Who (or what) can you trust?

[Protimenow]

I'm repeating a frustration that I sometimes run into when the issue of INR accuracy comes up.

A year or so ago, when I was still using the InRatio, I was getting results on the InRatio that were higher - often considerably higher - than the labs. When I asked them which I should trust (meter or lab), they told me to 'trust the lab.'

I've put my trust in the hospital lab that is associated with the clinic where I occasionally get a finger stick, instead of a monthly blood draw. I've had my blood tested at a general medical clinic and gotten a result that was many points higher than the hospital lab (and the tests were done within hours of each other).

For more than a year, I've tested with my meters within hours of my tests at the hospital lab, or by the Hemochron meter at the Anticoagulation Clinic.

I trust the hospital lab as the source of (probably) the most accurate INR. I trust my Coag-Sense meter to be slightly lower (usually .2 or .3 lower) than the hospital lab. I've been pretty comfortable with this scenario. When my CoaguChek XS strips expired, I didn't rush out to get new ones -- although it would be nice to compare the lab to the two meters, and the meters to each other between blood draws, I'm comfortable with just the Coag-Sense.

I recently got a Hemochron - and only one test strip. The Hemochron is one that is used in operating rooms to test INR and other blood factors. It's one that hospitals put their faith in. My Hemochron passed the electronic quality control tests.

Here's my dilemma:
I went to the Anticoagulatin clinic this morning, and their Hemochron gave a result of 4.0. This was surprising to me --- my Coag-Sense on Monday (four days ago) gave me a 2.8. Although it's possible for the INR to go from 2.8 to 4.0 so quickly (maybe through double-dosing on warfarin, for example), it seemed a bit odd that my INR would jump to a 4.0 so quickly. I assured the clinic nurse that I'll eat some leafy greens. Although I told her that 1/2 dose (or no dose - her first suggestion) wouldn't have any effect on my INR a week from now, I accepted her claim that it would.

I came home a few hours later and tested on my machines. Here's the summary:

Lab Hemochron - 4.0
MY Hemochron -- 3.4
MY Coag-Sense - 3.1

Assuming that the Hemochron is 'lab accurate' and that my Coag-Sense is often about .1 - .3 lower than the labs, the results on my meters are consistent with what I expect that my INR ACTUALLY is. The minimal spike of 0.3 from what it was four days ago (without any diet or acivity changes) is also more consistent with the big jump on the clinic's meter.

The clinic nurse insisted that their meter is calibrated and carefully matched to the lab's values.

Without a blood draw to answer this, the question is this:

Do I trust the Clinic's value, which seems somewhat illogical, based on my INR four days ago and my tests at home on a similar machine. Or do I trust my two machines as being more accurate?

Do I add a little stress to my life because my INR is in a somewhat uncomfortable range, or do go on living relatively normally with an INR below 3.5?

(Of course, if I was the usual patient, and had total faith and reliance in the accuracy of the clinic, I would blissfully follow their recommendations, and not even be concerned with the accuracy of their testing. Sometimes it's not easy to know enough to question the experts).

For myself, I haven't changed my life style except to probably add some greens and avoid situations that may result in bruising. I'll maintain my coumadin dose. I trust my meters more than I do the one at the clinic.

DO YOU AGREE with my conclusion? What would YOU do, given the same results?

 

[pellicle]

Hi

When I asked them which I should trust (meter or lab), they told me to 'trust the lab.'

of course they would

I trust the hospital lab as the source of (probably) the most accurate INR.

without an independent evaluation I would not do so unilaterally. My mate who's a pathologist has repeated to me in discussions on this that there are too many variables and that while they may have standardized on a reagent there are other factors that cause variance (as I've discussed here several times) so as to make it conclusive that it is not a gold standard.

I trust my Coag-Sense meter to be slightly lower (usually .2 or .3 lower) than the hospital lab.

I have a paper here which shows as much as (almost) 0.6 over and nearly 0.6 under on an iStat for a few outliers when comparing to the Lab reference (they used)

Interestingly the greatest variance seemed to occur around INR 3 ... for what's thats worth I'd warrant that needs 'more research'.

When comparing their same data set with the Coaguchek XS

the outliers on this is of enough significance to cause an alteration of dose.

This highlights to me that this is not an exact science and that reliability between sample groups is at best shakey.

When my CoaguChek XS strips expired, I didn't rush out to get new ones -- although it would be nice to compare the lab to the two meters, and the meters to each other between blood draws

yes it would have ... personally I am happy with my XS because it uses the most modern technology in making its INR assessment. What I read shows me that the more modern electrochemistry (rather than mechanical methods) is more trustworthy and Roche do standardise their reagents.

I'm comfortable with just the Coag-Sense.

well there's your answer right there.

I recently got a Hemochron - and only one test strip. The Hemochron is one that is used in operating rooms to test INR and other blood factors.

maybe where you are, but not here that's for sure.

It's one that hospitals put their faith in.

not in Australia ...

I went to the Anticoagulatin clinic this morning, and their Hemochron gave a result of 4.0. This was surprising to me --- my Coag-Sense on Monday (four days ago) gave me a 2.8.

too far away, you could quite easily have seen that variation within that time. I have done daily testing (as has Ola) and seen that sort of spike. So your methods will lead you to false conclusions if you have the premise that "INR doesn't change fast"

Although it's possible for the INR to go from 2.8 to 4.0 so quickly

quite

I think that you need to apply more strict rigor to your testing, graph and table it or you just can't conclude much from it and any decent peer reviewer would call it into question for validity.

HTH

PS

Point-of-care testing of the international normalized ratio in patients with antiphospholipid antibodies
Stephanie L. Perry1, Gregory P. Samsa2, Thomas L. Ortel1,3
1Division of Hematology, Department of Medicine, 3Department of Pathology, and 2Center for Clinical Health Policy Research,
Duke University Health System, Durham, North Carolina, USA

Summary
Antiphospholipid (APS) antibodies can influence the results of clotting
tests in a subset of patients, which can be a major obstacle in monitoring warfarin. The aim was to determine if point-of-care testing of the International Normalized Ratio (INR) is influenced by antiphospholipid antibodies.
...

In a subset of APS patients, the ProTime system will not yield an INR result and the HEMochron Signature (citrate and non-citrate whole-blood) INR results will exhibit elevated INR results. For this subset of APS
patients, we suggest using an alternative method to monitor warfarin.

{my underline}

 

[Protimenow]

I have been self-testing for more than five years, and recorded every test and every result during that time (although, with hard disk failures during that time, a few of the results may have been lost and unrecoverable). Since my 'event' a little over two years ago, I have been testing (usually) every week using one or more meters and monthly at a hospital lab. I went through an extensive period when I compared results of multiple meters to each other (Coag-Sense, CoaguChek S, CoaguChek XS, InRatio and InRatio 2, ProTime and ProTime 3, CoaguChek Plus) and to one or two labs. The hospital lab seemed to be the most reliable, although, of course, there's no absolute proof of this. I'm not sure how much more rigor I can apply to this testing -- if I wanted to, I could easily create a chart or graph of the results -- I just haven't yet.

I'm not quite sure about relying more on the electrochemical methods used by Roche, versus the mechanical one used by Coag-Sense. In my case, although CoaguChek XS was sometimes near to the lab results, and Coag-Sense was almost always slightly lower, I've seen some situations where the CoaguChek XS was MUCH higher than either lab or Coag-Sense.

As far as the current issue is concerned, in spite of the Clinic's insistence on the accuracy of its Hemochron results, I am more comfortable believing the results of my Coag-Sense and Hemochron (and not making any adjustments that are based on what I consider to be an inaccurate - erroneously high - result).

 

[pellicle]

Hi

I have been self-testing for more than five years, and recorded every test

yes, I know, and I know also we have discussed this before. But as we have talked about:
* you often do not test with all your tools at the same time (sometimes a day later)
* you do not always make the tests with the same subsets
...

so in my view if you want to compare them side by side, you must be thorough. I'd propose (again) that every time you do a lab test you then within one hour test on all your target machines. After you have a dozen or so readings you will have some data. Until then (to me at least) your process is too adhoc.

You asked for thoughts, I have given them. Digest and analyise them and discuss if you wish but equally reject them if you wish. In rejecting I would find it beneficial if you mentioned what is wrong with the methodology I have put forward.

Best Wishes

 

[newmitral]

Hi Protimenow,

Pellicle is "spot-on" in his assessment of the lack of scientific rigor in your approach to INR measurement.

Before adding my thoughts on that, mostly for the benefit of others, new to warfarin and anti-coagulation therapy, I'll address your question:

What would YOU do, given the same results?

I have posted before, and I give a different, but current, link here:

http://http://onlinelibrary.wiley.com/doi/10.1111/j.1538-7836.2009.03652.x/pdf

to a paper recommending a two-step algorithm for warfarin dosing. Given the results of a 4.0 (assuming you accept that as the baseline result), this algorithm recommends:

"Do not hold warfarin. If rising or high on two or more occasions, decrease weekly
dose by 10%; repeat INR determination in 7–14 days"
So, I would do nothing, except perhaps eat a few extra greens, and test again in a few days or a week to see if the 4.0 repeats or gets higher.

However, to give a more complete answer on what I would do (and am doing) about the issue of INR accuracy in general, I simply accept the fact that INR measurements from my home test meter, and even from the labs, are inherently inaccurate, and I don't obsess over trying to get an absolutely accurate value when it is unrealistic to expect such accuracy.

But, the reality of the situation is that it doesn't matter. Pick your pony and ride, as the saying goes. Use any one of the meters (or lab) as your guiding test result and try to keep your INR at its target value based on that chosen meter. Accept the fact that any given test result may be off by +/- 0.5 and live with it.

All you are doing by taking warfarin to raise your INR is to try to keep the odds of having an adverse event in your favor. If your INR is too low, with a mechanical valve, your odds of having a thromboembolism increase. But they don't increase much until you get below about 1.5 with modern heart valves. That's why staying above 2.0 gives you some safety margin.

Similarly, the odds of having a bleeding event increase if your INR is too high. Again, the odds don't change much until you get above 4.5 or so. Therefore staying below 3.5 (or even 4.0) still gives you a safety margin.

The most important thing to realize is that you can have a stroke or internal bleeding event even if your INR is perfectly at its target value.
- You can have a stroke or internal bleeding event if you have a tissue valve and are not on warfarin.
- You can have a stroke or internal bleeding event if you have never had any heart-related issues in your life and still have your native heart valves and take no medicine at all.

Being at your target INR does not make you invulnerable. Falling outside your target range does not guarantee a stroke/bleed. All you are doing is making the odds as favorable as you can.

I know you had a TIA and the hospital measured your INR below your target. I'm truly sorry that you had such an event, but you may have had that same TIA even if you had been perfectly anticoagulated at your target value. A "Post hoc ergo propter hoc" approach seems to have created an obsession with INR test accuracy as a result. Usually the doctors will raise your INR target by 0.5 after you have had one TIA to improve your odds against having another one (not a guarantee, just improving the odds). This increases your safety margin on the low side.

Again, to answer your "what you you do in the situation" question, I would pick one meter and use it as my "official" result until such time as a rigorous comparison to a statistically significant number of lab tests gave me a reason to change.

The reason it is not worth obsessing over the accuracy of the INR measurement is that it doesn't really matter much. The odds from all the statistical studies show that your risk hardly changes much at all until you get well away from the target "optimum" value for your particular medical condition. The safety margin is wide enough to accommodate meter/lab inaccuracies as well as variations in metabolism.


I don't intend this to be any sort of personal attack, and please accept my apologies if it sounds too harsh, but I want to get the message across to new valvers reading these posts that most of us don't have any issues with INR test result accuracy.

 

[dick0236]

The reason it is not worth obsessing over the accuracy of the INR measurement is that it doesn't really matter much. The odds from all the statistical studies show that your risk hardly changes much at all until you get well away from the target "optimum" value for your particular medical condition. The safety margin is wide enough to accommodate meter/lab inaccuracies as well as variations in metabolism.

.

Amen!!!!! Thanks newmitral for your post. You really put this INR stuff in its proper perspective. It is not "rocket science", nor should it be......and sometimes, forum chats, like some on VR.org probably scare the hell outa folks new to warfarin.....or considering a mechanical valve and ACT.

My range has always been 2.5-3.5 and I am OK with ANY INR between 2-4....altho, if I test around the upper(4) or lower(2) limit I make some minor diet alterations(more or less greens) and maybe test a little more frequently until I get back in the 2.5-3.5 area. It is not necessary to test, or stay, within narrow INR ranges and I like your idea....."pick your pony and ride it"......use the lab, or any of the various available meters and rely on it unless it really gives a strange result.

At the risk of really sounding "ancient" I was on warfarin when the available PT"pro-time"(now INR) blood test where really untrustworthy and the only true test was "if you had blood in your urine you PT was too high and if you had a TIA or stroke it was too low". This was probably when all the horror stories about warfarin began to emerge LOL. With the modern testing methods and available information on warfari, there is little reason to fear the drug....or obsess over INR.

We could, and probably have, posted enough on warfarin testing to fill a book. It is something that those of us on ACT have to do to minimize, but not eliminate, potential problems....but let's not make it out to be more difficult than it really is....."take your pill, test(your choice of method) routinely, and go about living a normal life"

 

[Protimenow]

Newmitral -- thanks for pointing out the obvious about INR and risks.

One thing that I may not have mentioned much previously is that I have done Master's Degree work on Biostatistics and Epidemiology and was one advisor (don't ask) away from a Master's degree (and probably, if I had continued to pursue it) a Ph.D. in Biostatistics and/or Epidemiology.

My situation yesterday -- being unable to test my blood for about four hours after the test at the clinic was atypical -- I didn't even get OUT of the clinic until two hours after my test.

When I did my meter comparisons, I tested with all my meters, and within a few minutes of each other and usually less than an hour from the lab tests. I haven't posted the results - either as a chart or as an extract from my spreadsheet - but I'm comfortable with my determination that, for me, the Coag-Sense (with a fairly consistent result that is slightly below labs and CoaguChek XS results) is the most reliable for me because it helps me to keep my INR above 2.0; and that the CoaguChek XS is my second choice (with its results close to, and sometimes above, those of the lab and my Coag-Sense). I ruled out the InRatio (classic and InRatio 2) because of a consistent history of seriously overestimating my INR. I ruled out the ProTime 3, basically because it was a bit more of a pain to use than the other meters were.

The clinic at first wanted me to skip a dose. Later, they relented and wanted me to take 1/2 dose then return to full dose the rest of the week. (At some clinics, they get a lab test for any INR at 4 or above - this one didn't).

My goal, as others have stated, is to stay in a range that reduces my risk of hemorrhage (INR too damned high) or TIA (INR too low). With what can now be considered an 'older' valve (1991), I am most comfortable with an INR between 2.0 and 3.5 (assuming that a 2.0 on my meter actually MEANS 2.0 and not 1.5 - 1.8). An INR between 3.5 and 4.5 or so just means more greens, perhaps a reduced one time dose, and a bit more caution than usual.

I'm not obsessing about my INR, but I DID want to determine which meter is to be most trusted. I've already stated which ones I trust the most.

Keeping an INR that is in range on those meters is what I'm trying to do. I realize that there are still risks of other events that are entirely unrelated to warfarin usage. I (we) all have to live with that.

 

[newmitral]

Hi Protimenow,

I gather from the tone of your post that you took offense at my answer. If I gave offense, I apologize. As I said before, this was not my intent.

Your original post, which started this thread specifically said:

Here's my dilemma: .... DO YOU AGREE with my conclusion? What would YOU do

I tried to give you the help and opinion you asked for. I'm sorry if you did not want to hear the answer to the question you asked.


I gave you a reference to a published scientific study on INR management, which you seem to completely ignore.

thanks for pointing out the obvious about INR and risks.

You're welcome. Sometimes pointing out, or restating what is obvious to you, with your years of experience on ACT, may be useful to other new readers of the forum who may not have that experience. Often restating the obvious is helpful, or called for, depending on how the question was asked, so that the reasoning behind the answer will be more clear.

I have done Master's Degree work on Biostatistics and Epidemiology and was one advisor (don't ask) away from a Master's degree

I really don't want to get into a comparison of credentials, as this only serves to polarize positions, but I actually have my Masters Degree from MIT, and was accepted into the PhD program at MIT but took a job offer for a positiion in industry instead (I went for the money over the prestige - my decision, not my advisor's).
I say this not to boast, nor to in any way diminish your own academic credentials, but to let the other readers know that my previous answer was grounded in a sound understanding of the mathematics and statistics behind the ACT risk studies. In addition, I have a good understanding of the technology behind the INR test meters, and how meter calibration and accuracy may effect warfarin management.

I am most comfortable with an INR between 2.0 and 3.5

At the risk of offering more advice that you don't want to hear, I am very surprised that you would accept a target INR range of 2.0-3.5 based on your knowledge and experience.
All of the published medical studies I have seen would indicate that in your particular situation - older generation valve and having had a prior TIA - your target INR range should be 2.5-3.5 (i.e. target INR of 3.0 with range +/- 0.5 from that). I am not a doctor and I'm not trying to offer medical advice, but from your original question "What would YOU do", I would speak with my doctor and ask him why he has given that 2.0 lower bound rather than 2.5.

I DID want to determine which meter is to be most trusted. I've already stated which ones I trust the most.

If you already have your answer, then the question is moot.

Nevertheless, for the benefit of other readers, I wanted to point out that the desire for measurement precision beyond what is typical, or even realistically available, is not necessary for the successful management of INR on warfarin. There is no need to make it seem like a complicated problem when for most people, all of the scientific evidence shows that any of the available FDA approved INR test meters will do just fine. Unless you have specific evidence to the contrary, ALL of the meters can be trusted for this purpose if used correctly. Just pick the one you like best and manage your INR based on that.


Sorry to again state the obvious, but your clinic is obligated to manage your INR based on their own meter results. It is a question of legal medical liability for them, so that's their choice. It's your choice to follow their recommendation or not. Most of the time my doctor's recommendation relating to warfarin dosage agrees with the protocol in the paper I referenced above. On the rare occasion when his office calls and gives me advice to change dosage in a manner I know to be incorrect, usually when another doctor is covering for my doctor, I just say "thanks very much" and then follow the dosage protocol that I know works best for me. No need to argue and try to convince them. By the next week's INR measurement I'm always back in range, and everyone is content. After some experience with my own personal INR management, I showed my doctor a few medical journal publications relating to how I would like to manage my INR, and also my desire to try vitamin-K2 supplementation for stabilization, and he was all in favor of helping me doing it my way based on that supporting medical evidence. Each of us can choose how easy or difficult we want to make it.

 

[Protimenow]

No offense taken.

Actually, a 2.0 on my meter is more like a 2.3 or 2.4 in the lab -- and I AM more comfortable at 3.0 +/- .5.

I've even registered concern with the clinic changing my range form 2.5-3.5 to 2.0 - 3.0 because I was concerned that others who had their own meters had higher risk of falling BELOW 2.0 if their meters reported higher than actual INRs.

As you noted, I let the clinic give its advice, and do what continues to work for me. I'm accepting the values from my meter(s), and managing my dosage based on that/those values. In short - I am not trusting the results on the 'lab correct' meter at the clinic.

I also plan to use low dose Vitamin K to help maintain my INR (and to give my body the other beneifts that Vitamin K provides), or will return to a vitamin 'pack' that includes 80 mcg of Vitamin K.

 

[pellicle]

Hi Protime

please don't take this the wrong way but ...

One thing that I may not have mentioned much previously is that I have done Master's Degree work on Biostatistics and Epidemiology and was one advisor (don't ask) away from a Master's degree (and probably, if I had continued to pursue it) a Ph.D. in Biostatistics and/or Epidemiology.

then you really should understand what I mean about the rigor and veracity of your data. I would expect my supervisors to tear it to shreds and this is not just about a paper, its about something close to your heart. (pun)

If you do seek the answers you wish then clarity will only come from a rigorous and methodical approach, and to be frank your data is often not up to that.

I admire what you do and why you do it, but I genuinely believe that without proper measurement approach (all done at about the same time, not this meter followed the day after) then your results are really difficult to use and rely on.

Lastly few here will actually encourage you (most will tell you you shouldn't bother), by contrast I actually take some time to carefully read what you post (often not simple) and then do my best to provide a critique.

Would you prefer I just gave the answer of "oh thats lovely, you'll do well, its great" : which says nothing and is basically unhelpful and really only mummy encouragement.

Best Wishes