- advertisement -

How do CGM's work?

Discussion in 'Parents of Children with Type 1' started by DavidN, Mar 25, 2013.

  1. DavidN

    DavidN Approved members

    Joined:
    Sep 7, 2012
    Messages:
    744
    I know the BG readings are "relative" to the inputted readings (calibrations), but am wondering if anyone knows more than that. For instance, are BG readings relative to the previous reading (5 minutes ago in the case of Dexcom), or are they all relative to the most recent calibration? Said another way, does being "out of range" for any period of time compromise CGM accuracy? My son was out of range for 2 hours the other day. When he came back in range, I assumed the number would be way off. But I pricked his finger and the CGM was spot on. Was that just a coincidence? If relative BG readings are based off of the most recent calibration, then it was not a coincidence. If they're based off of the most recent reading, then it was. Hope that makes sense. Just wondering if anyone has any insights. Thanks.
     
  2. hawkeyegirl

    hawkeyegirl Approved members

    Joined:
    Nov 15, 2007
    Messages:
    13,157
    I only really have knowledge of how MM works, but I believe they use a weighted average of the 4 most recent calibrations to arrive at SG (sensor glucose). They give more weight to the most recent calibrations.
     
  3. Sarah Maddie's Mom

    Sarah Maddie's Mom Approved members

    Joined:
    Sep 23, 2007
    Messages:
    12,521
    As I understand it, when out of range the sensor is still doing it's thing but it isn't storing, or transmitting the data. Those readings, made while the receiver is out of range are tracking along as they were prior to the range disruption. When the transmitter comes back into range and begins to transmit to the receiver, the data gathering just resumes and your accuracy will then reflect the relative accuracy that was prior to the disruption of signal. LOL, does that add anything to your thinking??
     
  4. Megnyc

    Megnyc Approved members

    Joined:
    Nov 8, 2012
    Messages:
    1,373
    I never managed to figure out how Dexcom does it but for Medtronic they are relative only to the calibrations. It is a straight conversion from ISIG to BG. So lets say the conversion factor is 10. If ISIG is 15 then BG would be 150. If ISIG then drops to 12 at the point at which the transmitter sends the signal then BG would be 120. The previous BG has no impact on the new BG.

    I am convinced (and someone please correct me if this is totally wrong or impossible) that Dexcom uses an algorithm that does account for rate of change. I found that Dexcom often "over-predicts" highs and lows. So for example if it sees the ISIG (or whatever they call the signal) goes from 15 to 20 and it was previously holding steady it assumes that BG is spiking and gives a BG that is higher then the actual straight conversion. This comes from my observation that Dexcom would often catch a high or low about 5 minutes before it happened based on the meter BG. I am not sure how well I am explaining this. Here is an example: Dexcom alarms high and says I am 220. I test and am 160. I test again in 5 minutes and meter BG is NOW 220. That is why I think the Dexcom uses some sort of algorithm to predict actual BG values rather then give the delayed BG from interstitial fluid. In this case the previous BG values (or rather rate of change of ISIG) DO influence the reading. In your case coming in from out of range it would have just used the current ISIG and whatever conversion factor it obtained from the last cal in order to get an accurate reading. So that works fine if BG isn't changing rapidly.
     
  5. DavidN

    DavidN Approved members

    Joined:
    Sep 7, 2012
    Messages:
    744
    I think you are saying that being out of range has no ill-affects on accuracy. No I will obviously not get readings when out of range, but once in range the accuracy has not been compromised because the sensor was still functioning/tracking properly. Is that right? Hulk head hurt. :cwds:
     
  6. DavidN

    DavidN Approved members

    Joined:
    Sep 7, 2012
    Messages:
    744
    I had always thought that Dexcom is trying to display a reading that is 10 minutes or so behind the actual BG reading. I say this based solely on our experience. If he is trending down and I re-calibrate (I know we should aim to calibrate only when he's flat but it's not always practical) then the new reading will always be above the actual BG reading. And the opposite holds if he's trending up - the new number will be below actual BG. If this is true then Dexcom is not trying to be predictive ... but I have no idea. This is another good question. So standing questions are ...

    1 - Does Dexcom give relative reading based on the last reading or recent trend? Or are readings based off of the most recent (or last several) calibration(s).

    2 - Is Dexcom attempting to be predictive in its readings?
     
  7. 3kidlets

    3kidlets Approved members

    Joined:
    Aug 3, 2010
    Messages:
    819
    Not sure on the explanation, but wanted to share that last week we were away on vacation. Hana would be away from the receiver for least30 minutes at a time because she was going down water slides. When she came back to where we were sitting and the receiver started reading again, every time the readings were spot on with the finger stick. I thought we'd see crazy readings on the g4, but not so.
     
  8. Megnyc

    Megnyc Approved members

    Joined:
    Nov 8, 2012
    Messages:
    1,373
    I don't have answers to these. But to rephrase. I think the dexcom is attempting to use CURRENT interstitial fluid glucose to predict CURRENT blood glucose. So it is not predicting in the sense that it is trying to determine future BG but it is trying to go from interstitial glucose to blood glucose more accurately by using some sort of algorithm to try to bypass the delay between blood and interstitial fluid.

    Do note that you probably have more dexcom experience then I do at this point. I was somewhat of a dex failure due to allergies :cwds:
     
  9. Sarah Maddie's Mom

    Sarah Maddie's Mom Approved members

    Joined:
    Sep 23, 2007
    Messages:
    12,521
    That's my understanding... and I had to stop and think it through after our first prolonged out of range situation. It's sort of obvious, but at the time I assumed that being out of transmission range somehow meant ... I don't know, "out of order" :p Silly, but true.:rolleyes:
     
  10. Tigerlilly's mom

    Tigerlilly's mom Approved members

    Joined:
    Dec 3, 2007
    Messages:
    3,492
    This is what I have found also. Tyler does not want to bring his receiver on the ice during hockey games, knowing that it will be alarming that he is high (due to adrenaline highs), so he will leave it with me before he goes into the rink...over an hour and a half later once he is back in the car, I will have him test with his meter, because I don't have the patience to wait for the Dexcom to pick the signal back up again, and once the signal is picked up, it is surprisingly accurate compared to the meter.

    (One game I did stand close to the boards to see if it would be enough to pick up the signal, but sadly it was not.)
     

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