Is the Livongo Blood Sugar Meter Accurate?

It occurred to me the other day that even though I wrote a couple of blog posts and even made a video about it, I still haven’t addressed one major component of my new Livongo blood sugar meter: its accuracy.

How does it stack up to my Verio meter? More importantly, do I think it’s accurate?

Well…the answer isn’t cut-and-dry because I think it can be accurate…as long as my blood sugar isn’t above 200.

How did I arrive at this conclusion?

I conducted a little experiment.

Bear in mind here, I’m no expert in experimental design, so I established a simple setup for this. During the period of approximately two weeks, whenever I checked my blood sugar, I used the same drop of blood on test strips for two different meters: my Livongo and my Verio.

And the results were interesting, to say the least.

Whenever I was about 80-180, my results from the Livongo meter and the Verio meter were within about 10 points from one another. One instance, I was 86 on the Livongo and 92 on the Verio. Another time, I was 105 on the Verio and 113 on the Livongo. The meters never reported the exact same number at any point during my experiment, but I was happy whenever they showed similar results.

Things got dicey, though, any time I entered the hyperglycemic range.

Add a heading (1)
That 163 and 165 show how at times, the two meters were in agreement with one another…but then there’s that 231 and 262. Those are just a little too different for my liking.

It was weird and I can’t explain it, but the Livongo would report that I was anywhere from 20-50 points higher than what was on my Verio at a given point in time when my levels were 200+. This really surprised me, because I’d suspected that my Verio skewed a little higher compared to most other meters, so I never thought that the Livongo would prove me wrong about that. At times, it was truly maddening: On one occasion, the Livongo said I was 251 and my Verio said I was 204. That’s the difference of at least 1-2 units of insulin for me in order to correct that high! Even more frustratingly, because I didn’t trust the result from either machine, I checked my blood sugar again immediately and the Verio said 242, while the Livongo said 228.

WTF?!

That particular example was extremely confusing because I didn’t know which piece of technology to trust. And that’s the big thing for me: I need to know that I can depend on whichever blood sugar meter I’m using to deliver accurate results.

So again, remember that I’m no good with numbers, and that this is simply an estimation…but if I had to guess how many times the Livongo agreed with the Verio, I’d say it was no more than 7 out of 10 times. And I’m super bummed about that, because I’d genuinely hoped that I could trust my Livongo meter and use it full-time whenever I was taking a break from my CGM. Given the fact that it seems to be accurate when my blood sugar is in range, though, I suppose I could use that as motivation to work harder to prevent hyperglycemic readings…but I’m not perfect and I know that they will still happen from time to time. And I deserve a piece of technology that will accurately report them to me so I can make the best possible treatment decisions.

In theory, I still like the Livongo: It has a great design and some of its features (e.g., test strip reordering) are totally unique. But in practice? It leaves a bit to be desired.

7 thoughts on “Is the Livongo Blood Sugar Meter Accurate?

  1. I was writing this when I heard the news about RBG. I am sorry I will not be able to comment until i get my gut back in my body. Yes elections matter.

    Liked by 1 person

  2. I have been using a Livongo meter for about 6 months. Recently, I wondered how much readings vary with the meter, so for the past week I have been taking 2 readings at the same time. The data suggest that at my current blood glucose level (~115) the standard deviation from strip to strip is about 10 – 11 mg. A two standard deviation range of about 20 for two consecutive readings would be considered within tolerances (i.e., the meter is not malfunctioning and the strips are reliable).

    The Livongo site provides accuracy ranges on their meters. They say the accuracy ranges are approximately +/- 15 to 20%. During the past week, I have seen absolute differences between strip-to-strip readings that were as low as 4 and as high as 20. It’s safe to assume that these strip-to-strip readings vary because the quality and quantity of the chemicals in the strips varies. Also, variation is not only due to variability in the chemicals found on the strips. The Livongo website says reading variability can also result from the way sugar mixes in the blood. It would be remarkable if the sugar levels were absolutely the same everywhere throughout the body or even between fingers.

    This means that you shouldn’t place too much weight on a single reading from any meter. We want these numbers to be spot on, but they won’t be. The meters are great tools, but we need to understand they have limitations in their ability to measure precisely. It’s not wise to assume that all the differences we see in readings are due to food intake.

    I am a statistician, so I know how to measure these differences. However, I don’t know everything and I really want to know more.

    Liked by 1 person

    • Thanks for the comment, Dennis! I am definitely not a numbers/statistics girl, so your perspective is very interesting. I like your point that we should understand the limitations in the ability of our meters to measure precisely.

      Like

      • It’s frustrating to realize that the meters are not very precise. In today’s world, we expect measuring devices to be precise. But, that is not realistic with some things. For example, suppose you wanted to measure your waist. That measurement will vary depending upon exactly where you place the tape, or how you are breathing, or whether your posture has changed. We understand that our bodies change, but unlike the glucose meters, the measurement device (i.e., the tape) doesn’t vary.

        I wish the doctor would have told me that I shouldn’t be concerned with daily fluctuations of 10 percent, because that appears to be about the precision of the meters. Fortunately, I am not adjusting my medications daily. But for people who are, more guidance would be helpful.

        As I learn more, I will share a few other thoughts. Right now, I think I will set a target number, then because I know my meter readings fluctuate I will set an upward boundary using that target number. If my daily measurements are consistently near or exceed that bound, then I will review my situation. For example, if I wanted to stay below 140, I would set my boundary at 150 (my target plus 7.5%). If over the course of a week, I would make an adjustment if I have three or more readings that are above 150, because the probability that I would see this many readings above the boundary while I was still on target is less than one in 20.

        This is a rather simple quality control technique that should not be too hard for a person to manage. Of course, if you get even one reading that is more than 20% above the target number, then that would indicate you should take immediate action.

        Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s