Insulet Corporation (PODD) Earnings Call Transcript & Summary

Good evening, good afternoon, good morning, wherever you are in the world. This is our post-ATTD or near-the-end-of-ATTD conference recap here with Dr. Trang Ly from Insulet. As many of you know, I'm Jeff Johnson from Baird. I do have one disclosure to read here. Please refer to the Event calendar, published research or Baird's website for important disclosures regarding the company's discussed during this event. And with that out of the way, Dr. Ly, you and I are just going to talk a little bit today about Insulet's presence at ATTD and maybe some bigger things we saw coming out of the conference. So thanks for taking the time.

Yes. I'm good. Great to be here, Jeff.

Yes. Great. Thanks. All right. So the first thing I want to talk about, not to make myself look older than I already feel with all this gray hair and that, but I think this is my ninth ATTD, and you're so much younger than me.

I beat you. I beat you.

I know. You're so much younger than me, but you'd still beat me. So you're more seasoned, we'll say that. But I think -- I was looking back at 9 or 10 ATTDs ago, and I remember pictures of a little kid hooked up to a laptop and a big pump with all these infusion sites go into the -- and all that, and us talking someday about having automated insulin delivery devices or what we called back then artificial pancreases. And we're there. And I think what really struck me this year at ATTD, more than anything, is not only are we there, but these devices are getting so much easier to use. The outcomes are improving so much. The form factors are getting so much better. I mean, we're just -- from a high level, where do you feel like we are at ATTD this year technologically than even a few years ago?

Yes. I think there has been really extraordinary advances in the AID technology and just the simplicity and ease of use with all the technology in both sensors and with pumps and, obviously, with Omnipod 5. And yes, so 10 years ago, we were presenting data from like 12 patients and 20 patients, really, inpatient studies. And now the fact that Omnipod 5 is on 250,000 people in the world is really extraordinary. So the scale and the reach of the product has been really incredible to see. And now we're talking really beyond AID use in type 1 diabetes, but now also AID use in type 2 diabetes really being that next frontier.

Yes. No, that's great segue to one of the questions I want to ask. So let me back up to type 1 before we go to type 2. And on type 1, you just published this afternoon or just presented this afternoon, your randomized controlled trial data on Omnipod 5 versus non-AID, I guess, open-loop systems, if you want to think about it that way. Tell us about the importance of that trial. It seems like time in range improved by about 16 points, if I -- if memory serves from that trial. And what kind of doors will that study open up for you?

Yes. So we have lots of different data sets, and we're really proud to be presenting the data from the Omnipod 5 003 study. So this was a study in about 200 adults with type 1 diabetes in the U.S. and France. And they were randomized to either continuing using the pump -- the stand-alone pump and CGM, all randomized with the Omnipod 5 system. And after 3 months, we saw a 17.5% increase in time in range, up to about 62%. And this is equivalent to 4.2 hours of additional time in range. And the reason why this is really important is that it is not a population that is commonly studied. So it was a population of people with an A1c that was greater than 8%. 80% of the subjects have A1c greater than 8%. And these are people who have suboptimal glucose control, so people who are not managing that target. And these are the people who are at the highest risk of complications of diabetes, and all the significance of the morbidity and the lost lives that comes with really diabetes. So really being able to show that our technology works in this population, that it's simple and easy to use, has incredible improvement in glycemic outcomes in that population. It is really important to show, especially for payers who are always trying to think about, "Okay, if we pay for this technology, how much benefit are we actually going to see in the population?" So really being able to show the value of the technology in large populations and where its nationalized health care is just really important to demonstrate that ongoing value. And that allows us to be in a much stronger position as we negotiate premium pricing for Omnipod 5. And that is what enables us to continue to innovate and bring better technology for our patients.

And I'm sure no coincidence, that part of the study, I think 1/3 or 40% of the study was run in France, one of the markets that will pay a premium for AIDs if the data is there?

That's right, yes. So we are still in negotiation with the French government. But if -- there's one thing to say you're coming, but to actually be in the country and have French patients on it and have all of our key opinion leaders advocate for this product and speak with such enthusiasm about Omnipod is just really incredible to see. And so that's why we've invested heavily in running our clinical trials outside of the U.S. and targeting specific markets such as France, where that is just that patient advocacy, the physician advocacy is really important, so for them to be able to offer the best care for their patients.

Yes. That's great. And then one last RCT question, then we'll come back to type 2. And I'm blanking on that question now, which -- oh, I did get a couple of e-mails right after the press release said the time in range of 61% in 005. Again, you addressed it or tied together for me the higher A1c starting point and that time in range of 61% as opposed to the 69%, 70% we saw in the old 005 pivotals.

Yes. So when we look at time in range and A1c reduction, it's really important to understand the population that you are testing the system in. So we -- generally, if you start with a very high A1c, it's your -- you generally get a very high delta in terms of time in range improvements, and you really have to really see. And what we understand from our conversations with physicians and payers is really that the magnitude of the difference and really being able to show that it works in a large population of users is really important to them. So the 61% is a vast improvement. It was pretty much a 50% improvement from the starting time in range. And that is what keeps people out of trouble in terms of the kidney disease and all the comorbidities that come with diabetes. And that's the population that we really need to reach. And we need simpler and easier systems to reach that broad population of users. And that's how we get pump penetration in the general population, and not just the patients who are already our goal. We're really trying to lift the entire population here.

Yes. No, that's great. And then, so let's go over to type 2. We've talked about increased penetration. This is probably one of the biggest opportunities to increase pump penetration over the next 5 to 10 years. I think the news coming out of this conference wasn't big news. But on your conference call, I think Jim had talked about the SECURE-T2D trial, sorry, was almost complete with enrollment. It is now complete with enrollment. And it sounds like we should get a first read of the data off that SECURE trial at ADA in Orlando this June?

Yes, that's the plan. Yes. We just wrapped up our last patient [ decert ], and we've been working hard to close up all the activities to get that out. So really important for Insulet to demonstrate that the product is safe in the type 2 population, and that data will serve as the key clinical data for our FDA [indiscernible]. So as a reminder for everyone, there is no AID system that is cleared for the type 2 population today. So it is -- it will be niche market. And so we expect a close review by FDA. And so we've said publicly that we'll be submitting at the end of 2024. And so we hope to bring Omnipod 5 out for our type 2 population beyond that time.

Okay. And maybe a couple of data points that did come out at this conference, and I don't know if I can mention Medtronic or Tandem, if you're going to bite my head off if I mentioned those companies. But...

I won't bite your head off, yes.

You won't bite my head off. Okay. Thank you. No. So they both showed some real-world data on AID use in type 2. Obviously, these type 2 -- or these AID systems can be used off-label. So there's data out there. And I mean it looks really impressive. It looks like we're getting time in range improvements and to levels that are pretty equivalent with type 1, maybe even superior to type 1. I always describe to investors type 2 diabetes, IIT type 2, is just very similar to type 1, only you kind of get it to yourself later in life instead of there being a genetic component. Maybe that's not the best description, but for us lay people, it works. But I guess, talk to me about the importance of this real-world evidence that truly is showing good improvement and what that might mean for AID penetration in type 2 over time.

Yes. So we've also shown real-world data in type 2 diabetes. I think we showed it at ADA last year.

I keep forgetting that. Apologies.

That's okay. In a few thousand people. And so we shared exactly that last year. And what we showed there was, in the population of type 2 users, very -- actually very similar time in range to our type 1 population. And that's an indication that it is highly effective therapy. And so it was a combination of our real-world data, the very real clinical benefits that we see with Omnipod 5 therapy in off-label use, but also the feasibility study that we conducted and published in diabetes care, I think, last year, the 24 patients that led us to invest in our indications to do with the SECURE-T2D study. So I think we have shown like the others that Omnipod 5 use in type 2 gets really good clinical benefits. And I agree with you. It is actually -- if I look at just the individual traces, it's actually hard to tell what is a type 2 trace versus a type 1 trace because insulin therapy is so effective and needed in this population. So really, it's really great to see lots of data from the companies showing the efficacy of AID use in the type 2 population.

Yes. And then one of my arguments over the years has been type 2 patients may -- kind of like children, if you look at the 2 ends of the barbell, older patients and younger children be really drawn to the form factor of a patch pump over a tubed pump. So it's maybe a softball question, but when we do start to see more penetration into that type 2 category, it seems like patch pump over tubed pump might be even more preferred than we're seeing in type 1?

Yes. Well, we've seen that already. And as a reminder, before Omnipod 5 was launched, we had Omnipod DASH, which is our stand-alone insulin pump. And in a world where there was AID and tubed pumps, 40% of our new patient starts were type 2 users. So it really shows a remarkable impact of having a simple and easy-to-use system on the market and how you'll really be able to influence care in that population. So I definitely think once we get the indication, we're going to be the #1 pump for the type 2, just like we are -- well, actually, we're already the #1 pump in type 2 with Omnipod DASH alone. But today, we are the #1 pump in the U.S., most prescribed pump in 2023 for type 1. And there's all the confidence that we'll continue to grow and really change the landscape and -- on a population level of the type of care that people are able to get with Omnipod.

And what would be your guess? I know it's -- we can look at how many endos exist in the U.S. and how many U.S. counties versus what U.S. counties have primary carriers in that. But I just find it interesting, if not insightful, the way Jim has talked now over the last quarter or 2 about increasing investments in the PCP channel. And now we're getting closer to type 2 where we know -- type 2 approval, where we know type 2 care is significantly delivered more by PCPs over endos.

Yes. So I think that it's sort of new ground for us, and so there's a lot of learnings to be had. I think on the call, Jim said we're up to 18,500 HCPs prescribing Omnipod, which is like mind-blowing. And I've seen the spreadsheet. It's like a really broad population of health care providers. So we're clearly reaching just a lot of doctors, a lot of nurse practitioners, PAs. So clearly, people are very confident to prescribe this product. And also -- and we certainly said this on the call, also just seeing the number of scripts per HCP also rising is just really incredibly encouraging. It wasn't that long ago that I was asking how many -- what's our top 100 prescriber look like? And we're like miles away from that. So it's really incredible just to say, look, if you've got as simple and easy-to-use product and people prescribe it for a couple of patients, I think their patients do well. Like the potential for growth for that product is really remarkable.

Yes. And do you have personal experience talking to these primary cares? I'm sure the answer is yes. But how hard is it going to be? Because I think a challenge 3 or 4 years ago, Dexcom and Abbott, I'm sure, had was, can these primary carriers get comfortable with prescribing CGMs? They have. They've clearly become comfortable with that. How comfortable can they become or how quickly can they become comfortable prescribing a pump? And how might a tubed pump versus a patch pump help them get there faster? For patch pump versus tubed pump, I'm sorry.

Well, I think simplicity and ease of use and -- is very, very important in the PCP channel. I went on a ride last year in PCP land, and it's like different medicine is practiced there. And you've really got to make it appealing. You have to show clinical benefits, number one. And so you need to give them a solution that's actually going to help their patients -- in their challenging patients. They want to give the best care for their patients. They don't want to just like refer on and never see their patients again. So they do want to offer better care. But your -- our solution, whatever we deliver, needs to be simple and easy to use. It can't be cumbersome. It can't take 10 hours to train, right? So -- and then what companies -- like how companies need to consider is like how do we best support these PCPs at scale in a way that will ensure that our patients get the best outcomes. And so I think still more to learn from the patients. So more to learn from that population. And Omnipod GO, our basal-only product, is in a pilot right now. And we're learning a lot on the doctor's perception of that technology. These are new-to-Omnipod doctors that we're talking to. And they're very interested in the whole portfolio of products. What does DASH do? What does Omnipod 5 do? And so really being able to have that conversation about technology and delivering better care is something that we're very excited to really get into.

Yes. And Deb is not on this call. I'm sure she would want me to say it now or you to say, don't put Omnipod GO in your models this year. This is a 2025 modeling event. So...

Yes, we're still in a pilot.

Exactly. We will declare that. All right. Maybe switch gears, I've got maybe 2 or 3 other topics I want to cover. In a presentation yesterday afternoon, it was in Bruce Bode's presentation, I think it was -- who's the guy from Minnesota, Lars -- works with...

Anders? Anders Carlson. Yes, yes.

Oh, yes. Sorry, not Lars. Yes, yes, works with Dr. Bergenstal, right, I think. Am I right?

Yes, yes, yes.

Okay. Sorry. So in his presentation, he showed some interesting data and a point I hadn't really thought of, and at some of these type 2 patients who go on a GLP-1, maybe the treatment protocol becomes they go from metformin to GLP-1 before they go on to insulin makes total sense. But as their weight comes down, they still progress to needing insulin, but now, all of a sudden, they need a lot less insulin. And their insulin sensitivity is up-regulated, if you will. And so an AID can actually get better results than if the patient was not on a GLP-1. So all of a sudden -- I don't want to steal the words Dexcom uses. And the word, I think, that has helped get investors over some of the concerns on GLP-1 and CGMs, but maybe there's a synergistic effect between GLP-1 and AID use here. And I think some of the examples shown in that presentation were patients who might have been using 110, 120 units a day, if not higher, down to 60 units. Now all of a sudden, at 60 units a day, they become a true candidate for an Omnipod. So I know that's a lot of kind of background, but talk to me about maybe the ability to synergistically use a GLP-1 and an AID system.

Yes. So there's nothing stopping anyone from using Omnipod 5 or an AID system with a GLP-1. And so it's exactly what we've been talking about since all of this happened. But it's really great to see people like Anders and actually Greg Forlenza and, yes, very professional. So he talked about this and -- so it is excellent therapy, but they're still on 60 units a day. And I think Greg actually showed a review article where you got it down with SGLT2, he's got total daily doses down from like 160 down to 120. But as a reminder, this is still a 120 unit. So there's definitely a reduction in insulin, if you're heavily insulin-resistant, because of all the great effects that we -- well-characterized from GLP-1. So with the reduction in insulin, it just makes that into work a lot more effectively. So what we've seen in some of our early feasibility work with type 2 diabetes is that when you add in that GLP-1, you get that additional improvement in time in range, just it makes it a little easier to get that lift in time in range. So we see them working together really nicely. And then the other thing with GLP-1s that we've seen is that, yes, they lose weight -- people with type 2 lose weight, but their A1c creeps up over time. And that's because it's not addressing the underlying disease progression, right? There's no reversal of underlying disease progression. People -- A1cs continue to creep up over time. And we really see that intensive insulin population really continuing to grow over time because it's not curing the underlying disease.

Yes. No, I think those are all fair points. And I don't know if you even saw yesterday, Cigna came out and said that they've entered an agreement with Novo and Lilly to not increase their spending on GLP-1s by more than 15% a year. So there's already some pushback now on how faster GLP-1 is going to be able to penetrate the U.S. if payers and really the employers who put these payment plans together push back on this rising cost of GLP-1. So that's good and bad. I mean, obviously, it's good for society that they lose weight and get on these drugs, but maybe a good data point for the pump companies, too, if we were concerned about GLP-1 and pump use down the road. Good. Well, maybe we'll show over to one last topic -- or 2, I guess, topics. One, looking forward, maybe it's forward a couple of years, maybe it's forward a certain unknown period of time, but you also showed -- or one of your physicians from New Zealand showed some data out of the -- early data out of the EVOLUTION study. And that study is on a new algorithm. It seems like that new algorithm is -- must be focused around meal time because part of the data that was shown, I think it was 9 type 1 and 8 type 2 patients, so a small data set, or 9 and 8 that completed. But out of that data set, these were patients who did not bolus at all over a 3-week period using their Omnipod 5 system and time in range with no boluses at all. But 61%, I believe, in type 2. And I've got...

65%.

65%, I'm sorry, yes. And up 57 -- upper 50s. You know the numbers better than I, but 57%, I think, 58% in the type 1 category. So one thing, and I talked about it earlier when we -- 10 years ago, we were seeing pictures of these kids hooked up to these big computers and that, I mean, we were actually -- I don't think close to closed loop as in the next year or 2. But I also think -- I'm starting to hear at this conference for the first time some talk about, "Hey, these can be closed loop if your patients are noncompliant, an 18-year-old who's just not going to bolus, they still get a benefit on these systems because these systems are pretty good." We see that in some Medtronic real-world. We definitely see that in some Tandem real-world of patients who didn't bolus for like 11 weeks on average, and they were still at a time in range in the 60s. And then, obviously, your EVOLUTION data, early stage, but kind of shows some of the same. So where are we going with closed-loop therapy with ability to not bolus at meal time and things like that?

Yes. So we -- that was one of our big data releases this conference, and we're really proud to be able to share that with the world. And it really is just the first glimpse into our next-generation work. And what I'd say is that the algorithm is one component of a successful system. And really, the remarkable form factor that we have with Omnipod and Omnipod 5 algorithm, a really solid algorithm with very favorable hypoglycemia profile is what has made Omnipod 5 so incredibly successful. And so with that system, because we get data from every patient using it, we can really tell from a population level how our patients are using the system, what leads to most optimal glucose control, where are people not doing so well, and what can we do as a company to really simplify their experience using our systems. And the reality is, even with the systems we have today, even with all the advanced closed-loop systems, there's still work that the user has to do in thinking about their diabetes. And we really want all of that to fade into the background. So what we've done here is really taking the data that we have from all of our patients and really leverage that cloud connectivity to analyze and make improvements in the algorithm and then get it on to patients as quickly as possible. And it's because of Omnipod 5 that we're able to make those changes and load it into the system and get it into the pumps and get it on to patients quickly. And we are -- Martin and I were talking about the study less than a year ago at ATTD. And I was like, yes, let's go. Let's do some studies. And so the type 1 population, the -- when you give more insulin, you run into trouble with hyperglycemia. So we really have to understand what are the risks with that. So you can give more insulin, but the outcome of that is more hyperglycemia. We actually see that with some of the other systems. When they go too high, people run into severe hyperglycemia. It becomes dangerous, and then they stop using the system. And that's something that we wouldn't want especially because of our dominance in the pediatric population. So we take a lot of care and consideration into what we release into the world, and we want it to be as safe as possible. And so I think that this enabled us to really understand in a very challenging population of teenagers, young adults, who were -- had a very high A1c really seeing how they did with the type 1 with our algorithm. And then we took the same algorithm, actually, just put them on patients with type 2 diabetes and, yes, small population, 8 users, and we got like 56% to 65% increase in time in range. And what was, to me, really interesting was that all of those patients were either on a GLP-1 or an SGLT2 or both medications and still really having suboptimal A1c and time in range. And so being able to see our algorithm, being able to deliver actually quite safely in the type 2 population insulin that is personalized for them and being able to see that improvement and also like the quality-of-life improvements. I mean, you saw all the quotes from all the patients about not having to think about their diabetes and like bringing them joy. Like that's the stuff that's going to change the way people perceive therapies. So it's -- I would say, it's the beginning of many great things that we're going to do with our next-generation system.

And from my side of the table, I want that at the end of this year, next year, or something like that. But what happens? You take these results. You then go back to the scientists and iterate the algorithms more, do a small trial. I mean is this something that's multiple years away? Is this in the intermediate term? And I'm sure Jim doesn't want you to answer exactly but...

Yes. Well, we don't know. The answer is we don't know exactly. And these are part of the things that you need great teams to figure out, right? Like when are you done? You could keep testing forever, and you'll -- will you ever be sure? But I think really understanding how it works in a challenging population of people is really important for our company, but also for any company out there that wants to bring technology to users. Because it really is about being able to improve care for the people out there who are being taken care of by PCPs. And majority of people out there who can't see and who are living with these conditions and the comorbidities that come with them, it has to be safe and effective. And so -- and people can't develop more complications from it. So the safety profile is something we take really seriously. And that, that might take multiple clinical studies to run for us to get comfortable with that. And the great thing is because we're #1, we have that revenue growth to be able to invest in the type of research and invest in next-generation products, and that's incredible.

That's 13% EBITDA margin or EBIT margin, let's go to 14% and 15% and 16%. Let's not invest it all the way. Mr. Hollingshead -- or Dr. Hollingshead, sorry. All right. Last 2 questions. One I skipped over and I want to come back to, and then one a little easier. So the G7-O5 integration limited launch in the U.S. kicked off in the last few weeks. Any early feedback? Any early data points? Patients love it? Patients hate it? What's going on there?

Patients love it. The -- as like -- I'm never sure like is it going to be enough when they go from sensor-to-sensor? Like what do people get excited about? They're very excited about G7, I have to say. People talk about a shorter warm-up period, being amazing and the smaller form factor, being really a fantastic user experience. So it's been incredible to hear that feedback. And patients do love it. So we're super excited. That's scaling really nicely. And we're looking forward to bringing that to the market.

Yes. And what has to happen to go to a full market release? What boxes have to get checked?

Yes. We want to make sure that the user experience is good and, obviously, that we make enough pods. So there's a lot logistically because it's all of -- a lot of our current customers are also wanting to get on, right? And so we have a lot more customer today, a lot more people are buying than we did a few years ago. So logistically, it's a big thing. And so really making sure that user experience is as good as we want it to be and making sure we can manage that, and the demand that will come with that product is really important to us.

Fair enough. The last topic, and this is the one I skipped over, but I want to come back to now. I think it was poster, 095, 95 or something like that, if I remember right. But it was an interesting poster. It really didn't show any data. It was more chart-driven. But it showed that at meal time, if patients didn't bolus, the system itself delivered a lot more insulin. I think I'm getting the details. Maybe I'm wrong, maybe they just bolused. I'm not sure. But I think I can say this, you wouldn't say it because you're a much better person than I am. But Medtronic has been out there. And then look, Medtronic has done some good work, right, with Klue. They've got their meal-detection bolus, whether that's real, whether it's not, but they've got some things to talk around -- about at mealtime. But they do tend to then also say, "Oh, but look, pod can only -- or Insulet system can only deliver micro boluses or whatever. The data in that poster, if I'm reading it right, really seems to show that there can be a big delivery of total daily insulin at one time if the system senses blood sugar going up. Am I right on that?

Yes, that's right. So when it comes to meal detection, all the systems do it. Our system does meal detection. All the systems do. There's nothing like special about it. It's -- basically, you're predicting a rise in glucose, and it's most likely coming from food. So we -- it's kind of been like so obvious that we didn't even talk about it. So for us, the reason why we did that analysis is we wanted to see how much could the algorithm deliver. And so we actually specifically looked for overnight traces, which were flat, so they would all start off with approximately the same starting glucose and then watching for the meal rise. And we had analytics -- and we had our own algorithm to figure out what looks like a meal. And so we looked at 500 meals without a bolus, 500 in the pediatric population and then 500 in the adult population. And then because we can see all the data where you can see the total daily insulin and then we can see how much of the -- like how much do all the -- does all the insulin micro boluses add up to over a 4-hour period of time, which is what we call the postpone your period which is the time after someone has eaten. And if -- so we searched for these patterns where people didn't bolus afterwards to see like how much could the algorithm deliver. And on average, in the pediatric population, we saw a 16% of TDI. So you can imagine someone's on 20 units a day, we were delivering 3 units of insulin over that period of time on average if they didn't bolus. And the same in adults, so 16%. But actually -- so that's the average. And this is the thing is like a lot of these -- a lot of companies talk about averages, but they don't talk about the range, right? There's like massive range. And so actually, the maximum of 30% of TDI in kids could be given. So that's double -- so that would be 6 units per 20-unit person. So that's a good chunk of insulin. And then up to 38% of TDI in the adults. So really showing that our algorithm really has a robust insulin response based upon patterns of meal detection and works as exactly like we designed it to. So just proving that out and really showing -- and so great to be able to show that beyond just the clinical study, actually, in real world, has been really marvelous to see.

And if I -- let me just make sure I take that down to levels I understand for the investors on the call. I won't throw them under. But when you talk TDI, that's obviously total daily insulin. No, no. So my point is, what you're saying is, if somebody misses a bolus, the system knows from the few days prior how much total insulin it normally delivers, and that's what it plans to deliver throughout that day?

Yes.

And now all of a sudden, it says, "Oh, wait, blood sugar is going up. I can see it. I need to give insulin." And it can, on its own, give up to 30% of that entire day's insulin into kids, 38% in adults. So it is thinking on its own and saying, "Wow, I need to do something."

That's right.

So that shows kind of the automation and just the algorithm behind trying to control these blood sugars. Is that right?

Yes. Absolutely. Because we have such a heterogeneous group of patients that we take care of. And you can imagine that delivering insulin on a 2-year old is very different to delivering on a young adult or a 70-kilo male is very different. So that's why we designed the product to scale with the patient to -- commensurate with their increase in total daily insulin. So you -- and that's why we consider, as a percentage of total daily insulin, to basically show that whatever the patient shape and size is, we're able to adapt to that user over time, which is actually a key differentiating feature for some of the other systems where it requires that fine-tuning basal settings to really get that optimal glycemic control. And so that's a key differentiator. And we've had a lot of clinicians say, "Look, we put them on Omnipod 5." You're not messing around with the basals. Like 2 weeks in, 3 weeks in, 4 weeks in, you just like put it on and get -- as long as they're bolusing, then they're getting really good results from the system. So -- and so this is just, again, helpful data to really show clinicians that being able to quantify that algorithm response.

Yes. I think that's all I've got. What am I missing? Am I missing anything?

Libre? Can we talk about Libre?

Yes. I wasn't sure if I could go there. So Libre 2, you're going into the U.K. and Netherlands in the first half of this year?

Yes.

Still plan to cover the vast majority -- or the majority of your user base in Europe this year with...

Omnipod.

With Omnipod 5. I guess that doesn't necessarily mean with L2 but...

Yes.

So I'm going to ask, I mean, L3 anywhere in the picture or L2 in the U.S. anywhere in the picture? But outside of that, say what you want to say about Libre 2 in U.K. and Netherlands, and then we'll get back to the point.

Well, it's really super exciting collaboration with, actually, both our partners, Dexcom and Libre. And we're actually in a clinical study in the U.K., in Belgium and France. And it's really incredible to be able to see the investigators in person and talk about the amazing effect of being able to give AID to people who've been on injections for so long and the results that they are seeing. So that study is underway and doing really well. And with the CE Mark and the launch coming, it's super exciting for us. So as you mentioned, you were spot on about Netherlands and the U.K. in the first half of this year. And we haven't publicly said when it's coming to other regions just yet, but these are projects that we're super excited to be working on.

Great. Anything else?

G7. Libre. I think we're good. AID everywhere.

Yes, yes. AID everywhere here, I would agree with that. That's all I've got in my notes as well. So thank you all for joining us. I understand there are a few other calls going on around this time in the states. Sorry, I'm looking at the wrong camera. You can tell I'm not TV-trained. But a few other calls going on at the same time. So for those of you who joined and those of us -- or those of you who stuck with us, thank you. We will say good evening. I'd try it in Italian, but that just doesn't sound good with my lack of an accent. Good evening from Florence. I am off to a cocktail, a note and Rome tomorrow morning. So are you heading straight back?

Heading back.

This weekend? All right. Well, safe travels back. Thank you so much for the time.

Yes, of course.

I know your time is so valuable to give us 45 minutes here. It was fantastic. Appreciate that. And everyone else, we'll catch up on ATTD takeaways whenever you'd like. Take care, and have a good evening.