Experience on G4 and Calibrations

My 3 yr old son has been wearing the G4 for two weeks, and we've had quite a few issues. Our first two sensors just didn't work, and our third one worked but it took four days to become accurate. However, our fourth and latest one has been working quite well (knock on wood).

In general, I have noticed that our G4 works better with more calibrations than the 7+, especially on the first day.

I have been having a lot of discussions with Dexcom about calibrations, and here's what I have been told so far. However, they admit they need more data to make actual recommendations (and they only have data in 18+ year old people who wear it on their abdomen to boot):
1. Don't calibrate on one arrow up/down or double arrow up/down. (We never did anyway.)
2. On first day, calibrate on a low BG and a high BG. (That's easy with a 3 year old.)
3. On first day, calibrate more than two times. (We've now been calibrating about 6 times on the first day, where on the 7+ we would only do 3-4 on the first day.)
4. If get a bad reading after first day, do 2-3 calibrations 15 minutes apart to get it back on track.
5. G4 can handle as many calibrations as you want to give it, as opposed to the 7+ where Dexcom suggested as few calibrations as possible.

I would love to know what other people's experience has been on the G4 so far, especially when it comes to calibrations and how to make it more accurate more quickly. Thanks.

Thank you for posting this! We are on the waiting list for the G4 for our daughter and expect to have it in the next 4 weeks. We are so used to the Seven+. I'm sure we would have had a lot of frustration with the G4 if I hadn't read this. Very interested in what others have experienced as well.

I just pulled up my Dexcom Studio program. It shows that I've calibrated over 12 times per day. Accuracy of the G4 is a difficult thing for the user to measure since we don't have an expensive standard with which to compare. Our only indication is if the receiver correlates well with fingerstick measurements in general.

I've been wearing the G4 for 10 days now. I'm on my original sensor; I restarted two days ago.

I was not aware of Dex's recommendation to stay away from calibration during one arrow up/down or greater. I remember many comments here back in 2009 from MM CGM users saying that they had to calibrate during flat BGs only whereas the Dex users said that their system was less sensitive to that.

Did Dexcom define a "bad reading?" Was it something like >10% off the fingerstick?

My Dex G4 readings seem reasonable to me but I haven't created a spreadsheet log to track and measure its performance.

Good luck with managing your little one's BGs. If I were the parent of a little T1, I would use a CGM too.

Thanks so much for this report! I have not used a G4 sensor yet, and this will help me get off to a better start. I am assuming that I can use a G4 sensor with my Dex 7 CGM. Am I wrong? My insurance does not cover CGM's, so I cannot afford to buy a new transmitter and receiver. The sensors will be used occasionally, as needed.

They did not define a "bad reading" - they just said whenever we felt it was "off". I am like you - I define it as when Dex is more than 10% off the fingerstick. That's when I feel it's out of whack.

G4 sensors require G4 transmitter and G4 receiver.

Thanks Helmut, I am very disappointed to hear this.

Interesting point. From what I understand with the "7", it is accurate to 20%. Meters are only good to 10%. So how do we judge? I use the 20% guide and it seems ok.


I thought meters were only 20% as well.

When I got the 7, they told me that it LIKES being calibrated when BGs are changing! Now I'm confused! I'm not getting the G4 until the end of December, so still using the 7, and for the most part, it's been really good, but this last sensor was crazy for the first day, and then settled down. I would be unhappy if it took 4 days to settle down -- even if you restart, that's a significant amount of time! Please continue to tell us about your experience, everyone who has responded to this thread (and others). The more we know, the better our own experience will be! :-)

I think they are better than that. If they are both standard deviation estimates then the combination of two 20's would be almost 30%. That's is too big! I guess I need to read all of those stats in the Dex manual to see what they really mean.

BUT, the G4 is much more accurate than the 7+. And the hypoalerts are louder. So far, it is working out so much better than the 7+ did for us.

I also was always told with Dexcom (Seven Plus and Gen 4 that it doesn’t matter when you calibrate in terms of moving blood sugars and that you want to calibrate at all ranges of blood sugar ranges early in the life of a sensor, ie low sugars, high sugars, and in range sugars. I was cautioned against over calibrating in that generally less is more.

So that is what I have always done with my Seven Plus and it worked well for me in terms of overall accuracy. I typically would calibrate 3 or 4 times a day for the first two days and after that typically twice a day, sometimes three times a day, usually in the morning, before dinner,and bedtime. I also would do extra calibrations if I had an unusually high sugar or mild lows. Severe lows I would treat first.

I’ve only had my Gen 4 since Friday so too soon to evaluate. Other than I hate the case.

ISO 15197 is the current FDA standard for consumer blood glucose meters.

This quote is from Patricia Bernhardt, a scientific FDA reviewer that determines premarket clearance and approval, at an FDA public meeting, clinical accuracy requirements for point of care blood glucose meters, conducted on March 16, 2010:

"The current FDA minimal acceptable system accuracy, and accuracy in the hands of lay users, is that 95 percent of individual glucose results shall fall within plus or minus 15 milligrams of the results of the reference measurement at glucose concentrations less than 75, and 95 percent of individual results shall fall within plus or minus 20 percent at glucose concentrations of greater than or equal to 75. This criteria is the current recommended ISO 15197, standard criteria for system accuracy."

To me this standard seems loose. While insulin is necessary for me to live, it can also kill me. Given that fact, I think ISO 15197 should be much tighter, especially for those of us that need insulin to survive.

While I'm a big fan of CGMs (I've used one continuously for three years.), using the above very loose standard to calibrate a CGM seems to be a system destined for failure.

I'd much rather have a CGM than go without one but we need serious improvement in the accuracy and precision of our hand held BG meters.

Sorry to hear of all this drama. I've just ordered my G4 as Ive been using the G7 with great success and aspire to the results found in flat liners group... In my experience 1st genration technology often has bugs to work out in the real world, having said this I am still hopeful of advanced benefits of my soon to arrive G4 cgm. Sending prayers of respectful compassion to you and your family.

I might be thinking of results from meter to meter. I also read that the nutrition labels on food are only required to be within 20% of actual. Sometimes I wonder if we shouldn't just toss a coin to calculate boluses! LOL


You're right about the food labels. That plus the vagaries of meters, not to speak of CGMs DOES mean that we're doing no more than taking relatively educated guesses when bolusing. It's hard enough for us adults -- how much harder it must be for parents who have to guess minuscule doses for children who vary in activity, eating patterns and growth! My hat's off to y'all! :-)

In the past I was occasionally involved with ISO standards. ISO is a quality control and verification program which simply makes sure that there is a consistency in the manufacturing process. It does not specify what a target should be, rather it establishes a target that has been shown to be attainable and makes sure that there is an adequate SQC program in place to insure the adherence to the correct practice. So, ISO doesn't determine the target... it just makes sure that the manufacturer seriously can deliver the best that the current technology can deliver. If you tighten the ISO document, nobody would be able to manufacture the device. I believe that the FDA actually determines what is acceptable.

That makes sense