How to estimate your BG meter's noise

BG Avg Deviation.xlsx (10.1 KB)

Try this @Nickyghaleb

1 Like

Thank you, gentlemen. :slight_smile:

I did it by hand, following your steps, Eric, but I’ll go put everything through to check it. It was disappointingly close to my Relion… I pull for the underdog and was hoping Relion would blow it out of the water. The good news, in reality, is that they were both very good. This is NOT what I remember my one touch doing, but that’s why I tested the numbers— and not just my memory.

1 Like

Well if you want to see a higher deviation you need to be at a higher blood glucose level. One would hope that the low 60’s would be mostly correct. Try it at 150 or 200 to see what it does at the higher measurement level.

2 Likes

EXCELLENT point. Didn’t think about that. It’s Friday night and I’m bound to eat something I shouldn’t and underbolus, so tonight might be the night… :dancer:

3 Likes

I thought I’d share… my ReliOn is still reigning champ.

ReliOn BG range 58-67 average dev. 2.284
One Touch Ultra Mini BG range 111-125 average dev. 2.86
ReliOn BG range 243-271 average dev. 7.13
One Touch Ultra Mini BG range 230-287 average dev. 11.53

2 Likes

Well, the average deviation is just a measure of precision. Of consistency.

From your tests, yes most likely your ReliOn is better than your One Touch.

But for what it’s worth, without comparing it to a lab draw machine, or a number of other meters, you don’t know if your Relion is consistently accurate, or consistently inaccurate.

Suppose you had:

Meter: Tests:
Contour: 78 81 88
FreeStyle 83 86 90
One Touch 82 84 88
ReliOn 65 66 64

In the example shown above, the ReliOn would have the best avg deviation when compared to only itself. When compared to all the other meters, it would be the worst!

To make sure your meter is great, it should be both precise and accurate.

best-bg-meter-precise-or-accurate

2 Likes

You’re like Joy Crusher man… couldn’t even leave me with that for a couple of hours… dang.

Okay… so if I want to take this further… I have to bring my meter where?? OR test it against “other” meters?? Let’s say I went with that. How many meters would i have to test it against, and how would i go about doing that?

Here’s my next question. What does this do to my Libre and Guardian data?? I’ve collected 30 days worth of data for both machines. I’ve finished running the numbers. A lot of work. A lot of good information… all based on my ReliOn. If I feel, based on these two tests, that my meter is consistent, then it would still show…

I’m not throwing away all that data… there’s just no way. Tell me I don’t have to do that…

1 Like

Sorry. :frowning_face:

I am not saying there is anything wrong with your data or your meter.

I just want to give the complete picture.

(If you are going to whine about mansplaining, I will stop now.)

If your meter is always perfectly consistent - numbers like 71, 72, 70 - but is always showing above or below your actual BG, that’s not great either. And because we at FUD love our Darling Nicky, we want her to know about that sort of stuff.


When you go to your endo, they probably do a BG test. Either with a machine they have in the office or with a vein blood draw. When they do that, take your BG a few times with your meter at the same time as the endo BG test and write it down. When you get your results from your endo, compare what your meter said at the same time. Now you know!

2 Likes

I really like learning… and you can’t learn if you are always right. SO. Mansplain away…

And the feeling is mutual… which is why i’ll not only tolerate the mansplaining but even come back for more…

I want to know as much as possible. So now, with my remaining 30 or so One Touch test strips, I’ll at least compare my next 30 or so ReliOn readings to see how the two of them line up. The picture part of it is important… to be able to at least get a picture. And what I think is that when my blood sugar is nice and stable, my Libre, guardian and Relion are usually within 2-3 points of one another. Maybe 5. And my ReliOn often splits the Guardian and Libre down the middle, with the Guardian hanging low and the Libre high. Now unless it turns out that my ReliOn is consistently running 20 points lower then my true BG, which would change the relationship between the three devices, then it would still mean the patterns I’ve seen are true. But that, now, IS the big question. Is my ReliOn consistency off. First two tests though have left me with higher values than my One Touch… so…

1 Like

By the way, I noticed you didn’t “love” when I called you “joy crusher man”… I hope you didn’t take that the wrong way or anything. :grin:

No, it’s fine. Just didn’t want the mushy stuff to get in the way of all the supreme knowledge.
:stuck_out_tongue_winking_eye:

1 Like

So since you called to attention the “accuracy” of my relion, my world has been turned upside down. Never pegged myself as the “blind faith” type, but I’m not sure I haven’t been doing just that…

Last night:
Guardian: 156 (triple arrow up)
Relion: 176
One Touch: 156
Libre: 178 (one arrow up)
I can deal with that… not happy about it, but at least there’s a pattern…

The from this morning:

Before my run (don’t judge… I got side tracked and took too long getting out)
Guardian: 292 (one arrow up)
Relion: 258
One Touch: 220
Libre: 278 (angled arrow up)

After run:
Guardian: 123 (double arrow down)
ReliOn: 114
One Touch 101
Libre: 94 (angled arrow down)

And a little bit later:
Guardian: 134
Relion: 132
One Touch: 102
Libre: 129
And then immediately afterward:
Relion: 139
One Touch: 135

I looked up on the internet how to test a meter’s accuracy… I was hoping to find something like the one I just did. The only thing I’ve found is taking it to the doc. If I were to try it up against other meters, how many would I need?? And, really, is that even a real measure if I don’t know WHICH is right?

After all of this, I’m about ready to bag up meters, sensors, and MAYBE this pump, and send them to the bottom of the lake…

1 Like

Don’t go crazy just yet.

Make it simple. First of all, for now and for the sake of calibrating, ignore stuff that isn’t looking at your actual blood glucose. Like the Libre. That isn’t a BG meter, it’s checking interstitial fluid, which has a delay. That doesn’t really tell you how accurate your BG meter is, because of the delay.

Just get a single meter to compare with. Why not get the meter that is consistently rated as the best for accuracy, and is also pretty cheap?

Get a Contour Next One meter from Amazon and 100 strips. You can get both meter and 100 strips for about $30. You can probably shop it around and get it cheaper, especially if you buy more strips.

Let that meter be your comparison meter. Trust what the Contour tells you, and see how your other meter compares.

Contour meter
Contour strips

Trust me on this Nicky. This doesn’t need to be overly complicated. The Contour is consistently rated #1, by lots of different people. And many people on FUD also think it is great.

10-popular-blood-glucose-meters-put-to-the-test

are-blood-glucose-meters-accurate-new-data-18-meters

accurate-blood-glucose-meters

1 Like

You’re like Siri…if Siri knew everything.

1 Like

Do you mean I am like this?

2 Likes

Kinda like what it feels like, yeah. :rofl::rofl:

1 Like

In addition to what @Eric has mentioned. Keep doing what you are doing, then when you have your next physicians appointment, right before or after your blood draw take a reading with both meters. Then when you have the lab value you will understand where the meter is for you.

2 Likes

I will definitely do that. I appreciate it. Just out of curiosity, is there anywhere else to get a lab draw that you can think of? I’m not due for an appointment for probably another 2 months. I’m not going to have any more blood in my body for that blood draw. :slight_smile: i’ll do the research, just thought maybe someone would know another way…

If you are on the good side of your physician, they might order a blood draw that you could do at a blood draw center close to you and get your bg tested. Honestly, would you really change your treatment if you know that your meter trends 10-15 points off your actual blood glucose. If the answer is yes, I would pursue the answer.

1 Like

I have taken so many numbers at this point that I’ve learned how to take them all with a grain of salt… pretty much. BUT there are a couple of things that I need to do because I am who I am (and that is a mentally ill person, I guess, because I don’t know who else would do this)… the MOST important thing is that I use these readings for my calibration factors. I actually use those to make important decisions. Now I do understand their limitations because they’re 100% dependent on my BG reading, and that’s just a problem anyway. But I would want to adjust my thinking… I’m still in the early stages of understanding this stuff, and, yes, I think that particular number would… okay I don’t know. Maybe not. But here’s the OTHER most important thing… I was getting ready to make a video about the Libre. And, honestly, I’m pretty impressed by it. (The Libre). Believe it or not though, I have a little integrity, and I want to put out thoroughly collected and calculated numbers. I collected 30 days’ worth of numbers based on my Relion… it not only showed the Libre to be more reliable, accurate, AND better equipped to handle the lag, but in showing all of that, it also showed just how WEAK the guardian is. The guardian is the only sensor that currently is involved in treatment decisions, and, well, it would be scary to look at these numbers knowing your pump is basing your dose off of them. I do want to make that video, too, though. There’s no doubt in my mind about it as it reflects what my experience has been since starting it, and I started on it using the One Touch. BUT a general difference of 15 really might have some signifance and might make those numbers look different. I have a good solution though, and that is to present my findings and to emphasize the fact I used my Relion. And then to repeat it with either my one Touch or the contour next one I just ordered. And you don’t have to respond to this very long thing here… I sometimes talk to myself through my posts to others, trying to talk myself into it out of a particular thing. :grin: