- advertisement -

PDM = endo Meter. But A1C >> CGM

Discussion in 'Parents of Children with Type 1' started by DavidN, Jan 22, 2016.

  1. DavidN

    DavidN Approved members

    Joined:
    Sep 7, 2012
    Messages:
    744
    So our CGM has always been very close to our A1C. At yesterday's endo appt they tested the accuracy of the meter built into the Omnipod PDM and it was spot on. So if the PDM is accurate (I know historically there have been Freestyle accuracy issues), and we always calibrate the CGM with the PDM, why was his A1C a full point above what the PDM would have predicted? Damn. Frustrated and bummed.
     
  2. njswede

    njswede Approved members

    Joined:
    Feb 9, 2015
    Messages:
    385
    Welcome to the club! We had a 6.6 predicted A1C and ended up with a very disappointing 7.4 at the endo's office.

    Three things:
    1) The point-of-care A1C assay has a 10-20% mean error, so an error by an entire point isn't unheard of.
    2) The A1C measures A1C and *not* BG. It's a proxy metric for BG. Correct me those who know the better than me, but I think the A1C-level itself has no clinical importance. The correlated mean BG is what matters.
    3) It's said that A1C represents the three month average of BG. That's not really true. For it to be a true average, every day across the three months would have to count the same. That's not the case. Readings older than 90 days don't just fall off the average. Instead, it's what you would mathematically call a exponentially weighted average. What that means in layman's terms is that recent readings have a MUCH higher impact on the average than older readings and the older readings kind of "fade away" rather than just falling off the average. So, let's say your kid had a cold with elevated BG the week before the assay, that's going to have a much higher impact than the perfect stretch you had a month ago.

    When I look at A1C, I try to spot trends. If they're constantly going up, I'd be worried. But a single "bad" A1C isn't a big deal if your GCM and fingersticks indicates you have your numbers under control.

    YDMV, I'm a newbie and all the other usual disclaimer apply...
     
  3. Theo's dad Joe

    Theo's dad Joe Approved members

    Joined:
    Jun 7, 2015
    Messages:
    802
    There was a thread discussing real reasons for false high A1C a while back like certain types of anemia. Also a high LDL or low HDL can lead to more glycation for the same blood sugar. Anyway, my thought is that if you are bolusing off of the PDM, and not getting crazy patterns then the PDM must be giving you fairly accurate information and I think I would trust 3 months of readings that are good enough to allow effective day to day management over one quick A1C test.

    StacyMM had an issue with DD of false highs and I think may have used the fructosamine test.

    I have had 2 fast A1C tests vary by half a point for ME when done on the same day.
     
  4. rgcainmd

    rgcainmd Approved members

    Joined:
    Feb 6, 2014
    Messages:
    1,379
    This sucks, plain and simple.

    But what in all likelihood is an inaccurate point-of-care A1c assay is NOT a reflection of the clearly stellar effort you make day-in and day-out to insure that your son stays safe and healthy.

    Try to view this as one of those WTH high blood sugars that come from out-of-the-blue: make a [small] note of it and move on, and continue being the diligent D-Dad that you have always been!
     
  5. njswede

    njswede Approved members

    Joined:
    Feb 9, 2015
    Messages:
    385
    ^^ This!!! ^^^

    +1,000,000,000,000 !!!
     
  6. dpr

    dpr Approved members

    Joined:
    Dec 17, 2013
    Messages:
    399
    As njswede said the margin of error is pretty high for the test. Also there are quite a few things that can also affect accuracy of the test. From my experience the last week and especially the last few days prior to the test have a FAR greater impact than the last month or two. About a year ago we had 4-5 weeks of uncontrollable highs followed by a week of incredibly good numbers. The last week was one of those lucky, crazy good never above 140 and never below 65 and settled around 80-100 just about all the time. By my calculations her A1C should have been in the high 7's because of the bad prior month but she pulled a 5.3 out her magic hat! So that's A1C for you, don't put too much stock in it...
     
  7. Cheetah-cub

    Cheetah-cub Approved members

    Joined:
    Feb 18, 2014
    Messages:
    241
    Well, since we have dexcom now, we don't do finger check as much as before. We only do finger checks before every meal, and at 2AM. The only other times that we do finger checks is to confirm dexcom's very low numbers, or to confirm very high bg readings. So, we know our pdm finger checks numbers can be skewed.

    But our A1C tend to track closely with our dexcom numbers. And these dexcom numbers are calibrated daily with our pdm finger checks.
     
  8. forHisglory

    forHisglory Approved members

    Joined:
    Jan 26, 2015
    Messages:
    382
    We missed our appt today because of Jonas, but our Dex software predicts things pretty accurately, usually. Just curious if you use any software to upload data? Basing the A1C on the PDM vs the Dex...would basing your prediction on the avg glu from the Dex be more accurate since it takes more readings? I don't get why more recent readings in the last week would bump it up either (other posts)? I thought it was based on the lifespan of the RBCs? Anyhow, it stinks...no matter what people say about it "not being a report card"...it always feels like one anyhow.
     
  9. StacyMM

    StacyMM Approved members

    Joined:
    Oct 22, 2010
    Messages:
    1,039
    Two thoughts. First, how did they test the accuracy? A blood draw done the same time you did a finger stick? Just a test with the control solution? Compared to an office meter? Second, it's not actually an average. It's much more science-y than that, unfortunately. And I'm as far from a scientist as it's possible to be, so don't assume my explanation is right! Here goes, though... Red blood cells typically last about 3 months so glucose stuck to the cells during a three month period would be measured as a percent at the end of that time. In my head, I think of it as multiple ways to get the same average... let's say you have someone with a standard deviation of 40 and an average of 120, their a1c would be different than someone with a standard deviation of 80 and an average of 120. And a third person has a high month in month 1, then brings it down in month two, and tehn runs low in month three, getting a 3 month average of 120. And a fourth person has two great months, then gets hit with an illness and has a crazy third month, ending with a 3 month average of 120. All have the same average but very different situations/histories so glycation (glucose stuck to those red blood cells) would be different. For me, an a1c test after crazy highs seems to result in a higher a1c than one after some steady months.

    Someone mentioned that there are people that the a1c just isn't accurate for. My daughter is having that issue and we haven't figured out the cause. Her a1cs are in the 8.5 and up range. Last one was something over 9. Fructosamine suggested an a1c around 7.0. Dexcom average at that time suggested something like 6.3. If you are curious, here is a link I like: http://www.healthline.com/diabetesmine/ask-dmine-follies-of-the-a1c-test#4
     
  10. DavidN

    DavidN Approved members

    Joined:
    Sep 7, 2012
    Messages:
    744
    Thanks everyone for your feedback. It's nice to have this group.
     
  11. DavidN

    DavidN Approved members

    Joined:
    Sep 7, 2012
    Messages:
    744
    Oops. That was a typo. I meant to type CGM instead of PDM. The CGM has historically been a very good predictor of A1C's but not this time. Thanks.
     
  12. DavidN

    DavidN Approved members

    Joined:
    Sep 7, 2012
    Messages:
    744
    To test accuracy they tested BG from the same drop of blood using their meter vs our PDM meter. It was very close. I was concerned our meter would be much lower which might explain the CGM / A1C discrepancy.

    Now that you mention your son's A1C discrepancy I remember you posting something about that before. Thanks for the info link. I will dig into that now.
     

Share This Page

- advertisement -

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice