I’m getting pretty annoyed, here! My Dexcom and my readings from my Omnipod PDM frequently don’t match - not even close - like 50 or more points difference. Is this a thing with a lot of people or am I just lucky? I would consider getting rid of the Dexcom but I think I’m beginning to lose my sensitivity to lows. Even so, my phone chirps at me for being low when I’m not and high when I’m not. It’s so hard to tell which one I should listen to - the CGM or the PDM? Any thoughts or words of wisdom? Thanks, y’all!!
How often do you calibrate your Dexcom?
Hi jpax,
First thing I would suggest is seeing if the problem is with the PDM meter or Dexcom.
Very first thing is to see if your PDM meter is accurate. A few easy things to try.
-
Maybe you are already doing this, but make sure you are washing and DRYING your hands before tests. Traces of water, sweat, or food on your fingertips will mess up your readings.
-
Take several tests, a few seconds apart, on the PDM meter. They should be very close. How close are they?
-
Use a different meter, and compare it to the PDM meter. How close are they? If you don’t have a second meter, you can buy them for a reasonable price. If you take tests with both meters at the same time, how close are they?
If your meter is not accurate, and your PDM is under warranty, you can get a swap from Insulet for free.
If you have an accurate meter, then it’s a problem with your Dex. There are certain things that make it less effective for me. Sweating a lot makes it practically unusable, until I have replaced the fluid. It totally flakes out on me with a lot of sweat. Placement of the sensor, and your tissue type will also affect how accurate it is. My tissue type doesn’t work well!
If you find it is a problem with Dexcom, reach out to rgcainmd on this forum. She is a Dexcom maven. A connoisseur in the art of CGM.
IMO, accuracy is all about thoughtful calibrations. I will copy and paste my calibration guide once I get to my computer.
1 Like
Accuracy is affected by the deviation percentage of the actual reading. Though Dexcom claims a MARD (mean absolute relative differences) of 9%, and Freestyle (one the Omnipod PDM) has a MARD of 4.9% - the actual readings can vary more (up to roughly 20%).
The MARD would suggest the Freestyle meter is more accurate than Dexcom, but Freestyle claims readings are +/- 20% - which means your reading of 100mg/dl on the Freestyle could be 80mg/dl, or 120mg/dl.
Some of these differences can be the test strips, how well you clean your fingers before testing, where your Dexcom sensor is inserted, and a lot of other factors.
The suggestions given on calibrating are important, and so is washing your hands/fingers with soap before taking a reading with the PDM. Surprisingly, many people do not know that dirty fingers cause inaccurate test results. But ultimately, they will never be as accurate as a professional meter or A1C.
here’s a link to Rose’s thoughtful tips on calibrating
@jpax Over calibrating will mess up your accuracy. It is best to calibrate when your sugar is fairly stable (no big drops or rises) or when there is a large difference when steady… I try to calibrate in the morning before eating when it is nice and steady. Consider getting a reference standard to make sure that omnipod is getting accurate readings. Site location is a big factor. Different sites will have different lag times. Dexcom is prone to lag regardless–but is notable more so during a large rise or fall. Coming up from a low is especially laggy.
2 Likes
I note we’ve yet to hear back from @jpax regarding calibration… Hard to say what the problem is until we know more. Calibrating too much? Not enough? Bad PDM? We need more info from you, jpax!
I’m here, I promise! I try to only calibrate twice daily and only when it asks me to do it. I haven’t had a chance to test the meter on my PDM against another meter, but that is next on my list. I’m getting that the important thing is to calibrate when the numbers are steady and not just before/after I eat. That makes perfect sense - I freak out a little when it asks me to calibrate and I don’t do it right away. I’m also pretty sure I’m never hydrated enough, so that will be a goal for me, too! Thanks y’all, and I’ll report back after I test my PDM against the regular meter. BTW, the Omnipod is the best thing since sliced bread in my opinion. I. Love. It.
Ok, I tested using my Omnipod just now and the result was 184. The result from the meter was 176. I feel like these are pretty close, what do y’all think? I also just replaced my Dexcom sensor and I calibrated it using readings 10 minutes apart. Hopefully replacing it and calibrating it that way will help. Any other suggestions? Y’all always have such great insights into what my issues may be. I appreciate it!
The higher the number, the less significant a difference. In your example, being 8 points different around 180 is not as big a deal as it would be when you are 90. I think in the example you gave, 8 points is fine.
But the other questions I have for this are: What other meter were you using? And did you do several tests? Doing several tests at the same time will give you a comfort level that the meter is ok.
Our CGM / PDM have been drastically different on many occasions. We calibrate only 2 to 3 times per day (typically now only after the blood drop appears) and only when the line is ‘level’. Just yesterday, for instance, the CGM read 176 double arrows down. We checked his BG with the PDM and his BG’s were actually 67! This happens to us often, honestly…and it’s frustrating, but the significant differences usually occur when we are seeing double arrows one way or the other. We now check anytime double arrows DOWN are showing, if he’s under or around 150, because when we see those arrows, it means it’s his real BG’s could be significantly lower than is being reflected on the CGM.
When I see ↓↓ I can FOR SURE expect a much lower actual BG value. Now I use the delta value to estimate how much lower. For instance a -20 is about 50-60 pts lower where -30 could be 100 or more. It’s just not accurate during big swings. Not Dexcom’s fault–just the nature of the tool. This nuance should be explained when for all new users–especially now that you are able to dose from it.
1 Like
The other meter I have is the FreeStyle Freedom. It usually matches the PDM very closely. If I take several readings at once from either one, they’re very close. This morning my Dexcom asked me to calibrate it and it was about 5 points off from my PDM and meter. They were all very close to each other this time, all around 110. Has anyone had experience with the FreeStyle Freedom - good or bad?
I really want to love my Dexcom, but maybe I have unrealistic expectations. Is anyone using it to dose their insulin? I know it’s not approved for that and I don’t think I can trust it to be accurate just yet.
Also, after I shower I lose contact with it for about an hour. Is that something anyone else has dealt with? Also, does anyone wear it anywhere but on their belly? If another area would make it more accurate, I’m definitely willing to try it.
SO MANY QUESTIONS!! 
FreeStyle testers are great!
As you are describing, they are consistent and accurate. Since the PDM uses the same internal chipset the FreeStyle uses, you will generally find them to be very close to each other - unless one is defective.
For many people, the Dexcom readings are close enough to dose. I don’t dose off of Dex, because there are too many times when it is far off. I’d rather do a BG test.
I’ve lost contact with the Dex in various circumstances - exercise, sleeping, showering. Dehydration can cause a problem. So can sleeping on it, lying on it, pinching it in the skin surface.
I’ve tried wearing it all over. Fatty areas are reported to work better. I think some people have had success on their arms.
Success may vary for you. For me it is not as great as many people have described, but I test a lot so it is no big deal. The biggest benefit for me is when I am unable to test, like when I am sleeping.
Thanks for the very helpful calibration guide. I’m on my first sensor so it’s very very useful. One question though - why is calibrating too often - assuming a horizontal line - a problem?
Maurie