Libre 2 - what 'in range' limits to check by finger prick?

Status
Not open for further replies.

Proud to be erratic

Well-Known Member
Relationship to Diabetes
Type 3c
Pronouns
He/Him
A thread on 23 Dec from @helli offered some excellent advice on awareness of the limitations of Libre 2. Within that advice was:
"it is calibrated to be accurate for “normal“ readings (between about 4 and 8mmol/l). Therefore, always check with a finger prick test before correcting a high or treating a hypo."

I can't find why, but in my mind normal is "between 4 and 10".
I've tried searching both within this forum and in Abbott's FreeStyle Libre 14 Day Indications and Important Safety Information [ADC-17661 v3.0 7/21] which includes:
"Check Sensor glucose readings by conducting a fingerstick test with a blood glucose meter under the following conditions, when Sensor glucose readings may not be accurate and should not be used to make a diabetes treatment decision:
  • When you are experiencing symptoms that may be due to low or high blood glucose"
This is just one of a list of 7 conditions; the other 6 are generally more closely defined. Slightly confusingly these 7 conditions are for all variants of Libre; there are further conditions and warnings under a Libre 2 heading. Also, again confusingly, none of these warning conditions seem to appear in the Libre 2 User's Manual.

I can't find any further guidance on whether low or high means to Abbott 4-8 or 4-10. Perhaps this goes back to pre Libre days when the medical advice was stay within 4-7 or 4-8, whereas my written conditions from the NHS when they offered me Libre 2 was that I was to keep between 4-10 for 70%, (ie time in target) and either reflecting the NHS recognising that as a T3c my circumstances were more volatile (yes, unlikely!) or a more liberal range is recognised as the new technology provides better BG visibility and better opportunities for diabetes control. Does anyone know of a more authoritative reference that ties this down?

I appreciate that this is also a subjective matter; individuals may well choose to use different in range criteria, depending on their circumstances and degree of control that they can realistically achieve.

For @helli, this is not in any way intended to be a critique of your excellent advice. I will be suggesting to any Newbies using Libre 2 that I am in dialogue with to read your advice. I have a further question on a sentence further down: "This is especially important when treating a hypo - always check your hypo has finished with a finger prick and do not retreat because Libre says you are still low". Is the word 'retreat' a typo, or did you mean something slightly to the contrary? I originally speed read this to mean: check your hypo has finished with a finger prick and trust that, even if Libre says you are still low (it will always be lagging actual BG). I personally think that I used to get big and rapid swings to hyper because I over-reacted to hypos; it was a sort of panic response, possibly worsened by misleading Sensor information.
 
I can't find why, but in my mind normal is "between 4 and 10".
There's not that much difference between "between 4 and 10" and "between 4 and 8" really. So likely it doesn't make much practical difference?

I presume you got the "between 4 and 10" from the idea of time in range, where the "range" is usually 3.9 to 10.0. (Though pregnant women (and women intending to become pregnant) should use a narrower range, I think.)
 
For @helli, this is not in any way intended to be a critique of your excellent advice. I will be suggesting to any Newbies using Libre 2 that I am in dialogue with to read your advice. I have a further question on a sentence further down: "This is especially important when treating a hypo - always check your hypo has finished with a finger prick and do not retreat because Libre says you are still low". Is the word 'retreat' a typo, or did you mean something slightly to the contrary? I originally speed read this to mean: check your hypo has finished with a finger prick and trust that, even if Libre says you are still low (it will always be lagging actual BG). I personally think that I used to get big and rapid swings to hyper because I over-reacted to hypos; it was a sort of panic response, possibly worsened by misleading Sensor information.
I did mean “retreat”.
The standard advice for treating hypo is to
- test to confirm hypo
- treat hypo with fast acting carbs
- wait 15 minutes
- test again
- if still hypo, treat again with fast acting carbs

My advice is based on the 15 minutes delay with CGMs. If you rely on Libre for the second test (after 15 minutes), Libre may still be reporting low when your levels have raised over hypo levels.
So, always check you have recovered from hypos with finger pricks - do not rely on Libre (or any other CGM) to decide whether you need to retreat a hypo.

I hope that makes sense.


Regarding the comment about Libre being accurate at “normal levels”, I commented between 4 and 8 based on my experience with Libre rather than what is meant by “normal levels”. But as @Bruce Stephens mentioned, there is not a huge difference between 8 and 10.
Plus, we are all different so some people may find Libre accurate up to 10 whereas others find it accurate only as high as 8.
 
I did mean “retreat”.
The standard advice for treating hypo is to
- test to confirm hypo
- treat hypo with fast acting carbs
- wait 15 minutes
- test again
- if still hypo, treat again with fast acting carbs

My advice is based on the 15 minutes delay with CGMs. If you rely on Libre for the second test (after 15 minutes), Libre may still be reporting low when your levels have raised over hypo levels.
So, always check you have recovered from hypos with finger pricks - do not rely on Libre (or any other CGM) to decide whether you need to retreat a hypo.

I hope that makes sense.


Regarding the comment about Libre being accurate at “normal levels”, I commented between 4 and 8 based on my experience with Libre rather than what is meant by “normal levels”. But as @Bruce Stephens mentioned, there is not a huge difference between 8 and 10.
Plus, we are all different so some people may find Libre accurate up to 10 whereas others find it accurate only as high as 8.
Thank you, now I understand: retreat as in treat again, as opposed to back off. Thank you.
 
There's not that much difference between "between 4 and 10" and "between 4 and 8" really. So likely it doesn't make much practical difference?

I presume you got the "between 4 and 10" from the idea of time in range, where the "range" is usually 3.9 to 10.0. (Though pregnant women (and women intending to become pregnant) should use a narrower range, I think.)
Yes, I appreciate that this difference isn't very significant and that in real terms most people, even Newbies, are less fussed about the upper limit. I wanted to pin this down since I found when I was a Newbie that I readily became mesmerised by complying with specific numbers and, even as a retired chartered engineer (and accustomed to rounding up or down to achieve a safe outcome) blinkered adherence would overtake common sense. Thank you.
 
I did not realise how reliant I have become with the Libre until today. I changed the sensor on Saturday, and it has been a bit erratic , out by 3, today it was at 7.3 and I was 4.3, so have been finger pricking today and now saying 10.3 finger prick 7.4. I will report it tomorrow. I have been lucky in that this is my first faulty sensor in 8 months of usage.
 
I did not realise how reliant I have become with the Libre until today. I changed the sensor on Saturday, and it has been a bit erratic , out by 3, today it was at 7.3 and I was 4.3, so have been finger pricking today and now saying 10.3 finger prick 7.4. I will report it tomorrow. I have been lucky in that this is my first faulty sensor in 8 months of usage.
Yes, I had drifted into this scenario. Then I had to correct a high, did a finger prick and found Libre was very adrift (low in that case). So I now more routinely do a periodic finger prick. I also use the Diabox app on my android phone and this provides me with CGM; this app can be calibrated against finger prick BG, when in a steady state. So that also has now brought about periodic finger pricking.
 
Upper limit is arbitrary, own is set at 9.
Thank you.
Since Abbott (Libre) specifically advise us to check by finger prick when low or high, I wonder what upper threshold Abbott had in mind. The low of 4 is self-evident - strictly could have been 3.9, since this relates back to US units.
But at what high point do Abbott believe their sensor needs verifying by finger pricking?
 
Does anyone know of a more authoritative reference that ties this down?

I would think the 70% 4-10 request comes from the International Consensus on Time in Range, rather than Abbott?

There was a thread about this a while back which I will see if I can find…
 
There was a thread about this a while back which I will see if I can find…

Here you go:


I’d forgotten I had added it to the ‘useful links’ thread 🙂
 
Well, personally, I haven't found the sensors are that ' out ' when higher anyway, unless I'm suddenly either soaring or plummeting.
 
But at what high point do Abbott believe their sensor needs verifying by finger pricking?
All I remember are more fuzzy guidelines. If you're low, you want to verify (after correcting) because of the ~15 minute delay, and otherwise you should verify if a reading doesn't match how you feel.

I guess there might possibly be some benefit if you're over (say) 10, but I tend to assume that finger prick tests will also be less accurate outside the normal range, so all any of them are saying is that I'm too high so I don't worry too much about exactly how high I am until I start getting in range again.
 
After a few years of Libre’ing, I have learned how different our bodies are.
The concept of testing when high or low is vague because sensors are less accurate at different levels for different people. Hence it being a guide not an absolute.
As an engineer, I understand the desire to be specific but human bodies don’t work that way so managing diabetes is a bit of an art form rather than pure science.

I definitely find Libre readings are a long way off finger prick readings when over 12 and drifting away at 8 or 9. As I correct at these levels, it is not good enough for me to just think of it as higher than I want. A correction dose for 15 is very different to a correction dose for 12, especially as I know my body becomes insulin resistant over 14 thus changing my correction ratios.
 
Yes with the correction ratios but there again the bolus wizard on my pump doesn't recognise that, so I always have to remember to tell myself to add a bit, so that's not very reliable. But nowt whatever to do with Libre.
 
After a few years of Libre’ing, I have learned how different our bodies are.
The concept of testing when high or low is vague because sensors are less accurate at different levels for different people. Hence it being a guide not an absolute.
As an engineer, I understand the desire to be specific but human bodies don’t work that way so managing diabetes is a bit of an art form rather than pure science.

I definitely find Libre readings are a long way off finger prick readings when over 12 and drifting away at 8 or 9. As I correct at these levels, it is not good enough for me to just think of it as higher than I want. A correction dose for 15 is very different to a correction dose for 12, especially as I know my body becomes insulin resistant over 14 thus changing my correction ratios.
Thanks again @helli .
I originally was trying to be clear in my mind what the limit criteria are for Abbott, in their caution about checking when low or high. Mike @everydayupsanddowns has effectively answered that question and I may yet email Abbott and ask them - but I feel sure they will tell me 4-10, whether that is their own criteria or derived from the International consensus.

I fully understand the principle that we're all different, even without the core differences that arise between T1, 2, 3c etc. So I support the concept that exactly when to test (ie at which number) is vague, even without the further variation arising between any sensor reading and actual BG.

BUT, when I was a Newbie (still am in so many ways) I needed distinct boundaries (ie actual nos) to reduce that fear of the unknown and confusion because it's not clear which no applies; also my wife, who had to bear the brunt of this confusion and who by her own admission is less numerate than I am, was particularly stressed by what was critical and when. Some of my early hypos, before carb counting never mind Libre, were extremely deep and made me very confused and confusing as well as irritable; they frightened her as well as me. So being specific to me as a Newbie was important.

I also fully agree most aspects of managing diabetes is an art rather than a science. Even carb counting (which implies tight nos and thus close to science) requires judgement: most packaging for oranges gives carbs per 100gm with skin on; are the carbs in the skin which I don't eat, the same as the flesh? No (9% vs 13.5% apparently); but are all orange types the same carb % - I seriously doubt it, however oranges are relatively low in relation to other foods and can be safely approximated with little consequence for MDI. But rice is a challenge: the packaging info varies; if microwaved in the bag, the packaging can be trusted (I think), but if boiled in a big pan of water how much starch is washed out by the water? So even carb counting is as much an art as a science. Insulin dosing might seem a science, but deriving ratios is more an art - certainly for me still after 18 months of carb counting. Adjustments for activity ... art; adjustment for medical stress ... art bordering on guess. So no doubt in my mind there is a lot of art needed - and I am at heart a Civil Engineer, used to calculating accurately, with elements of "art" for unknowns or vague knowns, and rounding up or down as appropriate to ensure a further safety factor. As a part time Structural Engineer the rounding off was criticised in peer reviews of my calcs!

I, like you, find Libre well adrift when out of normal range and like you I find my insulin resistance increases at some point above 15. Ironically that didn't happen last night; as I went to bed my BG raced from 7s to above 10 and at 15 I did a correction bolus; I contemplated increasing that correction but didn't and 30 mins later the alarms woke me as I plummeted to 3.2 by 01.30am. Diabox told me I spent almost 90 mins in the 2s, but is that reliable data? I doubt it; at no stage did I feel hypo. I have no idea why I had a surge, and presume some false aspect from something unknown hence why I plummeted. It wasn't a compression low, I was dozing much of the time and my sensor was completely out in the open. Unfortunately this happens a lot and why I consider I have very brittle diabetes; it goes with T3c after total pancreatectomy. I can't even account for this by a latent pancreatic surge!

I fully understand your using 8 as a threshold; I do corrections more frequently now than previously, but lack confidence to bolus stack, so always wait 4.5 to 5 hrs before correcting and my threshold has been 10, but often a little higher. This month I've been consciously using activity to control my rising BG. But this is very intrusive to my quality of life and not sustainable in the longer term.

Anyway, all that said, I liked your Dec 21 advice on the limitations of Libre 2 and intend to suggest those to T3c Newbies. I would have been happy to see those when I first got Libre 2
 
I, like you, find Libre well adrift when out of normal range and like you I find my insulin resistance increases at some point above 15. Ironically that didn't happen last night; as I went to bed my BG raced from 7s to above 10 and at 15 I did a correction bolus; I contemplated increasing that correction but didn't and 30 mins later the alarms woke me as I plummeted to 3.2 by 01.30am. Diabox told me I spent almost 90 mins in the 2s, but is that reliable data? I doubt it; at no stage did I feel hypo. I have no idea why I had a surge, and presume some false aspect from something unknown hence why I plummeted.
Can I ask if you double checked these readings? It sounds like you felt they were unexpected and I would certainly say that 15 was a hyper.
Also if you crashed to 3.2 30 mins later did you not double check with a finger prick and then again 15 mins afterwards to ensure that your hypo treatment had worked, in which case you would know that you were not hypo for 90 mins.

It seems extremely unlikely to me that even the fastest acting insulin would bring you down from 15 to 3 within 30mins of injecting it. I could understand a drop like that if you had injected the bolus insulin 15-30mins before and it was reaching peak activity but not from a dead start if you know what I mean so your numbers just don't seem to add up to me at all which is why I would be interested in knowing the finger prick results.

For me, it makes sense that a Libre reading which indicates that I need to take action, should be double checked, unless I can see a logical reason for that reading happening... ie after exercise. For instance my levels dropped earlier today after I ran up a steep hill to my stable yard to facilitate an unexpected haylage delivery. Half an hour later I was collecting eggs in the hen house and felt the hypo. Libre said 3.1 but it exaggerates 🙄 . I knew I was hypo because the run up the hill had likely caused it and I was feeling it quite strongly..... Ordinarily in that situation I would just have treated it without double checking because it was understandable and to be expected but I was curious to see how good my hypo awareness is so I did actually double check on that occasion and got a 3.9 (happy with that! 😎 ). The thing is that at 3.1 on the Libre I need to take action to treat a hypo, so a double check would be advised. If you are advised to do corrections over 10 then a reading of 11 or above on the Libre would constitute a double check with a finger prick before calculating and administering a correction dose. Personally I correct above 8 because once I get above 10 I am less responsive to insulin and it takes longer to come down and I like to keep my TIR stats in good nick because I treat them like a computer game.... but that is just me.

I agree for Newbies that clear rules are helpful and double checking before taking action is I believe the standard advice. Someone very newly diagnosed will usually not be instructed to do corrections at all and then after a few weeks, possibly correct over 15 and as your levels slowly come down then maybe over 12 and then 10 and now for me 8 is usually my personally imposed limit where I am happy to do corrections. Not everyone would be comfortable with that, so it is very hard to give definitive numbers.
 
Thank you.
Since Abbott (Libre) specifically advise us to check by finger prick when low or high, I wonder what upper threshold Abbott had in mind. The low of 4 is self-evident - strictly could have been 3.9, since this relates back to US units.
But at what high point do Abbott believe their sensor needs verifying by finger pricking?

No idea mate, pass on that one.
 
Can I ask if you double checked these readings? It sounds like you felt they were unexpected and I would certainly say that 15 was a hyper.
Also if you crashed to 3.2 30 mins later did you not double check with a finger prick and then again 15 mins afterwards to ensure that your hypo treatment had worked, in which case you would know that you were not hypo for 90 mins.

It seems extremely unlikely to me that even the fastest acting insulin would bring you down from 15 to 3 within 30mins of injecting it. I could understand a drop like that if you had injected the bolus insulin 15-30mins before and it was reaching peak activity but not from a dead start if you know what I mean so your numbers just don't seem to add up to me at all which is why I would be interested in knowing the finger prick results.

For me, it makes sense that a Libre reading which indicates that I need to take action, should be double checked, unless I can see a logical reason for that reading happening... ie after exercise. For instance my levels dropped earlier today after I ran up a steep hill to my stable yard to facilitate an unexpected haylage delivery. Half an hour later I was collecting eggs in the hen house and felt the hypo. Libre said 3.1 but it exaggerates 🙄 . I knew I was hypo because the run up the hill had likely caused it and I was feeling it quite strongly..... Ordinarily in that situation I would just have treated it without double checking because it was understandable and to be expected but I was curious to see how good my hypo awareness is so I did actually double check on that occasion and got a 3.9 (happy with that! 😎 ). The thing is that at 3.1 on the Libre I need to take action to treat a hypo, so a double check would be advised. If you are advised to do corrections over 10 then a reading of 11 or above on the Libre would constitute a double check with a finger prick before calculating and administering a correction dose. Personally I correct above 8 because once I get above 10 I am less responsive to insulin and it takes longer to come down and I like to keep my TIR stats in good nick because I treat them like a computer game.... but that is just me.

I agree for Newbies that clear rules are helpful and double checking before taking action is I believe the standard advice. Someone very newly diagnosed will usually not be instructed to do corrections at all and then after a few weeks, possibly correct over 15 and as your levels slowly come down then maybe over 12 and then 10 and now for me 8 is usually my personally imposed limit where I am happy to do corrections. Not everyone would be comfortable with that, so it is very hard to give definitive numbers.
Hello @rebrascora ,
Several strands to your response, but I'm happy to chat in case you can see a trend that I'm missing. This reply is, in retrospect, ultra lengthy and I won't be offended if you speed read and move on. At least doing this written analysis reassures me that within what I knew in the small hrs my routine was broadly OK.

Yes, my BG behaviour yesterday was sort of unexpected - with the caveat that I get lots of BG behaviour that is unexpected and at this 2 yr point I'm no longer so stressed or bothered about this aspect.

I started yesterday expecting to be active during the day, as I frequently am. So I reduced my breakfast bolus by 65%, which is a reduction I've found by trial and error over the last 2 months. (I'm stepping over whether I need to review my breakfast ratio for now, but I am alert to this possibility. The important thing is I've found a routine that normally works.) But "things" intervened, including heavy rain as I was changing to go gardening. So I was very inactive - for the first time for a while - and not surprisingly my BG rose during the afternoon into 10s by 3pm, 12 by 4.30pm and I gradually reduced this with modest indoor activity. Bolus for dinner + correction at 6.30pm and dinner at 7.30pm; my normal pre-bolus is only 15 mins. I rarely eat until I'm approaching 7 unless the domestic timetable forces my hand, when I'll eat while a bit higher! (Diabetes is in practice already a huge intrusion into our lives, but I'm alive, still getting all clear from the onc'y spec, so we're accepting the price, for now.) Anyway by 8.30 pm my BG was climbing then levelled in the 9s until at 9.15pm it more aggressively rushed to 11.8 by 9.33pm; stayed in the 11's for a further hr then climbed steadily until at 11pm as I went to bed it was on the cusp of 15. I checked my carb count was sound and concluded my body was reacting to the general lack of activity overall (but slightly wondered if my dinner pre-bolus had been faulty - my Echo pen confirmed the dose and timing @ 6.30pm, but had I been 'lazy' with my technique?).

Anyway based on a finger prick just after 11pm I took a correction and went straight to sleep; I sleepily cancelled an Urgent high alarm for above 15 (actually at 16.3) at around 11.30pm; I knew I'd taken the correction, so just needed to wait. Shortly after the 'normal' high alarm told me I had moved back into the 10-15 zone and a different "going down" alarm at some point told me I was falling. Then at 1.47am my Urgent low alarm told me I was at 3.9, finger prick said 4.3; this was my only finger prick in the small hrs - I wasn't hypo, so no need to wait and verify that I wasn't hypo. Libre misleading again? CGM told me after the event I'd been 7.5 at 1.37, ie an apparent drop of 3.6 units in 10 mins; there is no recorded reading for 01.42, which strongly suggests Libre had been overwhelmed by that rapid fall in 10 mins. This is pretty fast but I've fallen faster, so I wasn't unduly phased. At this point in my sleepiness I concluded my earlier high wasn't a solid high, just a momentary thing (of course with no identifiable explanation) the on board insulin was doing its stuff and was grateful that my 11.30pm correction hadn't been more aggressive - which I sometimes do when I'm above 15. I ate a 6gm biscuit to nudge myself away from the lower 4s (probably should have been 2 biscs) and even remembered to take a Creon capsule with a drink. I take Creon on an industrial scale, which is a slight throwback from the 21+ months I spent malabsorbing all food! I went back to sleep and was surprised to see later that Libre and Diabox had me in hypo zone for 130 mins, from 01.47 to 03.58. I don't believe I actually ever was: no night symptoms such as sweats or restlessness and I'm pretty sure I couldn't have caused a compression low - but who knows .....? If I'd been hypo it could only have been very shallow. I found this out at around 4.30pm when I needed my normal stroll to the bathroom and as I got out of bed saw my reading leap from low 3s to upper 4s. Then the dawn phenomena comes into play and dissipates 30 mins later, confusing the picture; frequently happens. Further scrutiny now tells me I was still wobbling around the high 4s and upper 3s until 06.30, so definitely needed 2x biscs and perhaps more. Hindsight is a great thing! Foot on the Floor at 08.09am took me from 4.4 to 6.6 by 08.39, then started dissipating over the next 30 mins before a shower (mild activity), prebolus and breakfast 40 mins later. Been high 10-12 a lot of today, until a moderate walk to a shop brought me back into the 9s.

I have an android phone so I have the Diabox app to give me CGM along with an array of alarms with sounds that are distinctive and tell me roughly what is going on; eg high pitch siren for the upper high alarm and lower pitched fire truck siren for the Urgent low alarm and rising string of notes when in rapid rise and falling string of notes for falling fast. So much of the time I hear an alert, merely glance at my phone (or watch by day) and see a reading and a trend. This has allowed me to intercept lots of hypos and while Libre records several Low Glucose Events most of those never occurred - just Libre reading low and providing false data. This is a nuisance, but manageable; I find it slightly more difficult when Libre is naturally reading high - can go towards hypo without due warning. Normally Diabox, which uses its own algorithm, is calibrated to reflect actual BG; but this needs a decent period of steady state to do the calibration and a second steady period at least 2 hrs later to confirm the earlier calibration. Diabox is, of course, dependent on Libre providing continuous output to convert into readings. Technology is wonderful when it works well; and when it doesn't ...*#*#*..!

So the night drop wasn't so rapid overall: 11.30pm @16.3, Urgent alarm 01.47am @ 3.9 (actual 4.3, but was this simply Libre lag?); albeit with an apparent crash from 7.5 to 3.9 in 10 mins. When watched, Diabox provides quantified rise and fall numbers as well as trend arrows; its theoretically capable of displaying 1.6 mmol/L/min and I've seen it show 0.7mmol/L/min

I do still genuinely, but infrequently, go hypo and I recognise the symptoms, but I've not had a deep "horrible hypo" for at least 3+ months. So I'm happy that I have reasonable hypo awareness. I finger prick quite a lot at the moment, partly to verify when I'm not hypo despite Libre saying otherwise, partly to calibrate Diabox and always before I apply a correction; if I'm high I don't trust Libre or Diabox (which was why I originally started this thread, to be clear about the official boundaries [4-10]!).

When I was discharged from Hospital in Feb 2020, the senior DSN gave me a standard formula for bolusing which included simple rules for correcting highs, so I was mentally equipped to do corrections from the outset. Because she personally dealt with all the HPB patients on their transfer from intravenous insulin to MDI, she was perhaps conscious that most of us were T3c and so likely to have more erratic (brittle) diabetes. So from my experience I wouldn't conclude that Newbies aren't ready to do corrections.

This year, I've been trialling using lots of activity to contain high periods, after a successful couple of months last year in almost stopping all hypos by diligently responding to falling alerts and alarms. Had some success with high control, but it's too demanding to be a long term solution.

That's enough for now, don't worry if this an overload!
 
Status
Not open for further replies.
Back
Top