Freestyle Libre - Coping with inaccuracies


Here is a site you can go to that explains why it is difficult to get accurate readings from glucose meters.

Glucose Meters: A Review of Technical Challenges to Obtaining Accurate Results

[Ksenia Tonyushkina], M.D.1 and [James H. Nichols] Ph.D., DABCC, FACB2

I have not studied physiology and I am not an enzymologist but in a former life I was a sugar chemist.

I have had Type1 long enough to remember having to recalibrate you meter with a chip or a calibration strip for every batch change.

Most companies are using glucose dehydrogenase as the enzyme reagent for their strips.

SO WHY do METERS give different readings?

It depends on the reagents composition and the strain of bacteria used to produce the dehydrogenase and also the algorithms used for the numbers display.

It is only since Roche patented “mutant 31” used in the Accu-Chek Performa strips that you do not have to recalibrate the meter.

The only time different company’s meters will give similar readings are when everything is stable and BGL are around normal levels.

If you have a background in chemistry, Google Fischers and Haworth Projections for Glucose and also its Chair Formation.

I have only given the Libre one trial. When the sensor was applied it hit a vein. Although the readings were usually lower, they were much lower when in the Hypo range. The trends and the graphics were the most valuable data that I got from the Libre.


That is what I love about this Forum. I have learned far more here than speaking with my Diabetes Counselor.

Thank you. Really.


that’s nice but I don’t have any trouble getting accurate numbers from my meter.


When I was first diagnosed I had to rely on the experts and it was a challenge to keep BGL stable. Within two years I had developed neuropathy in my feet to the stage where it was painful to walk when I first got out of bed. Although my HbA1c was OK, about 7, I had no control and BGL were fluctuating wildly. This was the cause of my neuropathy.

I did my own research, started testing with two meters and more often. My BGL stabilized and my neuropathy disappeared.

From this data collection, I discovered different meter were rarely agreeing when in the Hypo or Hyper ranges. They could be also very different in the normal range if there was IOB, if I had just eaten or doing strenuous exercises.

Although most of my readings with different meters are usually very similar, there are times when they can vary substantially.

We have a rule in Australia “Below 5 you don’t drive” and the other day when I tested before driving the Accu-Chek Performa read 5.2 and the FreeStyle Optium Neo was 4.3mmol/L. I had a tablespoon of Dextrose and within 30 minutes the readings were similar.

If I am close to 5 and have to drive I know which meter will give the higher reading under the different circumstances.

I am a MDI but if I had a pump I would not rely on the reading from one meter to do recalibrations.


What you just wrote flies in the face of all instructions about using just ONE meter to do calibrations. Docs and CGM techs hate it when patients mention the usage of multiple meters. To top it off, the best meter to use is any model of Contour Next. No point in using others when that is what is preferred, what has been proved by studies to be the best and what MANY of us can attest to as the most accurate meter.


Dave, here is the basic chemistry of Glucose which meters use to give a blood glucose reading.

Glucose is a simple sugar, a monosaccharide and the most important Carbohydrate in biology. It is used as an energy source by living cells through Respiration and is a major product formed by plants through Phytosynthesis.

Glucose (C6H12O6) is an Aldohexose because it has six carbon atoms and an aldehyde group which exist in an open chain (acyclic) aldehyde form and the ring (cyclic) form. With Aldehydes, the double bonded Oxygen atom and the Hydrogen atom are bonded to Carbon No.1; C-1.

The Anomers of Glucose are α -glucose and β -glucose and differ structurally by the relative positioning of the Hydroxyl group attached to C-1 and the group at C-6. With α -glucose the hydroxyl group is attached to C-1 in the trans position relative to the - CH2OH group at C-5, portrayed as a Haworth projection , while β means it is cis . Another method of identifying α -glucose from β -glucose is by stating if the Hydroxyl group at C-1 is above or below the plane of the ring. Confusion can arise with this method if the Glucose ring is drawn upside down or in the chair conformation.

Within the cyclic forms of Glucose , three Rotamer Conformations exist. Rotation occurs around the O6-C6-C5-O5 Torsion Angle called the ω-angle . The three stable staggered Rotamer Conformations are gauche-gauche (gg), gauche-trans (gt) and trans-gauche (tg).

In solution, the open chain aldehyde form of glucose exists in equilibrium with several cyclic isomers , each containing a ring of carbons closed by one oxygen atom. In aqueous solutions however more than 99% of glucose molecules, at any given time exist as pyranose (Glucopyranose) which is a ring of five Carbons and one Oxygen atom. C-1 to C-4 are linked to a Hydroxyl side group and C-5 being linked to C-6 outside the ring, forming a CH2OH group. The open chain aldehyde form is limited to about .25% and the furanose (Glucofuranose) exists in negligible amounts.

Alpha-D-Glucose has a specific rotation of + 113 and beta-D-glucose only +19 . If either pure form is dissolved in water, the rotation gradually changes until it reaches an equilibrium value of +52.5 . This process of the change in rotation is called mutarotation . In aqueous solutions equilibrium is reached in a matter of hours. Alpha-L-Glucose has a specific rotation of - 113 and beta-L-glucose only -19. L-glucose is not usually found in nature. At room temperature equilibrium takes about 3 hours. As the temperature rises mutarotation increases rapidly.

At pH 7 in aqueous solutions , glucose exists in solution in cyclic hemiacetal form as about 63.6% β-D-glucopyranose and about 36.4% α-D-glucopyranose , the proportion of linear aldehyde and furanose forms are negligible. The glucose oxidase binds specifically to β-D-glucopyranose and does not react with α-D-glucose . It is able to oxidise all of the glucose in solution because the equilibrium between the α and β anomers is driven towards the β side as it is consumed in the reaction.

Glucose oxidase has given way to Glucose dehydrogenase because it has over 25 times higher activity than Glucose oxidase enabling rapid glucose sensing.

As I have said my knowledge is limited to chemistry. When I have made 5mmol/L solutions of Glucose and tested my Performa and Optium meters I got constant but not expected results.

After equilibrium occurred the Performa read 11.8 and the Optium 7.4 mmol/L and an average 9.6 which is almost double the expected reading.

The only way that I can explain the readings compared to blood glucose meter readings is that as Alpha-D-Glucose is utilized and beta-D-glucose mutarotates quickly into Alpha-D-Glucose.

The liver only releases glucose back into the blood as Alpha-D-Glucose.

From my data I believe that the algorithms for the FreeStyle Optium favor Alpha-D-Glucose because the Optium tends to be higher when I have no bolus insulin, have dawn phenomenon, just eaten or I am not active for a period of time. My readings are usually higher during these periods.

The readings for the Accu-Chek Performa will be higher when I have bolus insulin, been physically active and in the Hypo range. If I have a Hypo and both meters give similar readings, I know I have to quickly treat with dextrose where as with a physically induced Hypo, I will eat a meal.

Dave, I mainly use the Performa but will always check with the Optium if I don’t feel right. If meter readings are very different at normal BGL, I will wait until readings are about equal before I decide how much insulin I have to inject.

I have never had a Contour Next meter but I am happy with my meters as I know how they relate to my 3 monthly blood tests.

I stand by what I have said, “When the concentrations of Glucose are changing rapidly or levels are high with limited removal of Glucose, the meter readings may be at the limit of its accuracy range of 15%.

May be when the IR (Infra Red) technology is perfected for blood glucose meters, we will have a more accurate result.


So I take it you have not read the studies that declared the Contour Next the most accurate meter?


The last report that I read was one, I think was conducted back in 2013. This report tested all the leading make and some of which I have never heard of.

To simplify the report, 100 tests were conducted per meter in the different BGL rang(High Normal Low). Blood samples were from people with BGL ranging from the Hyper, Normal and Hypo range. Because it was difficult to obtain 100 samples in the lower Hypo range, samples had to be stored. These samples were treated with per chloric acid to deproteinate the sample to prevent blood cells taking up the glucose in the sample.

I have not read the latest reports but I did hear that the parameters of the test had been changed. Please correct if I am incorrect in saying that “Blood samples were adjusted to a precise level then tested”. If this was the case, the variability that would exist from hundreds of people has not been taken into account.

Dave, I am not saying that the Contour Next is not the most accurate meter because all good meters will give very similar readings at normal levels when the blood is stable and what I mean by stable is that “There are no major glucose variation due to hormones, insulin and/or physical activity”.

I am very satisfied with both of the meters that I use regularly, the Performa and the Optium Neo. If I get different readings from the meters and my test has not been contaminated or the strip is faulty, I don’t think that a meter has given me an incorrect reading. I know that my glucose levels are not normal and I can make the right decisions to correct the situation.


IIRC, I’ve gotten 2 bogus readings over the course of more than year and I don’t know the cause, but when you figure I used to use 17 strips per day, on average, for many years, that’s not too bad. I knew the readings were faulty and retested. I’ve used well-known brands that were sorta “OK” but not nearly as reliable as the Contour Next models (I have 3 different models). I used a motorized Accucheck Compact and Compact Plus for far too long before it dawned on me that when my bg’s would get up around 150 (for real), the meter would start reading higher and higher as the actual bg level increased, so that by the time I was actually 200 it would read in excess of 250. I had more than 8 or 10 of those two models, so it wasn’t just a single meter issue. Doing correction boluses when you get bad meter data can ruin your day. :slight_smile:

PS. I’ve not used the meters you mention. I’ve used Lifescan meters–so many I can’t recall them all, but included One Touch II, Fasttake, Ultra, Ultra Mini, Verio and maybe another model or two Lifescans. I used Freestyle (briefly). The WORST meter EVER that I had the misfortune to try (had it less than a week) was a Medisense. I got my money back on that one, from Long’s Drugs after complaining that it could have put me in the hospital with it’s wildly inaccurate readings.


Dave, when I have my 3 monthly blood tests, my doctor or the nurse at the pathology room knows the procedure. I line up four meters, the FreeStyle Optium now the Optium Neo, the FreeStyle Libre and the Accu-Chek Performa and Guide. I take readings before and after the blood tests. The time difference between the before and after tests is about 7 to 10 minutes.

If my BGL is high but stable, “Lets say 10 mmol/L, (180 mg/dL in your language) all meters will give similar readings except the Libre which may be 9.5 mmol/L. As a rule, I have usually found that the Libre when used as a meter reads lower. The blood test will come back close to 10mmol/L.

Dave, you are correct when you say your Accu-Chek meters were reading much lower then your actual BGL and my before and after tests support this.

When my BGL are high and I am not active, a BGL of 9.0mmol/L for the Accu-Chek Performa and 11.0mmol/L for the FreeStyle Optium, will come back at about 10.8mmol/L for the blood test.

I know from experience when I see the Optium Neo giving a reading much higher then the Performa, that I have to increase my bolus correction.

Although I have just recently been given the Guide which gives similar results as the Performa, I don’t use it, as my hands are so big and the strips so small that I can not insert the strip into the meter without touching the electrode sensor. This could result with the contamination of the strip and give a faulty reading.

I never take much notice of blood test results when BGL are low and because of Murphy’s Law, I have ended up in the Hypo range several times prior to the blood test. These tests never agree with my meter readings.

Yes Dave, you have twisted my arm enough and the next meter that I will be asking for will be the Contour Next. I will have to make another 5mmol/L solution of Glucose and retest all the meters again.


Sorry Dave, I miss read your post. I though your Accu-Chek meters read lower than the actual reading. My Accu-Chek meters usually read lower then the FreeStyle Optium Neo when in the Hyper zone. On the odd occasion when the Accu-Chek reads higher it is just higher.


ah, I believe you mis-read my comment about my experience with Compact and Compact Plus meters. In fact, they read HIGHER than my actual bg, once my actual bg was roughly 150. The further I’d rise about ~150, the more those meters would read HIGHER than my actual bg. Not until I started comparing them to other meters did I realize they were so inaccurate at higher bg levels. It’s no fun thinking you are 250, bolusing for that, only to crash because you were closer to 200-215. This wasn’t a one-time thing. Once I compared those meters (I lined up a bunch of them) to other brands in the house, I stopped using them “cold turkey” and I’ve forgotten what I switched to.

Four years ago I got a pump that included a Contour Next Link meter. I liked its accuracy but it is very hard to see the screen outside so I bought a Contour Next EZ meter, and then I bought the same meter that Dexcom includes, the Contour Next One, before my first Dexcom order, so I’ve got several of them now. They are all similar to each other in reading-accuracy.

I’m totally confident you will appreciate the Next meter once you begin using it.


Dave your meters were probably reading within the acceptable range but you are correct. Making corrections adjustment based on higher than actual readings is dangerous. You have to made your decisions based on the faith of your meter readings.

Although I take notice of the amount of carbs, my diabetes management is mainly based on my physical activity and the period of time I have been working hard for. I go from using about 100 units of total insulin and struggling to keep BGL down when doing several days of bookwork to using no insulin at all for up to 20 hours and having good control while I am fishing. My endocrinologist blames me for the loss of his hair.

I don’t rely on Libre meter because the readings are usually always low in any situation.

From out of the blue diabetes will throw you a curve ball out of left field and you ask yourself “What the hell has just happened.” I once treated a Hyper of 18mmol/L with 2 units of bolus only to end up fighting a persistent Hypo within a hour resulting in a rollercoaster for the rest of the day. I did not cross check with another meter but a recheck gave a similar reading.

I worked out that I had been inactive for several days doing bookwork then cut cane by hand. From this experience, I know that unusually high BGL on the second and third days after starting hard work have to be treated very differently and the first port of call is not insulin. This is one of the reasons why I cross check with another meter when I have a unusual reading.


I appreciate the chemistry explanation, however, I continue to see a repeated theme in a lot of threads regarding inaccurate libre readings usually during times when the glucometer manual tells you the readings may be ±70 mg/DL when compared to lab work-during insulin intake, or a recent meal. The Dexcom rep (years ago when they still existed) always reminded me the reason to have a cgm is not to replace the glucometer, but to aid in decisions and control using the arrows and blood sugar trends. If you recently ate, exercised, or gave yourself insulin, the glucometer and the cgm may be a little off due to the delay in interstitial fluid, which can sometimes be corrected by algorithms, but the cgm is still guessing. What’s important to remember is during fasting, the 3 glucometers and the libre will be very similar. Meals and insulin intake can be administered and adjusted by using the cgm data and blood glucose arrows accordingly.


All good brand glucose meters should give similar readings if every thing is stable at normal BGL (4 – 7 mmol/L; 70 – 125 mg/dL). When I did cross check the Libre within the normal range there was good correlation with the other meters but it could vary outside this range usually reading lower.

The Libre is in the process of being subsidized in Australia for some people and when this happens, I will give it another trial. At a cost of $100 for the first sensor, it was far too expensive for me to continue with the trial. Yes, the arrow trends and the graphics are what I need to better manage my BGL but I will always rely on the meter readings to make the final decisions.

Due to my own misadventure, I have caused a few bad hypos. The worst was when a meter flat lined for me and the check with another meter gave 1.8 mmol/L. I know that all meters are not accurate in the low hypo range but it is all panic stations when a meter fails to read because your BGL is low.

Although it was only a two week trial with the Libre, when BGL fell below 3.0mmol/L the Libre always was the lower reading when I could or did cross check with a meter. All these low readings were caused by physical activity.

I will be giving the Libre another trial because I think it would be the most suitable unit for me considering my work environment.


Just wondering if you have given any thought to having your doctor let you use their Libre Pro. Yes you won’t get the info because these are downloaded when you go back in, but if you keep good records, you can compare your numbers to the ones on the report.
I wore a Libre Pro for my last study and while I didn’t see results, the doctors and I compared them to my Dexcom numbers and there is a difference. It was 40 points off the first 2 week session and I think 20 this last time.
But I do think before investing in this new system, try and take it for a spin. Good luck and sorry I am not into the technical chemistry talk. I just want to know if it will work for me. And again, what works for me might not work for you. Many, many have had great suits the Libre! If it makes thing easier and better, yeah!


Sally, very shortly I should have access to the Libre, as the process
of subsidizing CGMs in Australia for concession card holders and for
people who have regular hypos is being evaluated. I fall into both of
those categories.
I experience three different types of hypos when under 3.2 mmol/L.
When I was a gym junky and levels fell below 3.5 mmol/L, I would have
to stop exercising as eating carbs would not stop the fall. At times I
would inject 1 to 3 units of NovoRapid (if I had eaten carbs), to stop
BGL rising fast after I left the gym. This was to stop the rebound
(some times above 10 mmol/L) one hour after leaving the gym.
The most common hypo is the physically active one. I feel these hypos
at 2.8 mmol/L and I can drive a cane harvester and drive a tractor
straight while cultivating cane, at this level. These hypos are
usually treated by eating a carb meal or fruit. By stopping work, my
BGL will usually rise.
Although not a common hypo, when I am relaxed and watching TV I suffer
from a hypo that can disorientate me. At levels below 3.0 mmol/L, I
feel spaced out and I have to treat with glucose right away. When
checked with another meter, the reading is always similar.
All my first readings are with the Accu-Chek Performa. I will always
cross check with another meter, if the meter readings are not matching
how I am feeling.
I don’t see the different meter readings as incorrect but as a tool
that I can use to treat the different types of hypos that I
Sally, was the 40 point discrepancy an average of all readings or the
Libre graphics showing that it was always lower in all levels from
hypo to hyper? When I trialed the Libre, some readings were higher
than the Performa when BGL were high but almost all readings were
lower when in the hypo zone.
It can be confusing for someone who has no chemistry background. Most
people think of sugar as the white crystalline product used to sweeten
drinks or the foods we eat.
Known as refined sugar or white sugar, sucrose is a disaccharide
formed by a molecule of glucose and a molecule of fructose. Dissolved
in water, sucrose will not AND CAN NOT give a reading on a meter until
the glucose – fructose bond is broken. We have the enzyme, sucrase to
break this bond.
Sally, I am not at all concerned that the Libre may read lower most of
the time because it is the graphics that will be most valuable for me.
It is most important to have faith in the meter you use and know how
to manage your diabetes from your meter reading but remembering your
BGL are more likely to match the meter display when levels are normal
and blood glucose is stable.


The differences were in the two week reports. So it was the average. But there was the one during the overnight when the lab and my Dexcom said 81 and the Libre said 50. And the key was that I couldn’t run low going into the morning trial, so they were a little concerned until they confirmed with lab results (and side check with Dexcom).
I will again say, any CGM that can help you manage the ups and downs to diabetes, is a very, very good thing. And I never would say which system is best for you! The Libre has worked wonders for many!
And I must also say, we always talk about how these CGMS compare to our meters, but the reality is that meters are not always accurate either. So trying to find a happy middle ground.
I no longer stress about testing (haven’t done one in months) and I have let go to the concern that the system might not shut off like it should. There have been a few lows but my sensor alarms & I treat. And since my lab works is the same as it was before I stopped testing, I guess it’s working.
I hope this wonderful piece of equipment is everything you need! I must say as much s I love my pump, it would go before my CGM!