Freestyle Libre - Sensor failure rate of 40%

[Proud to be erratic]

Won't I be regarded as a PITA reporting it via Yellow Card?

No. If the device is not working, eg a message saying sensorcerror - replace; then you have no choice to replace and it can only help highlight awareness of the Libre unreliability issues if you report it.

For me, the sensors are still "usable" within limits, it's just that sometimes they are insufficiently accurate with the actual number, which is something I can work around,

Also, for me. I work with an inaccurate sensor, until the difference (or general behaviour) becomes unworkable. At this point my judgement is that it is no longer fit for purpose and I often unilaterally make that decision; from that point I contact Abbott and seek a replacement, which they have always given, not least because I have the record in my LibreView notes of the circumstances including the readings with excessive differential.

Once a sensor starts misbehaving, I promptly fit a new sensor - but don't activate it until I've decided that is necessary. I have worn a 2nd sensor for several days, in these circumstances.

and sometimes they fail, which is why I keep strips handy. The trends thing is great, and so are the alarms especially when used during physical exertion. I'm kinda queasy using Yellow Card in case they take it away.

I can't imagine that you could lose your entitlement to CGM which works properly. I think no need whatsoever to feel vulnerable to losing your prescription rights.

also I'm unsure what "too frequently" means in context. 50%? 25%? what is expected, what are the thresholds.

Since I have a 50% failure rate, ie sensors that stopped mid-term or are just no longer workable then I don't need to wrestle with the frequency question. But I think even a very low frequency of failures is too much and those responsible for approving the device as well as the NHS person doing the collective purchasing will know what the contract tolerance has been set at. I don't think we Users need to know that. What is reasonable for you?

 

[BadaBing]

Won't I be regarded as a PITA reporting it via Yellow Card? For me, the sensors are still "usable" within limits, it's just that sometimes they are insufficiently accurate with the actual number, which is something I can work around, and sometimes they fail, which is why I keep strips handy. The trends thing is great, and so are the alarms especially when used during physical exertion. I'm kinda queasy using Yellow Card in case they take it away.

also I'm unsure what "too frequently" means in context. 50%? 25%? what is expected, what are the thresholds.

Excellent questions.

I'm not suggesting that we should contact the MHRA on each and every occasion something goes wrong with a cgm/flashgm system. That would of course no doubt render a reportee as a PITA in the eyes of the MHRA.

We each need to exercise judgement about when and what to report.

As well as authorising and licensing the use of drugs and medical devices in the UK, it is also the MHRA's responsibility to ensure that once available to UK patients those drugs and medical devices continue to (a) work and/or are (b) safe.

A few weeks ago I had a run of five (5) successive Libre 2 sensors that failed over the course of ten days - one after the other - to provide me with low and high glucose alarms using a Freestyle Libre 2 Reader that Abbott had sent me. Abbott customer services diagnosed on each occasion from a code on my Libre 2 Reader that the sensor was faulty.

Four (4) of those five sensors had been sent to me by Abbott customer services to replace Libre 2 sensors that had failed previously.

Maths was never my strong suit, but even I realise that is too high a successive failure rate for a healthcare product.

I'm all for allowing someone who provides me with defective goods or services a reasonable opportunity to put things right. But how many chances must I give them? Particularly when those goods are health-related?

So on that occasion I filed a Yellow Card report.

Some people rely on these sensors to indicate and notify them by setting off an alarm (via a Libre 2 Reader or the Librelink app) when their glucose is going low or going too high so they can double check with a blood fingerstick test and make treatment decisions.

Some people rely particularly on the Libre 2 to alarm when they are asleep so that they or a partner or carer can wake up, double check and treat hypoglycaemia.

Over the last few months I've seen accounts of people on other forums who say they went to sleep relying on their flashgm to give them low glucose alarms. Some of those people woke up to find themselves being treated by paramedics for severe hypoglycaemia! With all the consequences that carries for any diabetic who finds themselves in a similar situation who drives a motor vehicle in the UK.

So it's a judgement call. it's not about thresholds. Dare I say, to quote various past and present members of HM government, it's about using one's commonsense.

 

[rebrascora]

Don't the Libre sensors (Libre 1 that was as well as Libre 2 and 3) contain a watch battery, that one would usually find in a watch that would power a watch for at least a year?

I am pretty certain it isn't the battery which governs the life of the sensors. They will have an expiry built into the unit itself, which will be because the adhesive or sampling system or perhaps the skin and tissue health cannot be relied upon beyond that short time frame, rather than the battery may die. From what I have read, the original US sensors were just 10day usage before they shut down, but then the ones here have all been 14 day sensors and our New Zealand member @HamishR above has just mentioned that theirs were originally lasting longer than 14days until Abbott took steps to restrict them. Not sure how much longer beyond 14days that was, but can kind of understand Abbott restricting it because I would imagine they can't guarantee its use beyond that and likely a legal issue when it's function is to keep people safe.

 

[BadaBing]

I am pretty certain it isn't the battery which governs the life of the sensors. They will have an expiry built into the unit itself, which will be because the adhesive or sampling system or perhaps the skin and tissue health cannot be relied upon beyond that short time frame, rather than the battery may die. From what I have read, the original US sensors were just 10day usage before they shut down, but then the ones here have all been 14 day sensors and our New Zealand member @HamishR above has just mentioned that theirs were originally lasting longer than 14days until Abbott took steps to restrict them. Not sure how much longer beyond 14days that was, but can kind of understand Abbott restricting it because I would imagine they can't guarantee its use beyond that and likely a legal issue when it's function is to keep people safe.

I'm sure you are correct. I suppose what I was trying to say (very badly) is that my understanding is that the batteries which are placed into these sensors are sufficiently powerful that, if necessary, they could power the electronics in the sensor for significantly longer than 14 days.

But then of course the wear time is limited for practical reasons. The adhesive Abbott use to make the sensor stick to one's skin would deteriorate. Wearing a sensor on a particular patch of skin for more than 14 days no doubt causes all sorts of dermatological issues for the user. There is no doubt an optimal time after which the sensors become hopelessly inaccurate.

And of course not forgetting the bottom line for Abbott and its shareholders.

 

[Eternal422]

currently have a failure rate of 5/12. By failure, I mean failed prematurely with an error code and had to be replaced. A couple of others that went on to work normally either started reading high (by high, I mean +4mmol) or low (more than 2mmol lower than fp 5 mins earlier) but became usable within 48hrs.

These are the columns. I was inaccurate in my previous note:

sensor started​

Supplier​

failure reason​

error number​

error code​

days before failure​

other errors​

lot number​

serial number​

date supplied​

expiry date on box​

the reason i capture lot number is to consider the possibility of "bad batch".

Looks good! I don’t have a column for specific error codes, so far I have just recorded in a comments column :

#

Serial Number

Date Started

Time Started (after 1h warm up)

Expired Date

Failure

Comments on Accuracy

Replacements from Abbott

Interestingly, I totted up failure rate and it actually isn’t as good as I thought, even ignoring a failure in the last hour or final day of life, I get 4 failures out of 15, I.e. nearly 27% failure rate! Quite shocking really! Being generous I could discount one of the failures which was failure to stick from the start (which may have been down to me being hot and my skin not being in the “right” state to apply it), but that is still a 20% failure rate. Hmmmm.

 

[Eternal422]

Once a sensor starts misbehaving, I promptly fit a new sensor - but don't activate it until I've decided that is necessary. I have worn a 2nd sensor for several days, in these circumstances.

I’m curious as to whether when the applicator fires the needle to insert the sensor filament, does this start the battery? In which case a sensor applied but not started may not last the full 14 days once started? Or is the battery always active and will last up to the expiry date plus 14 days, I.e. firing the needle does nothing other than introduce the filament into your arm?

 

[void]

[...about acceptable failure frequency]

I don't think we Users need to know that.

I think we do in order to intelligently use a facility like Yellow Card.

What is reasonable for you?

depends on the failure mode. It failing altogether is, for me, more useful and less inconvenient than it failing partly but still usable, modulo prompt replacement by the manufacturer. It's a bit more complex deciding what is "reasonable" with this tech as it's relatively new and relatively new tech can be relatively buggy.

"reasonable" is easier with eg test strips where the tech is well known and stable. A few test strips failing isn't an issue. Loads failing across multiple containers - yellow card.

 

[void]

Looks good! I don’t have a column for specific error codes, so far I have just recorded in a comments column :

#

Serial Number

Date Started

Time Started (after 1h warm up)

Expired Date

Failure

Comments on Accuracy

Replacements from Abbott

Interestingly, I totted up failure rate and it actually isn’t as good as I thought, even ignoring a failure in the last hour or final day of life, I get 4 failures out of 15, I.e. nearly 27% failure rate! Quite shocking really! Being generous I could discount one of the failures which was failure to stick from the start (which may have been down to me being hot and my skin not being in the “right” state to apply it), but that is still a 20% failure rate. Hmmmm.

that's quite good I think (27%) about half mine. My expectations are lower than yours though 😉

 

[rmk]

Hi Everybody,
My name is Renée from Sydney Australia. I was very interested in the many & various problems I've seen posted.
I myself have been a type1 diabetic for 35yrs & never have I had so much trouble as I have had using the Libre 2. I've been using them for about 4mths and dam so many problems, the first one the clip/pin had snapped off not usable, place another on and it did readings for 2 days, 3rd didn't work & 4th fell off.
Following instructions to the letter & still I have to purchase 4 at a time so at least 1 will work at the war they state in there information.
This system even stated that whilst having a 3 5 hypo that the level tested from my arm was at 6.2 & the only way other then deciding to test it the old way (finger pricking) was because I felt sickly & didn't trust the machine.
I am greatly so very disappointed they don't work well on hot days or if your in air conditioning to long so heat & cold another problem & then it's that they don't always stick well at all.
Well I've followed the application to the letter had my husband do it as well so I have proof to share information with my Endocrinologist. I wouldn't advise any person to utilise this crap it's extremely unreliable.

 

[Merryterry]

It would appear that some folk just don’t get on with the Libre. It’s not a systemic problem with the kit, otherwise many folk, statistically, would at least get the odd duff sensor. But there are folk like me on the forum who have never had a problem that wasn’t their own mishap.

The problem, therefore must lie somewhere along the line of constructing the applicator, and the application itself. The technique I use is first of all, clean the skin thoroughly with the wipes provided to clear any oil, or residual dried shower gel. A quick wipe is not really sufficient. While that is drying, build the applicator taking care that the lines are accurately lined up. The pressure needed when applying the sensor is the pressure needed to trigger the mechanism. Don’t hold the applicator otherwise, or do anything that might reduce the pressure. Don’t immediately remove the applicator - wait for about ten to fifteen seconds for the adhesive to warm up, then carefully lift the applicator off. I always give the sensor a little press to reassure myself it’s stuck down, but that’s a touch paranoid.

As I say, I’ve never had a problem following that routine. And the old sensors usually take a week or more to drop off after their expiry.

I’m sorry for putting the blame on the user, but that’s the only possible explanation, unless folk are so unlucky that they are the only ones who consistently get duff kit. How likely is that? And Abbott are probably thinking the same thing. They just don’t produce stuff with an intrinsic 40% failure rate.

Could it not be that Abbott are not carrying out Quality Control checks. If it was a user problem they would experience consistant problems. I have had 3 going wrong with 'Please Change the Sensor' messages on start up. Funny that the replacements never go wrong.

 

[Bruce Stephens]

Could it not be that Abbott are not carrying out Quality Control checks.

But some of us seem not to have nearly so many problems. (I've had to report only a couple.) I really think it's some individual variation. (Some people seem to get on much better with Libre than with Dexcom.)

 

[void]

Funny that the replacements never go wrong.

I've had both the nhs-issued and the replacements go wrong

 

[Proud to be erratic]

I have come to the same conclusion as @Bruce Stephens, there are lots of people getting great results from both the original Libre and now Libre 2. Yet I still have a high failure rate and my answer to my own question "is it me?" has to be "Yes".

But this thread clearly reveals that Libre 2 is a missed opportunity, in terms of technical development and progression and I find that both sad and exasperating.

Couple that with the extraordinary situation that Libre 3 is not really available in UK - incomprehensible really. Does UK have more rigorous criteria for medical acceptance than, say, Germany? If its caused by some "wrinkle" within the NHS contract - then it also makes no sense. Some people will choose to self-fund Libre 3 it it were available and whatever the rights and wrongs of self-funding it does alleviate, even if only by a tiny sum, pressure on the NHS budget. I frequently told my students and my peers if something makes no sense, it's usually because it is nonsense.

 

[void]

libre3 prob costs more.

 

[jgubits]

I'm new to Libre 2, after 9 days of using it, my first sensor start having problems, on the beginning in the middle of the night the alarm waked me few times showing a low levels of glucose around 60 mg/dl but a blood test was 97 mg/dl, then a series of lost signal and try again in 10 minutes, I didn't replaced it yet but in the last 24 hours it's on and off, if it continues I'll have to replace the sensor before 14 days, I spoke to the company they sending me another one.

So as a new user I'm a little disappointed hopping that the other sensors will be good.

 

[rebrascora]

I'm new to Libre 2, after 9 days of using it, my first sensor start having problems, on the beginning in the middle of the night the alarm waked me few times showing a low levels of glucose around 60 mg/dl but a blood test was 97 mg/dl, then a series of lost signal and try again in 10 minutes, I didn't replaced it yet but in the last 24 hours it's on and off, if it continues I'll have to replace the sensor before 14 days, I spoke to the company they sending me another one.

So as a new user I'm a little disappointed hopping that the other sensors will be good.

Hi and welcome to the forum.

The false low alarms during the night are likely what we call "compression lows" when you lie on the arm with the sensor causing the sensor to put pressure on the tissue underneath it and this causes it to indicate that your levels are dropping. It will rise again when you take the pressure off. Sometimes it will rise slightly higher and then settle back down to the level it was before you laid on it. This is one of the known limitations of Libre and I believe other similar systems are prone to the same issue. Gradually you either manage to apply the sensor to a spot where this doesn't happen or you learn not to lie on the arm with the sensor in a similar way to you learning not to roll out of bed, even when you are staying in a strange place as a guest.

Out of curiosity, are you in the USA because you are using different unit for measuring your BG than we use here in the UK. That isn't a problem, but we have to do a bit of mental arithmatic and divide by 18 to get an equivalent reading in mmol/litre. So if you read other posts here and the numbers don't make sense to you that is why and you would need to multiply by 18.

Anyway, good to have you here on the forum and I hope we can help you figure out how to make the most of your Libre.

 

[jgubits]

Thank you Barbara for your reply, now I understand that the false alarms are not related to the other problem of loss of signal that continues already for almost 48 hours it's constantly on and off and makes the Libre system useless, for 2 days I'm back to stabbing my fingers, I'll replace the sensor today before the and of 14 days, is there anything up to me to avoid that problem or as it looks to me a defective sensor.

Thanks.

 

[rebrascora]

@jgubits
I don't use the loss of signal alarm and so I just have it switched off and my reader works fine without it. I still get high and low alarms absolutely fine. I think you need to go into settings to disable it, which is the cog at the top RHS of the screen. Touch that and it opens up a menu and alarms are top of the list. Select that and you will see options for setting the low and high alarms and then the one at the bottom is Signal Loss and if you select "change alarms" below that it will take you to the next page where you select Signal loss Alarm and then turn it off and then select done.

 

[jgubits]

I disabled the signal loss alarm as you instructed but I also get a Sensor Error "Glucose reading is unavailable. Try scanning again in 10 minutes". sometime the reading is unavailable for 2 or 3 hours, then the reading comes back for a short time and unavailable again for couple of hours, that's the most disturbing issue that makes the system useless to me, is there anything you can suggest about that?

 

[rebrascora]

I disabled the signal loss alarm as you instructed but I also get a Sensor Error "Glucose reading is unavailable. Try scanning again in 10 minutes". sometime the reading is unavailable for 2 or 3 hours, then the reading comes back for a short time and unavailable again for couple of hours, that's the most disturbing issue that makes the system useless to me, is there anything you can suggest about that?

That is usually when your levels are changing very rapidly. Does it happen after particular foods? I know I ate some sweets and stuff that I shouldn't have the other day and I kept promising myself I wouldn't have anymore and then I did, so I had to inject more and more insulin and basically the Libre didn't know if my BG levels were rising and falling, so it threw in the towel and gave me that error message but clearly it can't count to 10 mins because it was out for half an hour or more and my levels crashed and I had to get the finger pricker out. Those are exceptional circumstances though and I don't think it was unreasonable that it got muddled. The problem is that it uses the previous readings to extrapolate in order to predict what your current BG levels will be to try to overcome the time lag with interstitial fluid, but if the previous reading are changing very rapidly either only in one direction or oscillating, then it can't safely predict what your next level will be, so it takes a time out in the hope that things will settle over the next 10 mins and it has a better idea of how things are panning out. So usually that will happen if you have eaten something very sweet without timing the insulin well or had too much insulin or injected it too soon or too late and your levels are changing fast or exercised with active insulin in your system. Ideally you want to prevent swings in your BG from high to low and back and try to keep things a bit more stable by improving the timing of your insulin or slowing those fast release carbs down with something slower... or avoiding them. I follow a low carb way of eating so I very rarely get problems like that.... unless I am naughty 🙄 , which doesn't happen very often.