CTO Peter Simpson on the Future of Dexcom CGMs

[00:00:00] Rob Howe: Hello podcast listeners, very excited to share a bonus tech episode with you releasing this Friday. I hope you guys are excited for a quick little background into what's been going on at Dexcom. Many of you have seen the social media clip that we uploaded last week with our guest Dexcom CTO Peter Simpson's response to the privacy and AI updates in the terms and conditions in the Dexcom G7 app. In this episode, we dig into what's coming with the G8, what's new with the CONNECT study, as well as with the type 2 diabetes study regarding results from Dexcom's landmark study that they just published and released during EASD 2026. So this is a quick tech update. I did not know that we were going to be able to have access to Peter to talk about the privacy update, and I am not an expert on terms and conditions. So were there questions that I wish I had asked? Sure, but we were fortunate to get the time and the quote that we did get. I understand that there's many in the community that are upset who are Dexcom users. And if I get a chance to revisit with someone from Dexcom to discuss that, I will publish it here on the podcast. So thank you for listening and enjoy this interview with Dexcom CTO, Peter Simpson. Guys, who was MVP of the chat? Rob Powell! Rob, what were you on? We need more of that. Peter, thank you for joining Diabetics Doing Things today. We have Peter Simpson, the Dexcom CTO.

[00:01:28] Peter Simpson: Thank you so much for your time.

[00:01:29] Rob Howe: Rob, it's great to meet you and I'm happy to be here. It's an exciting time. And we are having this conversation shortly after the 2026 EASD meeting in Europe, which was earlier this fall. And I'd love to talk a little bit about what you and Dexcom released and talked about during the conference. But I'd like to frame the conversation a little bit bigger first. As a technology leader, you've spent your career at Dexcom. It's an exciting time to be in the technology world. And I'd love to know your perspective on leading a team in the tech world in 2026 and all the things that are changing at such a rapid pace.

[00:02:01] Peter Simpson: It is a very excited time. Just a little bit of background. I've actually been with Dexcom for over 24 years now. So since it was a small startup. So I've been working on this technology through the years. And I'll say one thing. I'm more excited today about the opportunities and the products that we can make to help people with diabetes than I've been throughout the entire time frame that I've been here. So remain very excited about the opportunity. We're doing some very exciting things with our technology that can help a lot of people. And as I reflect to what's changed over the years, I think the pace of innovation is really picking up. And I think that's a key to what we've done. And if you look back in the early days where it's just let's get a CGM out in the market, let's improve its performance, let's make it accessible and reimbursed, and then combine it with the AID systems. Just as you've seen that technology evolve, some of those things took quite a long time. now you're seeing these things really start to speed up and move at a much faster pace. We're now seeing AID systems that are fully closed loop, even better CGMs, extended use, all the exciting

[00:03:05] Rob Howe: things that we've been working on. So when you say it's more exciting now than it's been in your entire 24 years plus at Dexcom, let's dig into the technology. You guys have a new sensor that's been announced and will be coming out shortly. Let's talk about what gets you so excited about

[00:03:19] Peter Simpson: the technology at Dexcom? One of the things that's really driven the adoption of CGM throughout the years is performance, accuracy, and the quality of the data. If you look back in the day, I mean, you actually have to, I don't know if you remember this, but you have to prick your finger before you'd make an insulin dosing decision on it. And you'd have to calibrate it multiple times a day. We've come a long way since then, but the performance of the system is really still a major priority to Dexcom and also the ease of use. So as we look towards G8, we're doing a couple of major things. So first of all, the size of the device is going to be about 50% smaller. So we're making it much easier to use and much more comfortable to wear. But the real innovation is on the performance and the quality of the data coming out. In terms of accuracy, we're actually, that's probably one of the biggest things, is we are incorporating a technology that we've been developing for over a decade now. And let me give you a little bit of background on how CGMs generally work. To remove the calibration using finger sticks, we put what we call a factory calibration system in place, where we measure the performance of the device in the factory, and then it gets a code, and then that code transfers to the device during use. Now, accuracy really has two pieces to it. One is the variables that you have in the manufacturing plant, the things that we can do to make the best possible quality. But the biggest challenge with CGMs is actually the variability in the biology of the different people that use it. And that can actually have a significant impact on the accuracy of the device. And that accuracy, then the factory calibration can't accommodate those variables because it only knows that bench top performance that it had in the factory. What we've developed is a technology that can actually look at the performance of the device every five minutes while it's during use and change and adapt the settings of that calibration routine to make sure it's the most accurate that it can be. So it's taking real-time data with a secondary signal that's not that glucose signal, and it's making the system more and more accurate throughout the use of life. So very excited about it. We've had a lot of great clinical studies that we've run so far, and we're working as hard as we can to get that to the market. But on top of that, you'll see some other improvements in terms of the wearability of it, the patch performance, the BLE performance. All of these things have been upgraded in this next-generation device.

[00:05:41] Rob Howe: That's really impressive. I think I want to talk a little bit about something we talk about frequently on the podcast is that diabetes affects all kinds of people. And so with that, you're literally down to the cellular level, like different biology that has to work for everyone. And it's one device. I want to talk about this. Is it software changing hardware? How does that second signal and what actually is happening inside the sensor to allow those changes to be made in real time?

[00:06:03] Peter Simpson: So I can't go into the super depth, but what I can say is that this is actually a separate measurement that we're doing. within the system. And then the algorithm takes that measurement and applies what it's learned to it, to the device. And, you know, maybe I'll talk a little bit more about what we see in our clinical trials. So whenever we run a clinical trial with say a hundred people or so, we'll get a distribution of accuracy. Some people will have a MARD. MARD is a number that gets used. And when you see the results of a trial, you'll hear something like, you know, Dexcom has an 8.0% MARD for that trial. But what it means is that some people will get a 5% MARD, some people will get a 10% MARD, the average is 8%. And so what we've seen in the clinical trials that we've run with this device is that not only does it move that whole distribution to be more accurate, but there are sometimes these outliers in performance that you'll see at the tail end, and we're able to eliminate a significant number of those outliers. So not only does the general population get better performance, but we've removed some of those negative experiences in terms of where the system is not moving accurately.

[00:07:08] Rob Howe: So you're able to just tighten the standard deviation or the variance of those numbers. Wow, really interesting. And general accuracy and consistency. And that's really important because I want to, as we go into the next question, talking about the type two diabetes report that was published at EASD and reviewed, there were three main barriers to optimal care. First was health literacy and tech literacy. So what do you know about your health? And then how do you interact with technology? So I love that you're able to help improve the experience for more people. And then the third was systemic access, which you touched on earlier as something that you're excited about. I'd love to discuss how you guys are thinking on the tech side when this research comes back. How do you implement that into the product or the product development?

[00:07:48] Peter Simpson: I think there's actually two significant things that we've talked about recently. One is this state of type 2 diabetes, and I'll talk about that first. But the other thing that's coming out within this last year has been the Connect study, and I think that's really compelling. Let's first focus on this state of type 2 diabetes. We've done this for a couple of years now. And what we do is we do research. We interview about 3,000 people across many different countries. And we try to understand how in type 2 diabetes is CGM being used? The people who are using it, do they get a lot of benefit out of it? And then what are the barriers for people that aren't using it? Or what would be the barriers to getting more people to use it so that they can get better clinical outcomes? And what we learned from this, and this is similar to years in the past, is that there are some barriers to use. A lot of it comes from education, some of it through access, but a lot of it is just getting people aware that the technology is there and getting the education awareness. And then, you know, finally access, meaning getting people to get reimbursed for the use of the device. I think that's an important key for access is just the affordability of it. The other thing that we found was that when people do use CGM with type 2 diabetes, the outcomes they get often exceed their expectations. It does have a big impact. It does help a lot of people. And so that is universally recognized that CGM does have a huge impact in type 2 diabetes. So that's exciting.

[00:09:08] Rob Howe: And it's important for the accessibility conversations for people with type 2 diabetes, for example, that studies like Connect show the 0.9% A1C reduction for patients using Dexcom compared to just the control group. And that was even in the study outpacing some of the other medication protocols, which is particularly interesting.

[00:09:26] Peter Simpson: Yeah, absolutely. The CONNECT study, first of all, is a huge landmark study. It just got published this week in the New England Journal of Medicine. So big journal. I think it's been 20 years since a CGM outcome study was published in that journal. So very, very big deal there. What it showed is what you're mentioning is that when used with people with type 2 diabetes, not on insulin, it has a huge impact, 0.9% against the controls of A1C reduction. And again, just put that in context, a level of 0.5% improvement is determined to be clinically significant and will drive reimbursement. So having a 0.9% reduction is just outstanding. And it shows what we've known a long time is that people with type 2 diabetes can really benefit from it. And a very similar thing that it does to people with type 1, it allows them to optimize their medication, it allows them to modify their behavior to get the best health outcomes, and really just be aware of what's going on in their body so they can do the best they can. I think there's one other really interesting thing about that study that we did have a portion of the people who participated in that study using GLP-1s. And as you remember in the history, GLP-1s were going to solve a bunch of problems. And what we found was that it was actually the combination of GLP-1s and CGM that had the biggest benefit. In fact, the people on GLP-1s were the ones that benefited the most from adding CGM to that. So they had already been using GLP-1s for a while, adding CGM, dropped it by another 1.2% or so. So huge impact, the biggest impact of type 2s using GLP-1s. And again, it's that same driver. You're optimizing your medication, you're doing behavior modification that allows you to live a healthier life and get better control of your diabetes. There's also the effects and knowing what your levels look like and thinking and taking in that data, that knowledge is power, whether you're on another treatment or therapy or not,

[00:11:18] Rob Howe: being able to see that impact is really important. I think it supports what a lot of anecdotal evidence or stories that we have here on the podcast of patients saying, you know, once I knew on a more regular basis, or I was always curious what was happening to my glucose during these times, and now I can see it. Now I can plan better, take better action, and get the better results in the process. Absolutely. I want to talk to something. I'm glad that we're having this interview at this time, because if we had done it a couple of weeks ago, I wouldn't have noticed this or seen this in the community. But I wanted to flag something that I've seen. Right now, there's a video or a series of videos going around and a lot of content creators and people with diabetes are coming in talking about a change in the terms and conditions in the Dexcom G7 app that was prompted to users, I believe, rolled out in the past couple of days. I'd love to ask you your thoughts and to tell you if you had a chance to address what this means for patients and where it's coming from. How could you help address the community and talk about that?

[00:12:12] Peter Simpson: Absolutely. And thank you for bringing that up. So first of all, at Dexcom, customers are always our number one priority. I think everybody has to know that. Patients' data privacy is a top priority. And we actually do significantly more than is required to ensure data privacy. And we are very sorry for the way that was rolled out. We recognize that we made an error there. We never intended to block people from being able to view their glucose data. We followed industry standards in this situation, but we really now understand the impact this had on users, and we've learned from this. So moving forward, anytime we have a change like that or an in-app messaging, it'll be a rolling update where we'll give users a window of time to acknowledge it so that we don't put their critical data at risk and that they can have time to get any questions they have answered to make sure that they are okay with the changes that are being placed forward. So that's a big change that we're doing. Now, one thing I want to make absolutely clear is that we do not track metadata across third parties. I think that was something that was suggested in the podcast. We do not track any metadata across websites. We do not track it across other apps unless you specifically give consent. and I'll give you the biggest example of what drove this is that we need that consent to allow Dexcom to share glucose data to our pump partners for AID systems and that's a technology that people with type 1 diabetes are using very important for therapy for many many people and without those types of consents we would not be able to share our data with our pump partners so that's the origin of a lot of these things that are coming across when you read the fine print They can sound a little bit confusing, but those are the origins. And I just wanted to underline that we do not track any metadata. We do not collect any of that information. And then I think the other thing that was a question was, how are we using AI with this? And for Dexcom G7 users, the example of how we're using it is with this advanced meal logging feature that we've recently added. And as we continue to advance that, we want to be able to bring nutrition information from nutritional databases into that experience. And to do that, we are leveraging AI. So that's the example that I would give in terms of how AI is used in that context. Just to touch one more time on the metadata, it's similar to anytime you download another app, there will be a secondary prompt from the app store that gives you the option to opt out of some of sharing that metadata or not.

[00:14:35] Rob Howe: It sounds like that's still very standard.

[00:14:37] Peter Simpson: Yep.

[00:14:38] Rob Howe: Great. Well, thank you for answering that. I know the community will be glad to hear it. And I'm glad that we were able to get that answer from you. So as we're thinking about the future of glucose tracking, knowing what you know from your 24 years, really growing, like you said, from Dexcom being a startup to a global leader in CGM. What are you most excited about for the next few years as it comes to patients having access to their glucose or the adoption of glucose sensors for type 2 diabetes population? What gets you excited as you get out of bed every day?

[00:15:07] Peter Simpson: Our core has and will always be with the type 1 diabetes community. And we are very excited about some of the systems that are coming out, really driving towards better safety, better clinical outcomes, and also ease and simplicity of use. So some of the things that we're doing with our partners to drive both the combined, you know, the CGM information with the AID systems, some of the advanced AID systems that they're developing, I think those are going to have a huge impact in people's lives. And we're excited about that. I think the G8 is going to help that. And when you have an AID system, having a reliable and accurate sensor is really key to make sure you're getting that best experience and getting the best outcomes from that. So I think from that perspective, it's really important. And there's also some opportunities within that space for multi-analytes that we're very excited about as well. So we're continuing to drive that. In terms of type two, I think the key there is getting access to a much broader patient population. We've seen over and over again the impact that this technology can have on people with a broad spectrum of diabetes and making sure that it's accessible to them and we get the reimbursement so that it's affordable to the broadest patient population possible. I think that's something that I think is very motivating and exciting for us because the amount of people that this technology can help is going to be just advanced through that. And I think the Connect study is one of those key pieces of information that was missing before to drive that reimbursement. So we're excited that may be coming in the future. And then just as we look towards the future, there's lots of opportunities to continue to develop to help people with advanced technologies and making their lives simpler and safer and healthier.

[00:16:40] Rob Howe: Peter, thank you for your time today. I appreciate you digging in and making yourself available for our listeners. And I agree with you. I think it's really cool to see a landmark study that's going to really show the outcomes that you guys have gotten from the Connect study published in such a large and reputable journal. And to see the downstream effects of that is really exciting. So thank you for your time. And I appreciate you

[00:17:00] Peter Simpson: being with us today. Thank you, Rob.