According to my continuous glucose monitor, I just had one of the best diabetes days ever– between 98 and 102 the entire day. The trend line on my monitor was gorgeous, a distant ocean horizon with nary a boat to be seen.
There was only one problem with this: my CGM throughout the day was not correctly reporting blood glucose measurements. It wasn’t entirely inaccurate or uncorrelated, though; when my blood glucose levels went up, the CGM’s output showed an increase as well, and when I went down, it followed. The problem was one of scale. Whatever rise in current was being detected in the interstitial fluid resulted in a reported rise in blood glucose levels that was only about a fraction of the corresponding finger-prick-measured rise in blood glucose levels. So, though I was centered pretty well around 100, I learned quickly to interpret the CGM’s “102” as “150” and the CGM’s “98” as “60.”
I have experience this before to a lesser degree– sometimes a sensor’s conversion of interstitial fluid current to blood glucose levels reflects a dampened function of the actual blood glucose values, such that I can be 200 according to a finger-stick, and the CGM will rise proportionately, but only ever reach 140. Usually, though, the scale results in numbers limited to a range of 70 – 140; 98 – 102 was impressive in how tight the scaled range of values was.
Now, I’m not sure what causes this dampening. Anecdotally, it seems to be related to the location of the sensor. I spoke to a Minimed representative about it, once, early on in my experience with the CGM; I was near tears in the bathroom at work, bleeding profusely as I tried to replace the sensor while waiting on hold, inserters and alcohol pads spread across paper towels on the counter. The representative insisted I was calibrating the device incorrectly– too frequently or at points of too much change. I tried to explain that I had been tracking as the day went by the current measured (this is displayed as “Isig” in one of the Minimed screens), and I could see that a 20 point change in blood glucose corresponded with no displayed change of Isig at all, so unless there were orders of magnitude more specificity available to the device than it was displaying, it couldn’t just be a problem of calibration. The call turned somewhat combative, and eventually I gave up. (It probably wasn’t entirely his fault; I was, after all, bleeding from a stomach wound in a public restroom, crying in frustration with my broken immune system, and not happy to be told, the problem is you, not our sensor. Not one of my prouder moments.)
I’ve had some success avoiding sensors with dampened trend lines by inserting the sensors higher and more towards the outside of the abdomen. Unfortunately, I have found that many of the places that seem most able to take the sensor are also the ones that have the least viable sensor performance. And even more unfortunately, there’s only so much space available, so sometimes I must start annexing territories with questionable promise as sensor sites.
And indeed, the sensor with a four-point range? In dangerous territory. So, though part of me wanted to keep that sensor in forever, reporting beautiful numbers, the loss of granularity was too much for me. I changed the sensor when I got home, cursing and crying at the pain and the blood and the scarred expanse of my abdomen.
Sigh. We can cure this disease, right? Like, soon?