First Libre 2 sensor and it seems inaccurate?

[bvnsheee]

Hi everyone. I was diagnosed with T1 last year and have been eager to get a Libre as I'm a new mum and can often suddenly go into hypo when running around after my son. I got my first sensor on Wednesday and it has been such a weight off my shoulders, getting alarms for my lows. My only issue is now that I'm starting to not trust it.

Since I started using it, my levels seem better than what they have been, by about 1-2mmol. This morning I woke up and it was saying I was 3.9 which hasn't happened before, but I'd slept on the sensor so figured it was just because of compression. I double checked my blood using a finger prick and it came up as 7 so I wasn't concerned. It got me thinking though and today I've double checked all my readings and they're consistently 1-2mmol lower than my blood glucose.

I've spoken with Abbott and they're sending me a new sensor but I'm worried now. Is it just bad luck that I got a sensor that was slightly off, or is this a common problem? If I'm going to be double checking my blood vs. sensor readings, it almost doesn't seem worth it (although I'm extremely grateful to have been prescribed it). I've seen some people saying they install theirs and don't use it for 2 days but that seems like a lot especially when during training we were told it was reliable after 60 minutes.

 

[Bruce Stephens]

Is it just bad luck that I got a sensor that was slightly off, or is this a common problem? If I'm going to be double checking my blood vs. sensor readings, it almost doesn't seem worth it (although I'm extremely grateful to have been prescribed it). I've seen some people saying they install theirs and don't use it for 2 days but that seems like a lot especially when during training we were told it was reliable after 60 minutes.

Different sensors can be a bit different. Remember that it's best to compare when your levels are roughly flat because there's a delay (Libre's raw readings are about 15 minutes after BG, and there's a predictive algorithm on top of that).

In that normal range 4-7 or so I'd say a difference of 1mmol/l is fine: that doesn't change what you'd do (well maybe at the lower end you would, but BG meters also have an error range). 2 is worse.

It seems that some people just find Libre not to work acceptably for them (for reasons that I think nobody knows) but do find other systems that work.

One thing I was told during the training is that often the best time to look at the graphs is the day after: look at what yesterday's graph is telling you (even if the actual levels aren't quite right). Similarly, once it starts showing them, look at the Daily Patterns page, again to look at trends. Those can provide value even if you can't quite trust the actual levels.

Also give it a couple of days and maybe it'll settle down, or (just as important) you'll get more used to the difference that the ~15 minutes delay can cause, especially combined with the predictive algorithm (which does sometimes feel rather aggressive to me).

 

[bvnsheee]

Different sensors can be a bit different. Remember that it's best to compare when your levels are roughly flat because there's a delay (Libre's raw readings are about 15 minutes after BG, and there's a predictive algorithm on top of that).

In that normal range 4-7 or so I'd say a difference of 1mmol/l is fine: that doesn't change what you'd do (well maybe at the lower end you would, but BG meters also have an error range). 2 is worse.

It seems that some people just find Libre not to work acceptably for them (for reasons that I think nobody knows) but do find other systems that work.

One thing I was told during the training is that often the best time to look at the graphs is the day after: look at what yesterday's graph is telling you (even if the actual levels aren't quite right). Similarly, once it starts showing them, look at the Daily Patterns page, again to look at trends. Those can provide value even if you can't quite trust the actual levels.

Also give it a couple of days and maybe it'll settle down, or (just as important) you'll get more used to the difference that the ~15 minutes delay can cause, especially combined with the predictive algorithm (which does sometimes feel rather aggressive to me).

Thank you, that really helps. I was wondering whether it's just a learning curve or whether the sensor was faulty, but Abbott didn't really explain and just said they'd send me a new sensor. I'll keep wearing this one til the new one arrives and check against my bloods just to make sure any hypos are genuine etc.

 

[Bruce Stephens]

Thank you, that really helps. I was wondering whether it's just a learning curve or whether the sensor was faulty, but Abbott didn't really explain and just said they'd send me a new sensor. I'll keep wearing this one til the new one arrives and check against my bloods just to make sure any hypos are genuine etc.

They do routinely differ (even allowing for the timing difference) by up to about 1 or so. But BG meters themselves are also not perfectly accurate (I think the permitted difference from a true reading (done with a vial of blood in a lab) is 0.83mmol/l in this normal range, and 15% outside that). So don't expect the numbers to exactly match except by accident.

We're recommended to check with a test strip when the scan feels wrong (usually when it's saying hypo when you don't feel hypo or the opposite).

I think most of us who really like them find they're good enough to be relied on most of the time. I think the BG meters giving readings to one decimal place has confused us a bit. If you think of 5.3 as being 5ish that's probably more accurate and really just as useful. And Libre can say you're 5ish and going down slowly (really that you were a few minutes ago, of course) or quickly (or other things) which is much more useful (even though it's inevitably about the recent past).

 

[helli]

I check my sensor at least once, usually twice, a day against finger pricks.
Typically, I check first thing in the morning before any food and last thing at night a long time since eating so no bolus insulin onboard and my levels are usually stable between 4 and 8 mmol/l.
This allow a quick spot check of the sensor and gives me confidence that it has not drifted.

 

[rebrascora]

Hi and welcome.

I am quite surprised that Abbott replaced a sensor that was consistently reading just 1-2mmols lower that a finger prick. Whilst being 2mmols out is not ideal, as long as it is reasonably consistent then it is workable and as @Bruce Stephens mentioned, firstly you can only really expect to get similar readings when your levels are nice and stable..... I usually wait until I have had a couple of Libre readings half an hour apart that only vary by about 0.2 to compare with my BG meter to assess how closely a sensor is behaving. I will normally do a couple of these "assessment tests" during the first day or two of a new sensor to get a feel for how closely it is mirroring my finger prick BG and I take that into consideration thereafter when I am making decisions about boluses/corrections etc. I find most sensors usually consistently read about 1mmol lower. I've had the occasional one that was about 2mmol lower which is a bit of a pain and one that was 3mmols lower which was unacceptable in my opinion and that of Abbott and was replaced. I also had one that read a touch higher and that was a pain as I am so used to them generally reading lower, but I just had to get my head around it as it wasn't far out, just different to what I am used to with them.
Also that your BG meter is not giving you an "absolute" reading either as it also has a margin of error in that reading and is not really accurate to the decimal place it gives you. It is however near enough and you will learn how to test and accept that the Libre readings is near enough too.... it's just that you need to adjust your mind set a bit and don't forget to listen to your body rather than rely too heavily on what it tells you and double check with a finger prick when how you feel and the Libre don't agree..... but always double check when it says you are hypo or hyper and recheck with finger prick 15 mins after treating a hypo as Libre will often continue to show your levels dropping after treatment when a finger prick will show you have recovered.
You should be aware that they are often less accurate at low and high levels, so you should always test with a finger prick before taking action when hypo or hyper. Even a sensor which consistently reads slightly lower than my finger prick in range can be 2-3 mmols higher once my levels get up to mid teens. It is just one of their quirks and you learn to accept that.

As regards applying it a day or 2 in advance of activating it, this really helps for some people. I usually apply mine 8-20 hours in advance and that gives me really good stable readings from the start. As with anything, people's bodies differ in how they respond and some people's tissue reacts to having a foreign object fired into it more enthusiastically than others no matter that the foreign object is sterile and it just takes longer to settle down again. There doesn't seem to be any downside to applying the sensor a day or 2 in advance of using it, so nothing to lose in doing that.

Libre is a fantastic tool but you have to understand and accept it's limitations and foibles and learn when you can rely on the info it tells you and when you need to double check. Good luck getting to grips with it.....

Just realised that I haven't done an assessment test of the new sensor I applied yesterday and my levels are beautifully stable at the moment so a good time to check.... 5.2-5.3 over the last half hour..... 5.0 on my BG meter.... I'll very happily take that!

 

[Proud to be erratic]

Hi everyone. I was diagnosed with T1 last year and have been eager to get a Libre as I'm a new mum and can often suddenly go into hypo when running around after my son. I got my first sensor on Wednesday and it has been such a weight off my shoulders, getting alarms for my lows. My only issue is now that I'm starting to not trust it.

Hi @bvnsheee,
I've had c. 25% fail rate of sensors. At first I found this stressful and a bit scary, but I reminded myself that I did my 12 month apprenticeship of finger pricking only and that I managed before I had Libre! Abbott have always replaced all of these failures. Nowadays, I mutter less than polite observations about a sensor failure and get on with activating the replacement sensor. I no longer get remotely stressed about the actual failure - its just a tiny inconvenience! Experience, thus knowledge, over time is a great help.

I fit my replacement at least 3 days in advance. My body still seems to react to the new foreign object. Also some of my sensors have failed 24-36 hrs early, so at least the replacement is already in place. The early fitting doesn't change the need for the 60 minute startup delay. Sometimes I activate the new one while the old one is finishing its defined duration. I can do this because I have a reader as well as a compatible android phone.
Using the reader for startup suits me (others might not want this) because there are times when the reader is quicker/simpler when busy or very active. To have both a reader and phone reading the same sensor means the reader MUST be the first device to activate the new sensor. Thereafter all alarms are provided by the reader, not the phone.

Also, invariably the reader only needs to be near the sensor to provide a fast reading, whereas my phone needs to be in exactly the right place for the NFC to pick up the signal and can take many frustrating seconds to 'whir' and finally give an error message. I have never succeeded in getting a phone scan successfully while on the move - just error messages. Whereas I do get the reader to give me a result as I'm walking into the local shops.

Since I have both the reader and phone in play, I put all input data onto my phone only and let the wi-fi transmit phone data to the LibreView website, for my and my Consultant's scrutiny. So I have never had to download any readings from the reader (which needs a cable and a PC).

Since I started using it, my levels seem better than what they have been, by about 1-2mmol. This morning I woke up and it was saying I was 3.9 which hasn't happened before, but I'd slept on the sensor so figured it was just because of compression. I double checked my blood using a finger prick and it came up as 7 so I wasn't concerned.

Well done on not over-reacting; trust your body signals as well as the tech.
The overall accuracy has varied for me from sensor to sensor. Some are constantly high with respect to actual BG, some constantly low and some start high and finish low! This is potentially confusing, but I try to stay calm and monitor the trends, not the precise numbers.
As @Bruce Stephens has said provided readings are broadly in range (4-10) then I just go with the sensor trend and I bolus dose with corrections based on what I'm seeing. I'm on 1x daily basal Tresiba and because that is relatively inflexible I rarely adjust my basal dosing. So I use my bolus NovaRapid insulin for food and corrections, along with simple snacks (5-10 gms carbs) to get a small boost when lowish. I find this system practical for me - only 2 variables: bolus and carbs. I let my alarms tell me when I'm low, high or rapidly changing - then look at scan and decide if I need to react.

It got me thinking though and today I've double checked all my readings and they're consistently 1-2mmol lower than my blood glucose.

When, as you describe, the sensor is lower than actual there is no real problem. But when the sensor is a modest amount higher than actual you can start to become hypo, initially unwarned - worsened by the natural lag (delay) of the sensor interstitial BG. If you are already alert to this problem, then its just the challenge of remembering that's how it is. I will set my low alarm relatively high, eg at 5 or even the max of 5.6, then get early warning and can take a small 5+ gm snack, such as half a cookie, or 1x Nairns oatcake(=6), as a precaution to fend off a potential hypo. It doesn't matter if that snack wasn't really needed - it's a form of insurance and won't push you very far up. A small amount of activity or exercise will quickly burn that off. But once really hypo it takes a more determined response to reverse that and often one over-reacts; hypos make one desperate to eat something (often too much) and then get onto the roller coaster of lows / highs / and lows again.

I've spoken with Abbott and they're sending me a new sensor but I'm worried now. Is it just bad luck that I got a sensor that was slightly off, or is this a common problem?

Yes, for me it's a 1 in 4 problem.

If I'm going to be double checking my blood vs. sensor readings, it almost doesn't seem worth it (although I'm extremely grateful to have been prescribed it).

And despite various limitations I'm also still very grateful for having Libre rather than just finger pricking. I used to finger prick c.12 x daily. Now, on a good day just once, sometimes 5 or 6 times on a volatile day. That increase is mainly because once sensor readings are out of range (which I interpret as 4-10) then even Abbott tell you to independently check with your meter.

I've seen some people saying they install theirs and don't use it for 2 days but that seems like a lot especially when during training we were told it was reliable after 60 minutes.

Keep in mind the Freestyle training videos are selling the better points of Libre. It's true that it starts transmitting readings after 60 minutes, how reliable they are depends on the wearer's body response as well as the reliability of the actual sensor - Abbott aren't going to tell you about the downside!

I now have the Diabox app on my android phone and get the readings continuously from L2 onto my phone and my smart watch. But that's a digression for now!

Hope this helps, sorry it's so lengthy.

 

[Loobyloo]

I found them very innaccurate (i had 4). Abbott did replace them as were reading up to 4-5 mmols out at times. They told me up to 2 mmols out was normal so you are lucky they replaced yours. If mine were continually 1 or 2 out i could have allowed for that but they were unpredictable, sometimes almost correct and other times 4-5 mmols out (reading hypo when i was in fact 7 or 8 on meter). I love the idea and ease of use but can't trust the readings so back to the BG meter.
I know some people get along with them and don't know why myself and others have these problems. Maybe just not good for some people.

 

[nonethewiser]

Not had any issues with libre 2 on accuracy or anything else, puzzling why some have problems & others don't.

Just apply sensor scan it then wait 60mins to use it, don't bed it in haven't done since start of using libre 1, do check occasionally with bg meter, mainly at start or in doubt about libre reading other than that its libre all the way.

 

[rayray119]

I found them very innaccurate (i had 4). Abbott did replace them as were reading up to 4-5 mmols out at times. They told me up to 2 mmols out was normal so you are lucky they replaced yours. If mine were continually 1 or 2 out i could have allowed for that but they were unpredictable, sometimes almost correct and other times 4-5 mmols out (reading hypo when i was in fact 7 or 8 on meter). I love the idea and ease of use but can't trust the readings so back to the BG meter.
I know some people get along with them and don't know why myself and others have these problems. Maybe just not good for some people.

oh my gosh mine when they worked were hit and miss too. athough they did fail consulllly.