• Guest - w'd love to know what you think about the forum! Take the 2025 Survey »

Improving Libre Accuracy?

Shannon27

Well-Known Member
Messages
292
Type of diabetes
Type 1
Treatment type
Insulin
Hi Everyone!

I tried using the Libre in July, when I received two free sensors and the monitor free from my local hospital, as a "taster" from Abbott. I loved it for all of 36 hours, after which it started giving me hypo readings of 2.9 when I had no hypo symptoms (for me these normally start at around 4.5). I used the BG on the machine and it read that I was 8.2! I then tested with my spare machine (Contour Next USB) which also gave me a reading of 8.0. If I had corrected believing I was 2.9 I could have been quite ill, which is why I lost a lot of faith in it and still haven't used the second sensor.

Thing is, after reading so many great reviews on here, I'm willing to give it another shot (especially as it may be free on the NHS in November!). I have read somewhere that there are ways to speed up the calibration process, so it gives you more accurate readings within the first 48 hours; apparently there are also ways to reduce the 5-minute lag time between the check and your actual blood sugar readings?

Just wondering if anyone can give me some advice or tips on this? It would be very helpful

Thanks!

Shannon x
 
My health insurance just recently started covering the Libre and I'm using it for about a week now.

What I learned is that, in my experience, the time-lag seems to be larger than five minutes: more like 15. This makes it very difficult in times of large fluctuations in your blood sugar.
I do not want to go so far as to say the Libre becomes useless in times of fluctuations because it is still better to see a curve and a tendency instead of only a single point in time. But I dare not depend on the Libre alone. It shows normal values when I'm low, it shows tendencies downwards when I'm already recuperating from a hypo. In these cases, it is better to test manually.
Other than that, the Libre seems to be pretty accurate for me (+/- 0.5mmol/l), taking the time lag into account, of course.

Where did you read about ways to improve the accuracy? This seems to be not possible to me. If possible, Abbott would boost the accuracy of the device at all times, wouldn't they?
 

I heard that one way to improve the accuracy more quickly was to do a fingerprick alongside using the meter to quickly calibrate it to your body, which makes a weird kind of sense. Another thing I heard (which seems less viable) was to use a combo of commands on the meter to essentially increase the functionality of it. I believe it showed predictions on the screen or something similar. This seems unlikely and also a lot like hacking, which I'd rather not do on something so closely tied to my health!
 
I don't think the Libre calibrates itself using a scan and finger prick result at all. As regards the first 24-48 hrs accuracy some people apply a sensor but don't activate to for 24-48 hrs, Abbot don't recommend this though. It sort of makes sense in that it gives time for your body to repair the injury around the filament before you rely on the sensor.

Then again Abbot clearly state the sensor does not replace blood tests and any unusual readings should be backed up with a blood test and not to bolus off a scan result.
 
There are CGM which need to be calibrated by pricking your finger. The Libre Link, however, is not one of them. The Libre Link is calibrated industrially, which is also the source of its inaccuracy (which was not so bad for me, so far). So measuring manually gives you more accuracy only in the sense that you now have the actual blood sugar. It does not increase the accuracy of the Libre Link, however.

Combo of commands sounds magic and is certainly wrong.
 
With respect the much quoted 'time lag' is a fallacy. I have proven this by doing finger prick tests and then doing a swipe my Libre 5 mins, 10 mins, 15 mins and 20 mins later. I have also cross referred and there is absolutely no correlation in either direction. There also is no constant degree of error between the Libre and my Accu Chek Mobile device either. I have used a Libre constantly now for 2 years and am more than familiar with all of its little quirks and anomalies. I am also very familiar with a lot of the no doubt well meaning but quite incorrect theories that people put out about it. If you want an absolute cast iron case for just how inaccurate they are then take a look on the first page of the report generation that the Libre Software has available with it. There is a predicted HbA1c result on there. Please compare it with your actual previous laboratory HbA1c and there is usually about 10 points difference.
 
Try again and you may find that your prior poor experiences were isolated.

There is no such calibration process with the Libre, so no way to 'speed up' anything. What @slip has mentioned is a good technique to allow your readings to be better just after activating the sensor.

There is also no way to reduce latency between your actual measured BG and the BG value calculated from your interstitial fluid. The lag is innate and cannot be changed by any external act by you. Remaining hydrated however will likely help keep the Libre's as accurate as possible, always worth bearing in mind.

I wouldn't be getting too excited about the Libre becoming available on the NHS in November. It will likely be extremely difficult to qualify for permanent funding and is unlikely to be readily available for the majority.
 
@Diabeticliberty Forgive me if I'm wrong but 10 points seems quite a hefty difference for a supposedly accurate monitor?
I'll look at this when I get my sensor set up.
Can I just say that the lag time is not a fallacy at all, as there have been medical trials to measure readings from blood glucose and interstitial fluid (which the Libre monitors). There is an approximate 5-10 minute lag, which can increase in periods when readings are changing rapidly. If you want more information see:
https://freestylediabetes.co.uk/freestyle-thinking/post/FreeStyle-Libre-system-measure
 
Last edited:

Thanks @GrantGam that's what I was hoping. And yes with the NHS I wouldn't be surprised if the Libre sensors did become free but not commonly distributed.
 
Let me put it another way. In trials that I have conducted myself where a finger prick test offers a result and Libre swipe offers a different result there is no constant time delay at which point the results are the same or similar by the same margin. This in essence suggests to me that the time lag is indeed a fallacy. I have as a matter of interest taken up this matter with representatives from Abbott Customer Services when I have been reporting one of my many too many to list faulty sensors. They become a little bit woolly when interrogated. Gently interrogated of course
 
"Your sensor glucose (SG) readings are taken from your interstitial fluid, and not from your blood. Interstitial fluid is the fluid that surrounds the cells of your tissue below your skin, and usually glucose moves from your blood vessels and capillaries first and then into your interstitial fluid." Hence the time lag.
I also said APPROXIMATE 5-10 minute, which would account for the differences in readings at certain times after original monitoring.
Personally I think there are too many posts and pages explaining this for it to be a fallacy.
https://www.medtronicdiabetes.com/c...y-sensor-glucose-does-not-equal-blood-glucose
 

The Customer Services guys are just phone monkeys, yes they'll have had training on the device and it's use, they won't be able to discuss any trials results etc with any authority.

I'm not surprised you found the interstitial delay hard to pinpoint as it's a moving target so to speak, biologically there will be a delay, how much depends on so many factors - a finger in the air moment, to say it's a fallacy is perhaps a bit OTT.
 
Regarding re-calibration, the Glimp app used on the phones/tablets can "recalibrate" the readings/graph. When you enter the results of the prick tests, Glimp will "recalibrate" the graph according to the prick test results. Glimp will not make changes to the sensor. I don't think the Abbott's app does recalibration - I cannot get it to work on my phone.

The sensor does take readings every minute, and then "averages" them every 15 minutes. If you took readings every 14-15 minutes, you would get the readings for every minute in the last 14-15 minutes and then every 15 minute averages for the past 8 hours. This is my observations when using the Glimp app.
 

I've never heard of this app @zbluebirdz but now I am curious. I'll look it up
 
Not used glimp @zbluebirdz but unlike Shannon I have heard of it. You say it recalibrates the graph - so re-jigs it to accommodate the BG result? and does NOT alter it's algorithm to make future results more accurate? (Not that I'm going to change phones to Android! )
 
slip said:
some people apply a sensor but don't activate to for 24-48 hrs, Abbot don't recommend this though. It sort of makes sense in that it gives time for your body to repair the injury around the filament before you rely on the sensor.
Click to expand...

I had one that read low. Mostly a red line along the bottom of the screen. I phoned Abbott who didn't want to know unless I had been wearing it for more than 24 hours. After that I learned to apply it, wait 24 hours and then activate it. Since then it has proven to be reliable and accurate when compared to my Freestyle Lite. I am not sure how I feel about comparing a machine with itself so I use a separate one.

When I phoned Abbott back they replaced it.
 


You have made a statement which you obviously believe in. I fully respect that. I did however earlier suggest that I have gone through a lengthy process to substantiate or otherwise the 5, 10, 15,20 minute, 1 day., 6 weeks theory in order to better understand how my Libre works for me personally. I once again state that I am unable to make any correlation in terms of time factor where a finger prick blood test and a Libre swipe synchronise. The Medtronic system to which you have posted a link is true CGM where the Libre is a flash test and I believe quite different. I have absolutely no knowledge or experience of true CGM monitoring so will decline to comment on it. With respect if you wish to post a further reply I am able to read it without your use of bold type. It is however very striking
 



Please do not call people 'monkeys' I find it rude to the point of being quite offensive
 
Sorry for that, I pulled it off google ha!
 
Diabeticliberty said:
Please do not call people 'monkeys' I find it rude to the point of being quite offensive
Click to expand...
If you are putting a racist slant on my comment then that is up to you, as it is not intended to be racist.

Look up the urban dictionary for the term 'phone monkey':

phone monkey
Anyone that works in a call-centre type environment or where the majority of contact with the customers is done through the phone.
But in particular anyone that cant do anything other than follow a pre-written script.
 
You must log in or register to reply here.
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn More.…