Note I think this was recorded in late 2025 so when they say “this year” they mean last year. Jake took over as CEO from Kevin in the middle of 2025.
Dexcom’s new sensor replacement policy Sensor Replacement Policy for Dexcom CGM Products | Dexcom
Old policy via archive.org Sensor Replacement Policy for Dexcom CGM Products | Dexcom
This podcast should be mandatory viewing for all who used CGMs or thinking about it. I think most who have hard words for Dexcom do not have my experience in R&D product development, There are so many places that unforeseen troubles can arise from supply chain, to biological issues.
It’s really good and needs to be posted at FUDiabetes site as well.
Oh interesting that the CEO is an engineer.
@Luis3 - Thanks for the reminder to look at this episode. I just finished viewing it.
I’ve lived with Dexcom CGMs starting with the 7 Plus model back in 2009. I now believe that my aging body has produced some problems with the CGMs over time.
Previously, I’ve always concluded the problems stemmed from the Dexcom manufacturing/supply logistic supply chain angle but now wonder how much my body’s individual ability to host a sensor may have contributed to failures.
I’ve held back with the G6 since my CGM forms the foundation of my AID system. I’ll switch over to the G7 this year and will focus on keeping a positive attitude.
I wonder if the chemical reagents and the sensor lead material has changed from the G6 to G7. I had regular G6 failure starting day 7. This was common enough to happen to several users. The G7 has been rock stable for me lasting 10 1/2 days. I did have a few mechanical failures where the sensor lead did not get inserted, and a few that failed to activate requiring placing and rotating applicator to get it to switch on.
This is on day 4 of my current G6 session. Note the three dips at 3 am, 4 am, and 5 am. They resemble compression lows (but are not) and wake me up. These false lows are irritating. This is typical of some, but not all, sensor sessions.
Your comments of your G7 experience offer me the possibility that I might do better with the G7.
“Past performance is no guarantee of future results,” 
I hop that the G7 works well for you
I’m seeing the same thing with my G6’s. Oddly the replacement sensors that come from Dexcom are the worst. I was at Disneyland with one, every time I sat down it said I was low. I got off the ride and walked around and my blood sugar “went up”.
I think we are paying a bit of a price staying with the G6 because low accuracy batches of of sensors don’t get out of the supply chain as fast as they used to. Also from reading the FDA reports and warning letters I’m pretty sure dexcom is using the G7 sensor wire in the G6’s these days and has scheduled the G6 end of life to coincide with when the FDA would make manufacturing adulterated products a problem for Dexcom.
Jake said currently 4 million G7 sensors are being used per week and they have a “single digit percentages of problems”. So each week they are producing 40,000-399,999 sensors with problems. “Single digit” can be a big number. i imagine volume shields must of us from dealing with too many G7 problems per patient.
At first I was excited when he talked about the G8 being a new manufacturing product generation because it is a chance to clean the slate of many of the problems that have crept in. But the more I think about it I realize dexcom is building at a time it doesn’t have someone, or a team, that is good at working with FDA and hiring QC staff with the necessary expertise to produce the product. I’m worried we are going to have to wait till the G10 to see the results of any problems Jake manages to fix within the company. That’s a long time to struggle and not be acquired in the current business enviroment.
There was something that kept me watching all the way through. I had a meeting to prep for and kept watching until the last second instead of just pausing it.
If you post it I’d really like to see the FUD conversation about it.
With your permission I will post it.
Luis, I think your comment is partially correct. However, I spent a good deal of time last year interacting with Jake over the geofencing issue, that put any traveling T1D at great risk for losing access to the app that was unrecoverable until they returned to their home country. They KNEW about this issue for a YEAR. We kept asking them to notify our communities and explain any workarounds. That’s when they discontinued talking for 6 months! So the open source community worked on developing workaround for travelers and speaking out everywhere we could and published the way to be as safe as possible.
TOTALLY irresponsible for a company selling billions of $$ of product for people with serious medical issues … and not to care enough to speak up! More concerned for their stockholders than the user base. I’ll repeat … they knew of the issue for over a year until we approached them.
There was also an issue when they changed the lubricant to use less (cost saving?) which made their sensors even more prone to failure.
Honestly, hard to make excuses for that behavior.
Dr. Anne Peters did a great piece today on Medscape … she outlines the problems with both Dexcom and Abbott. She makes suggestions on how to deal with the issues. Now that was helpful and accurate.
Nothing new to me in this video. As to the Libre 3+ deaths and injuries I would like to know more information.
Regardless of the delivery, MDI or pump, everyone using insulin from basal only to basal/bolus should have at hand meter, strips, lancing device and glucose. My kit when bicycling contains all those. The meter is the Contour Next which is small, the strips are individual foil wrap and the lancing device is small. To treat hypos - glucose tablets for quick and a pack of peanut butter crackers for sustainability.
I admit that the marketing of CGM devices has tended to hype “NO MORE FINGER STICKS!” When in doubt finger stick.
I use a Tandem T:Slim with C-IQ+ paired to a Dexcom G6 now 7 sensor. The algorithm of this pump is as automatic as I would allow. I got a new pump recently and the Tandem rep was trying to steer me to MODI. While it uses the same algorithm as the T:Slim I refused because it lacks the full control and display of the T:Slim. I would like to have full control with my phone app, but I still want to be able to control the pump from the pump. Phones have a way of not always being where they should and apps can fail.
I spent my whole adult life until retirement with electronic and digital devices. Those will fail, but unlike CGM sensors they did not have to deal with a hostile biochemical enviroment.
The closest to this would be diagnosing equipment failures that were connected to miles long lines and other companies equipment. Is the trouble in my equipment or the conductors or is it in the far end users equipment.
The point I am poorly attempting to make is that we gain great benefit from CGMs and pairing with pumps, but we should never be complacent, letting them do all the work and decisions. This is the problem I see with AI, it has a lot of knowledge, but doesn’t have the brains yet to separate the wheat from the chaff.
This law firm has as much information on the issue of Libre 3+ sensors failures that I can find. It does not really answer my questions.
Check out the MAUDE database:
Manufacturer and User Facility Device Experience (MAUDE) Database Manufacturer and User Facility Device Experience (MAUDE) Database
I am very grateful for the technologies that have made life and management better in my 61 years of T1D. Sadly to me, the 2 major CGM companies have lost sight of their social responsibility to the community in favor of profits and stockholders. I expected they might be better than that. And the FDA is so inbread that it is hard for new technologies to have enough money for the studies required … plus all the revolving door management at the FDA makes it hard for honest individuals.
I thought Dr. Peters, while not new to you, makes it clear to those who are hearing the scary stories and having the failures. At least there are some ways to deal with it.
