CGM inaccurate- beware Abbott Freestyle

[SEMKP]

Hi here’s a funny thing. I blagged a couple of Abbott Freestyle CGMs. They predicted an HbA1c of 42. That coincided with my T2 review at the GP- my HbA1c is 47. So I can’t trust that CGM’s prediction.

 

[Rokaab]

Did you ever try doing some blood tests to see how far out it was.
When was still using the libre1 (I use a different one that talks directly to my pump now), it normally read about 1mmol/L less than what it actually was - which makes about 8/9 (I think) difference in the HbA1C number so I always expected its estimate to be lower than what I actually was.

The libre without calibration is an awesome tool for seeing trends and getting a rough idea of where you are, but for most will not be pinpoint accurate

 

[In Response]

@SEMKP please beware of the impact of your comments.
You had 2 sensors which covers 1 month and are comparing something which uses a different method to estimate THREE months measured by "real" HBA1c.

There was a thread some time ago about Libre (and all CGMs) using a different method to calculate HbA1C because it does not measure blood but interstitial fluid).

Headlines like the one used for this thread can be potentially dangerous for someone with Type 1 diabetes who have been advised they can use Libre for calculating insulin doses.
There are tools to enable calibration of Libre .. there is also the good ole finger prick and compare method. Did you use any of these?

 

[Bubbleblower]

Hi here’s a funny thing. I blagged a couple of Abbott Freestyle CGMs. They predicted an HbA1c of 42. That coincided with my T2 review at the GP- my HbA1c is 47. So I can’t trust that CGM’s prediction.

Actually that is not even so bad, but you are right the FSL is completely unreliable, especially for type 1’s. In December my HbA1c was 36% higher than predicted. So then I compared it to fingerpricks and it turned out 89% of the measurements were more than 15% off.

Recently I did another test with 2 sensors at the same time and there were huge differences between the two. As I speak I am doing yet another test for the next 3 days with a sensor on each arm, which I hope will help to convince my DN I really need a real- and reliable CGM. I just took my first measurement comparing the 2 and there was a ~7 point difference:

According to fingerpricks my BG is still 2 points lower than the 8.7 and both sensors are lotnr. 6491448.

 

[In Response]

As I speak I am doing yet another test for the next 3 days with a sensor on each arm, which I hope will help to convince my DN I really need a real- and reliable CGM.

What CGM do you have in mind?
I trialed Dexcom and found Libre with xDrip calibrations far more accurate.

I believe it is about understanding the technology and its limitations then coming up with strategies to overcome them. It is very unfortunately that seems to mean we end up learning more about these "workarounds" than the HCPs we work with but as I have the capability to do so, I am happy to save our cash-strapped NHS money.

 

[sleepster]

but you are right the FSL is completely unreliable, especially for type 1’s.

I think you mean it's completely unreliable for you.

 

[becca59]

@Bubbleblower I’m a Type 1 and for me the FSL is bang on correct. I would now not be without it now. And the Libre 2 alarms during the night have been without doubt a godsend for correcting highs and lows which would otherwise have gone unnoticed.

@SEMKP I am sorry that you have that attitude to something that for many Type 1s has revolutionised their lives. Which is relentless day after day testing, testing, testing and more testing. Then injection after injection. Even during the night. No turning over and ignoring those high/low numbers for us. It may be inaccurate for you. Be thankful that you don’t actually need it.

 

[ROE100]

Agree Libre is a great tool but for me and after a long conversation with Abbott was told to stop using as it didn't work correctly for me so I was scanning but still doing the same number of fingers pricks as there were so different and couldn't trust the Libre as the Libre reported lower so constantly showing me on the low side when not I was actually in a great range. This also casued issued when telephone call consultantion came as i wanted to speak about specfic issues but they kept saying you are low too many times. Did consider the Xdrip solution but sometimes technology scares me so just stay away. But as I seem to say on most of my post we are all different!

 

[Seacrow]

Heck, if you go into the theory the HbA1c will measure something slightly different in everybody. The value found depends on how long your blood holds onto glucose, so anybody with sickle cell disease or with really high ferritin will get results considerably different than 'normal' blood.

Yet again, a test where the actual value is indicative not definitive (in spite of what the docs say). The change in an individual's values though should be accurate (as the lab can get).

 

[Deleted member 543160]

@SEMKP
I choose to rely on my actual finger prick tests as HBA1cs are calculated from 3 monthly averages, and it's quite possible for both stable glucose and erratic high and low glucose results to produce the same figures.

I used the Libre 1 for a while and found its HbA1c predictions were consistently 5 points lower than my actual HbA1c. My predicted HbA1cs from 90 days worth of my glucose tests are always about 2 points lower. But all their actual trends/patterns have consistently matched those of my HbA1c, and since all three measure slightly differing thing I've found these predictions accurate enough for my T2 purposes.

However the patterns my Libre sensors showed gave me extra insights of how my glucose levels were affected over long periods as opposed to my finger prick spot checks.

 

[Bubbleblower]

I think you mean it's completely unreliable for you.

Not just for me:

Assessment of factors that determine the mean absolute relative difference in flash glucose monitoring with reference to plasma glucose levels in Japanese subjects without diabetes

And most of us experience the difference between new and old sensors.
I suppose type 1's react more aggressively against foreign bodies, which makes that difference bigger.

coming up with strategies to overcome them..

You know my strategy, imagine how I would feel if I only had the new sensor.

 

[Peter03]

I'm still using the enlite sensers, lots of people do not like them but I find with 2 or 3 calibrations a day keeps it quite accurate they like all cgms and integrated pumps take a while to get used to, I found the libre inacurute at times so you need to look out for variations that you do not consider normal and do a finger prick test

 

[EllieM]

I used the libre 1 very happily until I started to develop and allergy to the sensor. Unfortunately that also meant that the results became wildly inconsistent. Abbott to be fair sent me numerous replacement sensors and even a new reader. Eventually I had to conclude that my body didn't agree with that piece of technology and went back to testing (8 times a day, because my hypo awareness was impaired).

For the last 2 years I have been using dexcom. The alarms make all the difference to my hypo awareness (I get warned before hypo levels so I get less hypos and hence my hypo awareness has improved). It's not perfect and occasionally I get really weird results, sometimes it reads low for no reason and I do need to check it against a glucometer occasionally, particularly of hypo. And I do occasionally get a bad sensor. But for me it's as life changing as my first glucometer back in the 80s.

I think you have to recognise that the sensor technology is sensitive to any reaction your body has to the filament and may not suit some people. I definitely need a glucometer as a back up, because some sensors are inaccurate, and if I'm having a serious hypo, the sensor time delay means I need the glucometer to check when I start to come back up.

As for hba1c predictions, @Seacrow makes a very good point about the lifetime of red blood cells meaning that people with identical bg results can have different hba1cs, and I personally think TIR (time in range) is more important, if you're using the technology that calculates that. I notice that my dexcom was under-predicting previous hba1cs but they have recently changed their algorithm, It will be interesting to see how my next hba1c compares.

 

[Jaylee]

Hi,

Understanding these things work by reading interstitial fluid in a layer under the skin & not blood.

Staying hydrated seems to help with consistency of Libre for me. Response or lag in my experience from low BGs coming back up into normal range for me can be delayed a good 20/30 minutes after the meter say yer OK. (Or even when I feel back in the zone.) but it's certainly on the ball giving me a heads up on a drop.

For me, predicted HbA1cs have been in the ballpark. & I've been pleasantly surprised when the lab results come in a little lower..

On the whole, this is the next step forward in tech after the memories of the early days peeing in a testube?
.. & I thought my first meter was "civilised."

 

[johnme]

Equally hba1c may not be accurate because the speed at which red blood cells die varies from person to person. We all know it’s possible for a stick and prick monitor to measure the last thing you touched. My cgm me gives results within 0.5% of my hba1c which seems reasonable. It shadows my intake of food and level of exercise within those limitations. It’s very useful.

 

[jonathan183]

I found the 5 Libre 1 sensors I used far too inaccurate to dose insulin on. Useful for trends except for one failed sensor but no saving on finger stick testing for me. Libre 1 also does not have alarms so did not provide any warnings of out of range readings either.
Glad CGM works for some ... but certainly does not work for all
I'll wait to see what gets offered with the due updates to guidance on CGM options at my next review.

 

[EllieM]

I'll wait to see what gets offered with the due updates to guidance on CGM options at my next review.

My understanding is that the libre 2 is better, If you are in the UK I would expect that they would be moving you to that version very soon. (Happy to be corrected by more knowledgable UK posters).

 

[becca59]

My understanding is that the libre 2 is better, If you are in the UK I would expect that they would be moving you to that version very soon. (Happy to be corrected by more knowledgable UK posters).

In England (do not know about the devolved nations) if you are already using 1 you can just request to be moved over to 2. The cost is identical to the NHS. From April every single Type 1 who wishes to can access this kit. No ifs buts maybes or hoops to jump through.

 

[jonathan183]

In England (do not know about the devolved nations) if you are already using 1 you can just request to be moved over to 2. The cost is identical to the NHS. From April every single Type 1 who wishes to can access this kit. No ifs buts maybes or hoops to jump through.

Self funding at the moment (don't meet the current criteria for NHS funding) ... updated guidelines should change that.

 

[EllieM]

Self funding at the moment (don't meet the current criteria for NHS funding) ... updated guidelines should change that.

I hope so.
However if you're self funding you might as well use the more modern product....