- advertisement -

Dexcom G4: When to Calibrate/When Not to Calibrate

Discussion in 'Parents of Children with Type 1' started by rgcainmd, Jul 22, 2014.

  1. rgcainmd

    rgcainmd Approved members

    Joined:
    Feb 6, 2014
    Messages:
    1,379
    I've heard a lot of anecdotal evidence surrounding optimal times to calibrate, how often to calibrate, and when not to calibrate your child's Dexcom. A lot of the information I came across was conflictual, so I called Dexcom and was connected with one of their CDEs who, IMO, sounded like she knew what she was talking about; she did not sound like she was reading from a pre-scripted "Answers to FAQ" sheet. Following is a summary of the information she gave me. I make no guarantees regarding the accuracy of this information, so proceed with caution and with the knowledge that YDMV.


    A. This is when you should calibrate:

    1. When you are prompted to calibrate by either:
    a. Two red blood drops that appear two hours after starting (or restarting) a sensor OR
    b. One red blood drop (these prompts occur once 12 hours have elapsed since your most recent calibration)

    2. Whenever a meter reading is greater than or equal to 80 mg/dL AND the Dexcom reading is greater than 20% higher than the meter reading or more than 20% lower than the meter reading.

    3. Whenever a meter reading is less than or equal to 79 mg/dL AND the Dexcom reading is more than 20 mg/dL (not 20%) higher than the meter reading or more than 20 mg/dL (not 20%) lower than the meter reading.

    Caveat: Do not calibrate even if prompted to do so by a blood drop if the trend arrow is pointing either straight upwards or straight downwards (one or more arrows). If this is the case, WAIT until the directional arrow is either horizontal, trending slightly upward, or trending slightly downward before calibrating. Be patient; you will not be "timed out" if you do not calibrate immediately after a blood drop icon appears.


    B. This is when you should not calibrate:

    1. If BG (via meter) is higher than 400mg/dL.

    2. If BG (via meter) is lower than 40 mg/dL.

    3. If any of the following appear on the Dexcom screen:
    a. question marks
    b. an hourglass
    c. the antenna icon (receiver is out of range of transmitter)
    d. one or more trend arrows that point either straight up or straight down (horizontal arrow or arrows trending slightly upwards or slightly downwards are O.K.)
    e. no trend arrow(s)


    REMEMBER: More is NOT better when it comes to calibrating the Dexcom. Calibrating more often than when instructed to do so by blood drop icons or when the Dexcom reading is not inaccurate ("inaccurate" as defined by A.2. and A.3. above) will not improve the accuracy of your Dexcom and will likely worsen the accuracy of your Dexcom readings.


    Anything you'd like to add to this is appreciated, but please do not make negative comments directed at me regarding the accuracy of the information I've posted; as I've said, this information was given to me by a Dexcom CDE who sounded to me as if she had a decent brain between her ears. Some, but not all, of this information is available on the Dexcom website: http://www.dexcom.com/sites/dexcom.com/files/dexcom-beginning-your-journey/story.html
     
    Last edited: Jul 22, 2014
  2. hawkeyegirl

    hawkeyegirl Approved members

    Joined:
    Nov 15, 2007
    Messages:
    13,157
    I would say that all of that tracks well with our experience.
     
  3. mamattorney

    mamattorney Approved members

    Joined:
    Apr 9, 2013
    Messages:
    1,076
    Just as clarification of what Dexcom told me - #2 and #3 trump the don't calibrate with a straight up/down arrow. If it's off, it's off and needs to be calibrated regardless of arrow status. At least that's what I was told.
     
  4. rgcainmd

    rgcainmd Approved members

    Joined:
    Feb 6, 2014
    Messages:
    1,379
    I could be wrong, but my understanding of what the CDE told me was to not calibrate any time the screen shows straight up or straight down arrows regardless of whether the Dexcom reading is inaccurate. I believe the reasoning behind this is that straight up or straight down arrows indicate a period of relatively rapidly changing Dexcom readings (and not necessarily rapidly changing actual BG because of the time lag between Dexcom readings and what is occurring in real life.) Calibrating at these times throws off the calibration algorithm built into the Dexcom system. As I said, I could have misunderstood this part of what she told me. But if memory serves, I recall asking her if I should ever calibrate when arrows are straight up or straight down, and she said "no." If I have time tomorrow, I'll call Dexcom again and ask for clarification on this point. (It will be interesting to see if a different CDE gives me different information. So I might end up calling a third time to see if their answers vary.)
     
  5. caspi

    caspi Approved members

    Joined:
    Oct 11, 2006
    Messages:
    5,134
    This has been my understanding as well. Calibrating when arrows are up or down is useless. We just test more often, if necessary, while we wait until it's "calmed down" as I like to call it :wink: . Then we calibrate.
     
  6. sszyszkiewicz

    sszyszkiewicz Approved members

    Joined:
    Dec 24, 2013
    Messages:
    842
    It was nice of you to call and share what they said.

    Around here if it allows us to calibrate when we have fingerstick data we calibrate, mostly so we have a record of the fingerstick, and then we don't worry when its been 12 hours. In 4 months this has not made any difference whatsoever (as far as we know.).

    I am looking forward to the day where it has your number without calibrating fingersticks!
     
  7. rgcainmd

    rgcainmd Approved members

    Joined:
    Feb 6, 2014
    Messages:
    1,379
    "Calmed down"--we use the same terminology! The Dexcom has been such an awesome game-changer that we have anthropomorphized it; some people even name their Dexcom. (I wanted to name our Dexcom, but my daughter thought doing so would be stupid. This didn't stop me from bugging her until she begrudgingly named it "Dex".)
     
  8. caspi

    caspi Approved members

    Joined:
    Oct 11, 2006
    Messages:
    5,134
    We're very original -- the PDM is called "Piddum" and the CGM is "Ciggum". :wink:
     
  9. Jordansmom

    Jordansmom Approved members

    Joined:
    Aug 23, 2008
    Messages:
    2,172
    That's the fastest way to mess with accuracy. At those times when you have a fast moving bg, most often the Dex isn't "wrong". The amount of glucose in the interstitial fluid is just different than the blood glucose. The readings are going to be different until the rate of change slows. The Dexcom algorithm accounts for the lag, but sometimes just doesn't keep up.

    I'd like to add to the list that there are times when the Dex has been really accurate and then suddenly you get a bg reading that is really off. Meters can give inaccurate readings. The Dex can be right and the meter wrong. Sometimes its smarter to trust the Dexcom and not put the meter reading in. That becomes an intuitive thing after years of Dexcom use, just like any other diabetes care adjustment you make just because you know better.
     
  10. Lori_Gaines

    Lori_Gaines Approved members

    Joined:
    Jul 10, 2013
    Messages:
    209
    Ditto about never calibrating when there are straight up or down arrows. Also, we don't always wait for the twelve hour blood drop. If it is time to go to bed, we just go ahead and calibrate rather than stay up later waiting.
     
  11. swellman

    swellman Approved members

    Joined:
    Jul 30, 2008
    Messages:
    3,544
    The only part of Dexcom's response in the OP that bothers me is that "more is not better". As a chemist who has calibrated countless instruments countless times, more is better providing it's an accurate and representative data point. Maybe they meant to say "more is not necessarily better" to which I would agree 100%. However, if your data point is not violating the caveats mentioned in the OP entering it shouldn't negatively affect a calibration.
     
  12. Ali

    Ali Approved members

    Joined:
    Aug 1, 2006
    Messages:
    2,223
    Agree, the programs just are not sophisticated enough yet to adjust to all the variables. I actually think the underlying programming of the different CGMS programs(Med & Dex) may work better for some people over others based on their own BG/insulin/food responses, I hope everyone follows me :). I just know reading this and other sites that the programs seem to correlate spot on for some individuals and not as well for others, I am one of those who loves my CGMS but it is off plenty of times, while others get much more accuracy. It is still miles ahead of individual finger pokes but still a bit of work ): for many of us.
    Ali
     
  13. jenm999

    jenm999 Approved members

    Joined:
    Apr 30, 2014
    Messages:
    855
    THANK YOU.

    Also, I don't think the Dexcom measures interstitial glucose. My understanding is that it EXTRAPOLATES blood glucose based on calibrations by measuring the conductivity in the interstitial fluid.

    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2903977/

    Current continuous glucose monitoring systems have the advantage of direct insertion of electrochemical sensors into the IF space rather than transporting the sampled fluid outside the body to detect glucose concentrations. Software programs have been designed to accommodate the lag in IF glucose readings. Despite the advances in the making of sensors with new and improved designs and materials, sensor insertion causes trauma to the insertion site. It can disrupt the tissue structure, provoking an inflammatory reaction that can consume glucose followed by a repair process.38–40 The interaction of the sensor with the traumatized microenvironment warrants the need for a waiting period for the sensor signal to stabilize, and that period varies depending on the sensor type.21

    Because continuous glucose sensor manufacturing has not progressed to the accuracy and precision of blood glucose meter strips, sensor glucose signals must be calibrated against corresponding blood glucose meter levels. Such calibrations transforms the sensor signal into a glucose value and assumes that the plasma-to-IF glucose gradient remains relatively constant.41 This assumption will not be valid if sensors are calibrated during rapid changes in plasma glucose, which is a major source of sensor error. The effect of sensor lag on performance is most obviously seen during periods of rapid glucose rate of change (either up or down). Sensor levels may trail glucose levels by 5–10 min during periods of rapid change, but the most important effect on lag is to introduce error during calibration, which affects long-term sensor performance. Moreover, changes in plasma–interstitial gradient in certain physiological conditions, like insulin-induced hypoglycemia, may be misinterpreted as sensor inaccuracy.41
     
  14. cm4kelly

    cm4kelly Approved members

    Joined:
    Apr 28, 2011
    Messages:
    547
    Interesting piece of information to think about -

    I listened to a guest speaker who has been in trial with artificial pancreas system and who wears a dexcom.

    After having multiple blood draws, he stated that his DEXCOM was more accurate with blood draw results than the blood glucose machine. INTERESTING - I bet we all think that our METERS are correct - but it could be the other way around.

    Food for thought! HAve a great day!
     
  15. mamattorney

    mamattorney Approved members

    Joined:
    Apr 9, 2013
    Messages:
    1,076
    Hmm, well it seems it's everyone vs. me on the don't calibrate when there's a straight up/down arrow. I was told that by a dexcom rep when I called in once for a crappy sensor. I don't even know if the arrow was straight up or not, but I remember saying "I thought I wasn't supposed to calibrate when there is a straight arrow" and he said - not when the blood drop appears, but when it's off, it's off and you should calibrate every 15 minutes up to three times in a row to get it back on track. If it's still off after three 15 minute calibrations, I was supposed to call and they would replace the sensor.

    I think I'll try not touching it with the straight up/down arrows and see what happens. Admittedly the main time I do that is during the first overnight when sometimes the sensor is just wildly inaccurate - trending up, up, up usually when her bg is steady based upon fingerprick measurements.
     
  16. jenm999

    jenm999 Approved members

    Joined:
    Apr 30, 2014
    Messages:
    855
    We calibrate first thing in the morning - he's usually steady from 4 am until he eats (weirdly - thanks, honeymoon) and again before bed after the dinner climb has stabilized. Rarely do we even see the prompt because it's just become habit. I also tend to enter the BG when it's right on just to confirm to the algorithm that it's correct.

    One thing that drives me nuts is the "confirm with a fingerstick" thing - and our CDE even prefers to see meter readouts when recommending carb ratio changes rather than the beautiful Dexcom reports. But given that BG meters are +/- 20% I don't see why that makes any sense. We treat lows off the Dex all the time. We never give insulin off the Dex alone, but I think of the fingersticks as more the corroborating data point rather than the holy grail.
     
  17. nebby3

    nebby3 Approved members

    Joined:
    Jun 5, 2007
    Messages:
    923
    I'm really glad to hear other people say they trust Dex more than the meter sometimes. I have thought this many times. I will do a second finger stick but even if it agrees with the first I sometimes wonder. It doesn't help that some of our meters are old. Our CDE bugs us about always checking with a meter too but for the most part Dexcom seems accurate for my dd and especially if I can see that there hasn't been a huge jump in bg I trust it. The one exception would be in coming up from lows when it lags behind significantly. Does anyone calibrate at those times of just wait for it to catch up?
     
  18. jenm999

    jenm999 Approved members

    Joined:
    Apr 30, 2014
    Messages:
    855
    I just wait for it to catch up. That can be hard when my son is feeling cruddy but has already had 15g and I know if I give him more he'll spike. I try then to give him water or something no-carb so he can get the psychological satisfaction of eating, even if it won't affect his BG. He's learning to trust me and trust the Dex.
     
  19. caspi

    caspi Approved members

    Joined:
    Oct 11, 2006
    Messages:
    5,134
    This is what we do as well.
     
  20. swellman

    swellman Approved members

    Joined:
    Jul 30, 2008
    Messages:
    3,544
    The sensors do, in fact, directly measure interstitial glucose by using a bio-electrical sensor wire using an enzyme that produces an electric current in the presence of glucose presumably in a predictable manner over a wide range of glucose concentrations. However, there are a crap ton of things going on at the insertion area that can affect the sensor's ability to perform as you pointed out. That AND the fact that blood plasma levels are not equivalent to interstitial fluid levels. The glucose just sorta makes it's way from the plasma to the interstitial fluid by diffusion. There are a bunch of other factors that affect the lag time between plasma changes and IF levels. The receivers are, however, programmed or modeled to account for these factors although not individually, because it has no idea what's going on in there, but as an average. This can explain why rapid rises and falls in plasma glucose levels causes predictability issues and why calibrating during these times introduces data that skews the model.

    At least this is how I understand it.
     

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