Libre II - Anyone Had Missing Data?

[MikeyBikey]

Late yesterday evening I had about an hour's gap in my Libre data and then this morning I found another two hour gap. Anybody seen this behaviour before? There were no "signal loss" alarms.

 

[everydayupsanddowns]

I’ve had that with various sensors where the information provided falls outside of what the sensor thinks is likely / possible, and it takes a few minutes to see if the information settles down.

Sometimes it’s when a sensor is failing, or perhaps slightly coming loose.

Libre will sometimes say ‘wait 10 minutes’. Dex has a ‘readings will be unavailable, check again in 30 mins’ message.

Perhaps you leant on the sensor?

 

[djdave]

I see this quite often with my Libre 2, when I do spot a gap like that I re-scan the CGM and it fills the gaps with the recorded on-board readings it saves on the CGM. Just annoying it chooses to miss some readings at times. It's got to be something to do with the bluetooth networking, but I haven't managed to track the reason yet...
I'm a software engineer and have my own scripts which get the readings from LibreLink so I can plot and see my readings over whatever period I want. When these gaps happen, I was also missing readings in my data. When you do a manual scan of the CGM the 'plot graph' data is also attached to the reading packet so I have adjusted my software to automatically read that when data is missing 🙂 Just annoying...

 

[helli]

I have wondered about these occasional gaps.
I do not think they have anything to do with Bluetooth drop out as the sensor memory stores 8 hours of data and the gaps are still there after a NFC scan.
My money is on the "weird" readings suggestion from @everydayupsanddowns - Libre thinks the readings are unlikely so misses them out of the graph. This could be due to a compression low, a steep fall or rise or a "glitch in the interstitial time continuum".
As you have your own scripts that extract the data from Libre @djdave it would be interesting to know what values are recorded. When I used to use xDrip+, there were times when LibreLInk decided the sensor had failed whereas xDrip+ continued to receive obviously corrupt data (bouncing between very low and very high). I guess xDrip+ doesn't do a "is this data reasonable" check.

 

[rebrascora]

Just want to reiterate that I don't get any problems like this with the reader, but I did with the phone app occasionally. Love my reader! So reliable! My TIR has improved about 10% just going back to using it from using the phone app. It was interesting to try the phone app for a few months again but it was also really frustrating at times and it has been such a relief to go back to the reader and return to getting TIR in the 90s again. Find it really hard to believe that the device you use can make such a big difference both to my results but also my confidence in managing my diabetes, but it really has.

 

[djdave]

@helli Yes there are marked differences between what data is shown on the app screen to what 'live' values are sent out.

The 'prediction' firmware on the CGM is a bit pessimistic at times. When there is a rapid rise the 'live values' actually reported do spike greater than the graph shows. I think the actual app graph is closer to a 30 minute rolling average.

I had one sensor a month ago which bounced back and forth for a day before setting down... I never did trust that one and swapped it out after 1 days rather than waiting for the full 14 days.

Here is the output from my graphing scripts...


And this is the app's view on life. Note app is normal short-term view, my graphs are 24 hour view...

 

[JohnS]

Like @djdave mentioned, if I get a gap in readings I just do a manual scan of my Libre and all the gaps get filled.

 

[Lucyr]

It happens sometimes. Just scan to fill them in.