Are INR test machines ever accurate?

[mile high]

I had an INR test yesterday at my cardio office and it was 5.9. At any level above 5.0 they draw a larger sample of blood and retest the INR. Well, that test showed the true INR was 3.9. Quite a difference. I inquired with the cardio who called me at home to tell me this, why was there such a difference in these two tests done minutes apart? Admittedly the second was with a much larger blood sample. The explanation was that he felt that above 3.5, the small machines do not necessarily provide accurate readings.

If that is the case, what good are the machines?

Just so you know, within the prior 24 hours of taking my INR I had taken numerous other drugs due to an excrutiatin painful neck condition. I had taken one 200 mg Celebrex, one Vicodan the night before so I could sleep with no pain, and also some muscle relaxer pills (5 mg) of flexirol.

They felt that this could also throw off the reading from the INR machine as opposed to a larger sample of blood.

What does anyone think of this? Are these INR machines that inaccurate?

 

[Ross]

Question, don't yell at me for asking, but when you say machines, did you test your own INR on a meter or are you refering to the lab draws work up? If it's the lab work up, I do not trust them at all. It depends on the technician doing the test, how quickly the test was performed after the sample was obtained, the difference in reacting agents used to conduct the test and on and on. I think this is what your refering too and I'd say the lab was in serious error on the first reading.

If your speaking of your own INR test machine. If you get two back to back tests that vary like that, something is wrong and it needs to be checked. It should only vary by perhaps .2 or .3 between readings.

 

[mile high]

Ross,


The original test was a finger prick test at the cardio's office by their technician who does this all day long. The second was at a lab next door that drew the blood from my arm and then do whatever they do to get my INR from that larger sample.

I do not self monitor. But this experience makes me wonder about the accuracy of the little machines people use to self monitor as well as the same type machine the cardio office uses to provide INR results. The question is is their something wrong with either of the two tests done on me or is their a problem with these hand held machines not revealed to us? I am not looking for a conspiracy, I just want facts to back up what maybe is taken for granted,ie that these hand held machinces are or are not accurate above certain INR levels.

 

[geebee]

I find my monitor to be a little less accurate when comparing it to a lab test (blood taken from arm vein) when it is above 5.0 (which has happened a couple of times recently because I was on Keflex. The monitor indicated 5.7 and the lab was at 5.1 (still not too far off in the realm of things).
However, when I did a "same time" comparison recently, my home monitor indicated 2.8 and the lab was at 2.7 so I am very confortable that, as long as unusual things do not occur (such as antibiotics, too much spinach, etc.), my home testing is accurate.
I am sure there are monitors that can fail (they are machines after all) which is why I will be lab testing every couple of months to make sure all is well.
BTW - I use an INRatio monitor.

 

[Guest]

Mile high - just got back from your rainy, snowy city - up and back today.

Not one of those things that the cardiologist told you has any basis in fact. The sample size has nothing to do with it. The checking test for INRs above 5.0 proved nothing except that the readings were different. How do you know which one was right. St. Agnes Hospital in Philadelphia killed several people a few years ago by depending on the hospital lab's INR as official. Someone forgot to program the sensitivity of the testing material into the machine. This cannot happen with the finger stick testers since the code is in each strip. Your doc just arbitrarily made a decision as to which one was "real".

There has never been even one well documented case report showing that Celebrex, Vicodin and/or Flexeril in any quantity over any length of time would affect the INR. Furthermore, if they did affect the INR, then why didn't they affect it on the "larger sample"? And since warfarin is a very slow acting drug, how did one dose the day before affect the INR to this extent? You have to accept the fact that your doc is just making up stories.

I have done about 30,000 tests and I'm sure that I haven't sent 30 people for a second test in the lab. The only time that I do this is when the machine is giving me an indiciation that there is something wrong with the person's blood -like it is literally too thick to go up the test strip.

To decide what to do with an INR of 5.9, you have to look at the person's history. Did your doc do this? If you have been getting good readings with the same warfarin dose for several months and then you get a strange one, the way to handle it is to skip one dose and then have the person take the same warfarin dose for another week or so and then retest. The chances are good that it will be back in the right range. If it isn't, then it is time to adjust.

I know my way works. My outcomes are as good as anyone in the country (world). I think your doc fed you a bunch of ..., sowed seeds of doubt in your mind and subjected you to a probably needless needlestick and there is no assurance that the results from the larger sample are going to produce a better outcome than my simple way.

 

[JimChicago]

One thing I learned from getting blood draw lab tests is that it depends on whether they put the right volume of sample size in the tube - if they overfill the tube then the result will be a lower INR since there is a certain amount of anticoagulant in the tubes and the test adjusts for that based on the volume of the sample. I learned that the hard way by having the lab tests come back a day or two later with the note **improper sample size*** or something like that.

 

[mile high]

Thanks Al and Jim.

I have a pretty steady INR, mostly between 2.5 and 4. No the Cardio did not ask about my previous INR's. He was not my cardio, just another in the group. I guess we are saying that the larger sample does not necessarily give a more accurate INR. I have many doctor friends in many areas of medicine and mentioned my experience to him. He commented that if you squeeze blood from a prick it can affect the amount of platelets which indicate the level of the clotting factor (I may not have described this quite accurately, but it is the best I can do). Kind of interesting tho.

Al, I am coming to Pueblo the last weekend of February for a soccer tournament for my youngest son.

 

[Guest]

I've heard that story about "tissue juice" too. The thing is that the thromboplastin used in strips is designed to only measure that part of the clotting cascade that deals with thrombin. This is why you can't measure the activity of aspirin with a INR meter. This field is full of urban legends.

 

[geebee]

Al,
I just want to make sure I understand. Are you saying that "milking" the puncture hole to get a larger blood drop will NOT affect the INR? Or is "tissue juice" referring to something else?
Thanks.

 

[Guest]

I haven't found that it affects things. I have done some pretty hard squeezing, particularly on older arthritic fingers where I was trying to not hurt them. If you aren't pretty aggressive with the stick, you can wind up milking the finger for a lone time. Remember that I have often said that I will compare my outcomes to any other clinic. It is the long-term outcome that matters most. Do I keep people out of the hospital or not? That is all that really matters.

 

[mile high]

For me and anyone else who has been following this thread, I am still puzzled by how two tests using blood that was taken minutes apart (one by a finger prick and the other with blood drawn from my arm into a tube) can yield an INR of 3.9 and 5.9. It would seem that one was right and the other wrong. Since I enjoy math so much I am tempted to add them together and divide by 2, yielding a 4.9.

I honestly do not fret about this too much because I am in range most of the time. From a larger perspective tho it makes me wonder about my finger prick tests and everyone elses too.

 

[Ross]

For me and anyone else who has been following this thread, I am still puzzled by how two tests using blood that was taken minutes apart (one by a finger prick and the other with blood drawn from my arm into a tube) can yield an INR of 3.9 and 5.9. It would seem that one was right and the other wrong. Since I enjoy math so much I am tempted to add them together and divide by 2, yielding a 4.9.

I honestly do not fret about this too much because I am in range most of the time. From a larger perspective tho it makes me wonder about my finger prick tests and everyone elses too.

One of the tests was certainly botched. Which one? If they use a Coaguchek, pehaps they started a new batch of strips and never ran controls to be sure of the accuracy. You don't have the problem with the Protime or INRatio units as controls are built in.

I suppose just about anything is possible, but before jumping the gun and sending you to the lab, I think they should have tested once more with the fingerstick to see if the wide variance still persisted. The machines are pretty accurate as long as the are programmed properly.

 

[Karlynn]

before jumping the gun and sending you to the lab, I think they should have tested once more with the fingerstick to see if the wide variance still persisted. The machines are pretty accurate as long as the are programmed properly.

That has happened to me. I brought in my Protime during a visit (usually do it once a year) to compare results with the Dr's Coagucheck. My machine's number was in range, hers was high. We redid both machines and came up with fairly close numbers and in range.

 

[Baldrick]

mile high,
I am glad to see your question as I had a very similar experience recently. I got a finger-stick test just prior to my TEE and I was very curious about the machine (I think it was a CoaguCheck) because my cardiologist recently told me that my insurance may now be covering these. The machine came up with an INR of 4.2 and the nurse said they wouldn't do the TEE if INR was above 3.5. I told her that I didn't think that could possibly be correct as my INR is almost rock-steady and I don't think it had ever been that high. Just to be sure, she drew blood & sent it to the lab. After waiting around for a few hours, she came back and said the INR from the lab was 3.0. As with you, it did not instill me with confidence about the little machines...
Just wanted to let you know you are not alone in your concerns.

 

[Ross]

I think if you both had your own and did what you need to do as far as running controls, your doubts would be cast aside. Your relying on someone else to make sure the machine is set up properly and they might well not have done so.

 

[ALCapshaw2]

I've seen 3 different versions of Coaguchek machines. My cardio's lab used the second version and I noticed a couple of 'suspicious' high readings compared with my Protime. I showed my historical data to my cardiologist and switched to another (hospital run Coumadin Clinic). A few months later I learned that the Cardiologist's lab had switched to a different manufacturers machine. Apparently they had some "correlation" issues with the second version Coaguchek machine. I've not heard of any problems with the original Coaguchek machines or the new S model.

I've forgotten the formula for INR now, but it involves an exponent which depends on the type of reagent used. Reagents which are associated with a high exponent are more prone to variation (and perceived error) at higher INR numbers. In the normal range, they are more accurate so a high reading may or may not be cause for concern but a normal reading is usually an accurate representation.

'AL Capshaw'

 

[Guest]

My clinic has done about 50,000 tests with CoaguCheks and I can't remember one oddball reading like this ever causing any problem that caused a person to need hospitalization.

I think that too much concern is paid to the number without considering how the individual is doing.

 

[Karlynn]

Not knowing how the blood samples are handled in the lab, or how long they sit before testing, I would tend to trust my machine or my doctor's first. Just my personal thought.

 

[Guest]

I added the sample size idea to my outrageous advice page.
http://warfarinfo.com/outrageous.htm

 

[Gail in Ca]

Excellent comments, Al.
I have double tested only twice and each time the results of lab vs protime were within .1.
I remember shortly after being placed on warfarin and having a reading from the lab at 4.0. The coumadin clinic nurse freaked out like it was a very high reading and of course I didn't know enough about it then to have an opinion.
She would have me hold a dose and then my INR would drop to 1.9! or lower.
Now, 4 yrs later I am so much more laid back and my INR has rarely been to 4 , almost always in range, or if a little high I don't worry at all. Of course I dose myself now and test myself too. If my reading was unusually high I would likely try to help myself, unless it was crazy high, which it has never been, even when I accidently took a dbl dose while traveling.
Gail