Freestyle Libre 2: Incorrect Readings

[altenam]

Hi everyone,

I'm a Type 1 diabetic of 23 years with good control (consistent HBa1c below 6.5).

I'm a relatively new user of the Freestyle Libre 2 system. Since getting it I have noticed a pretty major discrepancy in the readings being recorded on the Libre 2 vs my blood glucose monitor, and wondering if this is a problem other users might have had?

Basically the higher my levels the greater the discrepancy, whereby the readings on my Libre 2 show a higher BGL than my BG monitor. For example I recorded the follow readings: The Libre 2 showed a steady 11.5 and my BG monitor shows 9.5. Another time it showed a steady 15 my BG monitor showed 12. So the higher my levels the wider the gap. This same example was repeated over multiple devices, steady levels (not going up or down), left them all on for greater than 24hrs before recording, tried at different times of year, and different levels of physical activity, over and over again. At all times the results remained the same. The higher my levels the greater the discrepancy. This is particularly dangerous, especially for new users, as relying on the Libre 2 readings could result in an overdose of insulin and lead to a severe hypo!

I contacted Freestyle and my diabetes team. Freestyle sent me a replacement, same results, and no real answers. My diabetes team seemed to be aware there are sometimes discrepancies and just recommended I test as well as use the Libre, which kind of defeats the purpose of having it.

From what I've seen most people are getting accurate readings, or different discrepancies that are resolved after waiting 24hrs after applying the device. I seem to be a rare case with this particular issue, which to me seems like a pretty major one new users should be made aware of.

Anyone have the same/similar issues? Please comment?

 

[rebrascora]

Hi and welcome to the forum.

Actually this is a well known limitation of Libre and you may also find that it's accuracy reduces when you are hypo too. Certainly I often find that Libre reads about 1 mmol below a finger prick at low levels and at the other end of the scale with high readings, will read higher than a finger prick. I believe the advice has always been to double check any high or low readings before taking action ie hypo treatment or corrective insulin dose and it is also important to do your 15 min hypo recovery check with a finger prick because Libre will almost always show your levels have dropped lower 15 mins after eating fast acting carbs due to the fact that it measures interstitial fluid, but also because it works by converting the interstitial fluid reading into blood glucose using an algorithm which extrapolates the previous readings to predict the current reading and the previous readings will suggest that your levels are still dropping.

I don't think having to double check these high and low readings takes anything away from the usefulness of Libre which really has been a game changer for many of us, but it is important to understand it's limitations and know when you need to double check the values it gives you. I think at least some of this information is covered in the training/Libre Academy which used to be obligatory to complete before getting it on prescription, but seems to get forgotten now.

I think some of the advertising for Libre may have lead people to believe that they would no longer need to finger prick at all but there are times when it is still important to do so and I also tend to do a couple of double checks on the first few days of a new sensor just to get a feel for what sort of accuracy it is giving me when in range. Up to 2 mmols out, I can live with. It is very rare that it is consistently out by more than 1mmol and almost always below, unless my levels go into double figures.

You also have to bear in mind that your BG meter has a degree of error in it, so whilst more accurate than Libre, it is not giving you an absolute reading, and considering that our BG varies quite significantly throughout the day and night as a result of so many different factors, being 1-2mmols out isn't too bad. The huge advantage of Libre 2 is that if you do overdose a correction, it will warn you that your levels are dropping low and you can gently adjust it with a small carb correction to steady it into a nice mid range reading before you become hypo.

It is also possible to use a third party app to calibrate your sensor as well as convert it into a real time CGM instead of having to scan it to get readings. These are unofficial apps but many here use them and find them brilliant.

 

[mikeyB]

My Dexcom G6 can have variations like the ones mentioned in @alteman's post. So what do I do? Ring Dexcom? No, I go into settings on the App and calibrate it using the latest BG result. After that, it remains within 0.5 of the BG result, though by it's very nature, isn't that hot on rapidly changing BG. No sensor can do that.

As it happens, I''m switching to the new Dexcom G7 on January 2. Claims to be the most accurate sensor they've produced, and only has a 30 minute "warm up" period. Cheaper (or less expensive) than the G6 too.

 

[nonethewiser]

At higher levels libre sensors can be bit higher than bg test, but since switching to L2 find its fine at low levels.

Tbh find device pretty much as accurate as bg testing & even bolus from results, when doubtful of any readings I just do bg test to double check.

 

[altenam]

At higher levels libre sensors can be bit higher than bg test, but since switching to L2 find its fine at low levels.

Tbh find device pretty much as accurate as bg testing & even bolus from results, when doubtful of any readings I just do bg test to double check.

What amount do you mean when you say 'a bit higher'?

 

[altenam]

My Dexcom G6 can have variations like the ones mentioned in @alteman's post. So what do I do? Ring Dexcom? No, I go into settings on the App and calibrate it using the latest BG result. After that, it remains within 0.5 of the BG result, though by it's very nature, isn't that hot on rapidly changing BG. No sensor can do that.

As it happens, I''m switching to the new Dexcom G7 on January 2. Claims to be the most accurate sensor they've produced, and only has a 30 minute "warm up" period. Cheaper (or less expensive) than the G6 too.

Yeah I expect variations when my glucose is changing, but I'm referring here to discrepancies of 2+ when above 11mmol when my levels are stable. I don't mind small variations, like 0.5-1, but above 2 becomes concerning as it can be the difference between an ensuing hypo or not.

Do you get the Dexcom monitors via the NHS?

 

[altenam]

Hi and welcome to the forum.

Actually this is a well known limitation of Libre and you may also find that it's accuracy reduces when you are hypo too. Certainly I often find that Libre reads about 1 mmol below a finger prick at low levels and at the other end of the scale with high readings, will read higher than a finger prick. I believe the advice has always been to double check any high or low readings before taking action ie hypo treatment or corrective insulin dose and it is also important to do your 15 min hypo recovery check with a finger prick because Libre will almost always show your levels have dropped lower 15 mins after eating fast acting carbs due to the fact that it measures interstitial fluid, but also because it works by converting the interstitial fluid reading into blood glucose using an algorithm which extrapolates the previous readings to predict the current reading and the previous readings will suggest that your levels are still dropping.

I don't think having to double check these high and low readings takes anything away from the usefulness of Libre which really has been a game changer for many of us, but it is important to understand it's limitations and know when you need to double check the values it gives you. I think at least some of this information is covered in the training/Libre Academy which used to be obligatory to complete before getting it on prescription, but seems to get forgotten now.

I think some of the advertising for Libre may have lead people to believe that they would no longer need to finger prick at all but there are times when it is still important to do so and I also tend to do a couple of double checks on the first few days of a new sensor just to get a feel for what sort of accuracy it is giving me when in range. Up to 2 mmols out, I can live with. It is very rare that it is consistently out by more than 1mmol and almost always below, unless my levels go into double figures.

You also have to bear in mind that your BG meter has a degree of error in it, so whilst more accurate than Libre, it is not giving you an absolute reading, and considering that our BG varies quite significantly throughout the day and night as a result of so many different factors, being 1-2mmols out isn't too bad. The huge advantage of Libre 2 is that if you do overdose a correction, it will warn you that your levels are dropping low and you can gently adjust it with a small carb correction to steady it into a nice mid range reading before you become hypo.

It is also possible to use a third party app to calibrate your sensor as well as convert it into a real time CGM instead of having to scan it to get readings. These are unofficial apps but many here use them and find them brilliant.

Hi thanks for the detailed response. Very helpful to know I'm not the only one/this is a common issue. I actually completed the training you mentioned, but they never mentioned these discrepancies, only that there is a delay/lag as it is monitoring interstitial fluid. The readings I'm speaking of aren't a result of a delay though, as my BG was stable when taking them.

I don't actually mind a discrepancy if 0.5-1mmol, which as you say doesn't tend to happen unless in double figures. However, it's these double figure readings that are the problem for me. If I'm getting a discrepancy of +2 at 11mmol then +3 at 15mmol and the very rare +4-5 at 20mmol (if I've eaten out at Italian restaurant and take a chance on a pizza ) then that's pretty concerning. New users should be aware of that. If I had've put all of my trust in the libre 2 when I first got it instead of checking it's accuracy against my BG monitor then that could've been pretty serious. The Libre 2 is advertised as 99% accurate. Seems a bit misleading IMO.

Anyway, I do as you say and now confirm each +10 mmol or -4 mmol readings now and it's been pretty good. Takes a while to get your head around it though, and everyone is different so there's no one size fits all. Comes down to trial and error really. Not what I'd expected from something advertised as 99% accurate

Thanks again

 

[Tdm]

Hi everyone,

I'm a Type 1 diabetic of 23 years with good control (consistent HBa1c below 6.5).

I'm a relatively new user of the Freestyle Libre 2 system. Since getting it I have noticed a pretty major discrepancy in the readings being recorded on the Libre 2 vs my blood glucose monitor, and wondering if this is a problem other users might have had?

Basically the higher my levels the greater the discrepancy, whereby the readings on my Libre 2 show a higher BGL than my BG monitor. For example I recorded the follow readings: The Libre 2 showed a steady 11.5 and my BG monitor shows 9.5. Another time it showed a steady 15 my BG monitor showed 12. So the higher my levels the wider the gap. This same example was repeated over multiple devices, steady levels (not going up or down), left them all on for greater than 24hrs before recording, tried at different times of year, and different levels of physical activity, over and over again. At all times the results remained the same. The higher my levels the greater the discrepancy. This is particularly dangerous, especially for new users, as relying on the Libre 2 readings could result in an overdose of insulin and lead to a severe hypo!

I contacted Freestyle and my diabetes team. Freestyle sent me a replacement, same results, and no real answers. My diabetes team seemed to be aware there are sometimes discrepancies and just recommended I test as well as use the Libre, which kind of defeats the purpose of having it.

From what I've seen most people are getting accurate readings, or different discrepancies that are resolved after waiting 24hrs after applying the device. I seem to be a rare case with this particular issue, which to me seems like a pretty major one new users should be made aware of.

Anyone have the same/similar issues? Please comment?

Yes, i have similar. Libre is not to be trusted at high or low points. Nor, for me, in the night. It gets lonely. Wakes me up for company.

I've just ordered some dexcom g6 (a month supply) to see if it is more trustworthy at night. I think i may get a g6 on nhs but getting rather keen on a good night sleep asap, and not sure when nhs will get round to it.

G7 is supposed to be more accurate but doesn't seem available locally

 

[altenam]

Yes, i have similar. Libre is not to be trusted at high or low points. Nor, for me, in the night. It gets lonely. Wakes me up for company.

I've just ordered some dexcom g6 (a month supply) to see if it is more trustworthy at night. I think i may get a g6 on nhs but getting rather keen on a good night sleep asap, and not sure when nhs will get round to it.

G7 is supposed to be more accurate but doesn't seem available locally

Thanks for the input. Sounds like Dexcom could be the way forward. I'll look into it. Thanks.

 

[Tdm]

Thanks for the input. Sounds like Dexcom could be the way forward. I'll look into it. Thanks.

I'll see how i get on. The dexcom app isn't exactly popular according the the play store, and it looks like you may have to get a bit creative as the only phones they say are compatible are, like, pretty obsolete (mine is both obsolete and compatible, but not the model thats sold in EU, only the US model, so its a coin toss whether it will work. You can buy a receiver, snip at £250+)
Though apparently there is a way around it using APKmirror or some such dsrk, sinister and possibly virus ridden magiks....
Really, is it too much to ask for stuff to work without going 'off piste' with stuff that is, well, life or death?
Like abbot, dexcom, actually test your stuff?????
Ok rant over.

 

[altenam]

I'll see how i get on. The dexcom app isn't exactly popular according the the play store, and it looks like you may have to get a bit creative as the only phones they say are compatible are, like, pretty obsolete (mine is both obsolete and compatible, but not the model thats sold in EU, only the US model, so its a coin toss whether it will work. You can buy a receiver, snip at £250+)
Though apparently there is a way around it using APKmirror or some such dsrk, sinister and possibly virus ridden magiks....
Really, is it too much to ask for stuff to work without going 'off piste' with stuff that is, well, life or death?
Like abbot, dexcom, actually test your stuff?????
Ok rant over.

Haha, totally. I'd prefer they'd just wait with releasing them until they worked better than BG monitors, instead of offering false hope. I feel like all of this tech just gets in the way of living a normal life sometimes. Definitely thankful for the stuff that does work though.

 

[nonethewiser]

What amount do you mean when you say 'a bit higher'?

Double Figures we'll into teens.

 

[Tdm]

Haha, totally. I'd prefer they'd just wait with releasing them until they worked better than BG monitors, instead of offering false hope. I feel like all of this tech just gets in the way of living a normal life sometimes. Definitely thankful for the stuff that does work though.

Don't get me wrong, libre is great in a lot of ways, but it still has some serious flaws...i'd like to try libre 3, which i think will be good, but you can't just buy it online (unlike dexcom, libre 2) and when i asked my DSN she said it wasn't an option, whereas dexcom was possible.
And whilst i can afford a couple of sensors, i don't have the ££££ to permenantly self fund.

I just want accuracy and a good nights sleep. .....though BS on a watch would be good too (I feel a late xmas pressy to myself coming on)

I saw somewhere someone had a hypo detecting dog...at least a libre doesn't expect walkies

 

[helli]

Having tried multiple CGMs, I find they are all calibrated to be more accurate at “normal” levels. For me, this gives me more motivation to manage my levels and keep them in range as much as possible and check with a finger prick when high or low.
I do not see this as a “flaw” in the CGMs but more of a limitation of the technology. However, it is a flaw in the training we are provided.
If I want a good night’s sleep, I pay more attention to my basal testing and carb counting. No sensor’s accuracy (and, for me, Dexcom was the least accurate CGM I have trued) will beat better diabetes management. I also accept I make mistakes and unless I live the same day over and over again, I cannot replicate the job of a complex human organ.

 

[Tdm]

Not sure how i can manage my diabetes better when my a1c is 37 and time in range is 97% over the last week. On MDI.
Its not my fault the libre gives repeated, false alarms. Sometimes 5 a night.
It's libre's fault. And I am heartily sick of it.

 

[mark king]

I'm due to be moved over to the Dexcom G7 by my Diabetic team Jan 2023, with it being reported to be their most accurate sensor, only 30 mins to warm up, has alarms too. There were some other points but they have slipped my mind just now.

I'll report back my findings after I've worn it for a month.

 

[altenam]

I'm due to be moved over to the Dexcom G7 by my Diabetic team Jan 2023, with it being reported to be their most accurate sensor, only 30 mins to warm up, has alarms too. There were some other points but they have slipped my mind just now.

I'll report back my findings after I've worn it for a month.

To me the ability to calibrate the sensor to each individual makes way more sense, and it sounds like the Dexcom sensors enable that function.

Let us know how you go. Interested to hear your thoughts/experience. Are you getting it on the NHS?

 

[trophywench]

Brilliant at discovering trends you haven't otherwise spotted - and for 'just having a quick check' to see what your BG is doing just now (well, 10-ish minutes ago anyway) without the palaver of actual BG meter testing and finding somewhere to 'post' the used strip when you're out wherever - in a pub for drinks or a meal - flippin nuisance now they don't have handy ash trays to dump em in ! 🙂