Which is more accurate?

The dexcom or the one touch or accucheck or any of the other blood glucose testing machines. Has anyone compared both at the same time with the readings or counts? Have they both been the same number? What was the difference if any? Why is there a difference? Thank you in advance for any replies.

I posted these a while back:

A bg meter and cgm are measuring different things.

BG meter gives your bg from that one drop of blood, with a certain margin of error. Take a single large drop, apply to several strips and they might not all be exactly the same. Some meters/strips will be more accurate than others.

CGM is best used for its trending info, but also know that Cgm bg from interstitial tissue is not identical to blood glucose meter bg.

So focus more on the trends on cgm and use bg meter for dosing decisions, or when symptoms don’t match what your cgm shows.

Sometimes the first hours, days of cgm are less accurate, may react different if dehydrated or if bg is changing quickly.

When cgm is flat, it is more likely to be closer to bg meter check.


I find dex g6 very unreliable in the first 36 hours. I place it in my upper arm. When I calibrate it is with a contour strips & meter. I trust the contour implicitly. Always seems that g6 is working best the day before it get changed. I’m glad I have the g6 but have trouble trusting it. Makes one wonder about looping like control-iq or basal-iq.

I’m not sure i trust it fully yet either. It’s good, but it’s changing the way I’m thinking a little bit, like if I’m aware of my symptoms first or the machine is telling me, in in other words think about eating and such, worrying and such,…i dont know, it’s good though, i have to get used to it and understand it better i think.

I had a problem with the symptoms yesterday: alarm ringing LOW, it like dropped 40 every five minutes, but i was just sitting feeling fine talking and stuff. But it made me like worry and so i started eating some candy corn or whatever. And when i tested myself with the one touch it was 200. And then it kept going up. I was actually pretty mad. Was doing like great all day range of 100 pretty even not much change and then bam. And you know what happened also, the sensor had a problem and it went off for like an hour and a half. I was like what
Yeah the trends i hear you, but it is close. But one time i had a difference of 40. I read that if it’s within 20 points they are close and in the same range, so yeah i hear you, the ranges. But i’m trying to get used to it, like i have to constantly think of eating or something, which i dont like ,
Thanks everyone for your replies

The Diabetes Technology Society did a big study of all blood sugar machines. The study is older now and not as valuable, but can be found on their website. It showed a lot of machines did not meet FDA standards. I hope that a lot of that has been fixed now, but I have not seen any newer comprehensive studies.


1 Like

Your CGM should not be more than 20% different than your meter (for values above 100). For values below 100, the two should not differ by more than 20 points.

Dexcom is the most accurate of all CGMs, but you still need to verify accuracy off a manual check. It will be inaccurate sometimes. Know that you compare the two values when your sugar is not changing rapidly (you should see a flatline arrow). Also, if you REALLY want to verify, know that your Dex is reading your sugar 15 minutes ago, so you write down it’s reading, and then check a manual sugar 15 minutes later.

You can call dexcom and tell them that you want to speak to a nurse about how to measure sensor accuracy. I recommend you do this so you get all the details straight…in case I don’t recall them perfectly.

A sensor that is not accurate, might need calibration, or it might have failed. You cannot calibrate it if you have not set it up this way when you inserted it. That’s bad.

1 Like

Can you explain?

Whether you enter code or not, you can always enter a calibration. If no code, it prompts for calibration.

I always enter dexcom G6 code, and do bg checks for first few days if seems off. I keep in mind that sensor bg may lag fingerstick meter bg, especially when bg is rising or falling quickly.


Maybe your right, MM1. I honestly can’t recall the details. Thanks for verifying. Maybe I go “no code” when I see sensor accuracy issues, just because it demands that I do it. It forces an accuracy check.

You can definitely calibrate even if you entered the sensor code at the beginning (and, to my knowledge, there’s no switching to no code if you started with a code).
I calibrate sparingly. For me, Dexcom almost always rights itself and calibrating to fix a false low sometimes leads to sensors reading a little high which makes me more nervous than knowing my sensor reads a little low. I realize other people’s experience/philosophy differs.

I have a One Touch UltraMini and the Dexcom G6 (with my Tandem t:slim X2). I always start a sensor with a code. Around 4-5 times a week, I use the UltraMini in the morning when I wake up and compare it to the CGM. If the two don’t agree within the 20 units or 20%, I calibrate the G6 (because I can’t calibrate the UltraMini). The calibration splits the difference (unless you do two in a row).

My reason for doing it this way is I want both devices to agree (within the margin of error). Occasionally, I will use the UltraMini to verify (or dispel) a suspect CGM reading (i.e. I “feel” like its wrong).

I also have a Contour Next BG meter which is supposedly more accurate than the UltraMini but I use the UltraMini because the online t:connect portal will upload it but won’t upload the Contour Next. If I ever need a “third” opinion, I can use the Next.