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

Type 2 Are charts in Libre 2 app adjusted to blood test results

C

chrisnm

Newbie
Messages
2
Type of diabetes
Type 2
Treatment type
Insulin
Hi
I’m a type 2 that has been using Libre sensor monitoring for awhile, off and on. I just got The Libre 2 system. I often get very low (inaccurate) readings from from the Libre sensor. Maybe half my blood test readings. I typically enter my blood test results in the app data after I check Libre 2 low readings. I notice that the low Libre 2 data gets corrected in the dally charts (does not show the low readings). Is the software normalizing the chard data to reflect the more accurate blood data? I’m not sure what to think.
Thanks
 
Hi
The daily chart only shows the pattern and variability of your sensor glucose over a typical day. A minimum of five continuous days is required for the chart. Your low glucose events should appear in the low glucose events chart. If you are getting low low readings on the libre with a substantial difference from a finger prick test then I recommend you contact Abbot the manufacturer and consult with them. If the reading discrepancy is out of their limits, they normally send out a replacement sensor. They are having problems on Android phones on operating system 13 with sensors not emitting alarms. Hope this is of assistance.
 
The scans are predictions of where the Libre thinks you are, and are heading; the graphs show where you actually were and there is about a 15 minute delay. So although the scans should be relatively accurate, if your glucose is changing rapidly or not quite as the algorithm expects, you will find that the scan dots do not sit on the graph lines. There is no adjustment because you have added a finger prick reading, that will just remind you what that result was and may be of use to Abbott if you feel the discrepancy is too much.

I rescan after 15 mins if I think the last one was too high/low. I am not on insulin though, so I assume you will have been told how to react to high/low readings and whether to treat immediately or re-test, wait or whatever. Although I love using Libre, the first and last couple of days of the sensor are always a bit squiffy for me, but I have learnt to deal with this.

Hope this helps.
 
sgm14 said:
The 15 minute delay was for the Libre 1, but Abbott say the average delay for adults with the Libre 2 is 2.4 minutes.
Click to expand...
The delay is not in the Libre it is between the blood glucose reading and interstitial fluid reading.
As @Mrs HJG explains, the algorithm Libre 2 uses to covert to blood glucose takes into consideration this delay by predicting the current reading through extrapolation of the current trend. If the current trend changes direction in the last 15 minutes, the prediction is updated. Hence, some predicted lows may be lost from the historic data. Because they didn't happen.

I have not read about the 2.4 minute delay but this would be an approximation of timeliness based upon this trend extrapolation used in the algorithm. The delay is still there in our bodies
 
sgm14 said:
The 15 minute delay was for the Libre 1, but Abbott say the average delay for adults with the Libre 2 is 2.4 minutes.
Click to expand...
If you look at the graphs (not the one that shows after a scan, but the daily graph, it does not infill the last 15 minutes until the next scan - but scanning in under 15 mins since the last one won't update it - checking the logbook graph will show the difference between the scan and actual levels, with the same delay.

If comparing finger pricking with Libre, although measuring different things, it makes sense to compare a (second) Libre scan from after the finger prick a couple of minutes later, that's what I was taught. eg. Scan says 3.2 and dropping; do a finger prick, and it actually says 4.5, then do another scan and may be 4+ then decide if you need to treat, rather than downing the jelly babies after a single scan.
 
In Response said:
I have not read about the 2.4 minute delay but this would be an approximation of timeliness based upon this trend extrapolation used in the algorithm. The delay is still there in our bodies
Click to expand...

That is not my experience and it is also not what I would understand from reading https://freestylediabetes.co.uk/freestyle-libre/flash-glucose-monitoring-temp.

I know that the length of the lag can be different for different people, but my lag is certainly less than five minutes. I am sure of this as I usually scan immediately before getting up and then about five minutes later when taking my injection and there is often a significant increase by that time - the so called 'foot on the floor' syndrome . This (unfortunately) happens too many times for it to be a coincidence and can not be based on any extrapolation or prediction as it happens whether I get up earlier or later, so can only be triggered by the actual action of getting up.

There have been numerous other times when I have done something to cause my blood sugar to jump up which shows up quickly on my Libre 2 scans. Lately I have also been experiencing what I call 'foot OFF the floor' syndrome where my blood sugar jumps when I go to bed, so I usually scan before I get undressed and then about five minutes later when I am in bed and again this has quite often shown the increase.

As always, your experience may differ.

As for the original question, I am not sure whether Abbott have ever detailed exactly how they calculate the daily graph, but I would suspect that do smooth the graph. They have stated that they take readings every minute and then calculate the average of those readings every 15 minutes which is sent to their website. I suspect the graph line is based only on those averages with the individual scans added as dots which can explain why some manual scans are not on the graph line.
 
@sgm14 I think you misunderstand.
When you scan, you see the results of the algorithm applied to the interstitial fluid reading which has attempted to take into consideration the page between it and blood sugars. You do not see the value that the sensor reads.
What your experience shows is that the algorithm (not just the sensor) works for you.

The algorithm most often fails when our levels change different. For those of us who are at risk of hypos, this is most commonly apparent when we treat a hypo. Before treating, our levels are falling as shown on the graphs. When we treat a hypo with fast acting carbs, oir levels can rise quickly changing the direction of the trend in our blood sugars. However, because the algorithm uses interstitial fluid readings which change 15 minutea behind blood sugars, the sensor is not aware of the change in trend and the prediction reports a longer, deeper hypo that really occurs. When the sensor “catches up” with the recovery, it will update the history.
 
In Response said:
I think you misunderstand.
What your experience shows is that the algorithm (not just the sensor) works for you.
Click to expand...
Well I think I understand what you are saying, but I respectively disagree. At 7.00am my blood sugar levels have been steady for several hours and then I get up. Five minutes later my readings have jumped up. That can not be explained by any algorithm unless the sensor has already started to detect the increased blood sugar levels in my interstitial fluid.
 
Back
Top