I spent most of today thinking my 24 day-old sensor was 'way off', because the calibrations were rarely within 1 mmol; something I have come to expect from my trusty companion. When I wasn't scoffing at the result, I was resigning myself to pulling it - something I hate doing when I've managed to get almost a month out of it. For some reason I become attached to a sensor the longer it lasts and the better it performs. But I thought that this might be a good opportunity to test out a word of caution my consultant had given me in clinic some months ago.
When my consultant found out that I was using my Dexcom G4 sensors for waaaaaay longer than the 7 days they have CE approval for, he offered me a word of caution. Two, in fact. Firstly, doc was worried about the damage I might be causing to my might-be-needed-one-day cannula sites, as I was regularly getting 21-28 days out of sensors - up to four times longer than advised. The second concern he had was that over time, the peaks and troughs the trace would show me would be less sharp as the sensors response to glucose in the body becomes blunted, meaning you won't get the precision you are looking for and could even miss the highs or lows you bought CGM to catch in the first place
As I downloaded the data I was writing the 'watch out for this, folks' post in my head already. But as I looked at the picture, this is what I found:
The red dots are my blood glucose calibrations today, while the purple trace is the results my sensor was providing. I felt surprise as I looked at it, having been convinced this was 'the risky bit' my consultant had warned me of. But even though the calibrations felt 'way off', when you look at it though the objective eyes of hindsight (and some kick-ass software), the valuable trend information (that we have so banged into us is more important that the 'exact number') is still there. At my highest point the sensor was spot on, and at my lowest it was still reflecting the changes in my glucose trends. In fact when you really look at it, the CGM still appeared to be picking up a few more peaks and troughs than my BGs did, as it should.
Perhaps it felt that way because this is the kind of data you see would normally see on day 5 of week 1, when the sensor has had a chance to be calibrated numerous times but is still a 'fresh' sensor, in the long run:
This day was generally filed under 'not so great'. Any BG of 18mmol at 3am is always filed under 'not so great' (unless you are normally 24mmol at that time of night, in which case, go you!), but despite the crazy roller-coaster day, the only two calibrations I did (unusual, for me) were both absolutely spot on. Perhaps it is because of this kind of precision that I have become so accustomed to that threw me off when the calibrations seemed so 'far off' today.
But having seen the data from today and realised that the sensor was not as AWOL as I felt it was, I'm going to keep hanging on. Despite being 24 days in I've yet to see more than one set of ??? marks, meaning communication-wise the sensor is still going strong and speaking to the received without any problems - something that can falter in an older sensor. And although my calibrations are a little off bulls-eye, the trend data is still more valuable than finger sticks alone, so I will plough on until this one gives up the ghost for good.
Hopefully sometime in 2015.