0:02 There are many exciting things coming
0:06 out of Dexcom. A 15-day Dexcom CGM is on
0:09 the horizon. The app just launched AI
0:11 powered food tracking with image
0:13 recognition. And there's a new
0:16 integration between Stello and Aura
0:18 Ring, bringing glucose data and wellness
0:21 biomarkers together in one place. But
0:24 not everything has been smooth. Dexcom
0:27 received an FDA warning letter in March.
0:30 The FDA recently issued a class one
0:32 recall for certain receivers that failed
0:35 to alert users. And Dexcom is facing
0:38 growing frustration over users having
0:40 issues with sensor errors. From sensors
0:43 not connecting to the app to an issue
0:45 known as gooseenecking. I've experienced
0:48 both myself and in some cases these are
0:50 leaving people with no sensor readings
0:52 for days or if they're traveling having
0:55 no way of getting new Dexcom CGMs.
0:57 Welcome to the show. I'm Justin. I have
0:59 type 1 diabetes and on here I talk all
1:01 things diabetes tech news and research
1:03 with industry leaders, health care
1:05 professionals, and people living with
1:07 diabetes. Today I'm joined by Dexcom's
1:10 president and COO Jake Leachch to talk
1:12 through it all, the exciting updates,
1:14 the current challenges. We get into how
1:17 Dexcom monitors these sensor errors. We
1:20 dive into the FDA warning letter and the
1:23 recent recall. and Jake gives me a peek
1:25 into what the future of CGM technology
1:27 looks like. It's very exciting. Keep in
1:29 mind that this show is not medical
1:31 advice and is for educational purposes
1:33 only. Always consult with a physician
1:34 before making changes to your
1:37 healthcare. Let's get into it. Jake,
1:39 how's it going? Thank you for coming on
1:39 the show.
1:41 >> Yeah, it's great to be here, Justin.
1:42 Good to see you again.
1:44 >> Of course. Yeah, I want to start with
1:46 some of the most recent news uh
1:49 regarding Dexcom, which is this recall
1:51 over the receivers. So, I know that
1:54 there was a recall of about just over
1:58 700,000 receivers due to an audio issue
2:00 that could result in missed alerts. Can
2:02 you just explain to everyone what's
2:04 going on here?
2:07 >> Sure. Sure. So um through uh
2:09 investigation um of a number of uh
2:12 complaints we uh basically were able to
2:14 figure out that in very rare
2:16 circumstances there's a c certain set of
2:20 our receivers that um the audio speaker
2:23 component can fail. Um we had a it's a
2:25 particular supplier issue with that that
2:28 component. Um, and so while the chances
2:31 are very low, it's like 015%
2:33 chance that if you have that component
2:35 over time, and this is not like right
2:37 out of the box, this is over time that
2:39 speaker may fail. And so once we
2:41 discovered that, we wanted to uh
2:45 basically alert users and provide um a
2:47 free replacement if they have one of
2:50 those receivers that could potentially
2:52 in the future have that issue. Um, and
2:54 so what we have is on our website, it's
2:57 dexcom.com receiver check. We're we're
2:59 um asking users to go there, enter their
3:01 serial number, their receiver, and it'll
3:03 tell them whether their ser their their
3:05 receiver has that component in it or
3:07 not. And if it does, then they're
3:09 prompted to to call our technical team
3:11 uh who will get them out of replacement
3:13 to replace that receiver. So, the
3:14 chances are really low, but because of
3:17 the importance of um uh the audio
3:18 alerts, it's one of the most important
3:20 things our product does for users to
3:22 help keep them safe. It alert them when
3:24 their glucose is um out of range. So,
3:27 that is basically why we did the recall
3:29 is we felt it important to replace those
3:32 receivers for those they're impacted. Um
3:35 now, uh people who don't use their don't
3:37 have a receiver, don't use it. Um it's
3:39 not really going to impact anything
3:40 there. doesn't impact any of the pump
3:41 connections or anybody uses a mobile
3:43 app. Um but it but it is um if you had
3:45 the one of those receivers, we want want
3:47 to get you a replacement. There's even a
3:49 speaker test if you've used Justin
3:50 receiver. There's a speaker test built
3:52 in when you charge it that for this
3:55 exact um uh kind of to help mitigate not
3:56 knowing that the component was going to
3:57 fail, but just wanting to ensure that
3:59 people know that their alerts are
4:00 working. We actually even have that test
4:02 in every one of those receivers. Um as
4:04 you plug it in, it makes a beep and does
4:05 a speaker test. And that's why that's in
4:07 there is because it's so important to
4:09 make sure that these these receivers
4:10 would that are you know they're around
4:13 for many years um for our users and so
4:14 we want to make sure that they they're
4:15 always working.
4:17 >> Gotcha. And I'm sure there's like more
4:20 of an older crowd that are using the
4:22 receivers. I feel like these people tend
4:24 to be or or younger I guess that people
4:26 that don't have a phone, right?
4:27 >> It's all it's all the different mixes.
4:29 I've seen I've seen you know a lot of
4:31 our Medicare customers prefer I mean
4:32 they have a smartphone but they prefer
4:34 to use the receiver. For them it's
4:36 simpler. I think um I've seen folks that
4:37 use receiver but then you know their
4:39 loved ones really want them on the
4:40 mobile app so then they use both right
4:42 so the mobile app can share so they want
4:43 the their loved ones want to be able to
4:45 see their glucose so they're using the
4:47 receiver and the the mobile app so um I
4:48 think it's it's all around like there's
4:50 all kinds of different use cases but
4:52 yeah anybody who uses a receiver want to
4:53 make sure you check it and if if it's
4:55 one of those it's part of this group
4:56 then we'll we'll get you a new one.
4:59 Okay. And and these audio issues, this
5:01 wasn't something that was happening, you
5:03 know, out of the box, right? It was
5:05 something that I guess over time this
5:06 could happen.
5:08 >> Yeah. Over time that the basically over
5:11 time that speaker um component
5:13 connection um has the possibility of
5:15 failing. Again, like I said, it's a
5:17 really low percentage, but um but just
5:19 to be safe, uh we want to make sure
5:21 everyone um who has that particular
5:23 component in the receiver can get a new one.
5:23 one.
5:25 >> Okay. And what I'll do is I'll put the
5:27 link to this page in the show notes so
5:29 people listening if they have a receiver
5:32 they can go in and and put that number
5:34 and that skew. All right. Well, let's
5:38 talk about this Dexcom G7 app refresh. I
5:39 know there were some changes that came
5:40 to it.
5:44 >> Um, one of the maybe smaller ones that I
5:46 noticed from the instant I opened up the
5:48 app was when I was sliding over it,
5:51 there was just a change in like the
5:53 style of the reading. tell me a little
5:55 bit about why I guess this change was
5:57 made and what it is.
6:01 >> Yeah, so um the um version 2.9 of the G7
6:03 app for both Android and iOS, we
6:06 launched that globally. It has a whole
6:08 bunch of updates in it and some of as
6:10 you mentioned around the um uh the way
6:12 that we display the readings as you
6:14 scroll through. It's really a user
6:17 feedback around um you know wanting um
6:19 the look and feel of the app to be you
6:21 know kind of high design fun to play
6:24 with and work with and look at. And so
6:25 you know our design teams are always
6:28 looking at better ways to work the UI of
6:30 the of the mobile app. And so that was
6:32 one of the very small tweaks we made. We
6:36 also added in um uh photo event logging.
6:38 So super fun. You can now take a picture
6:41 of your hike or um uh your meal. And
6:43 actually on the meal, there's a new
6:45 feature um called smart food log where
6:46 you can take a picture of your meal and
6:51 it does a um uses um AI to identify uh
6:54 the meal and put a description um into
6:55 the text box. You don't have to type in
6:56 what you're eating. It actually does it
6:59 for you. Um it's pretty it's pretty
7:00 cool. I've been playing with it for
7:03 quite a while. Um and it it also puts
7:05 that information into clarity. So your
7:07 physician um if if you're having
7:08 challenges trying to work through you
7:10 know insulin delivery and and uh carb
7:13 estimation along with meals um you know
7:16 physician can see um the description of
7:18 of the of the meals and help you out
7:19 with the carb counting if if you're
7:20 having challenges. So I think it's a
7:21 good feature for users and for their
7:25 their HCPs um and it's a it's a great
7:26 way to log and it it's just kind of the
7:28 beginning of it. It's really let's start
7:30 getting those meals logged and then we
7:34 can start using um AI further uh use
7:36 that insight engine to to give more and
7:38 more feedback personalized around that
7:40 person's glucose and their their meal.
7:42 There's a cool you can also select um uh
7:44 kind of the meal reflection. It'll show
7:45 you the picture of the meal but also
7:47 captures your glucose for you can
7:48 program it for how many hours you want
7:51 it to capture your glucose and attach to
7:53 that meal experience. That's one of the
7:56 updates. We also um added in seven-day
7:58 GMI. So one of the things we're you know
8:01 so GMI before um we used a longer time
8:03 frame to estimate it and we had users
8:06 that wanted to see in their GMI
8:07 reflection of all the improvements in
8:09 their time and range that they were
8:12 making. Um and so we now have a 7-day
8:15 GMI um that's that's built in. And then
8:16 um one of the other things is the
8:18 customizable target ranges which is
8:20 helpful. something we've also user
8:23 feedback based HCP physicianbased
8:26 wanting to be able to set um different
8:28 glucose ranges for targets for certain
8:30 different individuals. Um pregnancy is
8:32 an example um where they wanted tighter
8:34 but they wanted a tighter range in there
8:36 so we now make it programmable.
8:38 >> Yeah. I think one of my favorite things
8:42 about using Dexcom G7 is the app really
8:44 does seem to get better pretty often and
8:46 that you are innovating a lot with it
8:49 and it's and sometimes I just it's me
8:51 update like noticing my app updated and
8:52 and seeing something completely
8:54 different before you even say it was
8:56 updated and it's just cool to to be
8:58 surprised and be like wow I can I can do
9:00 these new things. And when it comes to
9:03 the smart food logging, using that AI,
9:05 right now it it takes that photo, it
9:08 can, you know, use the AI to kind of
9:12 label it. Do you see this one day also
9:15 estimating carb count or even like the
9:17 fat content and things like that that
9:19 could prove really helpful?
9:21 >> There's um there's technologies out
9:23 there already that do that. not to the
9:25 level they're not validated and and
9:27 accurate to the level yet that we would
9:29 need to be able to provide these to our
9:31 users such as a carb count. Um but we're
9:33 going to get there. Um we already have
9:34 the systems in the background running on
9:35 those and doing those types of
9:37 calculations. We're just not displaying
9:39 it to users yet because we need to
9:41 validate that it's accurate. I think the
9:44 key there is um uh portion size is the
9:45 part that's the hardest for this. The
9:47 system is very good at identifying the
9:49 actual food. Uh I've been super
9:52 impressed. one of our uh team members um
9:54 took some was traveling in India
9:56 recently and took a bunch of pictures um
9:57 of different foods there and it was even
9:59 getting like the spices in the spice
10:01 rack in the tray and with the meal
10:03 getting those right. So it's like pretty
10:06 impressive. Um but uh it's the portion
10:09 size that is the the key to estimating
10:11 the the the actual amount of macros. So
10:13 we're working on it and we and we know
10:14 users want that and and we do feel that
10:16 that could really help help people in
10:18 their journey. are these images that
10:20 we're all uploading to this helping
10:23 train like your machine learning to
10:25 understand the photos that will come in
10:27 and yeah you're you're training on it really
10:28 really
10:30 >> we're training on it um we're the the
10:33 technology is based uh we um you know
10:35 our partner uh Google uh they've been a
10:37 partner with us for a long time um their
10:40 Gemini system is what we're using um but
10:42 we're certainly training it um with some
10:44 of our because you know you can use
10:46 their entire model but then you start
10:47 getting pictures of what actual our
10:49 actual users are are doing with it. And
10:52 so you can really start um fine-tuning
10:53 things. And so that's really what what
10:55 we're after.
10:58 >> This episode is sponsored by Omnipod. As
11:01 someone with T1D and with an incredibly
11:03 active lifestyle, it's important for me
11:05 not to get tied up, especially with
11:07 tubes on my insulin pump. I love to go
11:09 swimming, lift weights, and play
11:12 kickball with friends. And with Omnipod
11:14 5, I never have to worry about tubes
11:16 getting in the way. Plus, not having to
11:18 disconnect helps me to stay in range
11:20 during all of my favorite physical
11:22 activities. That includes sexy time,
11:24 too. See if you're eligible for a 10-day
11:28 trial by visiting omnipod.com/diabetic.
11:29 On there, you'll see full safety and
11:32 risk information and trial terms. That's omnipod.com/diabetic.
11:39 Now I know Stell your CGM that's for
11:42 people who aren't using insulin that has
11:45 an integration with the Aura ring. >> Yeah.
11:45 >> Yeah.
11:47 >> And it's taking all these biometrics and
11:51 and implementing Stello glucose readings
11:53 into the app. Can you tell us a little
11:56 bit about what these new metabolic
11:58 features on the Aura Ring app do? The
12:00 the meals and the glucose.
12:02 >> Sure. Sure. So, um yeah, as you
12:04 mentioned, um we have a strategic
12:06 partnership with Aura. If you uh acquire
12:09 an Aura Ring or an Aura Ring user, um
12:11 there's a new feature now that can
12:13 connect your your Stella directly, uh to
12:15 the Aura apps. You still run the Stell
12:17 app in the background, but it basically
12:20 is loading all that data um into the
12:23 Aura app. Uh and then so B it's it you
12:25 can always get your glucose in the Aura
12:26 app and you can look at meal reflections
12:28 in there. Um and then they're also
12:31 starting to link that um to sleep. Um
12:34 and the idea is is they're recommending
12:36 healthy habits um from you know being
12:38 able to see that glucose data. Um it's a
12:40 unique partnership. I mean uh Aura is
12:42 actually distributing Stella as well.
12:43 You can buy you can get Stella on
12:45 Stella.com. You can buy it from Amazon
12:47 but you can also get it directly from
12:50 Aura. Uh and so u we've definitely seen
12:54 an increase um in um uh basically or uh
12:57 stell users uh through that partnership.
12:59 Um and so very excited about the way
13:02 that they're they're innovating uh on on
13:04 the glucose data combined with their
13:07 their sleep um and um you know heart
13:09 rate uh respiratory rates all those
13:11 things. I think we're still early days
13:13 in terms of the guidance that the system
13:15 is providing based on the data that's
13:17 going into it. that the way that we you
13:18 know the way this starts just like you
13:20 mentioned around the food logging for
13:23 for the G7 app uh is that you start
13:25 generating the data sets that you want
13:27 to use to then further optimize things
13:29 and train machine learning algorithms.
13:31 Um also testing out what works for users
13:33 and what doesn't. One of the things I
13:35 love about Aura is they update their app
13:38 like almost every couple of days I can
13:39 see that there's an update there. Um and
13:41 so it's really been a nice partnership
13:44 for us uh to work with Aura and um bring
13:46 they're helping um you know some of our
13:48 thinking around um how to continue to
13:50 evolve um how we approach our users
13:52 because all you know some people call
13:54 Stella the the consumer product but
13:57 really all CGMs that we manufacture for
13:59 any person is a consumer product right
14:01 um you're using it it needs to fit into
14:02 your lifestyle it needs to be you know
14:05 it needs to work seamlessly um you know
14:06 on your phone and and the experience
14:08 that you use when you're on board so all
14:09 those things it's helping push us and
14:11 and we're learning things through
14:12 working with Aura that going to make all
14:14 of our products better.
14:17 >> Yeah, I was uh using Aura for a bit and
14:19 I really liked it. I I will say I don't
14:21 like that they charge you monthly to use
14:24 it um after paying $400 for the device
14:26 that so I actually returned it cuz I was
14:27 like I'm not going to pay you $9 a
14:29 month. But there are a lot of people who
14:31 will do that and and they have a ton of
14:32 customers and and one day I probably
14:35 will get Aura Ring and probably what's
14:37 what's really stopping me right now is
14:40 that I would love to see G7 integration
14:41 there and I guess that's kind of what I
14:44 was wondering is do you see this Aura
14:48 Ring collaboration with Stelling
14:51 ground for gathering all of that data in
14:53 just more of a consumer way people who
14:56 aren't making decisions on insulin you
14:57 know with this data
15:00 Um, do you see this as a way to gather
15:02 all this information and do you want to
15:04 see all of this like sidebyside
15:06 information of glucose with stress,
15:10 sleep, activity, etc. with uh on the
15:11 Dexcom G7?
15:13 >> Yes. Yes. Yeah. We we definitely want to
15:16 we we working right now on how to extend
15:19 it um into the connection between Aura
15:21 and G7. So, it's definitely part of the
15:24 the road map. Um I h I often get asked
15:26 about um when we're bringing it to G7
15:28 and we certainly will.
15:31 >> Would you need to get some sort of FDA
15:33 clearance for that integration? Like are
15:35 there different intricacies to getting
15:37 that integration going on G7 than
15:40 >> Stello? There can be it depends you it
15:42 really depends on what you're actioning
15:44 to the user and and how it how it works
15:46 what information providing just to do
15:48 the connection and the visualization is
15:50 is pretty straightforward. you know,
15:52 everything we do is is kind of um
15:55 purview uh for the FDA. Some things you
15:57 have to you have to have cleared before
15:59 you can launch, right? And so um but
16:00 some of the initial stuff we don't need
16:02 that. It's more around what do you do
16:04 with the data and how does it, you know,
16:07 um impact users. That's really where the
16:10 FDA is is interested. So yeah, it it's
16:11 coming. We're working through it. Could
16:15 you see a future where this biometric data
16:17 data
16:20 is being used alongside your CGM and
16:24 your automated insulin algorithm? I do.
16:25 I mean, I think one of the ones that's
16:27 um you if people aren't working on this
16:29 already, I'm I'm sure some of it may
16:31 already be done is just even the concept
16:32 of knowing that someone's in motion or
16:34 not, right? I mean, that's like the
16:36 simplest uh stream, right? You get an
16:38 activity, but knowing that someone is is
16:41 hasn't moved and they're likely sleeping
16:43 when also measuring glucose and having
16:45 an automated insulin delivery system.
16:46 Totally different than if someone's
16:48 highly active that you could that could
16:49 influence, you know, obviously the
16:51 insulin delivery algorithm. I mean today
16:53 a lot of the AID systems right there's
16:55 like the exercise button or you know you
16:57 can change it. So um I do think you know
16:58 just from that's like the most
17:00 simplistic way. I think um it's up for
17:02 the folks that are developing those
17:05 algorithms um to determine the impact
17:07 that something like sleep can have on
17:09 how you would tune an algorithm. I think
17:12 for the absolute best control uh with an
17:14 aid system I I do think that you know
17:16 sleep and stress and all those things do
17:18 factor into glucose um you know how much
17:21 insulin you need. So, um, yeah, I I I
17:23 think that I could easily see a day
17:25 where where that data is is impacting algorithms.
17:26 algorithms.
17:29 >> Yeah. You know, interestingly, the Kido2
17:32 pump from uh that in the from the
17:35 Netherlands, they put a gyroscope in
17:37 that pump.
17:39 >> So, there's no news on whether or not
17:42 that will be used with their with
17:46 Diaboloop, their aid algorithm. Uh
17:49 that's a separate company, but they are
17:50 putting it there because they know that
17:52 it could one day be used. So they they
17:54 created a pump that at least can fit it
17:57 in. So yeah, maybe at least if your pump
17:59 knows that you're, you know, going up
18:01 and down a lot. Yeah, that could be
18:05 interesting info. I wanted to get into
18:08 some um user feedback that I've been
18:12 getting. So, I personally had a sensor I
18:15 I've actually had a few sensor issues
18:18 when it comes to um applicating it. I
18:20 had put on a sensor and then it would
18:25 not connect to the G7 app. Then I took
18:28 that off and I put on another one and
18:29 the same thing happened.
18:31 >> It would not connect.
18:34 >> I called support. support wasn't able to
18:37 fix it, but I did put on a third one and
18:39 that one worked. This inspired me to
18:40 create a video and put it up on social
18:43 media across all of my pages just to be
18:44 like, "This happened. Like, is this
18:45 happening to anyone else?" Cuz I had
18:48 heard things, but I don't know. It it
18:48 happened to me and I was like, "Well,
18:50 okay, this is this is happening." I
18:54 received over 1,500 comments across my
18:56 platforms on this of people experiencing
18:59 the same the same thing, putting on a
19:00 sensor, it not connecting to the app.
19:02 So, it's obviously not an isolated
19:06 event. Is Dexcom aware of this? And what
19:08 is going on here?
19:10 >> Basically, um you know, as you
19:13 mentioned, there sensors can fail uh
19:15 when you put them on. Um certainly this
19:16 one sounds like it's not pairing. So,
19:17 you're basically getting a not it's not
19:19 pairing, right? So, generally when that
19:21 happens, um it's based on something
19:24 around um you know, the sensor starts up
19:25 right when you deploy it. And if there's
19:27 any kind of an issue with the battery of
19:30 that device, um it can get to the point
19:31 where it doesn't start up because we
19:32 have a bunch of safety checks in there
19:34 to make sure the battery has the energy
19:36 it needs to be able to, you know, last
19:40 the whole time. So when that happens, um
19:42 uh the device doesn't start up and
19:45 doesn't pair. Um and so yes, uh it is it
19:47 is a failure mode that can happen. Um,
19:49 and it's certainly something that we
19:51 constantly are working on, you know,
19:54 ensuring that, um, the devices always
19:57 work out of the box. Um, and I'm hoping
19:59 that when you called support that you
20:02 got free replacements. Um, you know,
20:03 that's something that, you know, it's
20:05 also something that, you know, we've
20:07 also continued to enhance the way that
20:10 our our technical support interacts with
20:11 customers and making sure they
20:13 understand that there's no cap on sensor
20:15 replacements for anything where the
20:17 sensor fails on its own like that. Um,
20:19 you know, any those are replaced without
20:21 any kind of uh restriction. I think
20:23 there was a lot of confusion over time
20:26 about that and I and you know, we take
20:27 responsibility for not communicating
20:29 that well to users. So, we want to make
20:31 sure it's super clear. Um but yeah that
20:33 that is that's something you know we've
20:36 seen those rates of that occurring um
20:38 you know have gone up and then gone down
20:40 um usually based on us you know
20:42 detecting the issue and then kind of
20:44 trying to determine exactly where it's
20:46 coming from. Um it's a small percentage
20:48 in terms of the millions and millions of
20:50 sensors that get used but um it does
20:52 happen and it's something that um yeah
20:53 we're definitely aware of it. I know
20:54 that's one of the things like are you
20:56 guys aware this happening? He's like yes
20:58 we see it we're working on it. We we are
21:00 making um updates all the time and and
21:02 chasing down issues all the time. We
21:04 have a whole team that does nothing but
21:07 focus on that um because as you scale to
21:09 you know um you know hundreds of
21:12 millions of sensors around the world um
21:13 things happen uh you know in different
21:15 places and so we're always working to
21:16 improve the systems to make sure that
21:19 doesn't happen and if it does the most
21:20 important thing is if it does happen you
21:22 know you need we need to make sure that
21:24 users can get replacements in a timely
21:26 manner. You know, one thing that the
21:28 reason I was so worried when it happened
21:30 to me was I was about to go on a 60-day
21:34 trip around Europe. Now, you know, I
21:36 already bring a fair amount of backups
21:40 just in case of anything, but the the
21:42 fact that two in a row failed and I
21:46 didn't know if my stash of the eight
21:49 that I was going to bring, if if five of
21:51 them failed, that means I would only
21:53 have 30 days of Dexcom and I'd be in
21:56 Europe where I'm unable due to
21:59 restrictions on the government and
22:01 insurance, whatever it is. And maybe you
22:02 can explain that
22:05 >> um I would be unable to get Dexcom. What
22:07 do you say to people who are about to go
22:11 travel for 30 days, 2 weeks, like should
22:14 they be bringing
22:16 four times as many Dexcoms that they're
22:18 you know cuz I' I've had people message
22:22 me saying four in a row failed upon
22:25 application. So, I guess like that's
22:28 where I'm not worried about as much um
22:30 replacements, although some people only
22:32 get a month's supply at a time, but I'm
22:34 worried about the people, especially
22:36 during summer that are traveling for for
22:38 a long period of time.
22:40 >> Yeah, it's it's a really important topic
22:41 around making sure you've got enough
22:43 sensors. And I would say that, you know,
22:46 multiple uh sensor failures in a row, it
22:48 can happen, right? It's clearly happened
22:50 to you, it's happened to others. Um, and
22:53 you know, I think it's it's certainly
22:54 less likely to happen now than, you
22:56 know, in the past when as we've
22:57 continued to improve the product, but it
23:00 it can happen and I think it is
23:01 important to have extras whether you
23:03 know whether it's you know an issue with
23:06 the sensor or maybe the sensor falls off
23:07 um while you're on your trip. It's
23:09 important to have those extras. Um I
23:10 wouldn't say you know you need to bring
23:13 a huge number. Um, one of the things
23:15 though is that we are working on that
23:18 the the complexity of replacing sensors
23:20 for users that are traveling. It's not
23:22 impossible to figure out how to do this
23:24 um because you are a US customer and it
23:26 is a US product. It's US approved. But
23:28 the way the current networks work is
23:31 that unless you have a European uh kind
23:34 of prescription or a European access, if
23:35 you're let's just get say you're
23:37 traveling in Europe, it's really hard
23:39 for us to like compliantly get you a
23:41 sensor with from with from the European
23:43 group, right? But there is a potential
23:44 opportunity for us to figure out how to
23:48 get it to you from a US um entity. And
23:49 so we are working on how because we do
23:51 understand how important it is. It's
23:52 similar to the when we had some of the
23:55 um geoloccation work that basically
23:56 guaranteed that if you were in a
23:59 location your your feature set available
24:01 to you if you installed the app was
24:03 based on that location right so people
24:05 were having issues traveling so we did a
24:06 lot of work to to resolve all those
24:08 things so that the mobile apps when you
24:10 travel or whether you even lose your
24:11 phone and have to reload everything
24:14 while you're in say you know Germany for
24:15 example you still have access to all the
24:18 things you need u uh for for your US
24:20 version there's still opportunity for us
24:23 to to continue to improve
24:24 >> all of these sensors. You see it on the
24:26 box and the applicator that they have
24:28 all of these serial ID numbers, right?
24:31 They ask you when you call up which lot
24:34 number was this. If Dexcom has all of
24:36 this data on applicators on which ones
24:38 they are, where they come from, the
24:41 specific factory, time of day, whatever
24:44 it was produced, is that information not
24:46 enough to know ex to know where these
24:48 sensors are coming from and then alert
24:52 people who have those lot numbers
24:54 because people like I'm on I'm a bit on
24:56 edge when I'm packing right now cuz I
24:58 have other sensors still and I don't
25:00 even know which ones cuz they're all
25:03 mixed up from that same lot. Like, does
25:05 that is that information not available
25:07 for you to kind of say, "Hey, anyone
25:10 with a Dexcom that has this lot number,
25:12 maybe that's not your backup for your
25:14 60-day trip across the world."
25:17 >> It's it's a we have all that data and we
25:18 do analyze it all the time to look for
25:20 any kind of trends or anything, but
25:22 generally this isn't like an entire
25:24 lot's bad or even though the the lot has
25:26 a higher failure rate. It's more like
25:28 these are generally component-based type
25:30 things. It's not like a whole batch of
25:32 things are bad. So, it's really more um
25:34 by chance you get something that maybe
25:36 doesn't that doesn't work. So, it's it
25:38 it's it's not really possible for us to
25:39 say like that because we wouldn't want
25:41 to say that lot's bad because they're
25:43 all perfectly good. It's more around um
25:45 continuing to get to the root cause of
25:47 it and then also making sure when people
25:49 do have an issue that they can get get
25:50 the replacements they need. I think
25:53 that's we attack it both ways. Um, and
25:56 um, that's basically what we've been
25:57 focused on since since the launch of the
26:00 the the first CGM, but you know,
26:01 certainly G7 in the scale, it's much
26:04 larger. Um, and so, you know, people,
26:05 you know, when they do have an issue, we
26:07 want to make sure they get that sensor.
26:09 >> Yeah. I had a lot of people reach out to
26:11 me saying that they they used a magnet
26:15 trick that the applicator on the G7, you
26:17 know, the magnet detaching from it,
26:18 that's what activates it, right? it
26:20 starts the warm-up time, which is so
26:21 cool about it because you don't even
26:22 have to link it to the app. It just
26:25 automatically is warming up. So, people,
26:27 what they've done, they found that when
26:28 their sensor doesn't connect directly to
26:30 the app, that if they kind of wave that
26:33 applicator over it with that magnet,
26:35 that it does allow it to activate that
26:36 there was some sort of mechanism that
26:38 didn't release it from that or
26:40 something. Are you aware of that?
26:42 >> Yeah, I am. And it can work. It's
26:44 actually quite um, you know, industrious
26:45 to the community. That's one of the
26:47 things I love about um the diabetes
26:49 community is um just kind of like the
26:51 creativity uh and and um we are aware of
26:53 it. That is something that can work. Um
26:55 it doesn't always work. Um and it's
26:57 basically just giving that sensor
26:58 another opportunity to fire up. Like I
27:00 said, we we check the battery to make
27:03 sure um it's uh you know adequate to
27:06 last the whole sensor duration. Um and
27:08 so it's allowing that check to occur.
27:10 Again, it shouldn't have to do that,
27:11 right? It's not designed to do it that
27:13 way, but it is it is a workaround that
27:14 people have figured out.
27:16 >> You know, a lot of people wouldn't have
27:18 heard about this trick unless me and a
27:21 few other people said something about
27:23 it. No, I don't think Dexcom's said
27:26 anything about it. And it is an easy
27:30 fix. Why did Why has Dexcom not put out
27:32 some sort of release letting people know
27:34 that if your sensor fails like this
27:37 could work, you know, to save aggravation?
27:38 aggravation?
27:39 >> It's a really good question. I think
27:42 it's because um while that can work um
27:44 the system didn't start up for a reason.
27:47 So we would really prefer to replace
27:48 your sensor than give you that
27:51 workaround. Um but you know again people
27:53 are absolutely free to do do um as as
27:55 they need. Um but yeah for us to
27:57 officially say that that we can't really
27:58 officially say that's the right way to
28:00 solve the issue. The right thing is to
28:01 get you a new sensor.
28:04 >> Um but yeah that's that's basically why.
28:06 >> Okay. Now, one other thing I've
28:08 experienced and and other people have
28:10 too, and I put this out, people
28:12 responded saying it's happened to them,
28:14 um, is what we're calling gooseenecking.
28:18 So, you put on the sensor and you can
28:20 actually see the sensor filament looped
28:22 and stuck in like the little hole like
28:25 it doesn't actually go into the skin.
28:26 What is happening with that?
28:27 >> Yeah, it's basically when the sensor
28:30 deploys, um, it the sensor doesn't get
28:31 deployed under the skin, it gets pulled
28:33 back with the needle. Um and so that is
28:35 a failure mode that can happen. Um it's
28:37 actually a failure mode that um earlier
28:39 this year we saw it start to creep up.
28:41 Um you know it's it's normally at
28:43 extremely low levels and it crept up a
28:44 bit and then we went in and we made some
28:47 changes in our manufacturing processes
28:50 to resolve it. Um and so um but users
28:52 were experiencing it. Actually we we got
28:55 an indication of that. Um and so it's
28:56 basically that's what's happening. The
28:57 sensor is not getting deployed properly
29:00 and so while it does pair it will
29:02 rapidly give you a sensor failure. And
29:04 so that's something that again we're
29:06 aware of. Um we've we've addressed it in
29:09 the in the manufacturing uh plant. Um
29:11 you know it was very it was it was small
29:12 numbers but the people that obviously
29:15 happened to um you know it's highly
29:16 impactful when your sensor doesn't work
29:18 out of the box. And so it's something
29:21 that like is not not acceptable. And so
29:22 we really want to make sure obviously
29:24 those get replaced. Um but we do feel
29:26 like we've we've um addressed that
29:28 issue. And while it may still occur to
29:31 folks, it certainly um we saw it happen
29:33 and we we made some changes to fix it.
29:35 >> Yeah. Well, it's great to hear that
29:37 you're aware of all this. And I will say
29:38 that I've had nothing but great experiences
29:40 experiences
29:42 um with support and getting
29:43 replacements. Whenever I've had issues,
29:45 I get that replacement, no questions
29:49 asked. And um so, you know, things are
29:51 working well over there, at least for my
29:53 experience. Um I have a couple more hard
29:55 questions and then I have some fun ones. So
29:56 So
29:57 >> yeah, no problem. I don't mind. I mean,
29:58 I think it's important.
30:00 >> Yeah. No, I appreciate you, you know,
30:01 giving us all this information because I
30:04 it's one thing for all of my listeners
30:07 and readers to hear from me, but I know
30:09 it does mean a lot to hear that Dexcom's
30:11 aware and that they'll give replacements
30:12 and that this is what they're working
30:14 on. And really, just hearing from you
30:16 directly is even putting me at ease
30:18 because it it does make sense when
30:21 you're producing these in the millions.
30:22 Like things can happen, right? and
30:24 you're seeing them uptick and you're
30:25 like, "Okay, let's figure this out." So, like,
30:26 like,
30:28 >> you know, it makes sense from so many
30:30 perspectives as to why you would, you
30:31 know, be aware of these and working on
30:33 them. But just hearing it from you, um,
30:35 it's making me feel more comfortable.
30:37 So, I I appreciate it.
30:39 >> Um, so earlier this year in March,
30:43 Dexcom received an FDA warning letter
30:46 due to a change in
30:48 um, a mechanism you were using from from
30:51 a third party to your own. And the FDA
30:54 said that this resulted in less accurate
30:57 sensors. Can you explain kind of what
31:00 happened here and and how the accuracy
31:04 was affected? Yeah. So, um I think what
31:07 is important to note is that um there's
31:10 no actual claim that the performance of
31:13 the sensor isn't acurate as the old one.
31:16 What the real issue was is that when we
31:19 qualified that new material, the FDA um
31:21 wanted to see us do some different
31:23 things to be able to qualify it. Um we
31:26 we used that material for quite a quite
31:28 some time on our G6 product and we used
31:29 it for a very short amount of time on
31:33 G7. Um and basically the FDA through a
31:34 normal inspection process, right, they
31:37 come in and inspect us all the time. Um
31:39 multiple facilities, factories,
31:40 everything. And so during one of the
31:42 inspections, they they um started
31:45 reviewing this and they said, "Look, um
31:46 we would have preferred you do this
31:47 differently in terms of the way you
31:49 qualified it and also we would have
31:52 preferred you tell us about the change
31:54 uh ahead of time as like a submission,
31:55 right? I talked about that idea of what
31:58 gets submitted versus what's um part of
31:59 this quality system and of course
32:01 reviewable." And so they basically came
32:02 back and said, "We'd prefer you do it
32:05 differently." And so while we still very
32:07 much um felt that what we did was
32:08 appropriate, we took their feedback and
32:10 said okay we won't make we won't
32:12 manufacture sensors with that material
32:15 um until we do um you know a redo the
32:16 validation in the way you want to see
32:19 it. Um but you know ultimately um in the
32:21 the product of the field it was not less
32:23 accurate by any means. Um it just was
32:26 they we did show them all the things
32:27 that they would want to see uh to be
32:30 able to verify it. So that was that was
32:32 basically it. Um, so we agreed and said,
32:34 "Okay, we'll next time we'll because we
32:36 do actually want to ultimately include
32:38 that material too in our sensors." Um,
32:40 and so we will um go back through that
32:41 that process with them and we're
32:43 actually already in the process of doing it.
32:43 it.
32:46 >> Okay. And none of those sensors that
32:48 were involved in that, they're not in
32:50 circulation at all, right? Anymore or at
32:52 least you're not putting those?
32:53 >> Yeah, we're not I mean we're not
32:54 manufacturing with it anymore, but
32:55 there's no reason there's nothing wrong
32:57 with those sensors. Um it was more of a
33:01 just that um uh we moving forward uh
33:02 we'll do a submission uh to get that
33:04 that material in there.
33:08 >> Today's episode is sponsored by T1D
33:10 exchange. You can directly make an
33:12 impact on diabetes healthc care
33:14 treatments and technology by
33:17 participating in the T1D exchange
33:19 registry. It starts with a simple survey
33:21 about your life with T1D and it only
33:23 takes about 15 minutes. After that,
33:25 you'll have a personal portal with
33:27 ongoing T1D study and survey
33:29 opportunities. Plus, some of these
33:31 studies even offer compensation. Signing
33:33 up with the link in the show notes helps
33:36 support my channel, and it allows me to
33:38 continue putting out free content. You
33:44 or click that link in today's show notes.
33:46 notes.
33:50 >> The G7 is getting a 15-day model.
33:54 >> Yeah. Yeah. So, super exciting. 50 50%
33:58 more days, uh, less to pack. Can you
34:00 tell us a little bit about this upcoming device?
34:01 device?
34:03 >> Yeah. Yeah, I'm super excited. We're
34:04 planning to launch it here, um, later
34:08 this year. Um, and it's a G7. It's the
34:11 longest lasting uh, sensor. So, it's 15
34:13 and a half days, and it's the most
34:15 accurate sensor uh, we've ever produced.
34:18 So it um comes with a new algorithm
34:20 within the device that not only extends
34:22 the duration but also improves the
34:25 performance of the product. Um and so
34:27 you can see you know less outliers um
34:29 you know more readings all all the
34:31 things that we're always improving.
34:33 Similar you know we've done this um this
34:35 type of algorithm improvement on each of
34:36 our generations of products after
34:38 they've been in the field for a while.
34:40 So G, we did it with G4. Uh we came up
34:42 with a new software release after G4 had
34:44 been out for a number of years. Did that
34:46 further enhanced the the accuracy of and
34:48 the reliability of the product. We did
34:50 it with G6 as well. Um after a couple
34:52 years, we made an enhancement. So now
34:53 G7's been out there for a couple years.
34:54 We're making this enhancement into the
34:57 15-day uh product. And so it it it
35:00 improves accuracy um and also extends
35:02 the the duration.
35:04 >> Yeah. It's like the iPhone 7, then the
35:06 iPhone 7S. This is the Dexcom G7.
35:08 announced the Dexcom G7s.
35:10 >> That's right. You learn I mean when you
35:12 when you have you know you launch a new
35:14 product um you know you do your best
35:15 when you're developing it. We do lots
35:17 and lots of clinical studies but you
35:19 know we're we're in the thousands of
35:21 people have have used the product before
35:22 it launches. But then once you launch it
35:23 and you have to have millions of people
35:26 using it um you learn about the the you
35:27 know infield performance and then you
35:29 start to learn okay well how can we make
35:30 this better? How can we further improve
35:31 it? And as you manufacture them, you
35:33 also learn things about how you can
35:35 tweak the algorithm to to really push
35:36 push the boundary on performance. So
35:38 does yeah, the most accurate ICGM ever produced.
35:39 produced.
35:41 >> Well, so how does the change in an
35:46 algorithm make the accuracy better?
35:48 >> It's um yeah, it's a really good
35:50 question, Justin. So it it's basically,
35:52 you know, when you when you're trying to
35:54 um produce the most accurate um
35:56 estimated glucose value for for a user
35:58 um on the on the display device, you
36:00 you're taking a signal from a sensor and
36:02 you run it through this fairly complex
36:05 algorithm that uses all kinds of things
36:07 including probabilities um as well as
36:09 the the life how long the sensor's been
36:13 in use. um as well as you know one of
36:14 the things is you're taking you know the
36:16 sensors are measuring interstitial
36:18 glucose but you're you want to reflect
36:21 glucose changes as fast as you can. One
36:22 of the things that um you know having
36:25 worn sensors for many years, the sensors
36:26 of today are much more responsive to
36:28 glucose changes than the sensors um of
36:30 yesterday, the older sensors. And that's
36:30 one of the things we're always working
36:34 on too is how do you uh ensure that when
36:36 let's say you've got a low and you're
36:38 waiting for that glucose to turn around,
36:40 you know, how can we ensure that we as
36:42 accurately as possible, we don't be, you
36:43 know, you don't have to be too noisy,
36:45 but you basically want to be able to
36:47 capture that that change in glucose or
36:49 rapidly changing glucose, right? I think
36:51 sensors today have have even less lag
36:52 time than previous sensors because
36:55 they're keeping up with that. So those
36:58 are all things that you do to to get to
37:00 a more accurate sensor as well as
37:02 outliers, right? We when you know we do
37:04 have the occasional sensor that you
37:05 start to detect, okay, this one's not
37:07 sensing glucose accurately. Um so you
37:10 further refine those algorithms too.
37:13 >> Interesting. So there's no physical
37:15 change, right? This is a software
37:18 algorithm change. the the sensors remain
37:19 the same
37:20 >> primarily the sensors uh you know we
37:22 we've continued to enhance the the
37:25 sensors themselves but this is primarily
37:28 a algorithm uh change uh that that is
37:29 and and a new clinical study that
37:30 supports it.
37:34 >> Okay. And now this sensor the 15-day is
37:38 for those 18 years and older. So what is
37:40 the thought process behind this age?
37:44 >> Yeah. Um two things. Um one was that um
37:45 but you know as we look at sensor
37:48 duration in in children um sensor
37:52 duration is is usually less the sensors
37:55 less last less time in the pediatrics
37:57 and it's typically things like the you
38:00 know patches get torn off um or um you
38:02 know the sensor probe um you know kind
38:03 of gets to the point where it says okay
38:05 I'm not sensing glucose more accurately
38:07 and lots of things drive all that but
38:10 when we look at our population of users
38:12 um we didn't feel that it was
38:15 appropriate yet to bring out the uh G7
38:17 that so many kids around the world use
38:20 to 15 days because we want more of the
38:22 sensors to last the full 15 days in
38:24 those kids than where we were at. So we
38:25 said in the adults we see good
38:27 performance 15-day stell's already
38:29 15-day we've seen great longevity of
38:32 sensors there um in the adult population
38:34 we ran our studies we're seeing good
38:35 longevity out to 15 days with it but
38:37 when you look at the pediatrics it's not
38:40 the same and so we felt it's important
38:43 um to have the pediatric 10day product
38:45 uh G7 10day available for the kids
38:47 because we just we want to make sure
38:48 everyone has I mean our number one goal
38:49 is to make sure you have the sensors you
38:51 need and you get you have the glucose
38:53 readings um all the time right we don't
38:55 want people to go without and so we felt
38:57 that if we took pediatrics to 15-day
39:00 we'd have more issues with that and so
39:02 um we will have a pediatric 15-day
39:05 sensor um but you know we we are the
39:07 leaders in in sensors for feeds and we
39:08 always have been and we always will be
39:10 and we felt it was the right thing to
39:13 do. So this is expected to launch later
39:16 this year in the US. Are there plans to
39:20 bring it to other regions that you are
39:22 already in? Like kind of what does that
39:23 look like?
39:25 >> Yeah, we we're looking we're we've in
39:27 fact we've been evaluating it. um all
39:30 every different region is um got
39:33 complexities to how you service um you
39:35 know uh CGMs for different uh age groups
39:37 and there isn't yet the precedent there
39:40 where we could do a you know a 10day for
39:42 pediatrics and a 15-day for adults there
39:44 isn't in the reimbursement channels and
39:45 how all that works there's some
39:47 complexity there that we're trying to
39:49 navigate into how to do it so that we
39:51 can get the 15-day duration um to to
39:54 those that that um could can use it. So
39:55 we're still looking at it. Um, we're
39:57 going to launch in the US first. Um,
39:59 there's a clear way for us to do it
40:00 here. So, we're going to work through
40:02 how we do it within the international
40:03 markets. We also have our Dexcom one
40:05 product uh internationally too. And so,
40:07 we're evaluating the 15-day um survival
40:08 there, too.
40:11 >> Oh, amazing. So, when it came to the
40:13 15-day, there was a press release that
40:16 had some fine print that mentioned that
40:19 roughly 25% of the sensors won't last
40:23 the full time. Is this typical for CGMs?
40:24 you know that there will be a percentage
40:26 that don't make it the full period.
40:29 >> I mean we I think having you you've worn
40:31 CGM before some people um their sensors
40:33 always last and I think it's it's just a
40:35 combination of lifestyle and a little
40:37 bit about physiology and so yeah it
40:39 can't sensors don't always last and
40:41 that's why um we want to make sure that
40:43 you know we have a robust process for
40:45 replacing your sensors. I think the vast
40:47 majority, you know, of users have no
40:49 problem getting their sensors to last.
40:50 And I think 15 days is going to be a
40:52 delighter for many people who wish they
40:53 could just keep wearing their sensor and
40:55 it when it times out at 10 days. >> Yeah.
40:55 >> Yeah.
40:57 >> Um and so I think that's that's what
40:59 it's for. But if um you know there
41:01 things can happen, right? And as you go
41:03 out to 15 days from 10 days, there's
41:05 more opportunities for something to
41:07 happen um you know, an adhesive issue or
41:09 or whatever. So I think that's where we
41:12 really um feel like as we've looked,
41:13 we've studied pretty closely. We studied
41:15 with Stella, we studied it a lot of
41:17 15-day product in type one and type two
41:18 people with type one diabetes, type
41:20 people with type two diabetes, lot of
41:22 clinical studies to be comfortable that
41:25 um we've got this right. Uh and so very
41:27 excited to get it out out the door. Um
41:30 and um you you should be hearing about
41:32 more soon about the actual launch.
41:36 >> Yeah, my G7s stay on like would stay on
41:39 way past 10 days for sure. I just I'm
41:41 like, I got to take this off. Um, so
41:43 yeah, I have really good luck with them.
41:45 I know everyone's skin type's so
41:46 different, but I'm I'm ready for the
41:49 15day. Yeah. Yeah. And I know it's gonna
41:52 stay on and I don't even always This
41:53 isn't label, but I don't put the over
41:55 patch on and it still stays on. It could
41:58 stay on for for 15 days. So, love that.
42:01 For me, a lot of people in the US, their
42:04 insurance only allows them to get one
42:06 month supply at a time for anything,
42:10 which I think is not
42:12 not the way it should be, right? Like I
42:13 get three month supplies with my
42:15 insurance and I have a backup
42:17 >> throughout all this time. So, if I have
42:20 a sensor that fails, I've got seven more
42:23 waiting for me. But what if there's
42:26 someone who gets their, let's say it's a
42:28 15-day, they have two sensors, they put
42:31 one on, works great, they put on the
42:35 second one, it falls off after 5 days,
42:37 something happens, right? They're out of
42:41 luck for the rest of the month until
42:43 currently Dexcom sends them a
42:47 replacement. Why is it on Dexcom
42:51 to do that when there is statistically
42:53 25% of sensors won't last the full time?
42:55 Why is it on Dexcom to do that and not
42:57 insurance companies? Why aren't
43:00 insurance companies held accountable for
43:04 the percentage 25% more sensors that we
43:05 may likely need?
43:08 >> So, um, a lot in there. What I would say
43:11 is um many people the the the 90day
43:13 supply, the three month supply is is you
43:15 know is a great way to go if you can do
43:18 it. Um and many many people do. Most
43:19 people do actually but there are those
43:21 like you said they only get two two
43:22 days. I think there's a couple things
43:25 I'd say. One is there are um you know
43:26 I've I've seen cases where folks are
43:29 writing prescriptions for more than than
43:32 sensors than just the the exact two or
43:34 exact you know three today for G7. They
43:36 write more than that just so people have
43:38 those extras in case they need them.
43:39 Some insurance companies cover that.
43:41 Right. There is an ongoing discussion
43:43 that we're working through with payers
43:44 in terms of how do we make sure this is
43:47 right. Um the other thing too that we've
43:49 launched too that is is helpful is that
43:51 we've got this uh voucher program going
43:52 and and we've been trying to you know
43:53 it's been going out across the states
43:56 where um if you do have a sensor issue
43:58 you can get a QR code from us and and a
43:59 voucher and take it into a pharmacy and
44:02 pick up a sensor that day. Um, and so
44:04 that's that's a that's been, you know,
44:05 we launched it in Canada. It was very
44:08 successful. We've launched it um in a
44:10 number of states uh in the US. Um, and
44:11 it's it's working quite well. We're
44:13 still working on getting it around the
44:15 entire US. So that gives another
44:17 additional option to be able to get a
44:19 sensor that day. And so um the other
44:21 thing too is for for anyone who's in the
44:22 situation where they they literally
44:24 don't have any more sensors and their
44:26 last one failed, we do expedite shipping
44:28 for those folks. So, that's important to
44:29 communicate that if you're talking to
44:31 one of our tech support reps to be able
44:32 to get you get you the sensors because
44:34 we we know how important this is and
44:35 we're always evolving. And I was
44:36 actually talking to some physicians in
44:39 Colorado the other day around how to do
44:40 this because the same issue came up.
44:42 It's like if if I you know can only get
44:44 a certain amount of sensors and have an
44:45 issue and I'm kind of don't have any
44:48 spares. Um what do I do, right? And so
44:50 the patients uh often ask for
44:52 prescriptions for extras. So I think we
44:55 we got it's one of those things we got
44:56 to continue to work on. Yeah, that's
44:59 great to hear. I actually that I now
45:01 remember when I was on the phone with
45:04 support recently, they asked if there
45:06 was a nearby pharmacy
45:08 >> and they were going to say, "Oh, we can,
45:10 you know, you can go pick it up." And
45:12 because I had my backups of supplies, I
45:15 was fine with getting it shipped, but
45:17 >> um that's a great option for people,
45:19 especially um if they're if they need it
45:24 right away. Uh I want to just end with a
45:26 little bit of a peak into the future for
45:29 Dexcom. Obviously, we're seeing 15 days.
45:32 That's fantastic. Uh the future would
45:35 could have a 15-day ped uh sensor as
45:37 well for you. You know, we're seeing a
45:39 lot of implantables. We're seeing ketone
45:41 monitoring. We're seeing this longer
45:44 wear time. Where do you see the future
45:47 of CGM going?
45:49 Um, I think the number one thing I see
45:52 is that the um, you know, there's still
45:56 so many people outside um, uh, of the US
45:58 that don't have the same type of access
46:00 to CGM that that we do here in the US.
46:01 We're very lucky. We still don't have
46:03 access for all people with diabetes in
46:05 the US. Some insurance companies cover
46:07 that, but um you know the data is so
46:10 clear uh around um the benefits that
46:13 people um can receive when they they
46:14 wear a CGM and and the lifestyle changes
46:16 that they make and that really helps
46:18 them manage. So we're working hard to
46:20 secure coverage for all people with
46:22 diabetes around the world. And I think
46:24 when I think about what's that's what
46:26 I'm most excited about is getting this
46:28 technology to more people. there's, you
46:30 know, we're still between us and and all
46:32 the folks that make CGMs in the world,
46:35 you know, we're at like a 2% use of CGM
46:36 and those that have diabetes, right? And
46:39 so it's just unbelievable um the the
46:40 people that could benefit that don't
46:43 have access. So I think that's one but
46:45 to your question around a big part of
46:47 making that happen is to continue to
46:49 enhance the technology in terms of you
46:52 know its affordability uh continuing to
46:54 drive uh you know awareness of the
46:57 technology u doing the clinical studies
46:58 that are required to show just the
47:00 outcomes and and some sometimes you got
47:01 to do it in certain countries the
47:03 economic benefits you know it's not just
47:04 a user benefit it's actually there's an
47:06 economic benefit to the health care
47:08 systems because they're spending less
47:09 money when people are managing their
47:11 diabetes or able to manage your diabetes
47:13 better, like less visits to the
47:15 hospital, less utilization of
47:16 healthcare, no the emergency room visits
47:20 go down. So all of that um requires us
47:22 to continue to innovate, right? And
47:23 we're committed to innovation across the
47:26 whole portfolio. Um you know, G7, Dexcom
47:28 1, Stella, we're going to keep pushing
47:30 the boundaries on all of it. Um you
47:33 know, smaller wearables, uh um you
47:35 looking at multi analytes, you mentioned
47:37 ketones as one of them. um adding that
47:39 into the sensor. I mean glucose is so
47:40 powerful, but as you start to add in
47:41 some of these other analytes that that
47:45 can help further the utilization and I
47:47 mean we're seeing in Stella, we're
47:49 seeing a lot of people in the
47:51 pre-diabetes category start to use the
47:54 product. Um and and while we don't, you
47:55 know, we don't have a claim for weight
47:58 loss, um I was literally in in the gym
48:00 this morning and one of the the folks
48:03 that was there comes in and says, "Hey,
48:04 uh I just lost 8 pounds." And he was
48:05 talking to a trainer. He's like, "I just
48:07 lost eight pounds over about a month
48:08 period." And they're like, "What did you
48:09 do?" He's like, "Well, I changed my
48:09 diet." And he's like, "How?" He's like,
48:11 "I used this Stell thing." And I'm
48:11 sitting there.
48:12 >> Oh, wow.
48:14 >> He doesn't know I work for Dexcom,
48:16 right? And I'm just like, uh, I'm like,
48:18 "Wow." I mean, that was like right front
48:20 and center. Uh, you know, he's like,
48:22 "Yeah, I learned that like, you know,
48:24 uh, sushi and the rice in the sushi was
48:26 wasn't great for me." And, um, some of
48:28 the drinks he was having, the the
48:30 carbohydrates in them were not working
48:31 well with his blood sugar. And he's
48:33 like, "Yeah, I just thought." So really
48:34 it had him reduce his carbohydrate
48:37 intake and he was you know feeling
48:38 feeling amazing. And so I I think
48:40 there's just so much opportunity for CGM
48:42 to make a difference in people's lives.
48:45 Um and you know CGM it doesn't really
48:47 have any side effects right it's a side
48:49 effect is better glucose as one of our
48:50 physicians said.
48:52 >> So yeah I think there's there's a lot of
48:53 lot of uh great stuff coming. And then
48:54 of course on the software side, we're
48:55 going to continue to innovate with um
48:57 the AI technologies is absolutely
49:00 incredible when you think about
49:02 supporting customers, providing insights
49:05 within the the mobile apps, the new data
49:07 streams we're starting to bring in um to
49:10 the platforms around sleep um activity.
49:12 Um you start to combine all that stuff.
49:14 I mean, there's just so much innovation
49:16 going on. I I haven't seen as much
49:18 innovation in wearables in a long time.
49:20 I mean, you think about the Aura Ring,
49:22 you think about CGM, um you know, you
49:23 think about the some of the activity
49:26 stuff. It just there's and you've got,
49:31 you know, um RFK, um you know, talking
49:33 about wearables and CGM for everybody. I
49:35 mean, it's really I think there we're
49:36 starting to recognize just the the
49:38 benefits that could come from from
49:39 people more people using these
49:41 technologies to better understand their health.
49:42 health.
49:44 >> Wow. Yeah. I'm excited to see all of the
49:47 new things that come with Dexcom,
49:49 especially with this AI stuff and all
49:51 the biometrics and Aura Ring. I'm really
49:56 hopeful that I get to buy another one.
49:57 >> Awesome. You will you will have Aura
50:00 data connect with G7.
50:02 >> Okay. Thank you. Well, Jake, thank you
50:03 so much for coming on the show and
50:05 answering all these questions and and
50:07 just giving some transparency to all my
50:10 listeners. I know I know it means a lot.
50:12 >> Yeah, happy to. Yeah, really I really
50:13 enjoyed the time and appreciate you
50:15 having me on.
50:17 >> Thanks for tuning in. If you're not
50:18 already subscribed to our newsletter,
50:21 head over to diabetic.info or click the
50:23 link in the show notes. You'll get
50:24 up-to-date info on the latest and
50:26 greatest in diabetes tech news and
50:29 research on there. Often things come out
50:31 over there before they hit my social
50:35 media, my podcast and our YouTube. And
50:36 if you're not already subscribed on
50:39 YouTube, all of these podcast episodes
50:41 go up on there in video form. Plus, I've
50:44 got episodes that release every Friday
50:47 or every other Friday on my experiences
50:49 using the technology, reviews, coverage,
50:51 stuff like that. As always, thanks for
50:54 tuning in to the Diabetic Podcast. I'm
