Libre 2 - BG difference from finger prick

U

UnexpectedDiagnosis

Well-Known Member
Relationship to Diabetes
Type 1.5 LADA
Hi all,

Recently I've noticed fairly large discrepancies between my libre 2 CGM and when I validate with a finger prick test. Yesterday I was reading 10.3 through libre but when I tested I was 7.1.

Now I am 4.1 but my libre is higher at 5.4.

I've never seen a difference of more than 0.5 but the last could of days my readings don't seem as reliable. Is it common to have such a large variation? Any way to make these things more accurate?

I have just requested a CGM reader as I currently use my phone but I need to collect it from the hospital.
 
UnexpectedDiagnosis said:
Any way to make these things more accurate?
Click to expand...
Unfortunately not unless you are willing to use an unofficial app like xDrip (there are other benefits).
Libre is "factory calibrated" and there is no way to manually calibrate.

If your sensor has not been newly applied (sometimes they can take a day or two to "bed in"), you can contact Abbott, explain the inaccuracy and they will probably replace it.
Even if your Libre is funded by the NHS, please report it - there is no reason why anyone (including a cash-strapped NHS) should pay for faulty goods.
 
helli said:
Unfortunately not unless you are willing to use an unofficial app like xDrip (there are other benefits).
Libre is "factory calibrated" and there is no way to manually calibrate.

If your sensor has not been newly applied (sometimes they can take a day or two to "bed in"), you can contact Abbott, explain the inaccuracy and they will probably replace it.
Even if your Libre is funded by the NHS, please report it - there is no reason why anyone (including a cash-strapped NHS) should pay for faulty goods.
Click to expand...
It's the last day of my sensor today... I am wanting to reach the end because I haven't had one last the full time yet. However again today it's still off by about 4. Which is slightly annoying lol

I'll pick the reader up and see if it's any more accurate. My others have been great but this one is a little off
 
UnexpectedDiagnosis said:
I'll pick the reader up and see if it's any more accurate.
Click to expand...
The inaccuracy is likely to be in the way the sensor reads your interstitial fluids rather than the way these are converted into blood sugar readings.
The way the sensor is read is unlikely to make any difference.

If I had more than half a day of readings out by as much as you, I would have contacted Abbott for a replacement. No one (especially not the NHS) hould be paying for faulty equipment.
 
Had my new prescription of libre 2 sensors 2 weeks ago, first one was way off. Even so much as saying I was at 4.2 with horizontal arrow yet I was hypo at 2.8 I phoned Abbott up immediately. I put another sensor on and that was a mile off too, 13.9 with vertical arrow yet I was 7.2 and dropping. That came straight off I didn't bother phoning them. Had a break for a few days and put new sensor on, its not great seems to be jumping around a lot, I really don't want to phone them up again. I don't have much luck with libre but I've never had 3 sensors in a row as bad as this.
 
AlfieT said:
Had my new prescription of libre 2 sensors 2 weeks ago, first one was way off. Even so much as saying I was at 4.2 with horizontal arrow yet I was hypo at 2.8 I phoned Abbott up immediately. I put another sensor on and that was a mile off too, 13.9 with vertical arrow yet I was 7.2 and dropping. That came straight off I didn't bother phoning them. Had a break for a few days and put new sensor on, its not great seems to be jumping around a lot, I really don't want to phone them up again. I don't have much luck with libre but I've never had 3 sensors in a row as bad as this.
Click to expand...
Are you giving them 24hrs to settle in? Sounds like you’re taking them straight off?
 
@AlfieT Libre and other CGNs are brilliant pieces of kit if you understand their limitations. If you do not know things like that it can take 24 to 48 hours for your body to get used to having an alien object in your arm, you will find us just frustrating.
The limitations have been mentioned many times (although it is uncommon for DSNs to know about them).
I hope the sensors you gave up with after less than 24 hours were not funded by the cash strapped NHS. The reason why it is recommended to report failures to Abbott is to get replacements so we are not wasting money.
 
helli said:
@AlfieT Libre and other CGNs are brilliant pieces of kit if you understand their limitations. If you do not know things like that it can take 24 to 48 hours for your body to get used to having an alien object in your arm, you will find us just frustrating.
The limitations have been mentioned many times (although it is uncommon for DSNs to know about them).
I hope the sensors you gave up with after less than 24 hours were not funded by the cash strapped NHS. The reason why it is recommended to report failures to Abbott is to get replacements so we are not wasting money.
Click to expand...

Sometimes I wonder why I bother with these forums. I don't abuse my diabetes stuffing my face with carbs, biscuits and cakes. I keep my TIR @ 100% My A1C is 37mmol. I work hard to achieve this so I know when things aren't right.

It's 2023 not 1980 if it doesn't work then it goes in the bin. Abbot says the sensor works for 14 days not 12 not 13, there is no "bedding in 24/48 hrs" in the instructions. As for the cash strapped NHS, the NHS gets plenty of cash. By the time the cash it gets to where its needed it's lined the pockets of way too many people.

I don't come here to be patronised, I was just reporting my findings. I'm done with this place.
 
AlfieT said:
Abbot says the sensor works for 14 days not 12 not 13, there is no "bedding in 24/48 hrs" in the instructions.
Click to expand...
No they actually say up to 14 days. You need to call them to report sensors that aren’t accurate before taking off, then they can check if it is outside their accuracy criteria and replace it free if it is, rather than the nhs paying for faulty sensors. If it’s within the first 24 hours they’ll usually tell you to give it 24 hours to settle in first.
 
AlfieT said:
Sometimes I wonder why I bother with these forums. I don't abuse my diabetes stuffing my face with carbs, biscuits and cakes. I keep my TIR @ 100% My A1C is 37mmol. I work hard to achieve this so I know when things aren't right
Click to expand...
Perhaps you miss the point of the forum. All we can do is relate our own experiences, and say 'have you tried this'
If you don’t feel the information is useful to you, then you can ignore it.
You speak of this being patronising. I find it patronising, no, worse, insulting, that you suggest that I don’t keep my time in range at 100% despite my best efforts, because I obviously stuff my face with cake.
 
To expand on what @Lucyr mentioned, sensor "bedding in" is one of the many things where we are all different with regard to diabetes.
Some people find their sensors are good from the first day and others find they are random number generators on the first day. This is why it is common to apply a sensor a day or two before activating it. That way you get 14 days of good readings.
 
helli said:
To expand on what @Lucyr mentioned, sensor "bedding in" is one of the many things where we are all different with regard to diabetes.
Some people find their sensors are good from the first day and others find they are random number generators on the first day. This is why it is common to apply a sensor a day or two before activating it. That way you get 14 days of good readings.
Click to expand...
Typically I leave my sensor in overnight and activate after either 12 or 24 hours... So far this reduces the time it takes for the sensor to become more accurate. The last sensor I had (stopped today) only started being off on the reading for the last few days. I haven't rang them as this was a free replace of a replacement :) They sent me 3 new ones for 1 broken oe for some reason but I wasn't complaining; so no cost to the NHS for sure.

I will try scrubbing my arm more before putting it on, if it's not accurate within by day 3 I will call them up again. Typically they go on the back of my arm
 
Much as its frustrating the libre ( or dexcom and no doubt others) arn't fully accurste from day 1, its still better than fingerpricking.
I'd be lost without a cgm
 
My new sensor has been on for about 36 hours and I'm getting about a 0.8mmol difference. Is this a normal variation?

I installed Wednesday night, activated Thurs morning and I've had that difference for 24 hours now.
 
UnexpectedDiagnosis said:
My new sensor has been on for about 36 hours and I'm getting about a 0.8mmol difference. Is this a normal variation?

I installed Wednesday night, activated Thurs morning and I've had that difference for 24 hours now.
Click to expand...
Remember that finger prick meters are not 100% accurate - they need to be within 15% accuracy 95% of the time.
For example, a "true" value of 5.0 could be reported anything between 4.25 and 5.75.
So you could easily see a difference of 0.8 between two finger prick meters.

In other words, the variation of 0.8 is acceptable. In fact, given the accuracy is expressed as a percentage, I am surprised it is a constant variation. I would expect the variation to be greater on higher numbers.
 
Yes, 0.8 is a perfectly acceptable variation between Libre and BG.
 
Brilliant, I can work with a fixed variation and can set alarms based on it.... I would expect a difference as the CGM does seem to be a little delayed; which is the "lag" that Abbot speak about with the interstitial reading opposed to the levels found in capillary blood vessels.

I will do another couple of finger prick tests to see if it remains the same

Thanks all :)
 
@UnexpectedDiagnosis Abbott talk about a delay between BG and interstitial fluid readings (ISR).
Whilst CGMs, like Libre, use ISR, they take the delay into consideration when converting this to BG by extrapolating the current trend. Therefore, you should not see a delay between the two.
There is a small caveat to this - if the trend changes direction (e,g. if someone treats a hypo), Libre will be delayed in seeing this, But, if your levels are rising or falling constantly, the difference in readings is very unlikely to be due to the delay in ISR.
If all you are seeing is a difference of 0.8, this is more likely to be due to approved inaccuracies,
 
As @helli says, there is an algorithm built into the Libre system which tries to negate the interstitial lag, which works well when your levels are stable or changing uniformly, but can get it significantly wrong in the short term if the rate of change of your levels is not uniform or they are changing direction completely and then it can often take Libre half an hour to "catch up" so to speak. So when your levels are dropping into a hypo for instance and you treat the hypo with carbs, the Libre will usually indicate that your levels have dropped lower still at the 15mins post carb treatment scan whereas a finger prick at that time will likely show that your BG has responded to the hypo treatment and come up. If you went off the Libre reading and had a second hypo treatment, your levels would then end up too high because you would have over treated the hypo and you would be on the BG rollercoaster :rolleyes: If you wait half an hour after hypo treatment and scan, the Libre will probably be showing your levels have recovered because the algorithm will have adjusted by then to the change of direction and if you had scanned at the 15 mins point and got a lower reading, you will likey see that your graph line does not actually go through that low point but leaves it hanging below the graph, because it "realizes" that it got it wrong. You will also see this when levels are peaking and the algorithm will often over estimate the peak of a spike in BG.

When you say you plan to factor this discrepancy into the setting of your alarms, can I suggest that probably isn't the best idea and smacks a bit of micromanaging, but also placing too much trust in any one measuring device.
I would recommend that you pick an alarm level which gives you time to treat a hypo before you drop below 4 on that device and just accept that in reality that reading is not totally accurate but somewhere near. None of our measuring devices are that accurate and the decimal place on BG meters and CGM actually gives us a false sense of that accuracy. If our meters all just gave us whole numbers, I think that would probably give us a more realistic understanding of managing BG.
 
You must log in or register to reply here.
Back
Top