Which meter to trust?




I don’t think you have to settle for +/- 12%

diatribe - blood-glucose–18-meters
Diabetes Technology Society Blood Glucose Monitor System

Sure use a single meter, but use a good one!


I guess I really don’t have much of an understanding of how the magic works, but if the strips are more important than the meter in determining the blood glucose value, then it makes sense that the strips are the expensive part of the package. And probably like a lot of people, I instinctively don’t trust cheap test strips, but I really have no reason to feel this way.

Just FYI, I just did another test of my meters and got this:

Dario - 87
CVS Advanced - 117

I skipped the OneTouch because I just don’t like it and don’t use it much. Regardless, that’s a 30 point difference between meters, which seems significant. But like I stated, since I mainly use this to calibrate my Dexcom, I will stick with one meter at a time. Plus, I would prefer to err on the low side.


Contour Next one, about 22 cents a strip on Amazon, and it’s on the top of everyone’s list…

diatribe - blood-glucose–18-meters
Diabetes Technology Society Blood Glucose Monitor System


Sure out of the box with fresh strips with a few tests averaged. Use a meter for a few months get test strips at the end of thier life, test in hot or cold, or eseciallly test high or low, results like diabetes varies. The number, is at best an estimate. A good estimate, but an estimate.

I use the contour (top rated in the Diatribe sample), but I have no illusion my 2 year old meter is 99% accurate. It is +/- 15% just like the company tells us to expect.


My experience is that contour next gives most consistent results when doing repeated test with same drop.

When my insurance covered Next strips, and others, I chose Next.
Now insurance only covers One touch, and I use verio, and find more variation when testing with same drop. Also, compared to my dexcom, One touch verio gives more unexpected results.

Recently I found I had expired contour strips, and bought some new ones on AZ. I did testing to compare, and even found the expired strips to be inline with the new ones. They expired in 2015-05.
At the same time (same drop), I tested verio. Most verio results were close, but in 1 of 10 it was 15 pts off.

So… I am going to use contour next for dexcom calibrations, and verio for other tests. If my insurance covered contour next, that would be my only choice.


According to FDA submissions for Contour Next Blood Glucose Monitoring System (BGMS) it is +/- 10% for more than 96% of tests. The FDA requires (at least) +/- 15% for 95% of tests. (Across the entire range the BGMS can measure.)

This does not mean the Contour is generally off by 10% as it is also listed as +/- 5% for almost 77% of tests.

I read this to mean that most of the time if the Contour says 120 that it is likely off by no more than 6 points.


That gets to two measures of statistics. THe first is confidence and the second is interval. Generally the greater the confidence the more the spread of the interval.

It works like this. You want to issue products with a rating that ia company can be confident in. THere are generally two confidence levels which are published. 95% and 99%,

So lets say i want to be confident that the interval i publish is within range. Again the higher the confidence the greater the interval. So I want 95% confidence that the interval i publish will be right. According to the company the contour meter at 95% has an interveral of 15% plus or minus. Now in actuality any single meter will likely be dead on, but occasionally one or more readings will be wrong. So if I want to be 99% confident that the reading will bein the published range than the interval will get bigger. It is a two part consideration

Like I said the contour meter will usually be accurate. but if it mis fires we can expect that it will be wrong no more than +/- 15%. It is a wide range, but it explains how a meter can test two fingers on different hands with wildly different results.

I once got a 123 on the middle finger of my right hand and a 101 on my left. Turns out both are in range of the meters published interval with a confidence level of 95%. Meaning the two readings are both valid,

If you use a dexcom when it calls for two calibrations it uses the two sticks to start the interpetation for the device. To that end I always used the same finge on two different hands for the calibration. It gives one the best picture of the Dexcom to act on.

We have to remember these are not absolute blood sugars. They are approximations. Even blood draws, are estimates based on a confidence level. No test is 100% correct all the time. Sorry for the spelling it is 3 AM and on this dam ipad the type is two small to really see and correct.


According to FDA submissions for Contour Next Blood Glucose Monitoring System (BGMS) it is +/- 10% for more than 96% of tests.


Comparative Accuracy of 17 Point-of-Care Glucose Meters
Journal of Diabetes Science and Technology
The study was performed September 2014 through December 2014 at Massachusetts General Hospital’s Diabetes Research Center, Boston, MA.
Article first published online: October 3, 2016;Issue published: May 1, 2017

Abbott, FreeStyle Freedom Lite
Abbott, FreeStyle Lite
AgaMatrix, AgaMatrix JAZZ
AgaMatrix, AgaMatrix PRESTO
Bayer, BREEZE ®2
Bayer, Contour Next
HemoCue, HemoCue Glucose 201
Nipro, SideKick
Nipro, TRUEresult
Nova, Nova Max
Nova, StatStrip Xpress
LifeScan, OneTouch Ultra2
LifeScan, OneTouch VerioIQ
ReliOn, ReliOn Micro
ReliOn, ReliOn Prime
Roche, Accu-Chek Aviva Plus
Roche, Accu-Chek Nano


In most cases, we don’t really care if the numbers we see are exactly the same as the lab. What we care about is how they change with different foods and what our particular meter says when we feel low.

So if you feel low at 65 with one meter and 75 with another, that’s what you should remember.

There are two factors involved here: Accuracy and precision. Accuracy means how close to lab values your meter is. Precision means how repeatable the measurements are.

So if lab value is 80 and your meter reads 79, 81, 80, your meter is both accurate and precise.

If your meter reads 89, 90, 89, your meter is precise but inaccurate.

If your meter reads 75, 82, 85 (average is 80), your meter is accurate but imprecise.

I was once told by a company tech support person, that smaller drops of blood are more strongly affected by a lttle interstitial fluid, and the amount of interstitial fluid depends on whether or not you milk the finger. After you’ve eaten, blood and interstitial fluid levels vary more than when fasting.

And even lab values aren’t always accurate. If blood sits around, the glucose levels decrease, and I once discovered that the office where I gave blood let it sit around for almost 7 hours before sending it over to the hospital lab. I stopped giving blood there.

Furthermore, labs are run by humans, and humans sometimes make mistakes. When I worked in a research lab, we always measured everything in triplicate to try to control for small procedural errors or dirty test tubes or whatever.

So I think our best bet is to assume our BG readings are pretty good approximations, but we shouldn’t panic if something changes a reading 5 points. It’s meaningless. If one reading seems wrong, then test again. The important thing is consistency, using the same meter the same way every time.


Hi Gretchen,
I agree with everything you said. But want to emphasize that if…

…if you are going to use the same meter, it should be a good meter.


@David48 - I was trying to find information on the CVS meter you mention. It appears this is the AgaMatrix JAZZ which is manufactured for CVS and then branded with the CVS name. Assuming this is accurate information, if you read the article I linked to regarding the Meter Comparison, you would see that there are different enzymes commonly used in the strips. GDH-FAD, GDH-PQQ and GOx. The GOx is the older enzyme with GDH being newer technology. The GDH-PQQ previously had an FDA advisory - not sure if that is still an active advisory or not. The GDH-FAD is the newest technology and has the lower MARD numbers as compared to the strips using the GOx enzyme technology which have the higher MARD numbers.

The AgaMatrix JAZZ (ie - CVS rebranded) is listed as using the older GOx enzyme technology.

It certainly is possible that CVS has switched the meter/strip under the cover of its brand and currently uses something other than the AgaMatrix JAZZ. But the CVS and JAZZ certainly look pretty dang similar (pictures below).

Summary - The CVS meter/strips you mention appear to be the older and less accurate technology as is the One Touch Ultra.

As I mentioned, we are switching to the Contour Next which appears to consistently be ranked as the best or almost best in pretty much every study and review I have seen. The Contour Next does use the newer GDH-FAD enzyme technology for its strips. If buying outside of insurance, the Contour Next strips are priced at about 21 cents on Amazon for a 300 count.




Tim, if we keep telling everyone and people start buying it, the strip price will unfortunately go up…

That meter is at the top of every single reputable meter comparison. If people want to use a worse meter, they are more than welcome to. We tried to tell them. I’ll be happy if the price stays around 22 cents.


A man with one clock knows what time it is. A man with two clocks is never sure.

All 3 of your numbers are within 11% of 113. And bg meters are commonly quoted as having 20% accuracy. If you repeat this experiment many times you many discover that one of your meters is systematically 5-10% low and another is systematically 5-10% high, and you will find many times when the random jitter of the measurement is that size too.


I like how this graphic illustrates your point.


Nice illustration.

BTW, here’s what the late David Mendosa wrote about the Jazz last year:


Nifty new way to look at Meter data:

The “Radar Plot”.
If really interested in this, all the little fine print at the bottom is actually interesting. While perhaps not readable by the average human in that crazy fine print, if you cut-n-paste the fine print into a word document than it actually can be read and really is interesting.

"A radar plot is a new way to show the differences between blood glucose meter readings and the laboratory reference values in meter performance evaluations. The radar plot uses polar coordinates to position data points, rather than placement within the x- and y-axes of a more traditional graph. Radar plots are a different way to graphically show data, similar to Modified Bland-Altman plots.

Accuracy on the graph is represented by how close the data points are to the center: the closer the point to the center, the more accurate.

Precision, on the other hand, is measured by how the dots fall on top of each other: the more spread out, the less precise.

Furthermore, … [clip]"



Has anyone used the Bayer Contour Next test strips from Amazon. The price is 1/4 of the cost at a pharmacy. (Bayer Contour Next Strips from Amazon)

I would like to know if these are genuine Bayer strips or if they are counterfeit. The reviews are fantastic, but when they are this good, I wonder if the people who wrote them are real.


I recently had to do this for the second time (both times because my insurance mail-order pharmacy refuses to sell me strips earlier than they “calculated” I should need them; nevermind that I am currently pregnant and need to test more frequently, and my doctor put in a new prescription to reflect my new needs). From this limited experience, the strips I got both times were completely legit and a long way away from their expiration date (my other concern), but the whole transaction still leaves me feeling icky.

Read these pieces for some more context of how these strips end up on the internet:


Just this year I have seen more local coverage on this from Chicago, Philadelphia and this piece from my state:

Lots of interesting info in all of them and it seems like “grey market” is the best description. Personally I hate I have to participate in it…