Well, it's certainly not complicated to fire in an FDA complaint, but it is significantly more complicated to fix the things which you want to be fixed.
Any cgm system, whether it's libre, dexcom, medtronic, medtrum, is operating in a hostile biological environment which doesn't want that sensor filament to be there. Biofouling, foreign body response, repair mechanisms will always be there and produce keech results from time to time.
That's true with dexcom as well. I've looked at American T1 sites - it's got more traction over there because of insurance. There's people who think it's the bee's knees, ultra-accurate, and people who think it's rubbish
CGM seems to work for some people and not for others.
When I first started with libre, I read William Lee Dubois's book, Beyond Fingersticks. He's a long time cgm user. He makes the point, which still holds true today even though the book is 8 yrs old, that one of the skills of using cgm is being able to figure out when to trust it, and when not to.
These things will never be perfect, and that's not because of Abbott failures or lack of FDA oversight - it's because of biology.
Most users figure out ways of dealing with these things. Sure, sensors regularly understate by 1 to 2, so I add 1 to 2 to the reading. I learned that in the first week of using it.
I then got a blucon transmitter which lets me calibrate it through xDrip+. Some sensors are really tight, others aren't, but there are ways of living with it.
With sketchy sensors, I'll test more as a cross-check. With good sensors, I'll test less - recent bg meter history from a few weeks back shows a grand total of 3 tests from Friday to Monday and none at all on Sunday, because I trusted that sensor. If it was a bad sensor, I'd have checked more.
The current generation of libre is batch calibrated. I've certainly had a few bad batches - it's why I only bought 3 at a time - but they've been ok replacing them.
Abbott has announced that for their new generation 2 (it'll send hypo/hyper alarms), they're going to calibrate each one individually. There's speculation in the #doc that they're already doing that as a test run. Some, but not all, of my recent sensors have been fairly spot on, more so than a year ago, so it's a watch-this-space thing.
