Dexcom G6 Calibrating - Help!

[JellyBaby21]

Okay, so I'm REALLY confused and wondering if there is anyone on here that can help me. Here's my situation...

I was, until recently, using the Libre 2 without too much issue. However, I have recently started a medical trial which requires me to use the Dexcom G6. Brilliant I thought, access to a proper CGM. The Dexcom G6s are provided through the trial and not on the NHS. So I am now 4 weeks into using the Dexcom G6 and am on my 3rd sensor. I'm pretty happy with it, although the Dexcom app is not quite a good a the Libre (lacks the notes option and for detailed info you have to go onto Clarity instead of accessing on the same app. But overall it's good.

The trial requires all participants to wear the Dexcom G6 on the stomach. Fine.
We have to use the transmitter (provided) but can run the Dexcom G6 app on our phones alongside this. Fine.

But, here's where I'm confused. The trial company have said that I am calibrating the Dexcom G6 too much. The nurses and cotors overseeing the trial are confused and do not know why this would be a problem. I only calibrate the Dexcom G6 when it says that I am low and I am in fact nowhere near. For example, it has just said I was 3.9mmol/L and my finger prick said I was 6.8mmol/L. So I calibrated. This kind of discrepancy happens quite for what it claims are lows. I've started using the 20% rule (but I have read this only applies in the US??) I have also noticed that if the margin is big (like this example) then the dexcom will only change it's reading to a point when I calibrate - so in this instance it would only read 6.1mmol/L and not the 6.8mmol/L I actually was. I have found that if I enter the calibration twice then it forced the Dexcom G6 to read the finger prick number (but this obviously looks like even more calibrations to the trail company!) Most of the time to reading on the Dexcom G6 is very accurate. I know if my bloods are changing fast it cannot keep up etc and that you should really only calibrate if your bloods are trending stable. I start the sensor using the code so it has never officially asked me to calibrate.

So my question is, what am I officially meant to do in these situation when the difference between the Dexcom G6 and the finger prick is well outside of the 20% range? I can't really see that the sensor is faulty as it is accurate most of the time. Should I be calibrating at all? Should I be calibrating? Should I calibrate every time the reading is over 20% out? Why would they be concerned I am calibrating too much?

Any advise on this would be greatly appreciated- I'm so confused!

 

[Tdm]

Is it like the below? I'm on my second sensor. My first was fine, but this second one dips for no reason whatsoever and all the alarms sound. It did it overnight (which libre did) but also during the day.
Anyway, if its like that, you can't calibrate your way out of it.

I'd say calibration works if the reading is consistantly high/ low - but its also best not to calibrate on the first day, appearently, which is always rather iffy. I think the general opinion is you shouldn,'t need to calibrate

 

[Sally71]

We were told that you should almost never have to calibrate it. The one time we did, it was then clearly wrong and we had to change it back again. What is happening with your blood sugars when you get the discrepancies? If your BG is moving fast then the sensor will be behind, it will always be more accurate when it’s stable. We also find that they can read a bit low when they first start up, but usually settle within a few hours.

 

[JellyBaby21]

Is it like the below? I'm on my second sensor. My first was fine, but this second one dips for no reason whatsoever and all the alarms sound. It did it overnight (which libre did) but also during the day.
Anyway, if its like that, you can't calibrate your way out of it.

I'd say calibration works if the reading is consistantly high/ low - but its also best not to calibrate on the first day, appearently, which is always rather iffy. I think the general opinion is you shouldn,'t need to calibrate

Yes that image with the dip is like mine and then I "force" it to recalibrate by entering the finger prick value twice, causing a sudden rise on the graph obviously. My readings aren't consistantly high/low though so I guess you're saying I shouldn't therefore calibrate it right?

 

[JellyBaby21]

We were told that you should almost never have to calibrate it. The one time we did, it was then clearly wrong and we had to change it back again. What is happening with your blood sugars when you get the discrepancies? If your BG is moving fast then the sensor will be behind, it will always be more accurate when it’s stable. We also find that they can read a bit low when they first start up, but usually settle within a few hours.

It is happening when my BG gets down to around 4.4mmol/L with the trending arrow pointing horizontal so stable. At this point I often eat something like a small biscuit to bring me back up into the 5s, 6s or 7s. But even after that and with the trend arrow still horizontal it suddenly starts screaming at me that I'm 3.9 or less. I do a finger prick within a minute of this and it shows I'm either still 4.4mmol/L but usually higher by then. If the trend arrow was pointing down or even part way down I would understand but the trend arrow is always horizontal in these situations.

Should I just leave it saying the low reading and wait for it to catch up then - even though it appeared my readings were stable?

 

[Sally71]

Yes I probably would, 3.9 is very borderline and neither sensors nor finger pricks are perfectly accurate. if It’s consistently out by about the same amount over a few hours then maybe it needs calibrating, but otherwise I wouldn’t bother.

 

[Sally71]

Forgot to add, if you've eaten something because you're a bit low then there absolutely will be a delay before the sensor picks that up, because the sensor isn’t reading blood. Any sensor is the same, always use finger pricks to check when you’ve come up again out of a hypo, sensors will always be a few minutes behind and quite often are still reading low when you aren’t any more. It's not a sensor fault, they should come back into line once things have settled.

 

[helli]

I calibrate my CGM twice a day. I admit this is partially as peace of mind. I do not calibrate it when it is grossly out; I calibrate it when my levels are in range and stable. This means I am not calibrating during times when the CGM is more likely to be less accurate or when my levels are changing faster than the CGM “delay”.
For me, the best times to calibrate are when I get up in the morning and when I go to bed. In both occasions, I have not eaten for more than 4 hours and I have not exercised. As I have a pump, my DP is minimised but with MDI, the morning calibration may be less suitable.

 

[Thebearcametoo]

We only calibrate if it asks us to and only when it’s stable and in range. There will always be some readings that are off but the overall trends will give the info the study needs and the info you need to know if you need to change ratios or correction factors etc. One of the temptations when you have all this info is to tinker with it so it’s as accurate as possible but you really don’t need to. Just double check lows and high highs and otherwise let it do its thing.

 

[mikeyB]

You should only calibrate the G6 when you know you are running stable, not when your BG is rising or falling fast. I now use the G7, but I remember that the G6 can be a bit inaccurate for the first 24 hours (not always) bur once calibrated, behaves quite normally.

I haven't calibrated the G7 yet, but interestingly Dexcom insist that you wear it on the back of your arm, like the Libre. In each pack there is a sticky pad that fits over the sensor to increase the sticky area, reducing the chance of knocking it off. But applying that extra sticky pad can be a bit tricky. I don't use it because I can walk through doors without touching the sides.

 

[everydayupsanddowns]

It is a frustration for sure. Especially when you are aiming for reasonable ‘time in range’ and don’t want apparent hypos recorded that were above 4

I think there are some things with the Dexcom algorithm that (in my case) seem to resist changes of BG direction, especially where BG has been rising or falling fairly sharply. I can see quite a few ‘overshoots’ where I turn a rapid fall around with jelly babies, and the Dex algorigthm plots more what it expects would have happened than what did.

It sounds like you are doing well by not calibrating when arrows or BG are rising or falling. This is important, and Dex will pretty much ignore calibrations with arrows in some circumstances from what I can tell.

All the calibrateable CGM I’ve used have done that ‘halfway’ thing where you needed to double-calibrate to get to the actual number you have seen on your meter. I’ve always suspected this was a workaround for the sensor lag. Because the sensor should be showing you what your BG was 5-10 minutes ago, which is likely to be not exactly what it is right now.

I wonder whether part of the challenges you are having might be the 5 minute updates? 5 minutes always feels like an eternity when your BG isn’t where you want it to be, especially if hypo, and while your levels can turnaround very speedily, the G6 will only change every 5 minutes (and from 5-10 minutes ago) so by the time you get notified it’s already been happening for a while, and will appear to continue for another 5-10 minutes at least.

For me it definitely helps to keep well hydrated, which seems to speed the glucose changes in interstitial fluid and reduce the really stubborn lagginess.