Huge discrepancy between the lab and CoaguChek XS

[brachikaa]

First of all hello, this forum and its members were incredibly helpful while learning about valve surgeries

I'm a 28 year old male, almost 4 weeks out of my mitral valve replacement, and I've had quite a lot of trouble setting my INR right in this period. We started with a low dose of anticoagulant medicine (Sintrom), and the doctors kept on increasing my dose slightly. I also got a Coaguchek XS in order to compare my lab values.

The thing is, even though my INR values were low they were steadily increasing. My initial values were 1.6 (and Coaguchek reported 1.8) on 30/01, then it started growing towards 1.8 (Coaguchek 1.9) 03/02, and then today/this morning my lab value was 1.56, while Coaguchek displayed 2.2...

I'm unsure what to think of this. I tested this morning with Coaguchek, and I was confident that my lab value will be relatively close to the CoaguChek value but it is quite far apart... (also keep in mind that I further increased my dose since the last chek 5 days ago). My diet is relatively stable, my exercise is also stable (40 min walks 2x a day), my sleep too. I have no explanation how this is possible.

One thing that I think could be possible is that the lab messed up due to some reason (simple mistake or even incopetence). I'm from a small european country, and I moved back to my hometown (10k people) with my parents while recovering, and the doctors in the town here aren't exactly known for being world beaters and the hospital itself isn't the best (all draws were done in a public hospital). Back in capital I could go to a private lab but here it is not possible as it doesnt exist, so Im stuck with this lab. I even called the lab doctor and I told her that my dose increased, my diet is stable, it was 5 days since my last test of 1.8, but she couldnt shut up talking about how food can mess up INR values. In the end she kept repeating, you ate something that messed up your INR.

I read a lot of papers saying CoaguChek is quite reliable with a small margin of error, and to be honest I trust the CoaguChek value more than I do the public lab here... I will definitely do another check in a private lab as soon as possible.

But could this happen? Could a lab mess up my results? Or even, could CoaguChek mess up my results if the lab showed the correct value? I know that INR testing isnt an exact science, and that many factors influence day-to-day INR values, but these two were taken at maximum two hours apart (my lab test was taken around 9 in the morning, while the CoaguChek was taken at around 11 when I got home). I got my lab results at 1 PM.


Any tips/help?

Thank you very much!

EDIT:
I think I'll go to a private lab tomorrow to check my values again in another city, instead of the lab here.

 

[Chuck C]

The thing is, even though my INR values were low they were steadily increasing. My initial values were 1.6 (and Coaguchek reported 1.8) on 30/01, then it started growing towards 1.8 (Coaguchek 1.9) 03/02, and then today/this morning my lab value was 1.56, while Coaguchek displayed 2.2...

You indicate that your recently increased your dosage and INR was steadily increasing. Both lab and Coaguchek were showing this. Then, the lab value dropped significantly, but the Coaguchek INR value continued to increase, as one would have expected with your recent dosage increase.

I trust the CoaguChek value more than I do the public lab here...

Based on what you have shared, I would trust my CoaguChek more in this situation as well. The lab's most recent value would appear to be an outlier, going against expectations for a recent warfarin increase, while your Coaguchek value would seem consistent with what would be expected. But, the data is limited, so impossible to say with much certainty.

I will definitely do another check in a private lab as soon as possible.

That is a good idea. More data may help resolve whether that was an outlier.

But could this happen? Could a lab mess up my results?

Absolutely. I'm not sure the level of quality control in a small town in the Balkans, but there could certainly be a lab error. Are there any other labs that you could use to compare results with your CoaguChek? Regardless of whether that is possible, more testing to compare the two would be a good idea. Personally, I have not yet had a discrepancy between my CoaguChek and a blood draw from the lab. It was spot on once and the other time it was off by 0.1.

Please keep us posted and it would be nice to hear if you resolve the discrepancy.

 

[sharky7]

You indicate that your recently increased your dosage and INR was steadily increasing. Both lab and Coaguchek were showing this. Then, the lab value dropped significantly, but the Coaguchek INR value continued to increase, as one would have expected with your recent dosage increase.


Based on what you have shared, I would trust my CoaguChek more in this situation as well. The lab's most recent value would appear to be an outlier, going against expectations for a recent warfarin increase, while your Coaguchek value would seem consistent with what would be expected. But, the data is limited, so impossible to say with much certainty.


That is a good idea. More data may help resolve whether that was an outlier.


Absolutely. I'm not sure the level of quality control in a small town in the Balkans, but there could certainly be a lab error. Are there any other labs that you could use to compare results with your CoaguChek? Regardless of whether that is possible, more testing to compare the two would be a good idea. Personally, I have not yet had a discrepancy between my CoaguChek and a blood draw from the lab. It was spot on once and the other time it was off by 0.1.

I home test, and because it is the VA from where I get my Warfarin, I am required to lab test with them. I find I am needing to get to a certain number, or else the VA test is out of range up or down too much for my likes! I find this difficult sometimes, although it is now getting a tad easier! And those that deal with me at VA allow for it, YES i told them I home test!
Example of what I am doing and have done, about 3days before testing at VA, I was at 2.2, [range being 2.5 and 3.5], and dose is, 4mgs daily, and 2mgs on tuesday and thursday. So my plan that worked for me, I need and should say, I took 1mg extra, next 3 days and was on test day a 3.2, right on target for VA testing! I found that if I limit the extra to 1MG it makes it easier to manage as it is going up slowing, and as long as I test to make sure I do not over do the extra Warfarin!
Now from time to time without changing anything, for some reason my number will go up and down and out of range, and there is no explanation or reason, it just happens! Even those at the VA admit this being possible, and it happens to everyone now and then! AND I am getting to think this is a more or less thing sort of like everyones blood pressure, its the numbers over time that count, not only one or two numbers! But of course your dealing with medicine that can cause HUGE problems and fast, and you would be bleeding and not even know it, and we need to be aware of this, and go to the ER if this happens, no questions about that! And I would add, for this to happen to any of us, well, I think we have a better chance of winning the LOTTO!

 

[brachikaa]

Hello Chuck, thank you for your answer!

Absolutely. I'm not sure the level of quality control in a small town in the Balkans

To be completely honest, I doubt even that they have quality control. For a fact, I know that bigger cities, including the capital, have labs that are quite reliable, but unfortunately due to me staying with my parents, I have to deal with this local lab...

Another interesting fact. After being released I was given a low dose of anticoagulation medicine if I'm not mistaken 1/4 of a pill every day. They thought that would be enough as I had the same dose in the hospital and they managed to keep me in the range. Well, it might be that after leaving the hospital I finally started to eat better or whatever the reason, so 7 days after my INR test in the capital was really, really low - 1.53 if I'm not mistaken (I think they do have a pretty decent level of quaility control, so this result is quite reliable). Then we decided to double my dose, and the INR was slowly increasing, and since this week I tripled my dose from the initial one which now increased the INR more drastically. So basically I got almost the same INR value from the smallest dose of medicine I've taken, and the tripled one. Something is weird here.

EDIT: Also when we bought the CoaguChek I think I read on the forums that we should test it. So my father tried it and he got 1.0, so I'm at least somewhat confident that the CoaguChek monitor works as it should.

 

[Chuck C]

So basically I got almost the same INR value from the smallest dose of medicine I've taken, and the tripled one. Something is weird here.

One thing to keep in mind is that during the first few months on warfarin, it is well documented that the amount of warfarin needed to keep the patient in range usually increases gradually. That could be a factor. Also, they often have you on several meds in the hospital at one time, and some might have been competing with warfarin clearance, making the small dose have more of an impact. But, at the end of the day, I would still question that one recent lab reading, which seems to be a outlier.

 

[LondonAndy]

I agree with both of you, and am more inclined to believe the CoaguChek than the lab. In my mind there is less to go wrong with the machine - the test result is immediate, and can be easily repeated if any doubt. There are more variables when testing at a lab, no matter how good it is.

If you are not already planning to do this, I would do a test on your machine at the same time as giving blood for the lab test. This is what I do when I go twice a year to my anticoagulation clinic, to check the accuracy of my machine is still good, so that there is no doubt about other changes affecting the two results.

 

[pellicle]

Hi

as a simplification I'm going to say "go with the Coaguchek"

The thing is, even though my INR values were low they were steadily increasing. My initial values were 1.6 (and Coaguchek reported 1.8) on 30/01, then it started growing towards 1.8 (Coaguchek 1.9) 03/02, and then today/this morning my lab value was 1.56, while Coaguchek displayed 2.2...

good background ... what is the difference between testing time (time blood sampled at lab and time blood sampled at coaguchek), hopefully its not hours.

'm unsure what to think of this. I tested this morning with Coaguchek, and I was confident that my lab value will be relatively close to the CoaguChek value but it is quite far apart...

myself knowing where you live I'd put the Coaguchek as the reference not the lab.

I've experienced this (and indeed larger variations of 1.5 at the lab and 2.8 on the Coaguehek) which came down to lab. I had to really push that one through because the person involved was not amenable to doing anything and very amenable at criticising everything).

So if you INR is 2.2 that's good, but we want it at 2.5.

I understand that you are strip challenged at the moment but I really do think on Sinthrom that you need to have a measurement just prior to taking your pill and a measurement at 12 hours after taking your pill. To see what the variation is.

 

[brachikaa]

Hey guys, @Chuck C, @pellicle, @LondonAndy
Thank you for your comments.

I went to a private lab this morning and my INR results were 2.39... So the results from the hospital lab must've been inaccurate unless something drastically changed in my INR results (my dose stayed stable the last week). So, at least this is good and a burden has fallen off my shoulders.

Now, I have a question that comes from a moral source - do you guys think I should report this somehow? I did the last three tests in this lab and all of them were suspiciously low, so I would guess that results weren't accurate for some time. I'm sure I'm not the only patient whose INR values they're testing.

 

[LondonAndy]

Personally I would raise it with the Lab as a query, and give them a chance to respond. You have set out a clear history in this thread, and at least one of their readings is not consistent with either your readings or another lab's. If they are making errors then they have a duty to take such comments seriously and act on them. If you are not happy with their response then would be the time to consider making a complaint.

 

[Protimenow]

As others have alrady said - trust the CoaguChek.

I've had tests done by a lab that were wrong - sometimes VERY wrong.

The CoaguChek has the reputation of the manufacturer, Roche, behind it Their meters are used in clinics, doctor's offices, and by self-testers. I have a meter that's designed for use in hospitals.

Roche can ill afford ANY errors in the accuracy of their test results. This is why they have a code strip in each set of strips - to provide a fairly accurate (I hesitate to use the word 'accurate' because of the inherent inaccuracy of INR testing) INR result - plus, they have quality control checks built into each strip.

Labs, on the other hand, probably do few INR tests. They have less to lose if their results are wrong than Roche does. They may not have fresh reagents or correct reagent values to correctly calculate INR.

Again - personally, I would have more confidence in the CoaguChek. You might even point out the lab's discrepancy to them.

 

[Chuck C]

I went to a private lab this morning and my INR results were 2.39... So the results from the hospital lab must've been inaccurate unless something drastically changed in my INR results (my dose stayed stable the last week). So, at least this is good and a burden has fallen off my shoulders.

That is great news. Glad that you were able to confirm that the lab result was an outlier.

do you guys think I should report this somehow?

As Andy suggested, I would bring this up with the lab first and give them a chance to investigate why their numbers are off. One would hope that they want to give their patients accurate results, as their lives depend upon it. It is no small matter to be significantly off with their INR measurements. If they seem not to care, then perhaps look to report them to whatever agency has oversight.

 

[pellicle]

If they seem not to care, then perhaps look to report them to whatever agency has oversight

Recall that it's in Bosnia, which isn't n EU nation and after the 1992 upheaval may not have much in the way of customer service mentality
https://en.wikipedia.org/wiki/Breakup_of_Yugoslavia

 

[LondonAndy]

For what it is worth, I had my six-monthly check of my CoaguChek XS machine results at my doctor's today. Although they have over 15,000 registered patients, only 3 other patients are on Warfarin and self-testing - others on Warfarin attend the nurse clinic there every 6 weeks or more frequently to have their test done there.

Since the test was against their "multi-patient" version of the CoaguChek machine, you will not be surprised to learn that my result (3.2) carried out on my machine was exactly the same as the result carried out on theirs a minute or two earlier. I did a separate finger-prick to get my blood sample, as I have read somewhere that it is not reliable to do two tests from the same prick, so to speak.

Given how tightly regulated our healthcare system is, although this was all a CoaguChek test of a CoaguChek, I feel confident the accuracy is comparable with a blood draw tested at a lab. The NHS would not allow CoaguChek testing if it was not, and of course I have regularly posted this report from our national standards agency which says the CoaguChek is the only recommended hand-held test device. When I first started self-testing eight years ago, the comparison tests were done against a lab test. Then my machine result was a maximum of 0.2 different, with the machine being below the lab's figure.

 

[Protimenow]

What may have been interesting is that YOUR XS used your supply of strips (and code chip), and the one in the doctor's office used ITS code chip and strips.

I'm not at all surprised that both meters gave the same result.

It would still be interesting (maybe even worth the cost and trouble) to compare it to a lab's results.

In my case, I also have three Coag-Sense meters - two of the newest model, one of the oldest. Tests done on any of the machines are close to, or match, the results of the other. And all machines typically deliver a value that is about 1 point (not .1, a full point) below those of the XS,

These are usually (though not always) lower than the labs, too.

My point is this - even if the meters are all in agreement with each other, this doesn't mean that they're accurate. It's like a 12 inch ruler that is actually 1 inches long, with divisions making it look like it's twelve inches. This may be the case with the Coag-Sense - although the results agree, they may still be wrong.

FWIW: I trust the XS, but would still, occasionally, compare it to a blood draw (assuming that the lab does it right).

 

[ATHENS1964]

Today I had blood tests done and INR done as well
LAB 2.93 Coaguchek 3.0

 

[LondonAndy]

It would still be interesting (maybe even worth the cost and trouble) to compare it to a lab's results.

Our health service did all that years ago, with hundreds of thousands of tests, and concluded the machines are similar quality results to labs. No need to keep doing the lab tests now. Good enough for me.

 

[sharky7]

I have a question about testing, and hope those here agree that after seeing my question, that this is a good place to ask, for no other reason then to find out, and inform everyone with an answer so everyone gains from it! I go and get the INR test done where I get my WARFARIN, the VA in the USA! Question is this, is it possible for the following to have any effect on the INR number that comes backs when and if this changes/etc., the time it takes for my BLOOD sample to get from where blood in drawn to the LAB, then the time it takes for LAB to do the test? And if anyone knows, does how the blood is "handled" from BLOOD DRAW, till the LAB gets and has done the test, are those variables also in the studies they have done? WOW, is this more complicated than it needs to be, or is it me? Although it sure seems that this has in fact been on most of our minds that need to deal with the INR issue, and I am not the only one at this point a bit confused with all the info and data coming out recently!

 

[pellicle]

Hi

that's a good question

Question is this, is it possible for the following to have any effect on the INR number that comes backs when and if this changes/etc.,

if you don't mind I'll just point you to where I've answered this in 2016

https://www.valvereplacement.org/threads/coaguchek-inrange.888161/post-908917
PS: I just dug out a conversation I had with a mate some years back when I was discussing this issue (2014) when I was in the early stages of developing my data model. He is a pathologist (headed a small tropical hospital in Auatralia because he loves tropical diseases and parasites {not the administration types})

In Australian hospitals the following procedure is not atypical​

​

Test within 2 hours of drawing the sample. If it will be longer than two hours, centrifuge the tube (3000 rpm), pull the plasma off into a second tube, centrifuge that tube, pull the plasma off (except for the bottom 0.5 mL or so) into another tube. Freeze. Transport frozen.​

​

Reid Hospital in the US (picked at random) has this:​

​

The specimen should be transported to the laboratory as soon as possible and centrifuged within one hour. If it is not possible to transport the specimen to the lab within one hour the specimen should be centrifuged at 8500 rpm for 180 seconds or 3000 rpm for 10 minutes. Tubes should remain capped until the test is performed. Patient plasma should be tested within 24 hours if stored at room temperature, or within two weeks at -20° C and within six months if stored at -70° C..​

​

So with basic handling protocols it seems that there is no rush.​

 

[sharky7]

Our health service did all that years ago, with hundreds of thousands of tests, and concluded the machines are similar quality results to labs. No need to keep doing the lab tests now. Good enough for me.

I did not know that, and will have more comfort that I have in knowing this to be true! I just seemed to me that it meant more then it really did, and now I know. What happened was this, My home test/3.0, the VA HOSP. test/3.4 Both done the same day, within hours. Was thinking of doing something my better-half was thinking of, test before and after getting the VA/lab test, and see if that shows anything that I need to know. Sure seems that going up or down .4 either cannot happen and then there is also the margin or error, whatever that is. And at this point I have no idea as to what the ERROR MARGIN may be, and then adding all the variables, it gets more crazy than I think that it needs or should be!!! [and for that reason, I am now going to be having the base number, the number I home test, and result from that! And the VA/LAB is whatever it is, and is going to be!]

 

[pellicle]

And at this point I have no idea as to what the ERROR MARGIN may be, and then adding all the variables, it gets more crazy than I think that it needs or should be!!!

some points on error margins.
1) error increases as INR does, which means a graph like this is possible

if you're unfamiliar with reading graphs let me add in one or two with guides

so in general (in that study) outliers tend to be off higher (red line was Lab = 3.6 while XS gave 4.6

Now, if you were (say) INR Target = 2.5 then either of those high figures would be reason to turn your INR around. However the XS would steer you back to the middle (the target) a little earlier.
Around INR = 2 the differences were less, which is good because 2 and lower is increasingly not where you want to be.
Around INR = 2.5 theyr were quite close

Also important to keep in mind "Clinical Significance" ... is there really any importance between being 2.3 or 2.7? Remember also that all of this is fuzzy. We have a range of INR's where its pretty safe and the aim isn't to get a number its to not have an event.

By being as close to our target as we can we minimise the risks of harm

HTH