Help-Libre Free / Finger readings very far appart

[byeazucar]

Hi all,

My wife's LibreFree just notified her of low sugar after waking up from a nap (last meal intake over 2 h ago). I tested and returned 3.3mmol then after 5 min 3.1mmol. I found it unusual so I decided to do a finger test as we are at home. To my surprise, the finder reads 7.0mmol!!

The gap is really big and really brings doubts to me if I can really trust LibreFree. So far we have been using it as our main reference, and on some occasions where we have tested finger blood to check reliability, the difference has been ~15% between the two reads. However, a difference of almost 4mmol really worries me and it makes me think when she is 7mmol in the morning spike, is she really 7, or could be an 11mmol reported incorrectly?

I will start comparing with finger reads more often but I would like to know if anyone has had this experience before and also what was Libre support response (planning on contacting them)

To add, my wife is currently starting her period and I believe sugar reads always are more "extreme" and change drastically in such times (would also be great to get people opinions on this)

5 minutes after, Libre is reading 4.8mmol, now 5.5mmol, I find this very misleading and I find it hard to trust the device and the readings, please some help to understand such situation

Thank you,

Adrian

 

[LittleGreyCat]

How long has shed had the current one attached?
Sometimes they take a few days to settle down.

Also, I have had an alarm over night when I was leaning (sleeping) on the sensor and it went very low.
Came back up to a reasonable value over the next few minutes.

The Libre cannot be trusted to be accurate (always check readings with a finger prick if you are concerned) but is very good at showing trends.
It is generally good to check each new Libre after a couple of days to see if it is reading low, high or just about right.

 

[In Response]

The most common reasons for Libre to be out are
- compression lows. Sensors will report low if you apply pressure to them. This is possible if your wife was lying on the sensor and something to consider when placing a new sensor - avoid placing it on the part of the arm that she lies on.
- bedding in. Our bodies can react to having an alien object inserted in our arm which results in inaccurate/random/noisy readings. I find my body takes at least two days to get used to a new sensor. Therefore, I insert it two days before activating it. So, for two days, I have an active sensor in one arm and another one getting familiar with my body in my other arm.
- when levels are rising or falling quickly. Libre (and CGMs) use readings from Interstitial Fluid to estimate out blood sugar levels. Interstitial fluid can be 15 to 20 minutes behind finger pricks. Abbott have updated the algorithm it uses in Libre 2 to attempt to take this delay into consideration by "predicting" the value. If levels are changing quickly or noisy, this prediction algorithm is less effective.
- highs and lows. Libre are most accurate between about 4 and 8mmol/l. It can over report highs and under report lows. Therefore, it is strongly recommended to always double check a high or low reading before making any corrections. And always keep finger prick meter with you - you never know when Libre may fall off or need a double check. In the UK, I believe it is a DVLA requirement to have a finger prick meter with you at all times when driving.
- faulty sensors. Some sensors are just faulty. This is not that common but some people do experience problem sensors that need to be reported to and replaced by Abbott. Generally, Abbott are pretty good at replacing faulty sensors.

 

[Dilly123]

I have also started on the libre freestyle and are having the same problem. The libre is showing low 3.9mmol or below but when i finger prick test it shows am in range. Very confusing indeed!

 

[byeazucar]

How long has shed had the current one attached?
Sometimes they take a few days to settle down.

Also, I have had an alarm overnight when I was leaning (sleeping) on the sensor and it went very low.
Came back up to a reasonable value over the next few minutes.

The Libre cannot be trusted to be accurate (always check readings with a finger prick if you are concerned) but is very good at showing trends.
It is generally good to check each new Libre after a couple of days to see if it is reading low, high, or just about right.

She has had it for 10 days now so should be the case, I did realize the first 2 days were a bit off hence why we will attach our next sensor 2 days prior to activating it.

If it can not be trusted...is there any point on wearing it? as the trends could also not be accurate

 

[byeazucar]

The most common reasons for Libre to be out are
- compression lows. Sensors will report low if you apply pressure to them. This is possible if your wife was lying on the sensor and something to consider when placing a new sensor - avoid placing it on the part of the arm that she lies on.
- bedding in. Our bodies can react to having an alien object inserted in our arm which results in inaccurate/random/noisy readings. I find my body takes at least two days to get used to a new sensor. Therefore, I insert it two days before activating it. So, for two days, I have an active sensor in one arm and another one getting familiar with my body in my other arm.
- when levels are rising or falling quickly. Libre (and CGMs) use readings from Interstitial Fluid to estimate out blood sugar levels. Interstitial fluid can be 15 to 20 minutes behind finger pricks. Abbott have updated the algorithm it uses in Libre 2 to attempt to take this delay into consideration by "predicting" the value. If levels are changing quickly or noisy, this prediction algorithm is less effective.
- highs and lows. Libre is most accurate between about 4 and 8mmol/l. It can over-report highs and under-report lows. Therefore, it is strongly recommended to always double-check a high or low reading before making any corrections. And always keep finger prick meter with you - you never know when Libre may fall off or need a double check. In the UK, I believe it is a DVLA requirement to have a finger prick meter with you at all times when driving.
- faulty sensors. Some sensors are just faulty. This is not that common but some people do experience problem sensors that need to be reported to and replaced by Abbott. Generally, Abbott are pretty good at replacing faulty sensors.

Perhaps compression could be the reason as she was having a nap. What has me more concerned is if the reads are not accurate overall, as I would not like to trust we are on the range and actually being above. Specially during mornign spikes, as she can reach 7.8-8.0mmol and I would not like to be on a realistic 10mmol thinking we are on the range.

I will definitely continue doing more finger tests to understand how accurate it can be and to identify if there is a pattern on when it reports incorrectly

 

[In Response]

@byeazucar there are alternative (unofficial) apps which allow you to calibrate and still provide alerts. But they do not link with standard HCP systems.
These are written by software developers for software developer so, if you are techy, they are great but f you are not they can be more of a pain than a hindrance.

I use xDrip+ on my Android and much prefer it to LibreLink but would only recommend it to likeminded techies.

 

[Jaylee]

Hi @byeazucar ,

Dehydration may mess with the readings.. Is your wife suitably hydrated?

 

[Antje77]

3.3mmol then after 5 min 3.1mmol. I found it unusual so I decided to do a finger test as we are at home. To my surprise, the finder reads 7.0mmol!!

5 minutes after, Libre is reading 4.8mmol, now 5.5mmol,

The Libre is always a bit behind a finger prick, so this looks like she was simply rising quite quickly and the Libre hadn't catched up yet.

 

[In Response]

The Libre is always a bit behind a finger prick, so this looks like she was simply rising quite quickly and the Libre hadn't catched up yet.

As mentioned previously, the Libre 2 algorithm is much better than the Libre 1 algorithm with regards to handling the delay. With Libre 2, it predicts the value 10 minutes ahead. Sometimes you can spot this when the value reported at the time of a scan disappears in the future when the trend direction changes before the peak.

 

[Antje77]

As mentioned previously, the Libre 2 algorithm is much better than the Libre 1 algorithm with regards to handling the delay. With Libre 2, it predicts the value 10 minutes ahead. Sometimes you can spot this when the value reported at the time of a scan disappears in the future when the trend direction changes before the peak.

Not my experience. For me the Libre2 still takes a long while to notice I've been out of a hypo for 20 minutes already.
If it senses a hypo, or a hypo coming on, it cannot know if I treated or not so it would be impossible to predict I'm going up when my interstitial fluid is still on the low side.

 

[Jaylee]

Not my experience. For me the Libre2 still takes a long while to notice I've been out of a hypo for 20 minutes already.
If it senses a hypo, or a hypo coming on, it cannot know if I treated or not so it would be impossible to predict I'm going up when my interstitial fluid is still on the low side.

Hi,

Yep, I agree regarding the coming out of hypo territory. The L2 is still slow to respond with the reading either regarding how I feel or the meter.

But I also have to agree regarding the drop into one. Or any other reading. It's 20 minutes ahead of the meter.
I've consistently had an alarm saying I've hit a low 5 or 10 minutes prior? At first I snoozed the alarm thinking, "like hell I am."
Then.. "Hello? Guess I is." As the symptoms hit..

In that respect in my experience it's a great tool for me to soften the impact of a hypo, if not head one off altogether..?

 

[In Response]

Not my experience. For me the Libre2 still takes a long while to notice I've been out of a hypo for 20 minutes already.
If it senses a hypo, or a hypo coming on, it cannot know if I treated or not so it would be impossible to predict I'm going up when my interstitial fluid is still on the low side.

Correct. It can only predict if the trend is following the same approximate line
As I mentioned, the prediction can be out when the trend changes direction such as when correcting a hypo.

The advice is to always check hypos with a finger prick - that is in terms of whether a Libre reported hypo is a real hypo and in terms of whether the correction has worked.

I guess I did not explain myself properly.

 

[LittleGreyCat]

As mentioned previously, the Libre 2 algorithm is much better than the Libre 1 algorithm with regards to handling the delay. With Libre 2, it predicts the value 10 minutes ahead. Sometimes you can spot this when the value reported at the time of a scan disappears in the future when the trend direction changes before the peak.

Thanks.
That may explain why some green dots (for readings) are outside the long term lines.
Guessed and got it wrong then adjusted the graph later.

 

[Antje77]

I guess I did not explain myself properly.

Reading back, I see I reacted to what I thought you wrote, and not exactly to what you actually wrote, so I wasn't reading properly either.
In any case, the different issues are more clear now