Freestyle Libre Accuracy

[SNW]

I have been using Freestyle Libre for about 2 years - paying for it all myself. I have trusted the system as it seemed to correspond with how I was feeling and hardly ever checked with a finger prick. However a few months ago I was sick with an infection and my BG readings went very high. Paramedics were called by my GP and I was put on insulin for the first time. In all this nobody took a blood reading and all were content to believe the Freestyle sensor. So I started to do comparisons myself. As an example I found that when the sensor said 22, a finger prick said 14. I now have dozens of readings, from 3 different sensors, that show they permanently over-read by between 20% and 40% even allowing 20mins between the blood and the sensor readings. This morning the sensor said 9.7 (which I regard as high and something I would deal with by medication) whereas 2 finger pricks with 2 different systems (Abbott and Accucheck) said 5.2 (which I would regard as normal - no action required). At no time over the course of the last month has the sensor reading been within 25% of the blood reading.

I am now very concerned and a little scared as I don't know what to trust - my GP has advised me to abandon Freestyle. I have reported every incident to Abbott and the only practical reaction is to offer to send me a replacement sensor. When I say that they have already sent me 6 replacements due to inaccuracies and I'd rather have a solution, the only response is to send a new sensor any way and deny that the system over-reads to this extent.

I am a cyclist and knowing my BG readings - especially before and during long rides - is essential; but I don't want to have to take a finger prick kit with me especially when riding on wet days with mucky roads. I wonder if anyone has had a similar experience with the system or any suggestions. I have previously tried the Dexcom system but found it too expensive and very easy to screw up the application of the sensor.

 

[urbanracer]

From experience your problem is not uncommon and there are many threads on the subject if you search (top right).

As far as 'trust' goes, for me it's the finger prick test every time, because it's an actual blood test. For me, the Libre is more useful for trend analysis than absolute BG readings.

I have usually found the Libre to be reasonably accurate in the mid-range, say between 5 to 9 mmol but outside this range the discrepancy appears to widen and the error is not linear.

I guess it suits some people better than others but we do have members with some clever tricks for calibration @Scott-C , @Mel dCP .

 

[Deleted Account]

My experience is
- The Libre is most accurate between 4 and 7. Outside if this range, it can go a bit haywire.
- The Libre is factory calibrated and my arm is not the same as a factory arm. I get closer results using the Glimp app which uses a different algorithm to calculate BG from interstitial fluids that takes into consideration recent finger prick readings. I find Glimp much better than the native Libre app (or reader)
- Libre is 15 to 20 minutes behind. Therefore, it is unreliable when my BG is changing quickly such as when exercising or soon after eating
- I use Libre but never go through a day without finger pricking at least 4 times.

 

[JMK1954]

I also had inaccurate readings and abondoned use of the Libre because of this, plus a skin reaction to the adhesive. It was closest to a blood test result between about 6 and 9 mml in my opinion.
Edited to add - It is disappointing isn't it ? I would rather use a straightforward blood test than rely on anything else.