Diabetic Testing Inaccuracy Part 1


I started selling diabetic testing supplies about a decade ago. When I started, meters that offered no coding were all the rage. Just years before, testing meters were monstrosities. They were huge and required a huge amount of blood. Samples could only be taken from the finger and they required a complex coding system to validate results and guestimate blood glucose.

Since then, the meters haven’t changed all that much. They’ve been perfected, but at their core they are still just the same simple system they have always been. As flawed as they are, they are still “good enough” for the use that they have been assigned.


I won’t want you to think this article is strictly an abuse on testing supply manufacturers. Instead this article is meant to help diabetic patients understand why their medical testing results swing so rapidly from test to test. Ever since I’ve been selling test strips, I’ve been asked “why is my meter reading differently that it did a minute ago?”

Let’s start with how testing meters work. Test strips contain an enzyme that reacts to the glucose in the blood. The test strips is exposed to blood glucose and administered a brief electrical impulse. Thanks to the miracle of quantum physics, the electrons are measured and an estimate is made of how strong or weak the pulse was on the other side of the circuit. Depending on the strength of that circuit, a corresponding number is generated and displayed on the screen to give you an idea of what your blood sugar is.

So far, so good right? The US requires these meters to have about a 20% margin for error about 95% of the time. Whoa! Hold your horses. My math experts out there are screaming. Josh, with a 20% margin for error, how can you call that accurate. You are telling me that if my blood glucose is 100, then I can get a reading between 80 and 120 and my meter will still be considered legally accurate? Bingo, my friend. Crazy or not, in the world of blood glucose testing, you can test the same sample back to back and arrive at an 80, or a 120, and both of those tests are correct.

That’s not all. The biggest problem there is the “95% of the time.” This allows the meter to give wildly inaccurate results 5% of the time. So if your blood glucose is 100 and your meter tells you that glucose is 600, as long as it only does it about 1/20 times, your meter is still accurate.

Okay, okay, so your meter isn’t very “accurate” but it’s accurate enough to do the following things with:

  1. Take diabetic medication(assuming you test regularly and take each test result with a grain of salt)
  2. Have an insulin pump delivery insulin to you based off its results
  3. Create a log of your blood glucose to determine how well you are controlling your diabetes

These are weighty tasks to assign to something so inaccurate, but fear not. I’m going to tell you why that’s okay soon enough, but first let me tell you why these inaccuracies exist, and a few other problems you may run into when testing your blood glucose.


*The author of this blog is not a medical professional and this article does not contain professional medical advice. This blog is not intended to substitute for medical advice, treatment, or diagnosis. Always seek the advice of your physician or other qualified health provider with any questions you may have regarding a medical condition. Never disregard professional medical advice or delay in seeking it because of the contents of this article. If you think you may have a medical emergency, call 911 immediately.