I’m sure someone’s asked this before, but when do you calibrate your Dex? I know that when it’s pretty level is a good time and that when it’s changing rapidly is a bad time…but what about if your finger prick reading is substantially different from Dex? Wouldn’t you want to alert the CGM to its error, or will that just screw up its algorithm?
I ask because the a1c reading doesn’t seem to correspond much with our Dex prediction, and am wondering if calibration issues could be part of it.
I find the best time to calibrate is before a meal, because I am generally sitting down and fairly stable BG-wise.
Today, I have been getting the dreaded ??? from the Dex. I tried two approaches.
Test and calibrate until the Dex bends to my will. Not very successful.
Ignore it for a while, then test to see if it’s back in range.
Was more successful
I don’t know whether over-calibrating annoys the Dexcom, but it usually works best when I let it sort itself out.
My rule of thumb has always been: Overcalibration = Inaccurate Readings
Calibrate only when you’ve “been horizontal” for at least an hour. Calibrate no more than twice daily (after the first 12 hours or so). If meter reading is way off from Dexcom reading, wash hands and try again. Allot extra time for Dexcom to recover from a low to avoid over treating. Repeated wonky numbers (we usually only get these after 14 or so days) = time to change sensor.
Bottom line: Don’t over calibrate! Almost without exception, nothing good comes of it.
Do you callibrate when you have a ??? displaying? I’ve been told by Dexcom technical support to not callibrate during these periods. Has it worked for you?
It depends on WHY my finger stick reading is substantially different from my Dex. If it’s within 2 hours of eating, or recently after exercise, no. There’s a lag between the CGM and the sensor, which tests interstitial fluid not blood. Also, if I’ve been in a hot tub or just taken a very hot bath I expect my sensor to report false highs - my CGM will routinely show a spike that drops off suddenly within about 4 readings of exiting the tub and cooling off, but during that time may be 50 points different from my glucose meter.
I’ve also learned to double check my finger sticks if they’re way off, not just automatically assume the Dex is the one that is wrong. A couple of weeks ago my Dex was reading in the mid 90s 2 hours after a carby breakfast, without having ever shown much of a rise. I thought that might be wrong and checked with a finger stick. My finger stick said 156. I went “AYIE!” and took a corrective bolus. Then something said “Hmm. It shouldn’t be THAT high, though…” and I tested again. Even though I’d just washed my hands, something was wonky with my finger stick, NOT my dex, as the second finger stick was within 5 points of my Dex. Whups.
I only calibrate once a day, generally sometime in the morning when I have a stable trendline. If my reading if way off, I give it some time to see if the Dex is just behind a rise or drop. Generally that is the case and fifteen minutes later, I will see the change in the Dex.
I think this is your answer.
I calibrate before and 2 hours ours after each meal (that’s 6 daily calibrations). I rarely (one in 4 months) get situations where the Dexcom varies significantly from the glucose meter. I have seen the ???, may once in the last two years.
I think location of sensor, alcohol etc. all play a part in how well the Dexcom tracks BG.
I usually calibrate when it asks me to, so that is twice a day. Although if arrow is flat and it’s off by more than 40 points and it happens often, I calibrate. It is why I don’t dose from my CGM readings. If it’s that far off and I have a sentisivity of 1:100, it could be a nightmare.
Interesting. So it seems the consensus is: calibrate when flat for a while, but that there’s some discrepancy on what to do if Dex is significantly off from the meter reading AND it’s flat. Whether more or less calibrations are better (assuming a flat trendline) also seems to be a matter of debate.
If it is stable AND significantly off–then re-calibrate. Over-calibrating DOES cause more inaccuracies (as confirmed by Dexcom rep). Site location makes a huge difference, individual people make a huge difference. We dose from Dexcom–but have learned about the caveats that come along with it. Pressure lows/highs, a delay from recovering from lows, and so on… factoring these variables we have found Dexcom to be very good.
Thanks everyone, lots of helpful advice here…and it came at just the right time. We got a super wonky sensor that has been reading all over the place, and I think we would have been tempted to calibrate like crazy in the past. Dex still did ask for LOTS of calibrations but we refrained from doing more than that based on this advice.
Now for the followup: When do you decide to give up on a relatively new sensor? This one is just two days in and was utterly useless for the first 24 hours and not much better the next 24. Will it get better in folks’ experience or should we just save ourselves the hassle, call Dex for a new one and switch sensors?
I just had an experience with “jittery” readings with abrupt highs and lows. Every BG check was way off. This was unusual and lasted for several days. We tried different locations and two different sensors to no avail. Desperate–we then changed the transmitter and now all is good. Here is a picture of what the jitter looked like:
This was a frustrating week and we were really happy once we figured out what the issue was.
Hey Tia - I follow the advice directly from Dexcom’s site -
“You should calibrate when the system requires it or if the CGM reading is inaccurate. You can calibrate when you are asked for the two start-up blood sugars, and whenever the sensor is inaccurate. Inaccuracy is defined as when the difference between your sensor glucose reading and blood glucose value is greater than 20% of the blood glucose reading and blood glucose value is greater than 20% of the blood glucose value for sensor readings > 80mg/dL or greater than 20 points for sensor readings < 80 mg/dL”
As others have mentioned, I only calibrate when my BS is stable. My last a1c was pretty much spot on with my Dex following above.
Twice a day. With more testing the numbers get wonky.
I calibrate much like @mikep does. I’ve also, upon Dex’s direction, repeated 2 or 3 calibrations on 15-minute intervals if algorithm maintains inaccurate interval from fingersticks, +/- 20% > 80 and +/- 20 , <80. I almost always find a blood-fouled at the root of the problem.