I prefer to hear from the horses mouth so I looked for the FDA 510(k) but it isn’t posted yet. So here’s the press release from Dexcom:
Dexcom G7 15 Day Receives FDA Clearance: the Longest Lasting Wearable and Most Accurate CGM System
- Dexcom G7 15 Day is now cleared in the US for people age 18 years and above with diabetes.*
Then way further down…
*A study was conducted to assess the sensor life where 73.9% of sensors lasted the full 15 days. When using the product per package labeling, approximately 26% of sensors may not last for the full 15 days.
Edit: Here’s the clinical trial Accuracy of the 15.5-Day G7 iCGM in Adults with Diabetes N=130
FDA used to take about 30 days to post 510(k) clearances on their website. Not sure if that has changed given the 3500 job eliminations. Either way it will take some time
So… how does this get implemented? I’m assuming when they refer to a “15.5 Day G7” they aren’t referring to a difference in hardware—afaik, these time limits live in the Dexcom app, not the sensor right? I only know from G6, but I used to blow right past the time limit when I used xdrip, cuz it was just a matter of programming; the sensor/transmitter didn’t care. So… just an app update? Their existing stock of G7s wouldn’t have to be yanked off the shelves and replaced. Same same for any in your own stock. Which would be great—I’ll probably bite the bullet and finally upgrade to G7 when this is rolled out.
T’other question, though, is how many of the 26% that end short of the full 15.5 do so because adhesive doesn’t last that long. For me that’s always been the limiting factor. I skimmed the article and they talk about adhesive a little, but just in terms of skin irritation not session-failure.
It sounds as if you are a G6 user. I personally notice a huge difference in the adhesion properties between the G6 and G7. G6 was horrible, always coming loose. But my experience with G7 is very different. The adhesion is much better, so much so that it is sometimes very difficult to remove the G7. I sometimes have to use oil to loosen it.
Not sure how the new 15 day timeline will be rolled out for existing sensors. If no hardware design changes were needed, then it may just be a software/app update. Hopefully this is the case as removing stock from the shelves is not ideal. If they leave stock as is, then they will have to differentiate between 10 day sensors and 15 day. That would be a logistical nightmare (new PNs, labeling etc)
I am pretty sure G7 sensors have never allowed you “to blow right past the time limit”. Unlike the G6, the sw shuts the sensor down once the grace period expires.
Has to be in the sensor firmware, there’s no other way to keep the app and a pump in sync when it comes to sensor EOL. Just change the expiration counter from 3024 readings to 4464.
The right way to handle it would be to create a new model with a new NDC but after the G7 tslim compatible shenanigans worthy of a scooby-doo episode I’m not betting on it. In the last investor call Dexcom said pump companies would have to validate the 15 day version before updating pumps to be able to use it. They also said, with emphasis, that they worked with pump manufactures to adopt a framework that allows those updates to be easier than G6 to G7. Reading between the lines I can almost hear them sighing about Insulet.
I don’t want to get into the weeds over my head, but I guess the question is whether the sw is actually in the sensor. Xdrip allowed you to do that because the time limit was coded into the Dexcom app, not hardwired into the transmitter, and xdrip was bypassing the app, so it didn’t have the limit built into it. My understanding is that whatever was in the transmitter unit itself was absolutely minimal. I assume same goes for G7 since it’s even smaller as part of making it integral to the sensor, but hey, reports of the death of Moore’s Law have been greatly exaggerated 
Likewise battery life with xdrip—I was getting a lot more than 3 months out of my transmitters. I stopped using it due to a combination of things but I don’t want to wander any further off topic.
Appreciate your good news re G7 adhesive, I heard a lot of negative reports on that early on, so that’s encouraging. People only tend to comment on things they’re having a problem with and that skews perceptions. It may be something they’ve improved since initial roll out. One reason I’ve held off so long making the switch is that my experience being an early adopter on this stuff has been… less than stellar. Happy to wait for early bugs to be straightened out.
On edit:
Interesting point, yeah. I was using G6 as a stand-alone CGM back when I used xdrip, not integrated into an AID pump (a Medtronic actually), so I don’t have an answer to this other than how it actually worked. Is it possible that one of the bits it passes at initiation is start time/date and the app or pump software does the rest? I’m not a real software engineer, I just used to write user manuals and hang out with 'em. (Always said the job of a tech writer is to force engineers to stop dithering and decide things cuz you got send something to the publisher.)
This was a very short-lived “feature” after G6 first released. Dexcom very quickly remedied that and put the time limit in the G6 transmitter itself, along with other features to detect a restart. That’s why anyone wanting to run a sensor for more than 10 days now has to pop the transmitter out without breaking the sensor, or otherwise disrupt the contacts.
There were only a few runs of transmitters that allowed restart. They were highly valued for a while because they were also the ones that could be rebatteried and made to run indefinitely. Basically if the second character in your transmitter is a letter, except “G” I think, it will hard stop at 10 days, no work arounds.
Interesting to note about the 15 day sensor, it has a SLIGHTLY improved MARD. Supposedly they tweaked the algorithm a little bit, but I wonder how much of that has to do with first day wonkiness getting watered down by other days. The sensor itself hasn’t changed any, to my knowledge, just the software.
I wouldn’t be surprised if pumps and the Dexcom app were already compatible with the 15 day sensor.
For me day 10.5 is for sure better than day 1.
Having said that I just put in a new sensor that is spot on after the warm up. Very unusual for me.
Most of my G7 sensors work better after the first 2 days. (I wish they would stop saying that you never need to calibrate, because it’s just not true.)
Without having a lot of technical knowledge, I’m betting on it being a sensor programming change … or maybe sensor programming plus a pump software update - heaven help us all.
I nearly always have to calibate at least once - often 2 or 3 times during the life of a sensor, and generally in the first 3 or 4 days. I keep checking the accuracy every few days, at my diabetes care manager’s recommendation. She knows that you can’t assume that they all work well.
Dexcom changed the G6 to use the same patch/adhesive combo as the G7 but in the G6 shape. I think I finally got G6es with the G7 adhesive in my 2024Q3 order.
Don’t believe the marketing, MARD is clinical trial statistic, not a device attribute.
From the G7 15 day clinical trial preprint, emphasis mine:
The MARD of 8.0% for the 15-day G7 CGM is the lowest reported MARD in a study designed to meet the FDA’s iCGM criteria.
The clinical trial saw the same thing. High five for being average!
Wear period Matched pairs (n) %15/15 (%) %20/20 (%) %30/30 (%) %40/40 (%) Beginning (days 1–3) 5708 85.5 93.4 98.5 99.8 Early middle (days 4–7) 5507 91.1 96.3 99.8 99.9 Late middle (days 9–12) 5175 91.3 96.3 99.6 100.0 End (days 13–15.5) 3920 86.0 92.3 98.5 99.6 Overall 20,310 88.6 94.7 99.1 99.8 %XX/XX: proportion of CGM values within ±XX% for YSI values ≥100 mg/dL or within ±XX mg/dL for YSI values <100 mg/dL.
I stopped calibrating Dexcom CGMs a few years back. Some sensors are wonkier than others, especially on day one, but they mostly sort themselves out by day two. Calibrating throws one more variable into the equation. To me, calibrating or not calibrating is more about the experience than the actual numbers. Similar to drivers that tell you how much better it is driving a stick shift car than an automatic transmission. Mainly in the beholder’s mind and an attitude of control over the tool (CGM or car). For the professional driver or patient on an automatic loop system, there may be a slight advantage one way or the other; however, for the rest of us, it is more of an emotional rather than a technical difference. The best way is to experiment with two to three sensors from different batch lots to see which method works best for you. For me, being MDI, no calibration saves me time and aggravation. And gives me the best long-term results. YMMV. Not looking for an argument.
I’m in camp no calibration (mostly), not because my Dexcom is always right but because it does seem to work itself out and calibrating sometimes introduces an error in the other direction. For example, in the first 24-48 hours my sensors have fits of false low readings. If I calibrate my way out of those false lows, the sensor sometimes starts reading a bit higher than my actual BG which worries me more (insulin not suspending until I’m already low / not being alerted to take action in time to prevent a low).
I also think the average MARD would be better over 15 days than over 10 simply because the first day or two represent a smaller percentage of the total.
I had a new G7 sensor once that had soaked for over 10 hours, but was over 60 points off. I tried to calibrate it, but it refused to change more than about 5 points. I turned off Control-IQ, and gave it a few more hours, then checked it against my meter, found it was still wildly off, and tried to calibrate it again - no luck.
So I called Tech Support and explained the problem. They had me try to calibrate it 3 times, each one separated by 15 minutes, but it didn’t work. They said that they would replace it, and I changed the sensor. They also said that calibration is usually a good thing, as long as you don’t micromanage. It tells the sensor that it is wrong, and the sensor takes that into it’s calculations and usually works better afterwards.
This sort of thing is very unusual, but I always check a new sensor against a glucose meter - just in case.
That time my BG was actually around 105, but the sensor gave almost 170. Imagine what Control-IQ would have done with that.
My experience mirrors the most recent posts; “Don’t calibrate, just wait.” This means, as you observe, that hybrid and closed loop systems require careful attention over the “wait” period and sometimes (the G7 I’m currently wearing for example) I calibrate because the thing is so far off that it might kill me.
It’s a matter of judgement. No customer service rep should ever have to deal with that; put the chief on the line on top of a pointy stick.