Unreliability of Many Blood Glucose Monitors

Previous to today, I never gave the reliability of blood glucose monitors a second thought. I assumed that if they were sold in Canada, they were reliable. Not all are as good as others, it seems!

Yesterday morning, as I always do, I tested my morning fasting blood glucose with my glucometer. As someone with Type 2 Diabetes, this helps me understand the effect that the food I had for supper may have had and also helps guide me as to whether I may begin the day with time-delayed eating. For Type 1 Diabetics or insulin-dependent Type 2 Diabetics however, the accuracy of this information is critical! They base the dosage of insulin they take on this data and count on it being reliable and accurate.

Accuracy is how close the reading on the meter is to the actual blood glucose value and reliability is the likelihood of repeating the measure with the same meter at the same time and getting the same result.

Yesterday, I swabbed by thumb with an alcohol wipe, let it dry and took my blood glucose reading at 5:27 am and got a reading of 4.8 mmol/L (86 mg/dl) and thought “that can’t be!“, as I know that is a blood sugar reading that I only obtain after more than 18 hours of fasting.


I got another test strip from the same vial (recently opened and not expired) and tested the same thumb in a location immediately beside where I had just tested and got a reading of 5.8 mmol/L (105 mg/dl) and thought “that seems more reasonable, but what’s with the meter?”.

Ironically, only several hours prior a physician-friend sent me the link a report from August 14, 2017 that indicated that only 6 out 18 blood glucose meters tested passed the standard for meter accuracy which is for them to be within 15% or 15 mg/dl (0.8 mmol/L) of the laboratory value in 95% of 100 trials. That means there was only a 1/3 pass rate!

Naturally, the first thing I did was look up to see how my meter – actually both my meters (which are identical) ranked.  It failed!

Even though I had brought my glucometer to the lab with me in July when I last had my fasting blood glucose measured and it matched the lab results exactly, my meter failed the test because when tested 100 times, it was NOT accurate 95% of the time.  

To pass a meter had to match or be within 15% or 15 mg/dl (0.8 mmol/L) of the laboratory value on 95/100 trials.

I only tested my meter against the lab value ONCE and assumed it to be accurate. It was accurate on that one occasion, but it was not reliable, because when repeating the measure 100 times with the same meter it did not produce results within the 15% acceptable variation.

At 5:27 AM my blood glucose reading was 4.8 mmol/L and 2 minutes later with a new strip it was 5.8 mmol/L – on the same meter. That is a huge amount of variation, although depending on what the lab value actually would have been at that time, the results may or may not have fallen with range (see box below).

NOTE: The average of the two readings, 4.8 & 5.8 is 5.3 mmol/L and a ±15% tolerance would be ± 0.795 or ~ ± 0.8, for a range of 4.5 mmol/L to 6.1 mmol/L, so the readings would be within that range, ASSUMING the AVERAGE is the CORRECT result. While 0.8 is +16.7% more than the lower result and -13.8% less than the higher result, the actual ± 0.5 deviation from the mean is +10.4% and -8.6% of the lower & upper results. If either one result was correct, then 4.8 x 1.15 = 5.52 mmol/L, while 5.8 x 0.85 = 4.93 mmol/L, so the other would be erroneous. But, 4.8 ÷ 0.85 = 5.65 mmol/L, and  5.8 ÷ 1.15 = 5.04 mmol/L, so if the laboratory serum reading fell between 5.04 and 5.65 mmol/L then the meter’s two readings would be accurate to within ±15%. Now ± 15% is 30% of the value which means that (a) A serum glucose of 3.5 mmol/L (low end of normal) could mean a glucometer reading range of 1.05, or 3.04 mmol/L to 4.12 mmol/L A serum glucose of 11 mmol/L (way too high!) would be a 3x larger range of 3.3, or 9.56 mmol/L to 12.94 mmol/L. [thanks to Dr. L De Foa for the calculations]

Unfortunately, I know that my device(s) are not reliable based on this study data and for people who are insulin-dependent Type 1 or Type 2 Diabetics, they rely on the readings from their blood glucose monitors in order to dose their insulin. When their meters have been proven unreliable, it is cause for major concern.

I am reproducing the main data from this study because it is imperative that people know whether the monitor they are relying on is indeed, reliable.

Overall Results of Blood Glucose Monitoring Systems – Diabetes Technology Society 2018

The full testing protocol and results can be found here.

The rated accuracy from Bayer of the number one rated meter above, the Contour Next USB is 100% within ±0.56 mmol/L for glucose < 5.55 mmol/L and 98.1% within ±10% and 100% within ±15% for blood glucose > 5.55 mmol/L and it was accurate 100% of the time in the tests.

As for me, I have gone back to using a glucometer that I had on hand (which also tests blood ketones), as it is one of the models that passed.

While I am left with almost 1/2 a package of new test strips from the unreliable meter, how much worse could it be for someone who is dosing insulin based on unreliable blood glucose meter reading.

Type 2 Diabetes?

If you have Type 2 Diabetes and have struggled to lower your HbA1C or achieve your weight loss goals please send me a note using the “Contact Me” form above about how I can help and I’ll be happy to reply.

Copyright ©2018 BetterByDesign Nutrition Ltd. 

LEGAL NOTICE: The contents of this blog, including text, images and cited statistics as well as all other material contained here (the “content”) are for information purposes only.  The content is not intended to be a substitute for professional advice, medical diagnosis and/or treatment and is not suitable for self-administration without the knowledge of your physician and regular monitoring by your physician. Do not disregard medical advice and always consult your physician with any questions you may have regarding a medical condition or before implementing anything  you have read or heard in our content.