Irl Hirsch, MD, MACP, and Diana Isaacs, PharmD, BCACP, BC-ADM, CDCES, FADCES, FCCP, authors of the recent publication, “Clinical Approaches to Managing Glycemia with the iLet Dosing Decision Software,” discuss the iLet Bionic Pancreas and its adaptive technology. They explore how the iLet shifts diabetes management away from pump setting adjustments toward behavioral education and support. The conversation turns to special considerations, including meals, exercise, alerts, and more. Learn how iLet’s tools can reduce the day-to-day burden of diabetes management and what the future may hold for automated insulin delivery systems. This episode is sponsored by Beta Bionics.
The article referenced in this episode, “Clinical Approaches to Managing Glycemia with the iLet Dosing Decision Software,” can be found here.
Patrick McMahon
Welcome to Diabetes Care Conversations, the podcast from the Association of Diabetes Care and Education Specialists, where we explore the latest evidence, innovation and practical strategies shaping diabetes care.
I'm your host, Patrick McMahon.
Today we are joined by two nationally recognized experts in diabetes technology and clinical care, Dr. Diana Isaacs and Dr. Irl Hirsch.
Together, they're the authors of the recently published paper, “Clinical Approaches to Managing Glycemia with the iLet Dosing Decision Software” in the journal of Diabetes, Technology, and Obesity.
You can find this paper in the show notes of this episode.
In this episode, we'll take a closer look at the publication, explore how diabetes care and education specialists can help people with diabetes successfully adopt and benefit from this technology.
Whether you're familiar with the iLet bionic pancreas or just beginning to explore emerging diabetes technologies, this conversation will provide practical insight you can apply into your own practice.
Today's episode is sponsored by Beta Bionics.
Dr. Isaacs, Dr. Hirsch, can you please introduce yourselves to our listeners and why it's important to write this paper now?
Diana Isaacs
Thank you so much for having me here today. My name is Diana Isaacs. I'm an endocrine pharmacist at Cleveland Clinic. I'm also the Director of Education and Training and Diabetes Technology. And I think iLet is just a newer idea of a pump in terms of there's not all of the settings to adjust. And I think that is great. But a lot of people have questions like, what do we do in the visit? How do we change things if there's nothing to change? And so we felt it would be really helpful, hopefully, to work on this paper so people have a great resource to know how to treat patients that are on the iLet.
Irl Hirsch
And thank you, Patrick and Diana. My name is Irl Hirsch. I'm a professor of medicine at the University of Washington in Seattle. I started the diabetes clinic here, and right now I'm involved in many different clinical trials. And I was actually involved in the pivotal trial for the iLet. And it was quite an important and actually fun learning experience for us as this is. very unlike what we've been doing with insulin pump therapy. I've been involved with insulin pumps literally going back to the 1980s. And this is unlike anything we've seen. And I thought this paper would be a great opportunity to sort of explain to everybody what we've learned because this is a little bit different than what we are doing now with other automated insulin delivery systems.
Diana Isaacs
I'll just add also, we at Cleveland Clinic were also part of that pivotal trial as well. And in that trial, I had the opportunity to be an insulin pump trainer and starting people on the iLet. So it's really been amazing to see how this has gone from clinical trial to like here we are in the real world available for everyone now.
Patrick McMahon
Thank you both. You both have lots of experience and insight that we can learn from.
Doctor Isaacs- what is an automatic insulin delivery device continuum or AID continuum? And what should listeners understand with system adaptation or autonomy?
Diana Isaacs
Yeah, I mean we have so many different AID systems now. There's actually five different algorithms out there. And when I think about the different algorithms on one spectrum, basically you've got your program settings and algorithms that are really working off of those settings and then making adjustments from there. So with systems like that, it is pretty important that you dial in those settings because if those settings are way off, the adaptation's gonna be way off. On the other end.
Fully automatic. And right now we don't have any system that's 100% automatic artificial pancreas. Because to me, that would mean you put it on, you don't do anything, you don't even need to announce a meal. It figures everything out, right? And then you've kind of got a spectrum in between where you have some systems that are a little bit more automated than others, some that work off of the existing basal rates that are programmed, some that don't. And really, I would say iLet is in a lot of ways, the furthest along because iLet doesn't have any programmed settings. It is automated. And there's actually three different algorithms. There is a background automated algorithm. There is a correction algorithm. And then there's also the meal announcement. This is still a hybrid closed loop. So there are still meal announcements. Now that's not entering in exact carbohydrates. It's entering in breakfast, lunch, dinner, usual less or more.
So it's definitely more simplistic on that, but there's still that hybrid component.
Patrick McMahon
Thank you. And thank you for explaining the 5 main algorithms and how the iLet fits in.
Dr. Hirsch, in your paper, you and your colleagues discuss management strategies for clinicians managing individuals using the adaptive iLet dosing decision software.
How does this iLet dosing decision software work?
And tell us the shift from adjusting settings to behavioral education and what that means for the clinician or patient.
Irl Hirsch
Yeah, this is really an important topic because coming into a room and seeing a patient on an iLet is quite a bit different than coming in and seeing a patient on any other system. And obviously that includes the open loops because we're dealing, as you say, more with behavior and we're not dealing with settings. It is very common when I am with an endocrine fellow, they feel like they have to make a setting change every time, no matter what. And even if it is with one of the pumps that you have setting changes, that's always not necessary because it may be a behavioral issue. It may be that they're not bolusing. It may be that they have severe depression. They may have been quite sick. There are all kinds of things. And what we teach the fellows and what I think is really important for everybody to understand is you don't always have to make an adjustment change, a change in one of the settings. And as a matter of fact, there's very little to change in the iLet. If you go to the Beta Bionics website and you look at the real-world evidence, what you see is all these people starting the iLet with double-digit A1Cs above 10%. And by the same token, you'd see quite a few with A1Cs down into the sixes. But what's so interesting is that even the people with A1Cs with this algorithm on the iLet, as a rule of thumb, everybody comes down into the sevens, the mid sevens, the low sevens. And when I first saw that from Dr. Stephen Russell, when he showed that, which is now on the website, that really blew me away. And one of the things when we talk about different populations, I don't think we talk about enough. And, you know, we have all these different populations that we talk about. But one population that is usually not included are patients with mental health challenges, patients with severe depression, patients who are not really interested in their diabetes. It could be an acute situation where maybe they're going through a change of life, the death of a family member, a divorce, transitioning to college, maybe. I mean, there's all kinds of things and diabetes is not the priority. And what is so interesting to me is when diabetes is not the priority and you have a double-digit A1C, often because they're not bolusing, they're not paying attention, they're not eating the right things, whatever it is, the iLet is able to adapt by these three different mechanisms that Diana talked about, and the blood sugars come down. Nobody's perfect.
But the blood sugars come down as a rule of thumb to A1Cs, averaging in the mid sevens, no matter where you start. And I've never seen anything like this. I've been doing this since the beginning of pumps in the beginning nineteen eighties. And this is really novel.
Patrick McMahon
These a big changes to management. And important in letting the system work to have that kind of impact. With education and support this seems like a great tool for clinicians and patients.
Dr. Isaacs, anything that you would want to add with working with special populations?
Diana Isaacs
So in the paper, we actually address several kind of special populations. But one thing we have in there is a case study from a real example of a person that developed a bacterial infection and was put on high dose steroids. And it was a short course over six days. It's so interesting because it showcases the time and range and the steroid escalation in the taper down and how it adapted so quickly. And I think it's really because of the fact that it is focusing mostly on that last 24 hours. Many systems, depending on the system, it may look at the last week, it may look at the last several days, but this is actually really focusing on the last 24 hours. And so when there is sickness, when there's steroids or additional medications, it does adapt very quickly. Nothing's perfect, but it's adapting very, very fast. And we also address several specific situations in the paper. There's a whole chart that kind of talks about it.
Like what are the strategies that you can do around exercise? Or a common thing that comes up is, well, what about gastroparesis? And how do you time this since there's not a quote unquote extended bolus? And so we address that kind of the different timing. Usually you don't need to make a change. That's what's so great about it. You don't have to remember all of this, like, let me put it in this mode, let me do this temp basal, this extended bolus. Like it's really designed to be automatic. Just wear it and let it take care of those things so you can focus on the other things in your life.
Patrick McMahon
Yes, if you are listening, and want to see the ways to use the iLet in special cases, there is a wonderful table that illustrates these special populations. You can access the paper in the show notes.
Dr. Isaacs- in your experience, what are the three keys- to success in managing type one diabetes using an AID system like this?
Diana Isaacs
You know, it's great when you put someone on and they're meeting their goals and great, you could focus on something else. But we know that doesn't happen all of the time, right? And so I would say key things, especially with this system, are keeping it filled with insulin and keeping it running. And I know that sounds like, that's so obvious, right? But you do need to charge it. And so the people that do the best are charging it a little bit every day, maybe when they're taking a shower, charging it for a few minutes. And then in addition to that, the insulin.
In terms of the insulin, this cartridge holds up to 180 units, or it works with the pre-filled Fiasp cartridges, which hold 160 units. So we know, you know, some people are on very low doses of insulin, some people are on higher doses of insulin. And so it's about developing a routine for okay, when do I usually go through it? If they are on low doses, fine, maybe it'll be every three days. But if not, it's really coming up with that routine of okay. this is when I need to change it so I don't run out of insulin or checking it before you go to bed so you don't run out of insulin, right? So that right there, those are are key things that will really help someone to be successful. The other thing, and this was something that I really learned through using the system more, is about the response to alerts. Actually, there's not a ton of alerts with this system. I mean, there's the alert if you're running out of insulin and you need to charge it, but otherwise, there's actually specific hypoglycemia alerts and hyperglycemia alerts. For example, the hypoglycemia alerts, there's one where if you're dropping below 75 or you're dropping quickly. And the reason why they are at these amounts is because it's designed that if you can catch it at that point, you can treat with like five to 10 grams of carbs, and you may be able to prevent going low at all. And so I am finding in my visits. That is what we're focusing on. Instead of making a specific change, like, okay, let's change your carb ratio because you're going too high or you're dropping too low. It's really more about, let's talk about these alerts. Let's see what's happened. I see that you did go a little bit low and then you went up a lot higher. So let's talk about maybe the strategy of recognizing that initial alert, treating it, and then preventing going low and preventing going higher.
And then the other thing I want to mention is the meal announcements. And, you know, the data shows that people that are able to regularly make meal announcements can generally have more time in range. So they're three big meals, or if they're not eating three main meals, whatever their main meals that they're eating. But just kind of getting into some good habits about announcing those, that can really further enhance the time in range. So it's really like I'm finding those are the things we focus on.
Patrick McMahon
Doctor Hirsch, anything you would like to add?
Irl Hirsch
There's so much. I wanna give a very quick case presentation that I think brings out these points that Diana was just talking about. It's a lady in her early seventies who was in our pivotal trial, but the trial was over, the pump wasn't available yet, and she had to go back to her previous pump, but she really missed it. And I think the reason why she missed it is because she's starting to miss some of her boluses. And she did better on the iLet than she did on the other pot. So eventually it was time for her. She was able to change up the pumps and she did very, very well. But then something else happened. She has early diabetic kidney disease and she came into a trial that we are doing where we are, of course, blinded to the drug, but she either received a GLP1 or placebo. And I'm convinced she received the GLP1. And as adaptation with everything else, the iLet with the GLP1, which technically is off label, but she was part of a clinical trial, she did absolutely great because of the adaptation of the pump. And it is my belief within the next two years we'll have one GLP1, and maybe within the next three to five years we'll have two incretins approved for type one diabetes, not to mention type two patients using more AID systems. But the point is for elderly patients and for the addition of a GLP1, we've seen firsthand how well this system does.
Patrick McMahon
When we're thinking about the clinician AND also the patient perspective.
How does the clinician's role, especially that of a diabetes care and education specialist, change with this new technology?
Diana Isaacs
I think that it's really instead of focusing on settings and worrying about all of that, it's really focused on the behaviors. And that's something that diabetes care and education specialists already are very familiar with and have usually a lot of confidence. There's something on the report, on the Beta Bionics report, called this insights. And I find that to be pretty helpful because it will actually point out some areas to focus on.
It actually focuses on the positive as well. So it might say, hey, this person's really in range overnight, or they're having a nice smooth recovery after Lowe's. But if there are opportunities for improvement, it will highlight those. And for example, one that sometimes comes up is the meal announcements. So are meal announcements being announced ideally before the person eats, or at least at the start of the meal? Or how are they being announced? Because usual is usual, right? So you would expect a usual meal occurring 50% or more of the time. And so if you see that usual is less than that, that then opens up the door to this discussion of well, what maybe is happening, or maybe reevaluating that usual isn't actually usual. Maybe usual is more, it's less, and kind of redefining what those look like so someone can really get the right amount of insulin to match what they are eating.
There's also areas in terms of like the timing of the meal announcements. You know, we see all kinds of things, right? Sometimes people announce meals, they're like, I'm high. So I'm just gonna announce a meal and that'll bring me down. But that's a behavior we generally don't want to encourage because that can mess up the meal adaptivity. And also it could cause someone to go too low because the amount that someone needs to correct a high can be very different than what they would need for their meal. And so a lot of it's just reinforcing the behaviors, even if you see on a report, they're running out of insulin. And this was something that I find the insights is very helpful with because when I first started looking at these reports, it wasn't always clear to me, well, wait, is the person out of CGM? Are they out of insulin? Did the battery die? But now, as you get comfortable with it and you see these insights and you read the key, you're like, wait, I see. Okay, they ran out of insulin. All right, we can address this and kind of trying to find out what led up to that. And then working with the person, really partnering with them to come up with strategies to overcome those barriers.
Patrick McMahon
Great. I love that. Doctor Hirsch.
Irl Hirsch
The behavioral issue can't be overemphasized. And it doesn't matter the age of the patient, but the one that I think needs to be addressed when we are doing the initial training and not waiting. A lot of things can wait, and you see what the behaviors are. But the one that may be the most important is how to deal with exercise with the iLet. And I say that because what people have traditionally done is carb up. I'm using air quotes here.
Carb up. And that's like the worst thing to do because the iLet will adapt, give you all this insulin on board, and you see hypoglycemia every time because there is so much insulin on board from carbing up. And so a very key behavioral point is you want that insulin on board, which for exercise includes both the basal and the meal announcement, you want that on the low side. And so you don't want to carb up.
And in fact, what we've learned is that in some patients to avoid hypoglycemia with the iLet, not only do they not carb up, not only do we maybe set the target to a higher level, but we may even pause the pump for 30 minutes prior to exercise so that we really do a good job of reducing that insulin on board.
Patrick McMahon
Great. Dr. Isaacs, you mentioned the insights report. Which can be used for management as well as partnering with the patient to help meet their goals. Where else can we find these special considerations?
Diana Isaacs
In the paper, I'm really encouraging people to go to this paper because there's a chart that mentions all those strategies for exercise. And I think it's so good to ask the person, like, what are they currently doing? Like what has been their strategy, especially if you see lows in their report, like that can really lead to this discussion and coming up with the best plan for it.
Patrick McMahon
Great, wonderful management tool.
From a patient's perspective, what would you say to someone who is interested in the iLet but nervous about giving up control of their insulin dosing?
Diana Isaacs
So I think control is an illusion. The reality is insulin is a difficult drug to work with. It actually has a very narrow therapeutic window, meaning that you just get a tiny bit too much and you can go low, right? And you don't get enough, you go high. And we know there's at least 42 factors that impact glucose levels. It is really challenging to self-manage this. I think if we had an automatic tool, if I mean when you think about it, when I go on a plane, I don't go and check with the pilot and ask him how he's gonna press all the buttons and like what's gonna happen. Like I just go and sit in my seat, open my laptop and do something else, right? Or go to sleep. So I think like if we could have that, I think that would be the dream. Now I get it, people have been used to having to micromanage their condition for a really long time.
So I get like it's hard to give up that kind of control. We get it, but I guess I just encourage people to give it a try because the headspace that you can free up and hopefully just focus on other areas of your life. I think, you know, is something a lot of people say really feels very good and and reduces all that diabetes distress.
Irl Hirsch
And I'm gonna say something that maybe not everybody will agree with. And Diana, I'd be really curious about your thoughts. But there are some people who are getting up in the middle of the night to see what's going on. This pump may not be for them. They want complete control. They can't give it up. Now that's a minority of patients, but we do see people like this where this may not be a good choice for them.
But with that being said, if you look at the big population, you look at the overall time in range in hemoglobin A1C, especially when we get down to the more challenging ages in the adolescent age group, that's not always the case. And so I think like anything else, we want to put the right pump on the right person. And there are going to be some people that don't want to give up their complete control of their diabetes. And this doesn't give up complete control, but like Diana said, everything is done in the background and the patient is able to step back and let the technology do the work.
Diana Isaacs
Yeah, I actually completely agree with that. I mean, that is why we all offer choice to our patients. And my whole thing is I just want people to be aware of the different options that are out there. But yeah, if someone wants to be able to micromanage it and be able to change every single setting, there are pumps that you can do that with. And so I think it's just being honest with what each pump can do.
Letting people know about the choices and then letting them try it. And fortunately, we're also getting in an era where it's easier to start pumps. Many are going through the pharmacy benefits now. And so people can try something and they can also try something else. And that's okay. And I think as we see, I know currently this pump is on label for type one diabetes, but as we see things expand and we certainly see more people with type two diabetes going on AID systems. Finding ways where it is simpler. And especially as I think about all of the primary care providers who may put people on pumps, the idea that you don't have to calculate settings to get someone started on it is also really appealing because there are pumps out there that are either running off of those preprogrammed settings or you have to start in manual mode with those settings and then you can transition to automate it. And so for a healthcare team that may not feel as comfortable with doing those initial calculations, that can certainly be more challenging. Or even if you are comfortable with those setting calculations, I have seen like massive differences where people needed like half the amount of insulin that they were taking or that we thought they were taking. So it's definitely appreciated kind of the ability to automatically calculate those settings.
Patrick McMahon
Thank you. So as we wrap up here.
Can you both share your visions for the future of AID systems?
Irl Hirsch
Well, I'll start I've been doing this a few years longer than Diana. Honestly, if we're going to look ahead thirty years, my own personal hope is we're not even using AID systems. We're using something even more novel, whether we're doing cell therapy without immunosuppression, intrathecal leptin, which I'm working on, there's all kinds of things. But having said that, not knowing if that's gonna happen, we hope it does.
And certainly looking over the next five or ten years, where these kinds of transformational therapies won't be available for everybody, what I see happening is moving to completely closed loop systems where you do not have to announce the bolus.
My father was so interesting. My father could not give up the control of a stick shift. And when everything moved to automatic transmission, my father was one of the last to change that because he wanted the control. And I think there are going to be people, especially with type one diabetes, who are going to be like that. But what we're going to learn over time is not everybody, but most people are going to want to go to the automated systems where the iLet is right now of the algorithms we have, the smartest of the systems, simply because of what Diana said, it's going to allow more space, more bandwidth for other things and not for the diabetes. And so I see us moving to more automation.
My father was one of the last people to move away from the stick shift. I guess I was probably in high school when he finally did that because my mother couldn't deal with it anymore. But the point is is that I see the automation only increasing. And I think we're just seeing the first pump of many that's gonna have less burden on the patient with regards to their blood sugars.
Diana Isaacs
Yeah, this is a really fun question. So I agree, like ultimately cure is is what we are really hoping for, however that comes. But in the meantime, with the technology, yeah, fully closing the loop, maybe through the use of more AI as well as faster insulins, so that it really becomes like an artificial pancreas that you put it on, but you're really not having to worry about it. I think we're gonna get to a place where we have tubeless options for everything pretty soon. And also probably smaller devices, you know, we'll we'll get even better CGMs, even more accurate, better signal connectivity that hopefully it's just it's like you barely even, you know, notice it's there. And I the other thing is that it really becomes the standard of care. I know in our guidelines we talk about how AID is the standard of care for really anyone needing meal time insulin, but we're certainly not there. Type one has improved dramatically in terms of the rates of people on AID, but there's still room even there to expand. And type two is a very untapped area in terms of, you know, a lot of people have never even been offered this technology. So that's where where I see things going.
Patrick McMahon
Dr. Isaacs, Dr. Hirsch, for joining us and sharing your expertise and vision for the future.
Today's conversation highlighted autonomous insulin delivery is not simply about introducing new technologies. It also changes how clinicians educate, coach, and partner with people with living with diabetes.
To our listeners, you can find the publication, Clinical Approaches to Managing Glycemia with the iLet Dosing Decision Software, along with additional resources in the show notes.
Today's episode of Diabetes Care Conversations was sponsored by Beta Bionics.
Thank you for listening. I'm Patrick McMahon, and we hope you'll join us for another conversation focused on practical, person-centered diabetes care.