Wild CGM readings

I have been on MM cgm for almost a year and so far haven't had any major issues. But since i started a new sensor yesterday night, my CGM has been giving me wild readings, constantly showing double arrows both up and down (even during the night when my blood sugars are usually stable). Two separate glucometers show stable BGs. I've had bad readings before but I was able to calibrate it correctly in a couple of hours and i could still rely on the arrow trends being accurate.

Has anyone had this problem? should I wait it out and continue recalibrating? or should i just get a new one?

I've gotten these types of wild readings if the sensor hits a vein or artery. Also got similar problems when the sensor had been on a shelf for more than 5 months before using it. (I got 3 months of supplies when I started, and used the sensors for 7 days, more than twice the 3 days recommended)

I would call MM. They'll help you troubleshoot it, and if it proves to be a bad sensor, they will replace it at no charge.

I had a similar experience when I took a shower with the sensor on my body, but without the transmitter. Apparently, water is not a good thing when it gets on the metal contacts on the sensor. The lesson is to always wear the transmitter when in the water or shower!


Do you have a lot of swelling by the site? Swelling can do odd things too.

thanks everyone for the advice, just as I was getting ready to give up and call MM, the sensor started acting "perfectly", giving me readings with 5-10 points difference from the glucometer. I don't know what i did "right", I guess sometimes you just have to wait it out.

i didn't have any swelling or bruising.

i also use the sensors for 7 days. Both my CDE and endo said it's ok as long as everything heals properly and i don't have scar tissue. The only reason i change the sensors is because the transmitter runs out of battery and has to be charged (although I always wonder if I could recharge it and reattach to the old sensor, if i could make it work even longer).

My cgm readings are often up to 100 points off - which is dangerous when my blood glucose is 40 and the cgm reading says 144! That happened yesterday - fortunately, I don't put my car in gear without checking the blood glucose, and I drank some juice and waited until that hit my system and I was safe to drive. But it is so rarely within 40 points of my blood glucose that I am thinking that I would never want to tie my dosage to the cgm readings. I have been using this device for over a month now, and I am really disappointed in the level of inaccuracy. 20 -40 points off would be manageable - 100 points off is not.

Rather than decreasing my need for the number of fingersticks in a day, it has actually increased it, which gets expensive since I always exceed my prescribed number of strips in any given month, and my insurance company only pays for the prescribed number.

I have talked to the teaching nurse in my endocrinologist's office and I am careful not to use a site which bleeds when I insert the sensor. Am I doing something else wrong, or is my body just not a good match for this particular technology?

I hav found that the more stable one’s blood sugar readings are, the more accurate the CGM readings are. For example, when I go to bed, almost all of my Bolus insulin is gone. With my basal rate set properly,I seldoM have lows in my sleep.

Dick, I am glad for you, but if my blood sugar readings were 'stable', I would never have gotten a CGM. It sounds, from what you say, that these are only accurate for the people who need them least. That is very sad.

My blood sugar readings were unstable for many years, probably like yours or
worse. I have worked out a way to have stable blood sugar readings most of
the time. There are a number of things that I have found that makes this
possible and it does work. Think about this; If a person has their basal rate
set properly and eats no food, they will experience no highs or lows in a 24 hour period. So, a high protein diet with a minimum of carb works best This minimizes bolus thereby resulting in more stability. Also, I know for sure how long my bolus lasts in hours so that I can time my dinner bolus so that when I go to bed there is only a tiny amount left. That makes it so that I am on my basal rate throughout the night and I rarely have lows.

Blood sugar stability is an ongoing problem and getting it stabilized requires desire and self discipline
Thanks to the pump and CGM and the test result printouts, one can stabilize their blood sugar.

I know this post is from awhile ago but did you ever figure out the problem. I am having the opposite problem where it wakes me up all night that I am low when my BG is fine. I also am having it be 100 points off. It also never lasts me for 6 days.

Nancy, I am frequently off by 100 points in either direction - I too get notifications for lows that are not really low. But I am more worried when I get readings that say I am ok and I am really very low - that is dangerous if I am driving, for me and for everyone else on the road. I also get "Sensor End" after a day or 2 at times, even though the sensor is supposed to last 3 days. This short life happens even on "good" spells, where the blood sugars are relatively stable and the range is not extreme and the CGM (welcome surprise) is relatively close to accurate. I just don't trust the CGM. I rely on finger sticks to manage my blood sugars. The CGM even shows sugars going in the wrong direction - double up arrows when it is actually going low and vice versa. Fortunately my endo was able to increase my prescription for test strips - otherwise, I would be having trouble affording this.

Part of the confusion which can follow resuming a discussion thread after it has been idle for a long time is not being sure which device is being discussed.

Since it was posted in the middle of 2013 I assume that the Original Poster (OP) was using the older Medtronic Sof-sensor (aka "Harpoon") CGM sensors. While the Enlite was available outside the US in 2011, the it was not (I think) available inside the US in 2013. I think it was not until October 2014 that the Enlite was introduced together with the Minimed 530G pump.

So ... I assume the OP was talking about problems with the Harpoons & the newer questions pertain to the Enlite. If someone is actually talking about problems with a Dexcom CGM ... well, heck, we all know that could just never happen, right? ;-)

-irrational John
T1 LADA ~1979; 1st pump: Minimed 507 (~6/1996); Paradigm 723 (4/2013) + CGM (12/2014)

I have only been using Medtronic CGM for about 2 1/2 months. (I started at end of Nov 2014). I am still frequently tweaking my theories of how to get the longest & most accurate use from Medtronic CGM. But I've listed below what I think I know at the moment.

While I'm thinking primarily of the Enlite, I believe some of the suggestions would also be pertinent when using Harpoons. (If anyone other than myself is still doing that ...)

  • Refresh your understanding of how to insert. I personally prefer the following YouTube videos:
    Get Right With Enlite (Medtronic Diabetes ANZ)
    How To Use An Enlite Sensor

    The first link is to a YouTube playlist I created containing 7 videos produced by Medtronic Diabetes Australasia (ANZ). These were recently posted and they are IMO unquestionably superior to the "meh" demos I have seen from Medtronic Diabetes US. What makes them better IMO is that they actually describe specific mistakes which can occur while inserting so you can know to avoid them.

    If you only have time to look at one or two ... and frankly they are still really tedious ... try looking at Chapter 4 Inserting the Sensor. After that possible also look at either Chapter 3 Loading the Serter and/or Chapter 5 Taping.
  • My guess of the moment is that many of the sensor accuracy problems arise from the sensor probe shifting back & forth in the insertion site. I think one needs to try to address this when applying the Enlite overtape.
    If you watch them, notice that in both of the demo videos (above) the presenter spends a lot of time on making sure the sensor is held snugly in place.
  • A Medtronic customer service rep I once spoke with used the term "pistoning" to describe this shifting. The sensor probe moves slightly up & down in the insertion site. This can lead to wide swings in the ISIG and, thus, big changes in the SG (Sensor Glucose) readings the pump reports. If the probe shifts far enough it can result in the transmitter "resetting" and your pump reporting some sort of sensor error or alert.
  • In an attempt to more consistently extend my sensor life to at least six days I am currently experimenting with applying small amounts of Mastisol or Skin Tac very carefully along the outside edges of the tape of my Harpoons before I insert them.

    If you try something like this remember to avoid the sensor probe.

    And, obviously take extra care if you try this with an Enlite since the protective covering over the introducer needle is automatically removed as part of preparing the Enlite inserter.
  • I'm only on the fourth day of my first attempt at "sticking" the sensor in place so it's too early to be postive if this helped. So far things have been "OK" with this sensor, but that could also simply be because I'm using a "better" insertion site. But my current gut feeling is that if you want to get both accuracy & a longer life from your Medtronic CGM sensor, you need to make sure it stays put.

In sum, the above is what I'm currently thinking about. Your mileage may always vary. If you do decide to try anything I mentioned ... or you already know it's wrong ... please reply and let me know.

-irrational John
T1 LADA ~1979; 1st pump: Minimed 507 (~6/1996); Paradigm 723 (4/2013) + CGM (12/2014)

I also make sure it is taped securely. Last one I used was perfect. This one at 22 hours was telling me I was low and I was not for a day and half. I always divide my BG by ISIG to make sure calibration is in zone of 3-8.33. First day usually okay, second day can never get a quality calibration and I turn sensor off. Next day sometimes I am able to calibrate after reconnecting old sensor. Perplexed by the inconsistency.

I don't know what to suggest other than to call Medtronic. As far as I know the FDA is still monitoring their log of help calls to see how well the sensors are working. I guess that might give us some "leverage".

I always thought I was taping securing as well ... which is why I'm now trying the mastisol. Also, I think even if I did have the sensor snug when I first inserted, it would later work loose over time as I wore it.

It might also be related to the location of the insertion site. But that's a very personal thing because the details can vary so much from one person to another.

My current sensor is on my left side "love handle" area. I decided to try it because I guessed I might flex/bend less often in that area.

Lots of guessing ... and, yes, it gets very frustrating at times. Still, I think I'm making progress.


Thanks !