Dexcom g5 accuracy

I have been using the dexcom g5 for a year now and it is rarely accurate, I am between 5 and 30 points off. I would never bolus using the the g5 numbers. I use it to basically see how I’m trending. I will soon be using the g6. Is this more accurate. Does anyone have advice. Is it me or the device?

Mary Ann,
First of all no matter how we try, our numbers are gonna be off from actual! Using the blood testing method can be off up to 20% and that’s okay according to the manufacturers and the Drs, those that don’t have diabetes! Then we have the G5 whose numbers are based off those finger sticks and well you guessed it, they can be off again as much as 20% too! Is it grand!
We are given a tool that has a wonderful way to keep up with our sugar levels and keep our overall numbers in check. And then we’re given nearly a 50% error factor when reading the data! Isn’t that special!
But even with all the flaws in this system, it is still better than it used to be! People used to have to dip litmus test strips into the toilet to make insulin dosage calculations.
Thank God we are way beyond that! Some of the things that most of our technology brings us is accuracy. We are used to things being very precise and accurate. However, because our bodies are not in stasis all the time our numbers fluctuate a lot! So this means we are always looking at trends and basing treatments off that, how we feel and experience ( coming from screwing up).
In most cases the reality is we are always guessing no matter how hard we try or how good the equipment is.

Everybody reacts different to different food or drink, so what works for me may not work for you. Everybody processes food and drink differently too. I might drink something and have no reaction for several hours. You on the other hand may show the influence of that drink nearly immediately. So agin we’re back to the guessing game!
Oh yeah forgot to mention our different absorption rates for insulin.

I for one feel lucky that we have what we have to help us thru the daily struggle. It “ ain’t” perfect, but it could be worse!
I guess that my real point here is this. Whatever we have to help us is always going to be guess. We are all too different to find a perfect solution for everyone. I think that this is our biggest challenge to accept and then work around it to make ourselves feel better and live better lives.
Hope I said some things that help

1 Like

Mary Ann, @Julliette

Your numbers sound similar, but in many ways better, than what I have seen with my G5.

The main reason I have returned to CGM use is to give my wife (and me) peace at night. (I refer to my G5 receiver as “little Hitler” - it can be a bit of a dictator at times.) It is nice to have the G5 alarm go off when my blood glucose is dropping at night; makes it easy to do something about it before it becomes a big problem. And when my wife gets up at night she can look at the receiver and know that my blood glucose is in an acceptable range. So we sleep better.

I have had mine give numbers that were off by as much as 100 points. But that was with a reasonably new sensor that hadn’t really “come to itself.” Most of the time it’s close enough to be helpful.

Best of luck!


Hi Mary Ann @Julliette, I have tended to rely mor on my G5 for accuracy although I will check what appear to be “erratic” CGM readings with a BGM that is said to be accurate within 9%.

How are you determining that your G5 is not accurate? Whenever I have a blood plasma glucose reading [lab-tested blood from a vein], I chech that reading against both my meter and the reading of that minute on my CGM and found the CGM G5 more closely matches the lab result. I have been told [by someone other than Dexcom] that the G5 has an awesome MARD [a scientific testing operation] rating; that was the basis for FDA to declare that the G5 is safe for insulin dosing.

I am using my BGM to compare. I have used two different BGMs and I just don’t get the accuracy. I am afraid to bolus using the DEX.

Mail](Outlook) for Windows 10

Thank you for your view. You are absolutely right. I use my cgm mostly to watch my trends and of course the alarms. The lows are about 20 points lower on the cgm than my bgm but I know to treat it right away. Do you dose insulin per
your cgm or bgm? I am afraid to dose insulin using my cgm.

Mail](Outlook) for Windows 10

Hi @Juliette. My fingerstick and G5 readings are rarely more than about 20 points off. They have almost never been identical but I don’t expect them to be, and there is a margin for error as at least one other person has said. If my numbers are still way off with a repeat fingerstick, chances are it’s because I left my sensor in longer than I should (I confess I leave mine in for a little more than a week to stretch them out - not saying I recommend this, just sharing what I do).
For me 20 points is about 1 unit of insulin, and since I tend to run a bit high after eating I’ll bolus for the higher number, unless I’m feeling low or am dropping.
Just chiming in.

Mary Ann @Julliette, I most often use my G5 CGM when calculating my meal insulin and for corrections. The few times I might dose based on my meter is if the particular sensor in use has appeared to be giving false readings or when my meter reading better reflects how I’m feeling.

Of course I pay attention to arrows knowing that interstitial fluid readings [CGM] can lag behind finger-stick blood [BGM] readings especially during a rapid “fall”.

Mary Ann,
I use my meter in a variety of ways. But I always listen to me! If I read a number that I know is off, I go with me. If I’m not feeling the number represented, I use the human factor of how do I feel.
I have had big readings from spot on(exact match) to off by 100 pts.

I use it for general trends and base it on how do I feel. I’m training a dog to so I’m really up on numbers. To be honest, the dog is better than the meter. He beats low alarms by 3-5 minutes. Blows me away every time. Even tho he’s training, he is spot on to my big readings or better.

Most importantly the cgm is a tool to help me determine where I’m at (generally) rather than specifically. I give my feelings on my body credit too! It seems to work, I have dropped from 5 or more crashes per day to 2 or less.
My dosages for treatment are kinda different than most everyone else. Go figure!LOL. But I have more problems with dropping low than with high #s. If I go above 200 it rarely lasts more than 30 mins.
Why ? God knows. But I absolutely do not treat for highs. Tried it once or twice and I was on that rollercoaster ride from hell! Stopped that really quick. Told Dr and he agreed right thing to do. Going up n down hard on the body.
For lows means I gotta eat something sweet. Carbs don’t help me. OJ not a good increase for me (roughly 10 pts) after an hour… A Pepsi and 4 mini candy bars works best. Normal increase around 30 pts. Not the best way, but it works. Tried juices, fruit, straight sugar, the glucagon tabs, you name it, I probably tried it. Found I need something solid to go with the liquid to hold me above say 100 for more than an hour. I usually stay 110 to 125 on average thru the day.
But when things go wrong they really go wrong in a heartbeat. I had to turn off the drop alarm signal cause I usually out pace it. Been known to drop 100 pts in less than 15 min.
I’ve found that if I can keep my basal dosages in the right ball park, I do better. I use much less of the fast acting at meal times. In fact I rarely dose above 6 units at any given meal. I don’t dose if I go high at all. I don’t give corrective dosages at all. It may be just my body chemistry that rules those things. I would be afraid to say try it and see, I only know it works for me. I stay in target range (dexcom weekly report) in the high 90’s.
In reality it’s a guessing game and you have to use everything you can when treating yourself. As you practice it you get better and better at it. Yeah I screw up a lot! But I learn from those mistakes and make adjustments along the way. Mostly in what foods to eat and how long they last in my system. I keep logs on foods, I try them 3 or 4 times before I say an apple will make me go up 40 pts and hold for 2 hours. Then I go back and test that 5 or 10 times to see if it holds true as a general rule. Just an example, but do to me being me, things don’t always work the way the books, Drs or expectations say it should. Males for a real fun ride sometimes! NOT! But it’s the best I can do for me.
I keep my A1C around 6.0, I have good weight, and my tests all show fairly normal every three months. Drs say keep up the good work and I think you gotta be kidding me. 90% of the time I’m good to go. Other 10 % it sucks.
You really have to dig in find out what works for you. If I eat this it does that…if you make big boo boo, don’t repeat it! Try something else. Gotta be flexible and free n easy with not treating immediately. Foods absorb differently and differently at different times. Gotta learn that if you just ate and you drop alarm goes off not to panic and slam down a 32 oz soda. But wait 5 or 10 minutes and see what the trend is. Am I going up or do I need to do something. Got a 50/50 chance doing it right. Do it wrong your in the stratosphere, or your in the toilet. And what you did last time may or may not work the same way it did. Practice and lots of patience! Slowly you’ll work out a routine, hopefully it’s one that flows with what you want and what the drs tell they want and you don’t gain weight like a freight train or loose like crazy( weight that is).

Hope what I said helps, but this is such a hard topic to figure out. If I do this will I do that. If I don’t do this what will it do? Time, patience and practice ( mostly screwing it up) and you’ll get the right things to do for you. But only you can do it, you just have to want to!

Short answer to original question - G6, which does NOT require calibration is (in my experience) within 10-15 of a BG check 90% of the time. It’s tends to be least accurate at lows, but is still WAY more accurate than the G5. We only do BG checks when something seems off, so we feel pretty confident in the numbers.

1 Like

Thank you for the reply. I was hoping the G6 would be more accurate. The G5 is not at all accurate with low blood sugars. I will be starting the G6 in the next week or two.

Mail](Outlook) for Windows 10

I switched from relying on my glucose meter to relying on my Dexcom 5, and my Hemoglobin A1C was virtually unchanged. I was worried, like you. And I have had occasional big differences. But there is usually an explanation, such as not quite getting all the sticky stuff off my fingers after eating grapes, or inserting the G5 sensor in some scar tissue. The readings work well enough. But if I feel fine, and the G5 is reading low, that is usually because of the delay in showing the increase of blood sugar in tissue, compared to in the bloodstream. And as another member said, you also have to take the arrows into account. I will start the G6 next week, so I am in the same position as you!

I never used the Dexcom g5 but have used the Dexcom g6 for just over 6 months and have relied on it for pretty much everything. Like many other people, I only use my meter when I feel significantly different than what the dexcom says. I have found that my A1C has gone down because I can see trends and can adjust dosages accordingly. While the g6 is usually off by a couple points, its close enough so I don’t really ever worry about it. Something I have found is that the trend arrows lag behind what the numbers actually say (i.e. I would be at 80, eat and dose, go up to 130 (with the arrow pointing up), then I would start to come down and it would take 5-10 minutes for the arrow to switch from pointing up to pointing down even though my bg is clearly dropping).

We have had the same success with G6. The G5 was a nightmare.