CGM Accuracy

I just started using the CGM, and up until this morning I haven’t had any problems. Last night before bed, I checked my sugar with my meter it read 124 and my CGM read 117 (no big difference), but this morning when I checked my sugar my meter read 333 and my CGM read 164 with no little arrow pointing up. I am curious to find out if the longer I where the CGM the less accurate it is. What I mean is this morning is day 4. The day I change my sensor. Did I leave my sensor in for too long? Should I have taken out my sensor earlier? AHH…so many questions. Help Please :slight_smile:

This is a post I started last week read through and see what you think. I use my sensors for about 14 to 21 days. I wouldn’t worry about leaving it in to long.

I use the Medtronic CGMS and I regularly upload my data to CareLink. Several times a week I will examine my data that has been collected to see where I can improve. Currently that would be around 10 pm at night when my BG seems to always be going up!

Many people have written about the pros and cons of meters and which is more accurate when compared to CGMS. While examining my data I found this to be very interesting. I ran a report that covers 8 weeks. The meter average BG was 129 and the CGMS was 123 a difference of only 6 points. Over those 8 weeks I used the meter 323 times vs. CGMS 13,569. Daily average was 5.8 sticks vs. CGMS 266.1. As I mentioned this was for 8 weeks, then I ran a report for 12 weeks and this is the data that was recorded. Meter 132, CGMS 124, 523–8 points different, meter sticks vs. 19,729 CGMS, 6.2 sticks a day vs. CGMS 266.6.

While using CGMS, it does a fairly good job at the current moment but when looking over a long period of time I would say that it is even more accurate and really can show trends and where improvement may be needed.

I would be curious to see if others who use CGMS are finding similar results to what I have found. In fact my last A1c was 6.0 which correlates to a average daily BG of 124, this matches the CGMS almost perfectly. Looking forward to other peoples experiences.

I have to agree with your data so far. This morning is the only real big difference I’ve noticed. I will have to keep you updated on my progress with the CGM. I wonder if other people leave their sensor in as long as you do? If so, I will too. I am not complaining or anything, but it would be nice to leave the sensor in longer when you already have to change your insertion site every 3 or 4 days.

On day 8 you will get a message that says weak signal. When that happens I remove the transmitter and I charge it. I learned this from someone post hear on tudiabetes. If you do a search you will find that other wear their sensors for at least 6 days if not longer.

When I get whacked out numbers like you are talking about I just put a calibration into the pump. Sometimes I will get a Cal Error and then it will as me to enter my BG again. After I do that it usually will reset it. Sometimes I just leave it alone and it will eventually right itself after a couple of sensor calibrations.

i usually leave mine for about 4-5 days average…i dunno i havent tried putting it on for long, coz i get pissed when it started goin crazy bout the readings… so i just take it off and let a day or two pass then start a new sensor again…but im gonna try what David said, leave it as it is and let it right itself after calibrations…

yeah one thing i like bout the CGMS is you’ll be able to see trends and patterns as far as blood glucose levels are concerned…if you can remember David, the first time i inserted the sensor myself i was profusely sweating coz of the needle size, but after all, it doesnt hurt that bad… and until now, i still insert it manually…havent used my sen-serter and i dunno if i ever will…

David, how often do you upload your data from meter and CGMS to the carelink…?do you have like a schedule (every 2 weeks or 4 weeks?) i usually do it every two weeks but i havent done it i guess for almost a month now… i keep on forgetting… my bad…

also, what do you use to secure your transmiter/sensor? do you use IV dressings or tape of any kind? i need to try a new one coz this IV 3000 im using right now i guess im allergic to it…



I use the sen-serter although I always look up when I push the button. I have gotten past the size of the needle for now!! As far as uploading data to CareLink I usually try to do it weekly, sometimes I forget but usually I am pretty good about it. I use the IV-3000 tape which so far I am not allergic to, it seem to work really well unless you get it wet for and extended period of time.

Diana: I don’t think you left it in too long. As others note, the sensor can stay in much longer. I always get situations where the meter and the CGMS are different. The longer the sensor stays in, in my experience, the more accurate it gets. I woke up this morning with the CGMS saying I was at 138 and the meter at 277. They have not been within 50 points yet today. But, over time, the sensor will show trends, and that is where its value comes in – when you see the graphs showing the sensor readings and the meter readings, they seem to be moving in tandem, even if they are not closely aligned.
I have never done an analysis like David has, but, even with the sensor, I still test 10+ times a day, and all of the information together gives a pretty good read as to how I’m doing.
One of the things to work with your CDE on is to not get obsessed with the incredible amount of data that you now have (and I’m not one to be talking about being obsessed with the data). I used to focus on the sensor every few minutes. Now, I just check when I’m testing and when I’m eating. I upload/download the information about every week and mostly look at trends to see what parts of the day need attention.
I’m not yet used to the needle (after about a year), but I manage.
A question for the group at large: when you leave the sensor in for a long time (7 days or more), do you find that the skin around the sensor site takes a long time to heal? My thighs seem like they are covered with “zits” from where the sensor has been inserted.

I find that if you can keep the sensor from moving a lot and have it well taped it heals faster and and goes away quicker.


Did you double check that high reading? Nothing on your finger, meter coded to strip, not a bad strip, that kind of thing?

Nope, I didn’t check my reading again. I probably should have, since one time I treated a low blood sugar with a spoonful of honey, and after I waited 15 minutes to check my sugar again my meter read 560. It turns out I had some honey left on my finger. My sugar was actually 87.
Thankfully, my meter and pump seem to be running similar numbers again. I think I will try David’s method and calibrate my machine when I get a weak signal.

My ways to do this:
*Do the original 3 day. At the end of the 3 days, if the site is not irritated or such, I ‘lie’ to the pump, saying ‘new sensor’.

After that 3 days, I lift the dressing off, keeping my finger on the sensor to keep it in. Disconnect the transmitter for a few seconds… maybe a minute, and re-connect it, starting ‘new sensor’. I’m on Day 8 and as these are VERY EXPENSIVE, I do all I can to get many days out of it. Thing is to watch the isig. If that goes below 9, you could start to lose accuracy. Mine today is 21.9 Very good actually.

Thing to remember: the sensor reading is from five minutes before your meter reading, so I wouldn’t expect a perfect pair in both readings.

As I mentioned above on day 8 the sensor starts acting squirrely and starts having weak signal warnings. At this point I disconnect and recharge it, plug it back in and have it relocate the lost sensor. Works like a charm. Personally I think it is totally normal for the BG values to be way off at times this is just how CGM works and one of the things that will improve as the technology gets better and more advanced.


how long does your charging takes? i mean on that day 8…? so while charging the transmitter, what do you do with your sensor to keep it in place…?

I have a thin piece of IV tape that helps hold the sensor in place. I carefully remove the transmitter and charge it. It take about 10 minutes to charge. Then I put it back on and have it look for a lot sensor. That’s it!!

If you were laying on the sensor site, that can cause it to read lower than the actual level. I dont know why, but that is a possibility. I don’t think its from leaving the sensor in too long, I usually leave my sensor in for 7-10 days.

I see TV ads offering a free meter. Of course, they want to sill the testing supplies. This one ad says their new meters are “more accurate.” Know this, accurate is an absolute. There is no such thing as more accurate. There is no such thing a less accurate. There is less that accurate, but there can’t be more that accurate. So, these companies selling “more accurate meters” should be saying, closer to accurate-- and admit that their previous meters were not accurate.