Freestyle Libre sensor Serious inaccuracies- Help us improve this!

[Zilsniggy]

That being the case pls report it. Only via your reports on the form 3500b can we get the Libre company to react. See the youtubes wheer i share unfortunately similar, if not worse, experiences and mind you i am tech savvy...

I've reported it to Abbott as long time ago, all they do is send another sensor. It's a bit of a lottery as to whether the replacement is accurate or not. Plus, I don't think it's a matter of being tech savvy or not! I'm an RGN and quite capable of siting the sensor properly, etc. I can only go on my own experience, but from various forums, I'd say the chances of a sensor being completely accurate are no better than 50%.

 

[TuviaDror]

I've reported it to Abbott as long time ago, all they do is send another sensor. It's a bit of a lottery as to whether the replacement is accurate or not. Plus, I don't think it's a matter of being tech savvy or not! I'm an RGN and quite capable of siting the sensor properly, etc. I can only go on my own experience, but from various forums, I'd say the chances of a sensor being completely accurate are no better than 50%.

Wow, i really hope and believe its better than 50%..... Incidentally, I to wasn't that impressed with the Abbott responsiveness (i mean i understand Abbott- they prefer attending to all the happy users and perhaps not invest too much in those few who are unhappy- unless they really have to...). That's why i filed a complaint by the FDA. If another one or two of us do this, i'm sure they will attend to it. Abbott will not ignore the FDA...

 

[tim2000s]

Actually, I've been using the 14 day version only. I never used anything else.... Does that still make sense?

Not really, no. The US 14 day version wasn't approved until 27th July this year and you have more than five weeks worth of data prior to that, based on the dates of your report! Are you in the US?

 

[Deleted Account]

I'm interested how people calculate accuracy of their Libre given
- the reading is about 15 minutes behind finger prick
- finger prick meters are also inaccurate
- any CGM experiences pressure problems recording low readings if you lie on them, e.g. when you are asleep so low night time readings without finger pricks to back it up should be taken with a pinch of salt.

I appreciate @TuviaDror found his Hb1AC was higher than expected.
But others have commented "the chances of a sensor being completely accurate are no better than 50%".
I'm intrigued to understand what the data is behind this fact.

I am just curious: as I stated previously, I follow the advice of my diabetes team and use Libre for trends not for bolusing so my curiosity is purely academic.

 

[TuviaDror]

Not really, no. The US 14 day version wasn't approved until 27th July this year and you have more than five weeks worth of data prior to that, based on the dates of your report! Are you in the US?

Ah, i forgot to mention, i got it in Israel, where i am based (where it was approved for Diabetes type I patients) from January 2018.

 

[Bill_St]

I would prefer to accept the FDA figures from 9075 professional tests comparing with laboratory standards than an individual using home BGM testing.
Over 90% concurrence within 20%/20mg in the non-hypo ranges above 80mg BG
Table 5

and Table 12 and 13 give the comparisons with day of wear.
Median Absolute Relative Difference of 7.6% or less after day 2

http://www.accessdata.fda.gov/cdrh_docs/pdf16/P160030S017B.pdf

 

[tim2000s]

Ah, i forgot to mention, i got it in Israel, where i am based (where it was approved for Diabetes type I patients) from January 2018.

Okay, let's just get one thing clear here then. The Israeli Ministry of Health is the regulatory body for the version of Freestyle Libre being used in Israel. Not the FDA.

The rest of this is somewhat speculative based on where you live and the timing of availability of devices in your region.

According to the import rules I can find, you will be using the Worldwide release device as AMAR requires one of :

• FDA 510(k) or premarket approval letter
• CE Marking certificate issued by a European Notified Body
• Certificate to Foreign Government (CFG) or Certificate of Free Sale (CFS)
• Proof of ISO 13485 certification

for approval of import of devices. There are additional checks for certain types of device but I don't think the Libre qualifies for those.

The timing of your device availability doesn't align with the 14-day US device approvals, so the most likely scenario is it uses the CE marketing certificate, which isn't approved for using to dose insulin. If you have issues with the device, you should be directing them at AMAR as your regulatory body about how Geffen Health is marketing the device, as the very clear marketing within Europe is that you don't use it to dose from.

The FDA doesn't (and can't) consider complaints about a product that is not licensed in its jurisdiction (which your version of the Freestyle Libre isn't). I'm also not sure whether it accepts complaints from non-US citizens (even though you can send a letter/on-line form).

 

[Zilsniggy]

Wow, i really hope and believe its better than 50%..... Incidentally, I to wasn't that impressed with the Abbott responsiveness (i mean i understand Abbott- they prefer attending to all the happy users and perhaps not invest too much in those few who are unhappy- unless they really have to...). That's why i filed a complaint by the FDA. If another one or two of us do this, i'm sure they will attend to it. Abbott will not ignore the FDA...

Since I am in the UK I doubt I could complain to the FDA, since they have no jurisdiction here.

 

[Zilsniggy]

I would prefer to accept the FDA figures from 9075 professional tests comparing with laboratory standards than an individual using home BGM testing.
Over 90% concurrence within 20%/20mg in the non-hypo ranges above 80mg BG
Table 5

and Table 12 and 13 give the comparisons with day of wear.
Median Absolute Relative Difference of 7.6% or less after day 2

http://www.accessdata.fda.gov/cdrh_docs/pdf16/P160030S017B.pdf

I can only say as I find, and my finding is that it is nowhere near as accurate as you say. There are people for whom this device does not work, and it is not ideal for. To continue to harp on about it's so called accuracy when Abbott freely admit that it's not effective for everyone is disingenuous. Perhaps it works well for you, it certainly doesn't for me, so, since I self fund it, I won't be buying any more sensors until their accuracy is guaranteed, which I think might be never!

 

[JohnEGreen]

Over 90% concurrence within 20%/20mg in the non-hypo ranges above 80mg BG

The accuracy quoted even if you accept the figures is for blood sugars in the normal range above hypoglycemic levels and does not speak to what the accuracy is in the lower than normal range.

 

[TuviaDror]

Okay, let's just get one thing clear here then. The Israeli Ministry of Health is the regulatory body for the version of Freestyle Libre being used in Israel. Not the FDA.

The rest of this is somewhat speculative based on where you live and the timing of availability of devices in your region.

According to the import rules I can find, you will be using the Worldwide release device as AMAR requires one of :

• FDA 510(k) or premarket approval letter
• CE Marking certificate issued by a European Notified Body
• Certificate to Foreign Government (CFG) or Certificate of Free Sale (CFS)
• Proof of ISO 13485 certification

for approval of import of devices. There are additional checks for certain types of device but I don't think the Libre qualifies for those.

The timing of your device availability doesn't align with the 14-day US device approvals, so the most likely scenario is it uses the CE marketing certificate, which isn't approved for using to dose insulin. If you have issues with the device, you should be directing them at AMAR as your regulatory body about how Geffen Health is marketing the device, as the very clear marketing within Europe is that you don't use it to dose from.

The FDA doesn't (and can't) consider complaints about a product that is not licensed in its jurisdiction (which your version of the Freestyle Libre isn't). I'm also not sure whether it accepts complaints from non-US citizens (even though you can send a letter/on-line form).

Tim
Thx again for the time.

So, in a nutshell, lets all help all diabetics WW and share with the FDA as well, that the libre is a GREAT pattern displayed, but requires help in the actual accurate presentation of the BG value (before making decisions on whether to provide a bolus or not). Agreed?

Details below

Thx for drawing my attention. I checked the issue. I read the OFFICIAL libre pamphlet (which is distributed with the libre sensor, in Israel, in English). I quote from the indications for use: "The freestyle libre Glucose monitoring system is indicated .... It is designed to replace blood glucose testing in the self management of diabetes with the exceptions listed below,,,"). I attach the approval. So YES the system is indicated for giving boluses, in Israel

Per your question about the FDA. I actually had a nice telephone conversation with the FDA branch chief handling diabetes (and i told him i'm in Israel...). Hence i filed an official FDA complaint.. . Yes, they want to hear from anyone, about any case pertinent to a device marketed in the USA as well. I'm afraid also there is no reason to assume that the current libre version marketed in Israel is different to the 14 days version in the USA.......

Per your perception about the FDA,... So, since the FDA is interested in every piece of information pertinent to the case which can help them protect and provide better health to their citizens. Similarly to any governmental health agency, i hope and believe. Therefore, if a company is withholding/ignoring WW (world wide) experience per a certain device which eventually is being distributed in the USA, you bet they are interested!

If a company isn't responding decently to a user (as i shared with all a recent USA post with the USA version, where the Abbott rep ignored the patient's complaint about the accuracy....), you bet they are interested! I actually approached the local Israel distributor who was great and provided me with replacement sensors but that didn't solve the issue. I want to know when the sensor is credible on me and when not, is that an unfair request? The local distributor himself had no reservations about the way i used it!

Since what i'm reporting is accurate, and a non negligible number of people from the forum reported similar, that the libre is GREAT for showing patterns but requires additional assistance for the accurate BG value, why not share with the FDA? Aren't we all trying to help all diabetics WW, not just those who succeed with the Libre at first shot? What about those new users who have difficulty, shouldn't we help them to?

Why can't we all together approach Abbott and share with them that while their device is good, it is causing some issues to users who get false readings from it? By them (Abbott) integrating gauging (by a less frequent BG test) would be very helpful, even if this is just an optional feature? Why don't we as a community share this as i suggested with the FDA? Is anyone in then community opposed to protecting and helping the other world wide (WW) Libre users, new users for sure but others as well, who need this help? If someone is strongly opposed to performing finger BG testing, fine but then that's the patient's sole responsibility and risk. Is is fair to deprive so many others from the possibility to integrate into the libre the actual gauging from the finger BG tests?

So, in a nutshell, lets all help all diabetics WW and share with the FDA as well, that the libre is a GREAT pattern displayed, but requires help in the actual accurate presentation of the BG value (before making decisions on whether to provide a bolus or not). Agreed?

 

[tim2000s]

So, in a nutshell, lets all help all diabetics WW and share with the FDA as well, that the libre is a GREAT pattern displayed, but requires help in the actual accurate presentation of the BG value (before making decisions on whether to provide a bolus or not). Agreed?

I'm not sure I do agree.

There are a few reasons for this:

1. The number of people showing the extreme variation that you've described, while there, doesn't necessarily stack up to a statistically important minority (sorry). In terms of statistical response, Abbott have more than 800,000 users of Libre worldwide. To be significant, you need to have nearer 5% of users having issues and reporting them (and even then, that's probably on the low end). With any system that interacts physiologically with a human, one size doesn't fit all. Even with calibration. Just look at Dexcom. You don't see it as much in Europe, but there are lots of complaints about the lack of accuracy of Dexcom in the US forums. Yet the FDA has not demanded change there.
2. Abbott already know about the variation in readings from blood that people are seeing. Every time that you run those reports, the data is uploaded to Abbott who are able to track the accuracy of anyone using the Libre blood testing system with the Scanner.
3. The people who are benefiting the most from this technology are those who are least likely to fingerprick (high Hba1C, testing very infrequently). Requiring fingerpricking doesn't help them (and they're a surprisingly high proportion).
4. You are working with an older version of the technology (it's four years old!). Why do I think it's not the US version? Because it was available in Israel nearly a year ago, and the only 14 day sensors available then were the European versions. The FDA had not approved the US version at that time, and the older version has a worse MARD than the newer version.
5. Abbott have already recognised that this is not ideal and is causing issues for some, and have rolled out a new factory calibration model for Libre that is done on a per sensor basis. There are also rumours that they have updated the algorithm. This is supposed to be more accurate and being rolled out in Europe now. Complaints are fine, but when the tech has already moved on, they are also irrelevant.
6. To satisfy that there may be an issue with the Libre on some people, all that Abbott have to do is include a statement in the packaging that acknowledges that there is a small risk that Libre may not be accurate for some people, as is true with all ISF based systems. Then they are covered.

I get the impression that what you are trying to do is use the regulator in the US to force Abbott's hand in making the Libre a fingerstick calibrated device in line with every other CGM (aside from Dexcom's G6) out there. I really don't think this is going to happen, or what Abbott are interested in doing.

 

[TuviaDror]

I'm not sure I do agree.

There are a few reasons for this:

1. The number of people showing the extreme variation that you've described, while there, doesn't necessarily stack up to a statistically important minority (sorry). In terms of statistical response, Abbott have more than 800,000 users of Libre worldwide. To be significant, you need to have nearer 5% of users having issues and reporting them (and even then, that's probably on the low end). With any system that interacts physiologically with a human, one size doesn't fit all. Even with calibration. Just look at Dexcom. You don't see it as much in Europe, but there are lots of complaints about the lack of accuracy of Dexcom in the US forums. Yet the FDA has not demanded change there.
2. Abbott already know about the variation in readings from blood that people are seeing. Every time that you run those reports, the data is uploaded to Abbott who are able to track the accuracy of anyone using the Libre blood testing system with the Scanner.
3. The people who are benefiting the most from this technology are those who are least likely to fingerprick (high Hba1C, testing very infrequently). Requiring fingerpricking doesn't help them (and they're a surprisingly high proportion).
4. You are working with an older version of the technology (it's four years old!). Why do I think it's not the US version? Because it was available in Israel nearly a year ago, and the only 14 day sensors available then were the European versions. The FDA had not approved the US version at that time, and the older version has a worse MARD than the newer version.
5. Abbott have already recognised that this is not ideal and is causing issues for some, and have rolled out a new factory calibration model for Libre that is done on a per sensor basis. There are also rumours that they have updated the algorithm. This is supposed to be more accurate and being rolled out in Europe now. Complaints are fine, but when the tech has already moved on, they are also irrelevant.
6. To satisfy that there may be an issue with the Libre on some people, all that Abbott have to do is include a statement in the packaging that acknowledges that there is a small risk that Libre may not be accurate for some people, as is true with all ISF based systems. Then they are covered.

I get the impression that what you are trying to do is use the regulator in the US to force Abbott's hand in making the Libre a fingerstick calibrated device in line with every other CGM (aside from Dexcom's G6) out there. I really don't think this is going to happen, or what Abbott are interested in doing.

Tim again thx for your VERY informative response.
First off, indeed the Israeli version is similar to the European one (not the USA version). So hopefully the USA one is better. Nonetheless, in any event, there are similar complaints in the USA as i shared with you.

You noted a very valid point: It seems like the Libre is helping primarily the Diabetics who don't have the best glucose control (HbA1c > 7.0). I don't think Libre manufacturer would be very happy for you to share this perception publicly.... However, by gauging/calibrating here and there with a finger BG test (even if only optional, not mandatory) this would be a tremendous help for us (diabetics with better control) and also help many other well controlled diabetics WW, to be even better controlled.

Isn't this what we all want? Is this such an awkward request? To make some adaptations? I mean this would increase allot the relevant exposure of the Libre

 

[tim2000s]

Isn't this what we all want? Is this such an awkward request? To make some adaptations? I mean this would increase allot the relevant exposure of the Libre

Fundamentally, no it's not what "we all want".

One of the biggest selling points is that it doesn't require fingerprick calibration. Dexcom's G6 has gone like a rocket, because it doesn't require fingerprick calibration. Most people don't want the option of fingerprick calibration. And to add it in after the event requires a completely new round of regulatory sign offs and approvals.

If you want to use a calibrated version of Libre, you could always try the Open Source programs like Glimp and xDrip, which allows you to scan (and calibrate) the Libre.

 

[jason720]

So this is the first ever comment on one of these posts I’ve made: I’m having a nightmare with the Freestyle Libre: Iv been using it for a while now, o knew there were fluctuations in the result you get and what your blood reader says, but his last few weeks has been a nightmare!
I first had a sensor that literally, once applied, even though I followed the instructions to the letter and my skin was dry etc etc.:: the sensor fell off within about 5 minutes. The stick on the sensor was unacceptable. So they sent me another sensor. I had another that was on the go while I awaited it. The one I had in last week was working fine, then it was a hot day and it kept saying I was LO. I didn’t think anything of it, wondered if I’d accidentally overdosed or was it the heat speeding my insulin up, well hours later I realised the machine was now only telling me I’m lO. So I pulled that sensor out and tried the new one Abbott sent me. So I put it in this morning, I left it for like 3 hours to ‘acclimatise’ or calibrate to me: well I know it fluctuates within the first day, but the readings I’m getting are dangerously different: Among the most different was the most recent, measuring 8.5mmol. I felt this sounded great, but I keep double checking on my blood glucose machine: in reality my sugars are 16.8!!! It’s been like this most of the evening with a disparity of between 5-8 mmol! I can’t see how this is a way I can continue using it if it’s constantly wrong: Iv got it prescribed to, instead of test strips but I think I’m going to have to revert because it’s so inaccurate! Any suggestions??! I was quite happy with it, but having 3 dodgy sensors with 3 different problems: and Iv had 2 previous problem sensors and Iv only had about 10 total!’

 

[Deleted Account]

@jason720 you asked for suggestions.
This is what I find works
- put the new sensor on at least 24 hours before activating it. It can take as long as 48 hours for our bodies to get used to having an alien o ject inserted.
- do not rely on any Libre readings less than 4 or greater than 9. Always tests these with finger prick.
- do a finger prick test at least once a day.
- check any night time lows as they may be caused by compression: lying on the sensor
- always use the alcohol wipes to thoroughly clean the area before applying the sensor
- if the sensor is likely to get wet for a prolonged period (e.g. When swimming or if sweating alot) use additional method to hold the sensor in place such as SkinTac, tegaderm or vetrap
- if you find sensors are frequently out, try alternative reader apps which allow calibration against finger prick such as Glimp.
- do not give up on finger prick. Always check any unexpected reading from Libre.