correlation of readings from human-calibrated vs feline-calibrated glucose monitors

My cat was diagnosed with diabetes 2 months ago, and has worn Freestyle Libre sensors ever since. Today, I took my cat to her "normal" vet for the first time since her week in the ICU at the specialty vet, and he demonstrated to me the difference in values between the Libre and Alphatrak 2.

The difference was pretty huge... according to sequential readings taken within the span of around 2 minutes (the two strip tests were made from blood drawn from the same prick):

• Freestyle Libre: 154
• Freestyle test strip: 131(!!!)
• Alphatrak 2: 191
• Freestyle Libre: 152

The impression I've gotten so far is that the values shown by human-calibrated systems (like Freestyle Libre) aren't literally "inaccurate" (in the sense of random hysteresis and error), they're just incorrectly-scaled... which implies that if you did something like compile a table of side by side values, you'd eventually be able to confidently translate a human-scaled glucose reading into a feline-scaled reading.

Has anyone actually compiled this kind of information yet, or tentatively come up with a formula/graph/table correlating human-calibrated values to feline-calibrated values?

Up to now, I've been panicking anytime her glucose falls below 70, and regarded brief surges to 180-200 as no big deal. Today has left me questioning whether I've actually been over-reacting to readings like 50 & needlessly subjected her to force-feedings of honey when her actual glucose might have been 80, 90, or more... and likewise, thinking occasional peaks above 200 for 15-30 minutes after eating were no big deal, when in reality she might have been way above 200 anytime the sensor said she was just 160 or 170.

I really don't want to start subjecting her to a half-dozen or more strip tests per day (I'm kind of obsessive about hovering over her with the Libre's reader and checking her glucose every few minutes any time she falls below 70), but I do feel like I need to get a better grip on reconciling the difference between the two.

 

There is no formula/graph to compare the meters. The Freestyle meter is a human blood glucose meter, the AT is for pet blood glucose. The Freestyle Libre measures human interstitial glucose. So all are designed to measure different things and I wouldn't expect them to be exactly the same. In addition to that, in the US manufacturers are allowed a margin of error on meters of 20%. The Freestyle Libre and Freestyle test strip are about 15% apart in your example.

Our dosing methods have been developed with human meters in mind. My vet used a human meter in her office and found it relatively close to what she saw from lab results. When I started here, I think there was only 1 person using at AT, they weren't common at all. The Libre is even newer. There are good reasons for using one or the other. The Libre gives you continuous monitoring and is great if you want a idea of what patterns look like, are not going to be around or have problems with manual testing. I've also seen recommendations that low Libre numbers should be retested manually, so good thing you are doing that.

And honestly, we are looking more for trends than exact comparisons. 152 and 131 are both in the first 1/3 to 1/2 of the 100's. We say that you should be giving higher carbs (or honey) if your kitty is below 50 on a human meter, or below 68 on the AT.

 

If you are numerically minded, your mistake is to assume it is a liner relationship with a slope of 1. All glucometers will have substantial errors, but they are all tested quite extensively and are designed to measure the same thing, they are just calibrated for different parts of the curve.

Like you, I measured with pet and human meters for a while. The trick is to look for their relationship. Here is my data:



That is a very good correlation, but with a 1.5 slope, you won't see a consistent difference between them. More importantly, as Neko said is not the absolute, but the relative data anyway. I use human meters because it saves me a lot of money, and that allows me to test more often.

 

Just note that my experience with a Libre sensor is that the comparison changes over time when compared with a regular human meter. See the following:
https://felinediabetes.com/FDMB/attachments/bg-comparison-libre-jpg-jpg.44029/
I can't attach the above graph as an image.

 

I also thought that there had to be some correlation. Even if it's not a linear one. So I started noting the numbers whenever I got enough blood to test two meters at the same time.

These were my numbers for Alphatrak2 vs the Contour Next:
562 AT vs 361 CN
401 AT vs 314 CN
521 AT vs 329 CN
468 AT vs 325 CN
451 AT vs 356 CN
618 AT vs 459 CN
449 AT vs 394 CN
456 AT vs 344 CN
303 AT vs 360 CN <- Note higher on Contour Next
428 AT vs 320 CN
414 AT vs 336 CN
474 AT vs 398 CN

I also compared the Alphatrak to the Relion meter when I was about to switch:
467 AT vs 493 Rln <- Note higher on Relion
516 AT vs 548 Rln <- Note higher on Relion
415 AT vs 363 Rln
408 AT vs 358 Rln
480 AT vs 403 Rln
325 AT vs 329 Rln <- Note higher on Relion
620 AT vs 585 Rln
441 AT vs 378 Rln
365 AT vs 360 Rln
476 AT vs 403 Rln
444 AT vs 414 Rln
479 AT vs 418 Rln
436 AT vs 350 Rln
464 AT vs 371 Rln
401 AT vs 332 Rln
482 AT vs 371 Rln
332 AT vs 295 Rln
376 AT vs 297 Rln
385 AT vs 358 Rln

 

Bottom line, there have been many members who have done comparisons, and no conclusions could be drawn. All you can compare is your AT to your human meter. There will be differences, and as Mariette posted, not even consistency between the AT and human meter. The suggestions we usually give is, pick one meter and stick with it, or the comparisons will drive you crazy.
@Hercule's mum Neko was the cat.

 

Exactly! and talked to me... oh wait, I am loosing my mind... I day dream with a good 10 hour sleep...I know you are Wendy, honestly! I am just really, really tired.... I wonder what kind of absurd things I am telling my students