Freestyle Libre Sensor: highly inaccurate??

[Fleegle]

If the sensor insertion can consume glucose, does that mean the level measured would be lower then?

All my readings are significantly higher than what they should be.

Also, if the manufacturer knows that the first day or so could lead to random results, why don't they just discard the first day of results? There does not seem to be an option to discard readings/periods

I think because it could does not mean it always does react. It is just advice.
Sometimes for whatever reason I didnt have time to do the 24hour warm up and I found more often than not the sensor was accurate as soon as the activation time ended. I suspect it is complex and largely individual.
I am pretty sure in my calls abbott did not advocate the 24 hour thing. Though they did advocate applying skin-tac-wipe to stop them dropping off which was my biggest problem.

Putting aside the inaccuracy you have - aren't they fantastic for measuring the rise and fall - PMP at 2 hour - impact of exercise at the time and after. Showers, DP the list goes on - really informative.

 

[TedTomato]

Putting aside the inaccuracy you have - aren't they fantastic for measuring the rise and fall - PMP at 2 hour - impact of exercise at the time and after. Showers, DP the list goes on - really informative.

Unfortunately, I can't put aside the inaccuracy I am experimenting, as the readings go up and down randomly (it's not just higher numbers than expected), without connection to what I am eating or exercise etc.

I had a reading of 9.1 at 12:00 today, so 4.5 hours after breakfast, and before lunch...

I had a reading of 5.9 at 20:00 today, 6 hours after having lunch.

My usual blood tests are at about 5.5 before a meal, and 6.5 after a low carb meal, 1.5 hours later.

I am not sure what numbers the sensor is providing, but I don't think those are my glucose levels!! Maybe lottery numbers?

Hopefully, the readings will stabilize after a couple of days, once my body has accepted that foreign object (if that's the issue), or the manufacturer will get to replace the sensor if faulty. Otherwise, this is completely pointless to me.

 

[Fleegle]

Unfortunately, I can't put aside the inaccuracy I am experimenting, as the readings go up and down randomly (it's not just higher numbers than expected), without connection to what I am eating or exercise etc.

I had a reading of 9.1 at 12:00 today, so 4.5 hours after breakfast, and before lunch...

I had a reading of 5.9 at 20:00 today, 6 hours after having lunch.

My usual blood tests are at about 5.5 before a meal, and 6.5 after a low carb meal, 1.5 hours later.

I am not sure what numbers the sensor is providing, but I don't think those are my glucose levels!! Maybe lottery numbers?

Hopefully, the readings will stabilize after a couple of days, once my body has accepted that foreign object (if that's the issue), or the manufacturer will get to replace the sensor if faulty. Otherwise, this is completely pointless to me.

Sounds like a bigger problem than the first 24hour thing - but fingers crossed.

 

[Scott-C]

If the sensor insertion can consume glucose, does that mean the level measured would be lower then?

All my readings are significantly higher than what they should be.

Also, if the manufacturer knows that the first day or so could lead to random results, why don't they just discard the first day of results? There does not seem to be an option to discard readings/periods

My post was intended to answer your question about why people generally attach a while before activating, not why your particular sensor is dubious.

The American version, which was only approved by their FDA a few months ago, has a 12 hour warm up time for that reason: letting it settle down to accomodate natural repair mechanisms and foreign body responses - looks like they've learned from their European forebears who've been using it for a couple of years now.

I've been using libre for about 18 months now. For the last 5, along with a small bluetooth transmitter which turns it into cgm and lets me iron out the sort of inaccuracies you're seeing.

I've used getting on for 40 sensors. I'm not going to pretend they are perfect. Most are ok, but some are plain sketchy. Abbott have normally replaced those. Maybe you've had bad luck and got a sketchy one.

It takes a few months of using them before you can make a judgment call on whether it's a "good" or "bad" sensor.

The science going on here is hugely complex: an enzyme on the sensor filament, glucose oxidase, is breaking down passing glucose into other chemicals, giving off electrons, which are measured as an electric current, represented as a number, which is then turned into a guess about what that number means as a bg level.

What could possibly go wrong in that situation? Obviously, lots of things.

I'm T1. I need to inject insulin several times a day to not die, and some of the judgment calls I make about those shots, if I get them wrong, might put me in a situation which would make a bad lsd trip seem pleasant.

Despite libre's imperfections, I've figured out ways to use it in ways which make my life significantly easier. Heads up on hypos, for a start, and quite a few other things.

I've got very little patience for non-T1s who use it for a few days and complain about it, "not being the same as my meter". They have failed to understand the complexities of it.

You might just have a rogue sensor, in which case Abbott will likely replace it. If you don't trust it, stop using it.

 

[hankjam]

My post was intended to answer your question about why people generally attach a while before activating, not why your particular sensor is dubious.

The American version, which was only approved by their FDA a few months ago, has a 12 hour warm up time for that reason: letting it settle down to accomodate natural repair mechanisms and foreign body responses - looks like they've learned from their European forebears who've been using it for a couple of years now.

I've been using libre for about 18 months now. For the last 5, along with a small bluetooth transmitter which turns it into cgm and lets me iron out the sort of inaccuracies you're seeing.

I've used getting on for 40 sensors. I'm not going to pretend they are perfect. Most are ok, but some are plain sketchy. Abbott have normally replaced those. Maybe you've had bad luck and got a sketchy one.

It takes a few months of using them before you can make a judgment call on whether it's a "good" or "bad" sensor.

The science going on here is hugely complex: an enzyme on the sensor filament, glucose oxidase, is breaking down passing glucose into other chemicals, giving off electrons, which are measured as an electric current, represented as a number, which is then turned into a guess about what that number means as a bg level.

What could possibly go wrong in that situation? Obviously, lots of things.

I'm T1. I need to inject insulin several times a day to not die, and some of the judgment calls I make about those shots, if I get them wrong, might put me in a situation which would make a bad lsd trip seem pleasant.

Despite libre's imperfections, I've figured out ways to use it in ways which make my life significantly easier. Heads up on hypos, for a start, and quite a few other things.

I've got very little patience for non-T1s who use it for a few days and complain about it, "not being the same as my meter". They have failed to understand the complexities of it.

You might just have a rogue sensor, in which case Abbott will likely replace it. If you don't trust it, stop using it.

That's a useful post @Scott-C, the science bit is good and I find your experiences informative. Many thanks.
Wish you good.

 

[TedTomato]

I've got very little patience for non-T1s who use it for a few days and complain about it, "not being the same as my meter". They have failed to understand the complexities of it.

You might just have a rogue sensor, in which case Abbott will likely replace it. If you don't trust it, stop using it.

To be honest, I don't really need to understand the complexities of it, or even how it works: as an end-user, it's either providing reliable (or at least consistent) readings/data or not (in which case, it's pretty pointless).

Surely, most people using such device for the first time will compare with their traditional blood monitoring device, and will be build confidence (or not) in the Libre sensor.

I am assuming many people may get reliable readings, very close to their blood monitors (otherwise the device wouldn't be sold anymore), and some people like me will get some random results, either because their body is rejecting the device (for a while), or they just got a "rogue" sensor... We shall see over the next 2 weeks.

 

[DCUKMod]

To be honest, I don't really need to understand the complexities of it, or even how it works: as an end-user, it's either providing reliable (or at least consistent) readings/data or not (in which case, it's pretty pointless).

Surely, most people using such device for the first time will compare with their traditional blood monitoring device, and will be build confidence (or not) in the Libre sensor.

I am assuming many people may get reliable readings, very close to their blood monitors (otherwise the device wouldn't be sold anymore), and some people like me will get some random results, either because their body is rejecting the device (for a while), or they just got a "rogue" sensor... We shall see over the next 2 weeks.

TedTomato (great name!) - I am someone for whom an extended bedding in period, before activation seems to improve sensor accuracy.

I am what my doc smilingly reefers to as a super healer, meaning I tend, to heal very quickly, but should I experience a splinter or an accidental foreign body getting stuck under my skin, I have a rapid reaction from my body to try to reject the splinter/foreign body. Part of that process is swelling and redness, which changes the tissue immediately surrounding the splinter/foreign body, but resulting in wounds healing quicly.

With the Libre, I tend to view this early period of inconsistency down to that, so I apply a sensor and leave it full 48 hours, prior to activating it. Yes, it's frustrating to have to do that, but it's better (for me anyway) that than rogue performance too often.

Abbott are not generally too excited by variances in the first 24 hours, but have replaced a number of my sensors where the issue continued for me.

If you do find yourself having to contact them, then I always prepare by collating some comparative measures - blood test, then 15/20 minute delay, then the scan reading to demonstrate the issue. That makes it easier for them to grasp the issue and takes the pressure off trying to find relevant tests scans whilst on the phone. Using this approach I have never had to request a replacement, the Helpdesk Analyst has always reached their own conclusion.

Fingers crossed for you.

 

[Scott-C]

To be honest, I don't really need to understand the complexities of it, or even how it works: as an end-user, it's either providing reliable (or at least consistent) readings/data or not (in which case, it's pretty pointless).

Surely, most people using such device for the first time will compare with their traditional blood monitoring device, and will be build confidence (or not) in the Libre sensor.

I am assuming many people may get reliable readings, very close to their blood monitors (otherwise the device wouldn't be sold anymore), and some people like me will get some random results, either because their body is rejecting the device (for a while), or they just got a "rogue" sensor... We shall see over the next 2 weeks.

There's a few companies working on these things, Abbott, Dexcom, Medtronic, and some others who are not yet on market.

They are all trying to make their products better but are working against huge biological problems which are inherent in all attempts to do what seems like a simple thing, measuring in vivo glucose, but is actually really, really difficult.

You're tag says you're pre-diabetic. If you're not even diabetic at the moment, testing with cgm is overkill. Wait until you're injecting insulin and then worry about it.

Blood sugar levels are a constantly moving target. Yours are still being managed by your pancreas. Mines aren't, so I have to do it myself. Cgm, despite what you perceive as inaccuracies, helps me a lot with that in ways you won't understand until you're in that position.

These things are cutting edge, they still make mistakes, but they're the best we've got for the time being, so we make the most of them.

 

[TedTomato]

You're tag says you're pre-diabetic. If you're not even diabetic at the moment, testing with cgm is overkill. Wait until you're injecting insulin and then worry about it.

Blood sugar levels are a constantly moving target. Yours are still being managed by your pancreas. Mines aren't, so I have to do it myself. Cgm, despite what you perceive as inaccuracies, helps me a lot with that in ways you won't understand until you're in that position.

I have some health issues, and my doctor is not sure what the exact cause is. We noticed a high average blood sugar, which would put me in the diabetic range, so measuring with CGM is very much relevant to further diagnostic, and I am on a low carb diet now.

I am sorry, but I don't understand what you say about inaccuracies. In your case, you must be getting accurate results, and the device is working for you.

I am getting random readings (yesterday, from 5.2 to 9.4 in one day, while only have low carbs meals, with spikes not even after meals) and also compared to my blood glucose meter tests, so it's not a "perception" issue. You wouldn't be able to manage your T1 diabetes with this level of randomness and wouldn't trust any of the readings anyway.

The fact that it's cutting edge technology etc. does not really matter, it's not like I am on a medical trial. It's now on the market, I bought the device at full price, and therefore expectations is that it should be working reasonably well (within an error margin).

Hopefully, it will stabilize over the next day or so, and start being useful, or I will get a replacement, so that I can test the device properly.

 

[etmsreec]

I've got an open case with Abbott in the UK at the moment, as this sensor and my previous one have had quite a difference compared to my Accuchek Aviva Expert meter. I've not had any response from Abbott since I gave them the data that they initially requested. Hopefully I'll hear something from them soon. My current sensor is reading about 2 or 3 units above what the Aviva is, which I find frustrating. I'll admit that the graphing is helpful, in between finger prick testing.
Maybe there's a case for almost discarding the units, per se, and just looking at the profile of glucose level? A bit like a musician following sheet music - they don't read whether it's a G or an A, they only look at the change from one note to the next.

 

[TedTomato]

I have called them today, and after a very long wait, managed to talk to someone.

He asked for three readings taken from a blood tester and the libre sensor, and calculated that the difference was more than 30% each time, so the sensor must just be faulty and will be replaced.

He said maybe the sensor had moved/lifted slightly after its installation, and this could affect readings too.

Apparently, dehydration and vitamin C (ascorbic acid) can also affect the readings. I did take some Vitamin C supplement yesterday, but I don't think I am dehydrated.

I have done various comparisons today between the Libre sensor and one of my blood sugar testers, and the difference went up to 2.5 mmol/L (but only about 0.2 mmol/L between my two blood testers, a Tee2 device, and a SD Codefree one), so something is seriously wrong.

Hopefully, the replacement will work better, and I will wait 24 hours before activating it after installation, as per the advice given on this forum.

I agree that a trend would be useful, without considering units, but this does not seem to be reliable either.

 

[HJROD]

My libre readings are actually lower than the glucose meter by 2mmole/L. 10 days of use. Wonder which one is right?

 

[TheBigNewt]

Try a Dexcom from what I've heard around here they are better than Libre overall. Plus they'll alert your phone if you go low.

 

[porl69]

Being able to see trends in your bloods and the effects that certain foods have on your blood sugars is a very valuable tool to T1Ds. I have been using the Libre for 8 weeks now and find it very accurate (always within 0.5 to 1 mmol. As long as they are within 15% then I am a happy bunny. Plus I am only finger pricking 3 times a day now (to bolus) and not 8 - 10 times. My fingers are having a well deserved holiday

 

[TedTomato]

Right.

So I have received a replacement sensor (was posted very quickly), and after wearing for a few days (and only activating after 24 hours), seems to now underestimate the real blood sugar levels by 1/1.5 mmol/L, which is quite a lot.

The first sensor overestimated most of the readings, and the new sensor seems to underestimate consistently.

I guess people are going to say that the trends matter, but those seems to be a little strange too: sometimes massive spikes for no apparent reasons, and sometimes a flat line even after a meal.

 

[Brunneria]

All my libre sensors seem to be around 1-1.5mmol/l lower than the prick tests. I just recognise that it happens, and watch the trends.

In your situation I would make a note of exactly what you ate when you get those large rises. For me, the Libre was fundamental in helping me discover that my problem is less about carbs/sugar and more about gluten. Turns out that by strictly avoiding gluten, I can tolerate a lot more carbs, since it is the gluten+carbs that trigger high blood glucose. Without the gluten I can eat the same amount of carbs and my blood glucose stays within range.

So it isn't outside the bounds of possibility that you may find specific foods cause much larger rises in glucose than their carb content would lead you to expect.

I would encourage you to use your first few sensors as learning tools - for how to make the Libre data work for you, and how to use that data to manage your blood glucose.

 

[novorapidboi26]

Some people are going to have high expectations I suppose but just think about what the Libre actually excels in.....

its patterns in glucose levels, not how accurate it is compared to an already inaccurate blood test.....

its the patterns you see that then go on to influence other day to day choices about the food you eat and the medication you take...

obviously some people need a good accuracy to help with hypo awareness, but the Libre isn't the right device if that's a real issue for people...(working alongside additional devices excluded)

@TedTomato, I think you need to establish what information you need from the Libre in order to solve the issue you are having to warrant use of the device in the first place....

Are you looking to identify foods you cant eat?

Are you looking to identify times of the day your insulin isn't doing the job of expected of it..?

 

[urbanracer]

Right.

So I have received a replacement sensor (was posted very quickly), and after wearing for a few days (and only activating after 24 hours), seems to now underestimate the real blood sugar levels by 1/1.5 mmol/L, which is quite a lot.

The first sensor overestimated most of the readings, and the new sensor seems to underestimate consistently.

I guess people are going to say that the trends matter, but those seems to be a little strange too: sometimes massive spikes for no apparent reasons, and sometimes a flat line even after a meal.

Reading this made me think - and something I intend to try myself if Abbott ever deliver sensors to me (I've been finger pricking for over a week now ). As the Libre is supposed to be 10+ minutes behind blood glucose, has anyone ever tried to compensate for this by checking the Libre at least 10mins (or more) before doing a finger prick?

 

[TedTomato]

@TedTomato, I think you need to establish what information you need from the Libre in order to solve the issue you are having to warrant use of the device in the first place....

Are you looking to identify foods you cant eat?

Are you looking to identify times of the day your insulin isn't doing the job of expected of it..?

I have borderline type 2, so I am indeed trying to identify what sort of meal/food is triggering a spike in blood sugar, what food to avoid etc.

I think the Libre would be useful to show a trend/graph (even if we put aside the actual units/measures), but I am not sure my current sensor can be trusted on even showing a trend. For instance, coffee seems to trigger a big spike on the Libre, but not on a blood tester.

 

[TedTomato]

Reading this made me think - and something I intend to try myself if Abbott ever deliver sensors to me (I've been finger pricking for over a week now ). As the Libre is supposed to be 10+ minutes behind blood glucose, has anyone ever tried to compensate for this by checking the Libre at least 10mins (or more) before doing a finger prick?

Yes, when comparing the Libre readings to a blood monitor, I always consider the delay with the Libre. It should take 10/15 min longer on the Libre to reach the same blood sugar reading.