Libre 2 Sensor Reliability and Accuracy Record

[Eternal422]

Full disclosure, I keep lists and spreadsheets But here goes . . . My Libre 2 sensor record since starting at the end of July this year :

8 accurate (*) throughout the 14 days
2 failed to apply (one didn’t stick, the other gave continual sensor errors and on removal the filament was bent)
1 read much lower than BG and failed after 4 days

(*) For me this means sensor readings up to 0.5 lower than BG, tested when BG is stable. Some readings have been as little as 0.1 different, so essentially identical.

All 3 failed sensors were replaced by Abbott, 2 by just filling in the online form (had to send back the bent filament one) and the 1 that read lower and failed after 4 days was replaced after a phone call to Abbott giving them examples of different readings.

Hopefully the above will help anyone new starting on the Libre 2 who may be worried and give some confidence in the sensors as regards their reliability and accuracy. Obviously everyone is different and sensor readings will always be slightly different to BG, but this does’t take away from the wealth of information you can get from the data.

Now just started my 12th sensor, fingers crossed it behaves well!

 

[Eternal422]

Thought it might be worth adding, I do occasionally check a sensor’s readings against BG finger stick tests, usually only when the Libre chart shows that BG is steady and flat. There is also a bit of guesswork in the lag between BG and the Libre’s ISF reading it converts to a BG reading, so taking a finger stick test then around 15 minutes later scanning the sensor should, if BG is stable, give a fairly close reading.

The first day of a new sensor for me always is not as accurate as during the middle part of its life, it does seem to take a bit to “settle” and I know others support this theory and have similar experiences.

A word of caution - don’t fall down the rabbit hole of continually comparing results between finger stick and Libre readings! They will never match all the time due to the lag with rising and falling BG. Of course test if how you are feeling does not match the Libre reading, otherwise I have found that it is best and less stressful to just go with the Libre readings, adjusting bolus from the readings As required.

Again, I hope this helps anyone just starting out with the Libre and will give some confidence in getting the best out of it!

 

[Tdm]

I insert my new sensor when i have 1-2 days left on the old one, and leave it to bed in 24hrs+ before starting it - that tends to work well

 

[tasty_morsel]

I came on here looking for information on the sensor readings. I've been trying to understand mine and what it all means and how it relates to how j feel and what I've eaten etc.

I knocked my sensor off today, it was feeling funny and then I was careless, but I noticed I'd bled a bit under it since it's been on, and the sticky out bit was also bent so I wonder if it initially went in badly. I didn't realise I could send it back.

 

[Eternal422]

Spoke too soon ! This sensor started reading about 2 mmol/l lower than BG yesterday and now today is continually giving the Sensor Error message, despite several loops through waiting 10 minutes and trying again I still get this, so just reported online to Abbott to get a replacement sensor. Of course, I am out this morning so this is really annoying! Will have to wait until I’m back home later to fit another sensor (luckily I have a couple in hand).

 

[Eternal422]

I came on here looking for information on the sensor readings. I've been trying to understand mine and what it all means and how it relates to how j feel and what I've eaten etc.

I knocked my sensor off today, it was feeling funny and then I was careless, but I noticed I'd bled a bit under it since it's been on, and the sticky out bit was also bent so I wonder if it initially went in badly. I didn't realise I could send it back.

You can get a bit of blood sometimes but whenever I’ve removed a sensor at the end of its life the filament is always straight. The sensor that gave me a continual sensor error had a bent filament when I removed it.

The Online sensor form has an option if you’ve knocked the sensor off and as far as I know they will send a replacement free of charge out to you.

What information were you looking for about readings?

 

[grovesy]

Has anyone else been told by Abbott to leave sensor for 2 hours , when getting the rescan in 10 mins message.
This morning I went to start a new sensor, then get getting the scan in 10 mins message. It then after repeatedly doing this I got the not working, change your sensor message. (reader). When I rang to report this, was told should leave for 2hours before rescaning, and to leave for 2 hours and if not working, ring back. At this I said would not be buying again. At this she went a way, when she came back she said she would replace, in the meantime I had removed the sensor and the filment was bent to a 90 degree angle. Again she said reiterated should leave for 2 hours when getting the 10 mi message.
P.S. I am a long time user and have had no problems for a long time.

 

[Eternal422]

Has anyone else been told by Abbott to leave sensor for 2 hours , when getting the rescan in 10 mins message.
This morning I went to start a new sensor, then get getting the scan in 10 mins message. It then after repeatedly doing this I got the not working, change your sensor message. (reader). When I rang to report this, was told should leave for 2hours before rescaning, and to leave for 2 hours and if not working, ring back. At this I said would not be buying again. At this she went a way, when she came back she said she would replace, in the meantime I had removed the sensor and the filment was bent to a 90 degree angle. Again she said reiterated should leave for 2 hours when getting the 10 mi message.
P.S. I am a long time user and have had no problems for a long time.

Sort of, the form says this :
The Sensor is unable to provide a glucose reading. Please wait 10 minutes and then scan again. If necessary, scan every 10 minutes until two hours has elapsed.
Were you able to scan your sensor and get a reading within the two hour period?

So not wait 2 hours before scanning but to try the 10 minute waits for a period of 2 hours and if you are still getting the message then report it for a replacement.

 

[grovesy]

That was not what she said. Before the 2 hours I then got the message, sensor not working. She did not seem to take that on board either .

 

[Eternal422]

That was not what she said. Before the 2 hours I then got the message, sensor not working. She did not seem to take that on board either .

I think they sometimes make the process up as they go along. However, that said, they have always sent out replacements to me without any issues whenever I have had problems.

If the filament was bent then that seems like the one I had that continually gave sensor error messages and wait 10 minutes before rescanning. I think with that one it was bent when the applicator mechanism fired it into my arm as it didn’t work from the off.

 

[grovesy]

I think they sometimes make the process up as they go along. However, that said, they have always sent out replacements to me without any issues whenever I have had problems.

If the filament was bent then that seems like the one I had that continually gave sensor error messages and wait 10 minutes before rescanning. I think with that one it was bent when the applicator mechanism fired it into my arm as it didn’t work from the off.

It is the first time they have been reluctant to replace, I have been self funding Libre 2 for a couple of years, and prior to that Libre for a number of years. I have had the odd one that has failed on day one as this one did, she was not going to offer replacement, till I said I would no longer buy, then she came back and offered a replacement.

 

[Eternal422]

Spoke too soon ! This sensor started reading about 2 mmol/l lower than BG yesterday and now today is continually giving the Sensor Error message, despite several loops through waiting 10 minutes and trying again I still get this, so just reported online to Abbott to get a replacement sensor. Of course, I am out this morning so this is really annoying! Will have to wait until I’m back home later to fit another sensor (luckily I have a couple in hand).

Just an update on my 12th sensor, as soon as I got home it started working again! It showed a huge spike which didn’t seem right as I hadn’t had any carbs to cause it (unless it was the stress of the sensor not scanning!) and then quickly settled into some accurate figures once again. It has been working ok since then and due for replacement tomorrow evening at the end of its 14 days.

So who knows what happened? I sent all this info to Abbott but have not heard anything back from them, I did say I would continue with the sensor assuming it stayed ok so there was no need for them to send a replacement.

 

[tasty_morsel]

Hello, sorry, I missed the last replies.

 

[Dave-T]

Makes me wonder if there is a dodgy batch of sensors doing the rounds currently. I've just had my second sensor in a row pack up after 10 days preceded by two days of bizarre readings. Have never had this before in using the system for 8 months.

 

[Tdm]

I wonder if the wider prescription if libre has stretched their capacity resulting jn issues..been on libre 2 fircyear now but getting a not more false night lows...i feel i should be able to run around 5.8 at night without getting woken up multiple times ( its not compression as its done it whilst i was wake from prev alarm and definitly not sleeping on it). Hope to get dexcom prescribed but won't be til jan if then - considering self funding a 3 month supply decom 1 to get some darn sleep!

 

[Proud to be erratic]

I wonder if the wider prescription if libre has stretched their capacity resulting jn issues..been on libre 2 fircyear now but getting a not more false night lows...i feel i should be able to run around 5.8 at night without getting woken up multiple times ( its not compression as its done it whilst i was wake from prev alarm and definitly not sleeping on it). Hope to get dexcom prescribed but won't be til jan if then - considering self funding a 3 month supply decom 1 to get some darn sleep!

@Tdm ,
My initial thought was that unluckily your interstitial BG is hovering close to your alert setting (or vice-versa). Since our BG is always changing a little then multiple alerts mean that you are actually very well managed by your MDI regime; you are rising and falling around that alert threshold multiple times, but overall your BG is pretty stable. So just reset your alert lower, say from 5.6 down to 4.8. There is still enough time to respond if you are woken again at this lower setting and head off a potential hypo with a biscuit, rather than relatively short lived glucose, such as JBs. Then this would also potentially mean your basal is a little strong and needs lowering. But this needs 2, or better 3, nights in a row to verify, I suggest.

However its never quite that simple with D! When you are getting woken up multiple times do you ever either do a finger prick test to confirm all is actually well or (and) do you ever have a small snack to nudge your BG up a little?

The finger prick should (nuisance that it is) reassure you by confirming what your true BG state is - then you can make an informed decision whether to: ignore, even switch off, the alert; to adjust the alert threshold; to take a snack.

If you are already taking a snack after each alert then your basal is a little too strong; but if those snacks are just a JB then that pure glucose doesn't endure, indeed further demonstrating that your basal is incorrect. A lower GI snack, such as a 6gm biscuit, should simply lift your BG a bit and since you have no rapid insulin left on board you will then sit above the alert threshold - unless, again, your basal is too strong and dealing with your extra carbs regardless.

The beauty of the CGM is that you can interpret the results so much more easily than by repetitive finger pricking - which was all one could do before having CGM.

Each night you are actually conducting your own fasting basal check - providing you are not taking any extra insulin as a correction just as you go to sleep.

If your BG should steadily rise, or spike, each night - then it could mean your basal is a little low, but first look at what your last meal was. If a fatty meal then it might not have been fully digested before your NovoRapid has dissipated, so once the digested carbs are late into your blood then insulin is needed; ideally from bolus rather than basal. Options here are to look at pre-bolus timings (or even a split bolus) and even basal timings.

[I'm on Tresiba, which has a 40 hr profile so timing of my daily basal injections is irrelevant, since every day I'm just topping up yesterday's basal. It's one big advantage of the seemingly inflexible nature of Tresiba. My NovoRapid has to be juggled to cover food, corrections and my irregular exercise/activity/routines]

On my recent DAFNE course we were encouraged to describe the 2 insulins as basal and quick acting (QA) to reduce the potential confusion between basal and bolus for those relatively new to MDI. Made sense really, most of us don't know why it's called a bolus anyway; I think, but never bothered tocheck, bolus means packet, presumably originating from Latin or Greek the foundation languages for modern medical terminology.

 

[Tdm]

@Tdm ,
My initial thought was that unluckily your interstitial BG is hovering close to your alert setting (or vice-versa).

If only! This (screen print) is a night when i got 4 false alerts. I finger pricked on each occasion and got 5.8, or 5.9. My alarm is set at 4.5, sometimes I set it lower but that does not help much.

You can see the algorithm hasn't even bothered to try to pretend the readings are on the line.

Most of my bonkers readings are overnight but you can see one at 8pm as well.

 

[rebrascora]

Looks to me like those overnight ones are compression lows.
The 8pm one looks like your levels were dropping fast and the algorithm over predicted the drop and then corrected itself leaving that particular reading hanging below the graph because it was an erroneous prediction.

 

[Tdm]

Not sure re compression lows as they happen when there is no pressure on the sensor.
Early on a wore a band close to the sensor with a item attached that made it uncomfortable to lay on the arm. It made no difference.
And whilst i can't guarentee i don't sleep on the sensor, there have been occcasions when a false alert has happened whilst i was still awake from a previous false alert and when no weight was on the sensor.
I have bog all hypo awareness, and live alone, so turning alerts off overnight is a last resort. I'm hoping to try a dexcom and hoping that will help.
Generally, sensors are getting more accurate...i think i heard one youtuber say that the libre 3 was less prone to compression lows. But the libre 3 isn't an option at the moment. But at least my new diabetes team seen to be more solution focused than my previous ones.

 

[Proud to be erratic]

Because you diligently finger pricked, you do at least have good evidence of correlation: the actual BG is c.1.3 above interstitial. So in that scenario, after the first or at worst the second alert, you could reset your alert threshold to 4.0 or lower (as an exception) and probably get no further interruptions, or simply turn alarms off.

Turning alerts off does step over the question of whether these are, or not, compression lows. Just got to remember to reactivate them in the morning. Honestly, even though you say these are not possible compression lows, the dots below the joined up line imply even the algorithm has ignored these and think they are outliers - which would suggest they are compression lows; or a truly faulty sensor. You could email Abbott for 1st thing Mon am and ask for proper technical guidance; I wouldn't hold my breath for a swift reply, but emphasis that multiple low alerts, against such consistent actual results makes the sensor unworkable for practical purposes = not fit for purpose! Your 8pm outlier reinforces the possibility of a faulty sensor - unless you'd dozed off in a comfy arm chair (one can wish) and created a compression low! You could gently try to blackmail them and say there is currently a strong business case for any of the other 3 CGMs now available from the NHS.

Personally I would have had a 6-10gm biscuit, no chocolate, and gone back to sleep - to allow my BG to get a bit higher and hopefully resolve that night's disturbance. I actually did that last night because my basal has had to be adjusted up and now downwards, following a minor surgical procedure and the antibiotics for a real infection - and currently my basal is still too strong and needs a further tweak less. If that hadn't worked for me I would definitely have turned my alarm off, knowing I was safe for the rest of the night.

I understand your caution, living alone and thus your diligence. I might add that in my email, to reinforce the aspect of "fit for purpose". Because I have no pancreas, my D can be exceedingly brittle (rapid rises or falls, no other pancreatic functions to help with BG regulation) I try to keep myself above at least 6.0 overnight. Given your "bog all" hypo awareness, you might consider giving yourself a larger buffer and set yourself overnight targets at c.6.5, which is still extremely good and not setting yourself up for longer term problems.

About Dexcom One: I want to change to this but my GP feels constrained by current local Bucks guidelines. I will be challenging this tomorrow. But various comments about how basic the Reader feels is slightly deterring me and currently I can't afford to replace my 12 month old mid range smart android phone. Have you, by any chance, done any research into this? If so I'd be grateful for your thoughts.