My dexcom is pretty good overall. However in the middle of the night I have had 3 consecutive readings of like 50… So I test and I am like 100. Most of the day it is pretty accurate except at night. Anyone else have this phenomenon. It’s happened with 2 different sensors
Strange, No my readings have read 50 and I’ve been 40! Ouch! 
I’ve had that happen a couple times as well, but very sporatic. Nothing like getting woke up in the middle of the night, go check and find out it’s off by 40 points! I don’t know if I have an answer to why or how to correct, but I can sympathize.
That is a little strange as well to me, I have had some bizarre readings but they came at any and all times. I dc’d using the Dex b/c of false alarms too many nghts in a row, (My wife and I were getting NO sleep). It happened for almost an entire week EVERY night, I would get a LOW alarm and my bg was about 150 (my LOW setting was 80) and I often got readinngs during the day that were at least 100 points off. (just too much error there!)(I did call dex, and they told me the only reason one would get such bizarre errors was the use of tylenol, or any drug with acetominaphin.)(which I hadnt been using). Good luck! and Peace!
And…just to clarify, this would be the case for all sensors. (I went through about 13 before I quit).
Don’t know if this has anything to do with it, but i was getting a lot more readings like this when i would calibrate more often than needed. sometimes it would be 100+ off and i would re-enter my bg a second time to get it to adjust the reading, which i later found out causes more inaccuracies…
I’m with aa, entering more readings in the Dexcom only confuses the calculations done by the box. There will always be a resonable error, as there is in the fingerstick, so be sure you calibrate properly, which means that when the Dexcom is asking you for a BG reading while still operating normally, wait until you have a stable BG before you enter any figure in it. Entering a number while the BG is changing rapidly only confuses the Dexcom…
Having said that, reading 50 when you are 100 is clearly beyond any acceptable error, so I would think that calibration was not done properly.
Ciao, Luca
You need to give Dex good information. Resist the temptation to give it a bunch of readings all around one bG level when you first start up a new one. Instead, give it a set of good, stable numbers at your typical lowest and highest points, with a few more stable ones in between. In my experience, I get the most erroneous readings during the first 24 hours of a new sensor, when it seems to go thru a “training period”. The second week of a healthy sensor seems to be the most accurate–like it’s been through “secondary school”!
The other thing I’ve noticed that really helps with accuracy is inserting the sensor at least 2 hours before I start it up. From what I’ve learned from those in the know, the sensor will work better when it has settled in one stable spot and doesn’t get jostled around. More than once, I’ve felt mine move out of place and soon thereafter, the ???'s start coming on.
I heard from Dex reps that what Etta v is saying is the best way to start up a new one. “Insert, then wait about 2 hours before initiation” I did that quite often my last few weeks on the dex, and it did seem to help.
I’m confused, you can leave the sensor on for more than 7 days? I thought you had to change it, because you say the second week. Also don’t you have to wait 2 hours before starting a new sensor anyway before it gives readings? Are you saying wait an additional 2 hours?
Actually, Rich, you can (although recommendations are NOT to) leave the sensor in for as many days over 7 as you are comfortable, I would sometimes leave mine in for up to 10 days (dont tell anyone!). Yes, I meant an ADDITIONAL two hours, and that would be BEFORE you do the first CALIBRATION. Hope this helps!
Adding to what Steve said, most times I squeeze another full 7 days from a sensor, and the Dexcom is even more precise after the second calibration on the same sensor. I might get some bad readings towards the 12th/13th day, but it’s not the norm. I’ve once used a sensor for close to 3 weeks, but that just to see if my skin would get any irritation, which it didn’t.
Ciao, Luca
I agree with Luca…it almost seemed to me the second few days the meter was even more accurate…but I thought that must be my imagination. I guess not!
Well Steve, definitely not your imagination! 
I’ve been told this is due to the fact that the initial wound due to the needle insertion takes a few days to heal completely, and thus readings during the initial 2-3 days might be a little off. But at the beginning of week 2, the sensor is working nicely and the Dexcom readings are matching often to the unit my fingerstick readings, obviously during times in which the BG is flat.
Ciao, Luca
Because the Sensor was reliable on both the day before AND the day after, you very likely rolled over and “squished” it into the bedding for a while (while you were sleeping). This can reduce the rate of glucose transfer into ISF around the Sensor (from capillary blood), but nearby cells continue to suck glucose out: And so, a very localized “glucose shortage” often develops. It clears up rapidly after you get up and start moving around…
But it’s a real effect, so DON’T enter those 100-ish bG readings as calibrations. Dexcom will try to match them up with the too-low raw voltage reading, and screw up the calibration curve completely. Just use your valid bG tests to make your decisions, and maybe modify your alarm settings (and time between repeat alarms) to allow sleep if you aren’t in trouble. Also try to wear your Sensor in a spot which you can keep up and away from the mattress.
Maybe this was the cause. Or maybe not (YMMV, IMNAMD, etc.)
If this is a new sensor I would suspect it is bad. If it is a sensor that you have worn for a while, that performed well, this may be a sign that either the sensor is failing or you are doing something in your sleep to disturb the sensor.
Aaron
i agree with the ‘squished’ idea.
i too have had this strange “you’re at 50… no i’m at 150 you moron!” incident. i was also thinking slight dehydration would cause wacky numbers that don’t match up?
I get this occasionally. It’s usually because i slept on that side and the ISF pooled around the sensor causing the false reading. I’ve been putting the sensor just under my rib cage, about 1-2 inches off the center line of my chest. I don’t sleep on my stomach and this has solved the problem.
I find that calling my Dexcom names is self satisfying, however, rarely solves any problems and tends to make the Dexcom angry!
i agree. the only name i usually call it is “Dex” when i’m searching for it…" hm… where’s Dex?"