed the same. That's the SC standard of care group. But in the experimental group, what we see is that we basically cut off the right half of the tail, the group that's likely to develop complications, and we kind of clustered everyone around the center at seven. And so we also lowered the average by 0.5%. And so the average lowering is nice, but as a provider, what I think is even more exciting is the reduction in the extreme high group. And so the takeaway from this for me of kind of working with these systems is that you basically run a clinic where you almost never see anyone with an A1C above nine. And so that is pretty cool that you know you're gonna walk in the room, the average patient you see is gonna have an A1C of about seven, and even the people who are struggling the most are not even gonna have double-digit A1Cs. And that is a phenomenal difference to me in distribution, especially as you kind of exclude the high-end outliers. And then here we see the main thing that I like to have as a takeaway, this is what I talked about at ADA, is the higher your A1C was at baseline, the greater the delta you're gonna see. And so this gets to a big point of not limiting who you prescribe these systems to. People that have higher A1Cs at baseline stand to gain more from technology, they should be the ones we're driving the technology in, not the ones we're limiting the technology for, and that's what we see. The people who started out at seven tended to stay around seven, but the people who started out at nine tended to have much greater deltas in their A1C. Following that up with the pediatric data, just because this is my world, I think it's cool, we actually had 35% of our population that were in a group other than non-Hispanic white, much more diverse than we've recruited in previous studies, and then we had 58% of the kids with a hemoglobin A1C of 8.9, and then 10% with an A1C of greater than nine. So a much more representative population than we've seen in a lot of our previous hybrid closed-loop studies. And what we see here, this is my other favorite slide, is kind of a squinching of the A1C distribution, a cutting off of the tail on the right, and kind of centering everything around 7.5. So again, not quite where we wanna get it, we wanna take some moves and lower it even more, but a massive improvement in what we're seeing at the full population level for hemoglobin A1C. Similar data here for kids also, the higher A1C was at baseline, the greater the improvement that you saw. These are the patients we should be fostering technology use in, not limiting it. And then the other study that I wanna present is one that I'm very excited about that we're doing some work on now, which is development of a fully closed-loop algorithm by the University of Virginia group, and this was published in Diabetes Care last year. And so this is basically a randomized crossover study of the new Rocket AP system versus the Legacy system, which is essentially the Control IQ design. And so what we see in this pilot study is that with use of a fully closed-loop design, you can basically achieve time and range values here. Study is showing of 83% after unannounced meals and 100% after announced meals. This system design is kind of early design, it may be too aggressive for what we're gonna see commercially, but it shows that the concepts can work and it's something that we can be developing, that we can basically utilize meal detection and aspects of meal automation to enable us to have higher floors if you don't interact with the system. And so I think for like adolescents and young adults, that's really powerful that you can say, okay, well, it won't beat you doing what you're supposed to do, but if you decide it's Friday, I'm not taking care of my diabetes on Friday, I'm just gonna wear the system, I'll respond to alerts, I'm just gonna go out and do what my friends do, which is what every adolescent wants to do, every young adult wants to do, then the system can still manage that in a way that's acceptable. And then if you kind of get back, you get back to kind of more close monitoring, you can achieve even better results than we're currently achieving. So here we kind of see the distribution after meals that announcing still beats not announcing, but in my opinion, automation of unannounced meals is something that can be developed in the next few years and be feasible. So what these studies show us, I believe, is that meal announcement with machine learning of pump settings can be successful. And that's what I think that the beta bionics islet studies have shown us. They also show us that fully closed loop is something that can work and that we can develop and bring to market within the next several years, and that it provides us, like I said, with an acceptable floor. So where we are, I think between 2009, the initial roadmap and today, we've seen numerous hybrid closed loop systems come to market. For me as someone who develops these systems, it's really exciting to see them go from something very kludgy that works on one tablet, one model of tablet, one model of phone, you need the engineers who developed it there with you to get it to work. And then in four to five years, to basically take that to something that people are using at home in their daily lives, that's improving their care, is really cool and really powerful that we have so many people using these systems. When I'm out in public, especially at these meetings, I notice when people are wearing these systems and I'm like, that's so cool, it's a person I've never met who's wearing that system. My wife will say to me, they don't know you're a pediatric endocrinologist, they just think you're weird. And so I try not to point it out anymore because they don't know, they just think I'm weird. But it's really cool to think that these systems are out there and people are using them in their daily lives and additional systems are coming to market every year. So where are we going? This is my opinion, but fully closed loop will work. We can develop ways for fully closed loop to be a safe and effective technology for people with diabetes that beats what they're currently doing now. And that's a big engineering concept that I learned at Georgia Tech is don't let perfection be the enemy of the good. If we can develop a system that improves things, we should compare that against what people are doing now, not against what it would perfectly be. And then there are some of these systems that are being developed now, and I kind of think that this development is similar to where it was in 2016 for hybrid closed-loop designs. It may take a little longer than that because the bar we're trying to clear is a little bit higher, but I believe this is something we can very successfully do. And then these systems will probably do better for people with higher hemoglobin A1Cs, and that's something that we definitely want to target in future studies going forward is kind of reaching additional rungs in the technology use spectrum to continue to further improve the lives of people who are doing worse at baseline and thus have greater area to gain. And so this is my final slide that I was thinking of with future being in my talk is pump settings where we're going. We won't need pump settings. And I think that that's a really important idea is that we're going to remove the numeracy from diabetes. We're going to make it so that every meal is not a math problem, every activity is not a math problem, that we'll be able to automate our way out of this. The analogy I use for this is when I was six years old, my dad actually worked at Georgia Power right about five minutes from here. We had a compact computer that he brought home with a little green screen on it, and I could load video games on that in DOS when I was six years old. It was obviously a very high reward getting to go to the video games, and I could use the commands to load a video game in DOS. I couldn't load video games in DOS now if I needed to, and the reason for that is I haven't needed to. I haven't needed it. Windows uses DOS for me, and I think that we can get to the point where our engineers use diabetes numbers for us, and we can start to get away from the numbers. So thank you guys for listening today, and we're going to do questions at the end, but thank you for coming out so early this morning and listening this morning. All right. All right. Yes. Thank you. We are going to be doing questions at the end with all three speakers, so next up, we will have Mary Lauren Scott. Mary Lauren Scott is an associate professor and fellowship director at University of Alabama. She is PI for the T1D Exchange and director of the High Risk Clinic, and her work is focused on clinical core access for technology, so we're very excited to have her here. Thank you. Thank you guys for having me, and it is not lost on me that I have the short stick here, and I have the least data to present, so do not shame me for that, but I am also wearing a hybrid clothes loop, so that's my other disclaimer. So this is not based on personal preference, but basically what we can do to help our patients. So Rock'em Sock'em Robots came to mind, but we really don't have to stick technology against one another. I do think that, again, getting back to patient-centered care, what works best for your patient is going to be what controls their diabetes, and we need to learn how to listen and have these open, honest discussions with them, so that we can all achieve the end points that we want. So we're one of the lead centers in the NPN quality improvement project, where we assessed barriers and ways that NPNs can improve care in diabetic providers, and so we kind of looked to see what were challenges in clinic to getting the program started, to use of them, and what things it actually helped with. So we did have funding through T1D Exchange by Medtronic for that. So our clinic experience, my colleague Jessica Schmidt created this document that has been quite useful in advocating for patient-centered care, as well as equitable care. I feel like it gives our patients a voice. Dealing with kids, typically this is best used for adolescent patients who can complete the paperwork themselves, but we have this My Diabetes Journey handout, and the first question asks them, what do they think they're doing well with? So they can brag on themselves, and you know, what technology, what change they've made that they think is helping them. This next section, we put a lot of, you know, circle this, you can pick the phrasing, but if they're over diabetes, I met one patient who said he didn't have burnout, and a year later he did. So really, all of our adolescents kind of go through phases of this, but it's a way for them to express that to us easily. It's a way for them to admit failure with adherence, and I think that is such a barrier in the conversation with care, that if we can open it up to be a part of our model and what we do talk to our patients about, so that we're able to advocate and let them know that we are here to help them figure out how to overcome their challenges, and we're not judging. And so they can tell us easily by circling something, if they're skipping doses, if they are struggling with math, if they need more help, which is unusual to select, but some kids do ask for more help, if they feel like they're having a hard time with their basal insulin, their bolus insulin, et cetera, if they're tired of blood sugar checking, that kind of thing. The next section is what they think might help what they would like to try with their diabetes care. This is interesting because there are some patients that you hear, and I've heard a lot of anecdotal tales from people over the years, that feel like they've never been presented with certain technological options for their care that they actually think would be helpful. And so as providers, it sucks to have to learn so many different things, and to have to educate our patients like this, because we don't get reimbursed well for that time. But this is, again, what helps them with their care. And so patients who may never have been offered an insulin pump, they can say, well, I'm interested in that. And maybe if they scare you, and you've never seen them check a blood sugar, and you're not sure about them being on a pump, it at least opens the door to a conversation about a pathway to get there. So patients can kind of give you an idea of what they think would help. And then the last thing is they think what has helped most improve their self-management. So sometimes they put their provider, sometimes it's their parent, sometimes it's their pump or their CGM. So our data from one of our providers' clinic, this is a high Medicaid population, one of our nurse practitioners runs. And she handed this out to all of her adolescent patients for about four months this year. Patients can obviously circle multiple things, so this is not one answer for each kid. And some of the patients are already on pumps that are already on CGMs. But 27% of her patients wanted a pump that weren't already on it. 15% wanted CGM. A good chunk thought maybe alarms would help, fixed meal dosing, 10% there was a parent or patient that didn't want to be on the pump for whatever reason. A lot of times they're citing body image issues or they read a horrific story on a mommy blog somewhere. The rare 4% asked for more parental supervision and less supervision at school and a dose calculator or sliding scale. These all came up as options from our patients. So in kind of morphing this into thinking about SmartPen use, which patients in those groups could be helped with a SmartPen? Well when you don't know a lot about SmartPens and how they work, there's a lot of things we think they can't do that they actually can. And so you can do set doses. You can do sliding scale. You can do time of day dose changes. You can have a patient who's on eight different carb ratios be on eight different carb ratios. Not that that's always necessary, but you can do that on a SmartPen. And so you can have settings that are very similar to how you would use in a pump or with your MDI patients who have numeracy issues. And so these patients who wanted more supervision by their parents, the parents can check the downloads just like you can ask them to do with the pump and they can see how often they're bolusing. Same thing for you adult providers. I'm forgetting I'm at a majority adult conference, but for you adult providers who deal with the elderly and so patients who may even be aware that they might forget their dose. Can I go back and look and see if I remembered a dose for breakfast an hour ago? Yes. That's great. And so you can kind of follow behind someone or follow behind yourself. It does work with CGM and it works best with CGM. So again, Gregory's point that CGM makes the biggest impact is not lost on SmartPen companies because most of them have their best functionality with patients who incorporate CGM into their day. So the patients who either the provider or the parent or the patient doesn't want a pump, we see this a lot in young adults and adolescents, body image issues, not wanting to wear something. Some people have beliefs that they can't do it with certain sports and that kind of thing. And so if that's your patient's preference, this is a great backup. I had a long-term pumper, child of a doctor friend of mine who I followed, who was headed off to college and she was like, I just really don't want stuff attached to me. And she had had a wonderful A1C on a pump CGM system and she opted to do CGM alone for six months before she left for college. And her A1C went up to 10 and she was very disappointed and she was trying as hard as she could, she felt. And we decided to try the N-Pen, she'd never heard of it. She came back before she left for school with an A1C back in the sevens and she was so happy with being able to do those calculations. We train people on pumps in some ways to be dumb about their math, right? Because you input numbers, it does all the math for you and then you have to go back to scratch with old school MDI if you choose to come off. And so for that patient, it was a game changer and it really helped her to fly and do well as she moved away alone to college. The question comes up when you're talking about fixed meal and sliding scale, if a patient either does that or wants to do it, depending on their reasoning for wanting to do it. If it is burnout and feeling as though it takes too long to calculate their doses, they may do just fine on carb counting and a correction dose with the SmartPen app because all they have to input is their sugar and their carb and it'll do the math and it'll be more accurate than a set dose. And so there's that sector of patients you could potentially get them to use it because the app speeds up their care and helps with that burnout feeling of dosing taking too long. The patients on fixed dose who can't conceptualize carb counting or won't do it can still do it. So you're not robbing them of that. So again, the data is not a lot. So if we look through some of the data that is out there, in 2019, these guys did a review of the publications that were out there at the time. They found only nine of 286 met criteria. There was a low quality of evidence. A lot of the studies that had been done were based on patient questionnaires and interviews to collect data on satisfaction and how they felt like things were going. There were five that used control groups, but really most of the data at the time looking back at 2019 was qualitative and there really was not a lot of control group data, randomized control trials, that sort of thing. There probably won't be. If you can imagine doing a randomized control trial with SmartPens, how do you tell someone they're not on one and make them believe they are? You can't. But they did find when they were reviewing the data that a lot of the studies indicated it could be abuse in patients with memory concerns or those in need of supervision. These next two studies were both done out of Sweden. They were done using the Novapen 6, which we don't have currently in the U.S., but they found that they were single arm, the patients were compared to themselves at baseline, there was no control group beyond comparing to themselves at baseline, and it was a prospective study on a pen about a year for both groups. The adult study had about 94 patients and they showed increased time and range of almost two hours a day, which is significant. So the two hours more a day that you're in range, which is great when we're aiming for that 70%. They found reduced hypoglycemia in general of about almost, again, two hours a day, which two hours a day of less lows, wow, that actually feels like a lot more than the p-value looks. When you look at the severe lows, the lows, the grade two, level two, less than 54, that's almost 20 minutes less of those severe lows a day, which is big. This is data I would not have just thought of when I thought of, I'm going to put a patient on a smart pump, that it would really protect from lows that way. Now, caveat being, these patients were wearing CGM, and so the two together helped their dosing accuracy work to decrease lows. This is not something that can do threshold suspend or predictive suspend to do that. This is just their doses being more accurate. And then they found fewer patients were missing doses for boluses, about 43% improvement there. In the pediatric arm of the study, again, there was not a huge population of patients, but it was, I believe, around, I think, 64. They found a significant reduction in mean daytime hypoglycemic events by about 31%, and then nocturnal hypoglycemic events by about 24%, which, again, huge. Those are game changers. Decreased severe lows less than 54 by a little over half of a percentage point, which, again, anything that decreases severe lows, we're happy with, but not, again, the most all-striking data. When we look at a couple of the other papers that have come out, this one is a single-user perspective. So this patient, back in 2018, about a year after the first SmartPen hit the market in the U.S., decided he was burnt out on his pump. He was a runner. He was quite active. And he went to MDI and was quickly disappointed, and then learned of a SmartPen option, and he found that he was able to rely on the insulin on board, active insulin time feature that a lot of our pump athletes really get into using through the SmartPen, and that that really helped him to maintain his lifestyle and enjoy his pump hiatus and maintain control. He looked at some of the insurance payer coverage at the time, and in 2018, about half of insurance that he looked at covered the in-pen product with about a $75 or less copay. Now I think a lot of patients can get it for $35 or less per year if it's not covered on commercial insurance. And then he put this out there. I don't know if the company wants it out there, but he said he found that the app continued to work after you had paired it once with a SmartPen. So even though if he ran out of an in-pen or it broke, he was able to still use it for the math. You cannot do that with any of the SmartPen products approved in the U.S. if you haven't used it with a product. And so, again, it's useful if he just went to get a regular short-acting pen to still record his data. And then he enjoyed the ease of emailing or printing the reports for visits with his health care provider and found that that emulated what he had done before on the pump. The quality outcome and cost outcome in Sweden, I did some conversions just so you kind of get an idea of what the dollar conversion would be for these cost savings. But it's something around, I think, $17,000 a year of direct cost savings and about $38,000 lifetime. They estimated a 60-year period, $38,000 for a 60-year period looking forward for the same group of adult patients that was in the study on the last slide. So they basically looked at their baseline data, they input it into the IQVIA model, and they got this data. Again, it's all based on a formula. It's not real world. It's not case-controlled and that kind of thing. But it's interesting numbers if it does hold true long-term. They found decreased frequency and delayed onset of complications might be something that you would see with long-term use. And so I'll take it, but it definitely is not the kind of data that we love to see. And then one group has looked at pediatric information. This is more recent, back in 2021. And again, the literature still at that time was limited. They're not widely used products yet. The benefits were these, like we talked about, of calculated doses with personalized settings and tracking data just like we can on pumps. When they looked at the data, the Bigfoot product had just come out, and then the N-PEN was the one that was on the market since 2017. And so not much beyond what I've already told you. So my personal thoughts and our experience in trying to implement more of these in clinic, we're getting to where we have about 6% to 8% of our patients who are on smart pens. We do enjoy the improved documentation for these MDI users, especially in the setting of telehealth. And Jennifer's going to speak to telehealth, but that has been one of the big holes in telemedicine I think for all of us is that you don't have a lot of data. They're going based on patient-reported outcomes, which we all know you can't trust as far as you can throw them sometimes because people hate to admit that they feel like they're failing, right? And so if you can see legitimate data to hold up, are they bolusing with meals? Are they not? What do we need to work on? It really does improve their care, and it opens a conversation free from judgment. You can use the active insulin time and dose calculators, which I think appeals to a lot of my athletic patients, patients who really are trying to avoid lows in the context of sports, especially school sports with kids. And the dose calculators helps, especially in the setting of burnout and patients who have been guesstimating and making inaccurate doses. People who, again, have burnout or have numeracy issues, it can be quite helpful in patients who are used to pumps and are on a longer break. People who are absent-minded, elderly, forgetful, that kind of thing can be helpful. I describe it that way to my teenagers without trying to tell them their parents can stalk their doses. I try to tell them, hey, if you're not sure you dosed for the Twix that you ate after school, you can look back and see. So can their parents. So can I. But it's one of those things that I think helps them to kind of be empowered with knowing what's happened. And then the downloads, again, help to guide discussions about adherence, given that they're all recorded, if they're using the product. So things to consider if you're implementing use of this in your clinic. It's learning about a new technology. I have not personally found it to be quite difficult to learn about. Pardon me. I think you can download at least the Npen app and play with it yourself and get an idea of how the doses are input. The Bigfoot one you can't unless you have a prescription. And it seems as though that product works if you have a hub set up in your clinic. So you need to learn the pharmacy needs in your area, insurance coverage, which the reps can help with. And then teaching your patients on use of cartridges to refill the pen is important if they've never used them, because they are glass. And they can crack or break them if they don't put them in correctly. Teaching your staff how to obtain the downloads. Our staff, we try not to email to our direct emails because of HIPAA, but we do have a HIPAA safe educator email that it can come through. There's about a one hour firewall. So trying to get patients to email early, that's the best way to see the downloads in color unless you have a color printing fax machine that can fax it to and or get them to print it from home. It does work with gluco as well as tide pool. So if you have these systems in your clinic for downloading data, it can be a great way to just add that to what your staff does as the patient comes in, just teaching them to ask. Our educators will walk through setting up the app for them when we prescribe it so that they have it. And if you're following all the JCO regs about having the doses written by the provider, the one change in workflow for us was having to add active insulin time and maximum bolus dosing to our doses so that our educators could legally input those into the system. Kind of like when you're doing pump doses, you have to kind of sign off on those. And so I just added a little academic spander to include those and usually start with a generic four hour active insulin time unless I knew what they were on when they're pumped before. And so these things I think are workable and they're doable. We spend, again, my only plug here, we spend so much time educating ourselves on our complicated pump devices, which we love. I love. I wear mine every day. It's great. But if it's not what's right for our patients, we can spend a little bit more energy learning about the other things so that that gap between our MDI users and our pump users is less if we can also potentially convince them to use CGM. So that's my two cents, and we will do questions again as the group at the end. Thank you. Thank you so much, Dr. Scott, for that wonderful overview. Our final presenter this morning is Dr. Jennifer Raymond from the University of Southern California. She's an associate professor there, division chief, and the chair of the virtual care committee at the Children's Hospital of Los Angeles, which is a perfect segue into her talk on telemedicine. Yeah, thank you. Okay, so hi, everybody. I am really excited to talk to you today, and I'm gonna talk about what virtual care can look like. This says, the title is post-pandemic. I think we're still very heavy into the pandemic, so what it looks like moving forward. Okay, so I have no financial interest to disclose. However, I wanna give a huge shout out to Dr. Stephanie Crossing, who's actually sitting right up here. She's a colleague of mine who also focuses on virtual care, and she and I have given multiple presentations together on telehealth, and I'm using a variation of some of our slides, so I wanted to acknowledge her expertise and also thank her for our work. Okay, so this is the QR code. I think this will come at the end, too, but this is how, if you all wanna take a picture right quick you can use this for, I think, for evaluations as well as for questions at the end. All right, I'll wait until it looks like phones come down a little bit. Okay, cool. Okay, so I'm gonna kind of follow the path that Greg took and really look at data pre-pandemic or look at the history of virtual care, and then also look at it during the pandemic and then talk about what it can look like moving forward. Now, before I say that, I wanna take a quick minute to just really focus on the power of our words, and I want to explain the words I'm going to use and why I'm going to use them. So I'm gonna use marginalized, historically excluded, or people of color versus the term minority. I'm gonna use Latinx to be inclusive of queer folks, and I'm gonna use clinicians versus providers because there's actually a historical context to the term of providers. So I also want to acknowledge that these words are not inclusive of everyone's identity, and I just want you to appreciate my intent. Okay, so for virtual visits, we're gonna talk about clinician to clinician or clinician to patient or group type visits. There is so much in the telehealth and telemedicine world, but I'm not gonna touch on all of that. Okay, so let's look at data pre-COVID. So first, I'm gonna just talk about the different studies that have been done, and then the next slide is gonna summarize all of that. There is much literature in this space, so apologies for all of the references at the end. I also have multiple references slides at the end, so if anybody is really into this, please let me know, I'm happy to send them to you. But basically before COVID, we were using virtual visits or telehealth for consults to rural areas, both for patients to receive care as well as for clinicians to connect with other clinicians. We were using them for supplemental visits for elevated hemoglobin A1C, virtual group visits for young adults with diabetes, school-based care, behavioral health support, family-based therapy, and transition-focused care. So what did we learn from all of these studies? So both in individual studies as well as meta-analyses, we saw improvement in glycemic outcomes for those that were being seen with telehealth. We saw increased satisfaction in time and cost savings. We also saw increased visit frequency and visit completion rate, and then we saw improved patient engagement, psychosocial outcomes, and self-efficacy. So what were our concerns pre-COVID, or why were we not all running to telehealth at that time? So we definitely had a lack of training. If you think about the fact that we've all trained for decades or centuries in in-person care, to really think about transitioning the entire healthcare system, our model to telehealth was a lot, as well as we didn't necessarily have staff or workflows to be able to make that transition. There were concerns for reimbursement as well as state licensing about where you would be able to see patients, and then concerns about secure or safe platforms, really being compliant with HIPAA concerns, and then accessing patient data, which my previous speakers have talked about, and then the need for A1C. So how would we be able to do those things? So what happened in March 2020? Well, basically we just knew we had to make that transition, which kind of helped us. In the United States, there was the public health emergency as well as waivers related to licensing as well as reimbursement, and those things definitely helped us move forward. And then the things that I think are also really important that I think for those of us that have done telehealth for a while, actually knew pre-COVID, is that our patients can actually learn to access the data, and or we can consider platforms that really allow them to access it. So I think when I started this 10 plus years ago, and was realizing like, I work with predominantly adolescents and young adults, and to have people that either wear data or have data around them but can't access it, and to think that they're managing a chronic disease, I think is quite a lot. So learning that we can actually teach our patients and families to be able to access that, I think is important. And then really looking at access to CGM as well as GMI, and what that can look like to be able to take place, or such as smart pins or insulin pumps. Okay, so what did we learn during COVID? And I'm gonna go through several sections as far as this goes too. So one is that we saw an increase in virtual visits. And these are data both from California FQHCs, as well as diabetes centers within the US. And we saw a dramatic increase in use of virtual care during the pandemic. We have seen a decrease over time, but the levels still remain higher than pre-COVID area times. And then there's a variability between video and audio visits, depending on where you're receiving care. So if you refer to the left-hand side, the table is looking at, or the figure is looking at primary care visits in California FQHCs. Your x-axis is months and years, and then the y-axis is patient visits. The light blue is all visits, and then the orange is in-person care. So we can see in March 2020, in-person care decreases, and then the dark kind of black to blue is audio visits. So in FQHCs, we saw that telehealth was being delivered more by audio visits than video visits. And then you compare that to diabetes centers, which is on your right-hand side. And this is looking at months and across the x-axis, and then percent of total visits across the y. The in-person is the dark blue. And so you can see that there's a transition from in-person to light blue is telephone, and then the yellow is video visits. What we see in diabetes centers is that video visits predominantly are the way that people receive telehealth care, and that's been consistent. It has dropped off over time, but video visits have still played a huge role in diabetes care within diabetes centers. So the other, now this is looking at data from 89 countries. I wanna point out that this is March 2020 to May 2020, so super early in the pandemic. But for the three figures that I'm gonna go through, it's really looking at age and then gender differences based on a specific metric. I also do wanna point out, they use gender and sex interchangeably, and it's a binary of just male, female, which is also not representative of everyone, but that's kind of what we're looking at here. So for A, that's telemedicine by, I guess I could use my pointer, by gender and then percent of respondents. The light gray is no, you haven't had an appointment, and there's not one planned. And then the intermediate color is no, but have one planned, and then the dark is yes. So basically what we can see just overall here is that there were definite differences by age, and then also a little bit by gender, with males being less likely to use it. That continues to B is looking at usefulness by age, and it's almost the same trend. So the younger adults and then males more than females found it less useful. And then the other thing I think for us to point out and think about as far as our patient population is that in this data set, A1C also made a difference with the thought of usefulness or not. So the higher A1C was less likely, with either gender, was less likely to see telemedicine as useful. Now these are data looking at satisfaction. So we talked about use, and now looking at satisfaction in that experience. Now this is through fall 2020, so people had had a good six plus months of telehealth use. So these are opinions of the online survey, about over 1,400 looking at virtual visits. Now this was just for type one diabetes. So many reported saving time, stress, and money compared to in-person care. And then if you look on the left-hand side, this is looking at, I feel video care is as effective as in-person care, and then is more effective is the yellow. So well over 60% felt that it was just as effective, if not more effective compared to in-person care. Now then this is looking at preference on your right-hand side. So the blue is preferred to have video care for some of my appointments, and the orange is preferred to have video care for all of my appointments. So well over 80% are feeling that they would like video care to be an option for some of the appointments, or all of the appointments, compared to 17% wanting in-person. I think this highlights, and I'm gonna talk about it on another slide too, is that we have to be really person-centered with our use of all things in medicine, but certainly also video care, because it's not going to be the right answer for everyone, but it can be a helpful answer for many. This is looking at successful visit completion. So we see an increase in trend, improved satisfaction, and then we also found, both during COVID and then data before COVID, that we see increased successful completion of video visits. So these are data from, the top is from University of California Davis, and then this is from Children's Hospital Los Angeles. So the light gray is before COVID, and then the dark, the black is after COVID, 16 weeks before and after. And basically what we can see is that the successful completion of office visits after COVID went down for in-person visits, but the successful completion of telehealth visits went up post-COVID. Now, that's 16 weeks before and after. There were many things that were impacting our ability to do successful visits at that time. So this one is looking at a little bit longer term. So this is March 2020 to November 2020. These data are from University of Florida, and these are from University of California Davis. And basically, so the gray is in-person, and then the black is telehealth visits. And we can see the successful completion is higher for telehealth visits. As we get to fall 2020, that difference becomes smaller, but also the same for University of California Davis. So we see that telehealth was more likely to be successfully completed over that time when compared to in-person. And again, the data pre-COVID that we have seen in our studies and through all data really shows that the successful completion of visits is higher for telehealth. Now, the other thing I wanna point out is that I think is with all things COVID, what we found was that it really highlighted disparities and inequities in our care, as well as the model of care. So I wanna talk a little bit about what increased disparities look like thinking about this from a telehealth standpoint. So patients with public insurance, older age, rural location, or non-English language preferences were less likely to use telehealth for diabetes care. Those maybe are related to lower rates of smartphone ownership, broadband access, which I'll talk about looking at the Pew data in a second, private space. So as someone who works in Los Angeles, I care for many that live in multi-generational homes that also have very few rooms. And so everyone may live in the same space at all times. So really thinking about what that looks like for the patients and families and talking to them about that in advance is important. And then digital health literacy and also having that discussion may contribute to this. So there's some evidence also that telehealth is seen as less beneficial by people with diabetes with lower education levels or higher A1Cs. And that was on the Scott data that was shown looking at the 89 countries. We also saw that the higher A1C group was less likely to see telehealth as beneficial. So these data on your right-hand side are from the Pew Research Center. The blue is smartphone and the green is home broadband. And these are US specific data. But basically what we can see is that the older an individual is, the less likely they are to have smartphone as well as broadband access. The populations that are either black or Latinx populations are also less likely to have smartphones, but more predominantly less likely to have home broadband, which when you think about that for either uploading data or during virtual visits, that can make a big difference too. The lower education attained in a formal manner and then the lower income are also associated with being less likely to have smartphone and then specifically broadband access. And then also rural communities are less likely to have access to either one or both of those. Okay, so we talked about increase in visits overall during COVID, increased successful completion, satisfaction and increased disparity. So what do we wanna think about like moving forward? So I would like to think that we could all just take a minute and then start again. I feel like virtual visits have much to offer, but doing it in a global pandemic is a little rough. So I think for us to think about as we're redesigning it or hopefully restarting is that we've done decades of in-person care and it's also really uncomfortable to do new things, or at least it is for me. So thinking that we would be able to rapidly transition to something that's completely new to us is a lot. I also think for all of those reasons, it's not our best effort. Now that's not because any of us are not able to do it wonderfully. It's just that it wasn't implemented in the ideal state. So how would we redesign it or what would it look like moving forward? And I'm gonna go through five different areas to talk about and then also share some resources. So quick acknowledgement, and I said this before, but I think when we're really looking at person-centered care, we need to acknowledge that virtual care is not for everyone. It is also not for all situations. So there are times where we're going to want to see patients in person. There's a time where patients are gonna be want to be seen in person. And there are also situations where it's not gonna be an option and we need to see them in person. And I think all of that can be true and there's still much good to be offered. So we need to think about how we move forward with virtual care. So as far as video requirements, and this is something that I think many of us that had done work in the past, it's been a lot of time thinking about. Some of this has been addressed potentially during the pandemic as institutions or clinics found answers. But as we're thinking about restarting it, I've talked to some institutions or clinics or clinicians who are acknowledging that maybe what they did in that rapid state was not the best choice for them. And so it could also be whether or not it was HIPAA compatible. So there was some leniency during the public health emergency, but some of that is going to be changing. So really thinking about what does it look like long-term. It's tough to give recommendations because I think it really looks different based on each clinic as well as EMR. So there are some options to run virtual platforms through the health record, and you can launch from your schedule, someone could join through the portal, but that's not gonna be an option for everyone and everyone's electronic medical system might be very different. I do think, and this is again, pre-COVID, I spent a lot of time thinking about the cost and trying to convince people that we really wanted to use it. And I do hope that the successful completion of visits and the increased completion of visits shows that there is an economic reason to actually use virtual care, and it also increases access. So then if we think about A1C. So, and my previous presenters both talked about these points too, so they set me up very nicely for thinking about ways to really consider the data from an A1C standpoint. So previous to COVID, I think many of us, again, that did work in this area would either think about getting A1Cs with the annual screening labs, or some research studies actually sent home A1C kits. I was asked about doing this early in the pandemic and really the workflows as well as the reimbursement for those things are really tough, right? It's gonna take more people as well as more work on the patient's part to be able to go and get those labs. Now, I still think if you are doing screening labs or you have a super awesome clinic that actually gets labs in advance, so you have them for your visits, I think that's a nice option. More importantly, I think we can really move towards and consider time and range as well as the glucose management indicator and use that for discussion in our visits. So again, that's something for us to think about as access in devices that individuals are using. It's more supportive for virtual care too. And then the Glycemia Risk Index, this was just many of you probably in here were on this manuscript. I think everybody was. But the Glycemia Risk Index is something that was published in the journal Diabetes Science and Technology recently. And it's basically looking at taking the percent hypoglycemic and hyperglycemic and really looking at that as a risk. So again, having access to CGM so that you could do that could also be used in place of the A1C. So then what about glucose data? And again, we've talked about this and the other speakers have reviewed. But I think considering glucometers, CGM data, endpins or pumps, what are the ways that individuals are gonna get that information to you? And then could you actually make patient-centered choices or person-centered choices that would make that process simpler for them? So choosing things that could be Bluetooth versus web-based will make that process of them sharing data back to you a little bit easier. Also thinking about platforms. So again, this is very clinic or clinician-specific. But some choose to use like one single platform where everything is kind of uploaded, whether that be Gluco, Tidepool, or whatever it is that individuals might use. And so there would be one thing for clinicians to log on. Now, some use, have 20 cables and upload every single device. And sometimes that's necessary. But that could be, there could be differences as far as that goes in the ease for the clinicians as well as the patients and families. And then really thinking of our patients and what can we do to make it easier for them? So passively uploading data through a mobile device where you can actually log on, of course, is going to be simpler for them. But that's not an option for everyone. And also acknowledging insurance limitations. Everyone doesn't have access to some devices. And so really thinking about what's required of the patient and also acknowledging that not everyone has access to a personal computer. Not everyone has access to broadband internet. And that requires a lot to be able to do that. And I will tell you, I've had young adults who were like, I'll just fill out those logs that my mom did when I was like diagnosed and that'll be fine. And I was like, okay. And then I've had actually a couple when we talked about like the plans for the visits. They were like, you know, I learned a lot when I interacted with my data. So I'm gonna go ahead and just keep writing on the logs for a while, which I thought was kind of funny. But actually I have some that do that. So it's never a bad thing to ask the patient their preference because they might actually be cool with using that. Okay, so then thinking about virtual visit workflows. So I think the ideal is that you have clinics that are blocked. So you have virtual care versus in-person clinic versus combining it. Now I'll tell you recently, as there have been more people that have either been getting sick or kids spending more time in school and then making everyone at home sick. I've had more families reaching out and being like, we really don't wanna cancel our visit, but could we see you online because everybody in the house is sick. And so when that happens, I'm like, that's cool. And my clinic is not going to be running on time. So if you're cool with me joining you late, then I'm fine with that. But we just have to acknowledge that my in-person clinic doesn't run on time and telehealth does most of the time. I also think that there could be an ideal state that if we move forward and we're starting to onboard people to virtual care, we could consider doing that at an in-person visit. So where you're talking to them about what it would look like to use that, you could actually have them upload the platform on their phone or whatever it may be to really answer questions at that time. There are new data showing that engaging staff at check-in, there's a new study showing that medical assistants supported onboarding of patients at the beginning of the visit actually helps address and decrease disparities between populations, which I think is huge. So we can use this to really be mindful of the way that we onboard and support our patients to help address inequities in care. Interpreter connection, prescription reviews, or whatever else the clinic staff would be doing, there are workflows to be able to do that for virtual care, which I think can be really helpful. Looking at multidisciplinary clinics or education, if that happens in your clinic, group appointments virtually I think actually work well. There is much to learn about doing them in-person versus virtual, but I think it can be an efficient way to do that. If you have a clinic where there might be a rotation of team members, that's a little bit more challenging to do on a virtual platform, but at the same time, virtual typically runs in-person, or I mean on time, so in some ways it actually can be more efficient. And then I think also ask our patients what that looks like and then how they would want to do it. So with those things in mind, I'm gonna share a couple things on person-centered care. We use things very similar to what Mary Lauren shared about person-centered care and shared decision making for the beginning of visit. So what do patients want to focus on? I think deciding that on related to doing virtual care is really important and also using that to set goals in the role of virtual care in someone's medical care. So you decide on the frequency of in-person versus virtual care and how they want to use that. I also think we have to be transparent about what we're going to ask of patients and families prior to visit. So if we're going to ask them to do downloads or get an A1C or check labs, some patients might decide, I'd rather just come in person versus trying to figure that out or I don't have the ability to do that at home. And I think that's great. One thing I was going to say is we're also working on a toolkit of how you would help onboard patients and families as well as clinicians. So if anybody has an interest in that or you have knowledge you want to share, please let me know. We're hoping to put something together that would be free for everyone to use. And then thinking of person-centered care during the visit. So engaging your patient in reviewing data and asking their thoughts. So when we did this previously for research, I had some clinicians that were like, you know what's cool is when you pull up data, they're seeing it just like you do versus upside down like you do in clinic or whatever. And so you could talk through it a little bit better. I also think we have to acknowledge that reviewing glucose data might not be someone's number one thing for their visit. And that's true person-centered care of thinking about what actually might be in the way of even that glucose data. And then thinking of sharing victories as well as challenges at home. So for some, I'm a pediatric endocrinologist. So pets and art and things that they have at home can be really wonderful things to connect during visits. Also doing prescription drawers or closets or whatever with people as you're doing refills can be helpful. And then I think some of us have talked before, in this setting you're actually able to meet family or support individuals that are maybe not always able to come to visits for multiple reasons. And so you can really expand your reach in that way too. I do think having the private space, which we already kind of talked about is something that we have to acknowledge and really being transparent with patients and families about how they want to set that up. And also acknowledge that some people don't feel comfortable sharing their space. And that's okay, right? They can be in a space where they either blur their background or have it just up against a wall so that they don't feel they have to share things at home. Okay, so virtual care skills. So I think starting with this, we just have to acknowledge our discomfort and anxiety. Using scripts or checklists or patient handouts, I think one of the things I like about virtual meetings, not like this, but like at work, is that I can have things pulled up for me to remind me of what I want to review. And I think that works really well for virtual visits. This is actually my desk at work, which I realize now, I'm like, oh, I should have cleaned that up a bit. But you can see all of the stickies around. Those are the things that I was working on during a study for like person-centered care, right? So like elicit, provide elicit, and then check for understanding. Those are things that I was working on and I needed to change my own behavior. So like a memory thing helps me. I think involving our trainees and learning new skills together. So virtual care is going to probably continue long-term. So those that are training right now really can learn with us. And we want to involve them in that process. And then share best practices within your group. And so review with your practice, talk to your colleagues. And then there are also many telehealth resources. And so if you haven't used either of these, the Center for Connected Health Policy and the National Consortium of Telehealth Resource Centers has a great deal of information, both about approaches, but also from a legal standpoint, our policies or what types of things are used by state is excellent, so I would highly recommend them. There are also regional areas, so you can even connect within your region. So kind of conclusions and then to summarize. So as far as challenges, this was maybe not our best effort, but that's cool, we can do it again. Discomfort with new things, I think that's true for everything, but we really did virtual care fast. And so to think about the opportunity to kind of do that again. And we have to be mindful of disparities, but they can improve with attention. I think the benefits, we have so much data pre-COVID to really encourage us to keep moving forward with telehealth, so increased access, higher satisfaction, improved health, mental and diabetes outcomes, as well as cost and time savings. And now there's new data to show that with attention, we can actually address inequities and disparities. And then kind of just some take homes. So one is to give hugs, and I say this, like I know that's not supposed to be what we do all the time, and I'm very much a hugger. But I think what's important is that for all of us, we need in-person and human connection. And especially over the last two plus years, where maybe you haven't even, for some people going out of their house or even being able to go to the doctor is definite connection, especially because they often see us every three months. So acknowledge that that is really important and something that we should include. And then be kind to virtual care. So screens all day every day is not virtual care's fault. So we can think that there is still much that is positive about that, that we could use moving forward. And we have data both before and then during COVID to show that. And then I think for all of us, whether it be advocating for NPINS or CGM or integrated systems, as well as telehealth, we have to advocate for access coverage and addressing disparities so that we can help, have that be something that's accessible to everyone. And that's it. I have all of my references. I will go to the, this is for questions. Is that where I'm supposed to leave? Yes, thank you so much. So we're gonna, thank you. What a lovely group of speakers. So real quick while we make the transition, this is a reminder, I'm gonna move here. So yes, so now you can navigate and tap on the session title or just point your camera to the QR code directly to join and fill out your evaluation. Also, we would really love people to come up to the podium with questions for either Dr. Ferlenza, Dr. Raymond, or Dr. Scott. We have a question and we also have a few questions coming in through virtually. So we'll sort of prioritize in person and then the chairs, my co-chair and I will be switching to ask some of the virtual questions as well. So please speak your name, where you're from, and then if your question is for one of the panelists or all of the panelists in the question. Thank you. My name is, let's see, is this microphone working? It's working. Okay, great. My name is Sue Brown, I'm with UVA and that was fantastic all through those talks. I love them. Now I am certainly tech biased, but they're really right on, so I really appreciate it. This is a question for Greg. I wanted to know what your perspectives are about the challenges of the future. And you shared with us two really important areas, fully closed loop systems and the islet, which will be phenomenal once that's available, that will make the ease of use, both for clinicians and people using these devices, will be much better. But I wanted to understand a little bit of the challenges and very specifically the things that I'm worried about in this field have more to do with whether or not we might be increasing the rates of hypoglycemia specifically. And fully closed loop, and just as a background, fully closed loop, I think even since you presented that data, what we've done this past year is even much better when trying to get to home. But what is concerning me is about whether or not the hypoglycemic rates might be different. In the islet data, there's severe hypoglycemia rates are absolutely clearly much better than what the background T1D exchange data is. But it's much higher than kind of the pivotal trials from where we were. So I kind of wanted to understand a little bit about what your thoughts are on those challenges. Yeah, thank you, Sue. Great question. So I think the first challenge is actually kind of related to education. I'll get back to the hypo in a second. But unfortunately, each system that's come out and the others that are continuing to come out are going with drastically different designs. And so when we talk about the technology research crowd, we're able to keep track of what those designs are. But the first challenge is, now we have three different systems that are on the market that all are tuned differently. And if you're just like, I always say, I only have one interest, it's diabetes technology, I can barely keep up with everything. And if you have a variety of interests, as we were discussing, then how do you know what the settings do? And so I've had patients come in who are on control IQ, who have seen a provider who is familiar with clearly 670G, who then come in and say, oh, my provider said the basal rates on this don't do anything. And when I've shared that with my friends at Tandem, they were horrified, because of course, as you all know, having developed it, the basal rates are very important to control IQ, but the provider got educated about one form of hybrid closed loop, thought that idea was translational across other systems, and it obviously was not. I was also telling you before, I had a mother of a kid who was on Omnipod 5 in the study, and she changed the basal rates every two weeks. Even though I told her the basal rates don't do anything in this system, you don't have to mess with them, she couldn't let go. And so if you're used to tinkering, and you wanna tinker, and a system doesn't let you tinker, that is also a challenge. As far as hypo goes, I guess what I would say there is expectation management and education. And so I don't know that a system that maintains population level risk of hypo, that significantly decreases A1c, is necessarily a worse design than what we're currently doing, as long as you understand those are the risks. I think part of what you're describing with the beta bionics islet study is that in that study, compared to some of the other trials we've done previously, we recruited a more real world representative population, which is part of the reason that the hypo risk goes up. I agree with you, we can't bring a system to market that increases hypoglycemia risk over where it is in the population, but is more hypo with lower averages or less burden better than less hypo with somewhat higher averages and more intervention? That's certainly what FDA made us do with first gen systems, but I don't know that it's necessarily gonna be our only option in third gen systems. But ideally, I wanna make something that's engineered well enough, these problems don't happen, so I'll put that back on Mark. Hi, I'm Mick Davidson, I'm from Eli Lilly, really appreciated all of the talks. My question is, as the algorithms get better, I think the improvement in overnight control and during periods of relative inactivity have improved dramatically, but we still continue to see real challenges with postprandial hyperglycemia, and then the other issue is managing exercise. I guess I'm wondering about what do you see as limitations with regard to the insulins that we're using, because I still get, as a user of a hyperglycemia, I still get frustrated by postprandial hyperglycemia and rage bolusing, and as I mentioned, exercise as well, so thank you. So thank you for that question. So exercise is one of the first things that we studied, and very early on, we tried to use body harnesses and the kind of things that high-performance athletes use to try and detect exercise, and the challenge with that is that by the time your heart rate goes up, by the time the amount that you're perspiring goes up, by the time that your accelerometry goes up, your step count goes up, whatever marker you're using, it's too late. We had to announce it 30 minutes ago, and so that's a challenge of the speed of the insulins, is even if we have a very, very, very twitchy system that thinks that everything is exercise, by the time you're showing any kind of physiologic sign of exercise, it's too late. We have to have announced it at least 30 minutes in advance. So for adults with planned exercise, sometimes some of the newer designs have allowed us to hard-code that into the systems to either set a basal rate that's lower during that window or to set a target that's higher during that window. For responsible people who are very on top of things and plan out well, that works. I focus a lot on toddlers, and so I have two toddlers myself, plus one recently former toddler. And so they go from zero to 60 in about five minutes and then back from 60 to zero. And in that age group, it's really hard to plan or announce anything because are they eating exercise and you're doing both at once? Sometimes, yes. And so it's a challenge, and this gets to some of what, obviously, I've heard Lily talk about before, which is when we talk about faster insulins, you're not talking about one element, you're basically talking about three. You're talking about time till it's on, time till it peaks, and time till it's off. And so it's a whole different form of co-kinetics talk, but I think that really for exercise, we're really talking about the third piece, the time till it's off. And so that would definitely benefit from better drugs because that's gonna be one of the things that we're latest in automating our way out of. And part of the reason that I tell a group of different educators and providers is we're not gonna automate our way out of that. We're gonna need to still educate and support people through it. And it's the reason technology is not gonna put us out of a job. I don't know what else to say. I'm gonna jump in with a quick virtual question for Dr. Scott here. So questions about how best to use the printout to review the SmartPen data, and then how to learn how to use the SmartPen. And we may need some IT help to get back to her slides if we can do that. Yeah, I can pull it up on my slides if we can open the presentation. So I put a bunch of pictures at the end for that very reason. And what's nice about the printout, and these are product bias because they are of the InPen, which is the one that we can use in our clinic currently. Some pictures of the app, how you can enter everything. You can pick carb counting, meal estimation, or fixed dose. So they do have a small, medium, large meal option if that is your patient. And so again, all the different options for dose calculations, which was not the question. And so this is an example of some of the printouts you can get in particular through the InPen program. I think you can get similar data from Gluco and Typool, but it looks at their CGM data and it can incorporate that. It can give you ideas for how much insulin they're using with each meal and percentage of basal versus bolus. Again, same kind of CGM charts you get looking at those. And then the day-to-day variation, which I think is the most helpful part. This is what the patients can also view in their logbook. Just picking a couple examples from one patient I had in clinic. You can see when insulin was given. It teaches them, which is educational for the patients, about how insulin functions. The kinetics of it in a way that patients can understand with the blue bars or graphs being the insulin given and how it wears off. It teaches them about insulin board if they didn't already know that concept. And you can get an idea of potentially why she's high, did she miss a bolus, that kind of thing, so that you have particular facts and ideas to talk about. You also get how many carbs she's eating. There's one time she went low after having 130 gram carb dinner. Probably more about what she was eating or how she counted that was the issue. Very helpful, I find, and you can incorporate that with their CGM data. As you're talking to them. All right, thank you. And I'd like to, with this particular slide, I think I need help even interpreting it. So, for, all right, so next, next speaker. Hi, Seth Charitz from Rochester, New York. So, on most of the studies or the studies that you've done, I mean, you're looking at A1C, which is probably like an old marker at this point. Do you, I think, or is there patient satisfaction studies? Because I think that's really, at least when I'm talking to my patients, I mean, that's what everybody cares about. Mm-hmm. And this is for? For both of you guys. Both, yeah. So, that's a great question. We were actually just talking about this right before the session. Patient satisfaction surveys have been done as part of or after every hybrid close-up study we've done, and they haven't been as dramatic as what we hear in clinic, which has been very surprising to a lot of us. With 670G, there were obviously, as we've somewhat said, a lot of warts with the system. I still think it's remarkable the barriers that it broke down, but it was a first-gen system and suffered some consequences as a result. Part of the issue is, as we've been criticized, rightly so, we also over-screened the hell out of our participants to participate in pivotal trials, and so it's possible that they might have unmeetable expectations for being the kind of people who participate in pivotal trials. I think now that we're getting to a broader initial use case and we're getting to systems that are olfactory calibrated and focused more on burden reduction, we'll start to see some of that, but one of the things, even in the studies that have looked at it, some of which have been published, some of which haven't, which is part of the reason I'm not quoting them here is because I have trouble remembering what's available and what's not yet available, is that it wanes. We see a little bump initially, and then people say, well, that's old now. I've been using that for six months. What have you done for me lately? And so we don't see it being a main outcome. The final thing is, sadly, as it's been discussed to me with industry, payers don't care. They only want to see the things that they associate with hard outcomes, not silly things like whether or not people are happy. And I would like to make a comment about that. I think for everyone in this room, we really need to make a push for, really, the patient's voice and patient-centered care. I think we have stakeholders, so policy, payers, as well as the technology companies. There's a really wealth of data on patient-reported outcomes, and I think putting actual funding behind patient-reported outcomes and disparities work in technology use is just really gonna be the future that we have to do, whether or not there's payment this moment. I think all three of our speakers really alluded to that that cost-benefit may not be in six months to a year, but actually several years from now. Go ahead. I think for the patients in regard to the SmartPen use, anecdotally, all the kids I've started on it love it. They find that it's very helpful. It's kind of getting the clinic and the provider on board with how to use the data that you're given. And if it's improving quality of care, that's great. I think our MDI old-school diabetes patients are the ones that have been overlooked with a lot of the research. If they are curmudgeons about change and they don't wanna wear a pump or something, we have to still think about them as we advance this other technology, which we do love, but we have to incorporate what works for all of our patients. And I agree. I was joking a little bit about, because industry has kind of been focused on that, but I very much think it's important that we measure patient-centered outcomes, that we have the right tools for it, which has been the other piece is somewhat that the tools that we use to measure it have not been in place at the time we started some of the studies that you're familiar with. Yeah, and I think as several of us are clinicians, I think we're really beholden to others, our patients and ourselves, in order to really give technology to everybody. This will be our last in-person question. Go ahead. Okay, thank you. Theodore Freedman, Los Angeles. I'm the head of the Diabetes Work Group at Los Angeles County Department of Health Services, so I wanna introduce myself to Jennifer. And we're a cost-starved system, so many of the things we talk about today are beyond our limits. We're trying to get CGMs, but the thing I'm really trying to get is these remote glucose monitors where the patient can hook it up to a smartphone and get us the data. Most of our patients have smartphones, very few have computers and broadband access. But I haven't really had much success with that. The typo one doesn't, you need a hard computer. The glucose ones, it's pretty expensive. They're not giving us any discount. So do you have any ideas about how we can get even these sort of simple things like remote glucose monitors to our system? Yeah, so nice to meet you. Thank you for that question. I think just the glucometers, sometimes I think the easiest thing is just having them download the app. Now that's putting something else on their phone, but downloading the app and then emailing it to you or doing that type of thing, I think is probably the most straightforward approach. LA County and then the state, we can talk more about California, but we've really been pushing for CGM and then glucometer access, and that should be improved, at least in January now, kind of. But I think as far as the meters go, really thinking about the ones that can do it. And then the simplest approach is just doing it through the app for that specific meter. We currently don't use a universal platform because it's a lot for everyone and you have to pay for it. And we definitely don't have the opportunity for that. So I don't know if you all have answers. No, I mean, I agree. I think there are options for meters that have their own app. There are some other apps that can connect with meters, whether it's like MySugar or SugarMate and some of those other things that you may be able to import through. iPhones can incorporate data through Apple Health, through the cloud. So it really depends on the product, depends on the meter, and unfortunately, again, for us poor endocrine doctors who are underpaid, it means educating yourself about all the particular brands of devices out there and which ones work best for your patients and then advocating for coverage when it's available. I also send a lot of people to YouTube. That's not maybe appropriate, but it's a good way to learn from other patients. Well, thank you so much to Drs. Ferlenza, Scott, and Raymond for the wonderful presentations and for everyone for joining us. I'm sure the presenters are happy to stick around for a few more minutes if there are additional questions. Enjoy the rest of the meeting.

automated insulin delivery

artificial pancreas systems

type 1 diabetes

hybrid closed loop

fully closed loop

continuous glucose monitoring

beta bionics islet system

fully closed-loop algorithm

smart pen technology

diabetes care

diabetes technology

virtual care

patient-centered outcomes