First of all, I know that the best scenario for calibration is when bg’s are steady. Now that I’ve gotten that out of the way…
Let’s now discuss how to best handle a calibration when one’s bg’s is moving up/down at a reasonably moderate rate, such as 1 mg/dl/min. Given how interstitial readings are behind finger stick readings by APPROXIMATELY 15 minutes, would it be useful to input a reading 15 mg/dl higher than one’s fingerstick reading (during a period when, from successive readings, you notice you are rising at about 15 mg/dl/min), so that when 15 minutes elapses, hypothetically, the sensor will read quite close to the expected actual bg at that time? Make sense? Or is there an algorithm built into the G5 system that would keep such a “cheat” from providing a more accurate calibration?
The Dexcom is pretty forgiving and usually quite accurate. Yes, it can be squirrelly when you’re BG is transitory, but those are the moments to remember that you have a BG meter for right NOW.
The beauty of the Dex is that it really can help you avoid disaster because it shows TRENDS. Seeing or is often WAY more helpful than a number!
On those occasions where I’ve “had” to enter a calibration when there was something other than a flat arrow, I entered the numbers as shown on the meter. That has usually workout out well for me. I am sure that the Dexcom algorithm is already attempting to calculate for that 15-minute lag time - as evidences by all those times that Dexcom and the meter are virtually the same - as such, I wouldn’t want to second-guess the algorithm on my own.
Not, though Dexcom has started recommending making corrections based on the eventual BG (when those auspicious arrows are showing) - I’ve found that the arrows reflect the change in BG until the current reading, which frequently does not (for me) necessarily indicate a continuing trend - so, if I correct for 15-50 points higher/lower, I’d end up in trouble (and have!). So… like with calibrations, I tend to correct for the number I see, rather than the number I expect, and make additional corrections later, if need-be.
We then follow that up by watching the CGM and waiting for a moment of opportunity for a flat line in a decent range at which point we will do another calibration without regard to whether the Dexcom is asking for a calibration or not.
Diddo with Thas and Tim35. If it seems off, it can be adjusted. Although, its always better to cali on a flatline. That can be a pain, but I try not to take insulin 2.5 hours preceding calibration.
We always enter the meter value. I feel like I read somewhere along the way that the algorithm does indeed attempt to compensate for any lag time.
Incidentally, I don’t find lag time with G5 to be a significant issue overall. With prior versions of Dexcom it was definitely something that we were thinking about and compensating for when making decisions. Perhaps it’s more a matter of greater accuracy of Dexcom, or Caleb’s bg variability is less (I don’t think it’s this one), but I only think about the lag issue in times of double arrows. Otherwise, we don’t feel an impact from CGM lag.
