Coag-Sense versus CoaguChek xs and labs

[Protimenow]

Anyone who reads this thread knows that I've been a strong advocate of the Coag-Sense. Years ago, when I compared many meters to blood draws, my focus went to the Coag-Sense and CoaguChek XS. At the time, the Coag-Sense was usually slightly below the lab, and the CoaguChek XS was slightly higher. If my INR somehow rose above 2.5 or so, the difference between Coag-Sense and Labs from the CoaguChek XS was larger, with the XS being increasingly higher than the labs.

I chose the Coag-Sense as the meter that I would trust with my life. I would occasionally have blood draws.

Over the last few years, though, I've questioned the blood draw results, which were often 1.0 higher than the Coag-Sense. Coagusense (the manufacturer) swore that their meters were accurate. They still do.

It was a bit troubling that the labs - even the hospital labs - were always considerably higher than the Coag-Sense.

I recently starated using the CoaguChek XS (I was able to get some strips) and I began testing both against each other. As the INR increased, so did the difference between the two meters. The XS results seemed to be closer to the labs than the Coag-Sense (not SEEMED - WERE).

So, although Coagusense insists that their rsults are accurate, that they carefully test the accuracy of their strips, etc., etc., I get a strong feeling that the Coag-Sense is consistently reporting values that are too low - and that the difference from the lab results and the Coaguchek XS increases as the INR increases.

(I'm stuck going to a 'Coumadin Clinic' on Wednesday. I'll check their CoaguChek XS to my own).

Here are a couple questions:

If you use a Coag-Sense meter, and have gone to a lab or hospital for INR testing, how closely did the results compare to each other?

If you use a Coag-Sense meter and also test (or have tested) with a CoaguChek XS, how close were the results to each other?

I'm not only trying to verify that there IS a significant difference from the Coag-Sense and other methods, and also curious about whether there's something in my blood that somehow confuses the Coag-Sense (my guess is that maybe their values for the reagents are inaccurate - because the method detects an actual clot).

Any feedback would be appreciated.

 

[Pete81]

This is so incredibly difficult, I hear you say “significant“ and “(in)accurate” which are terms used in statistics, I think it is very difficult to try and say anything statistically meaningful about the differences you record that could then be used to correctly describe the performance of these machines. How big are the differences you have typically observed? You mention 1.0 between the lab test, do you mean 0.1?

 

[Protimenow]

I took graduate courses in biostatistics. I'm aware of statistical terms.

No - I mean differences of 1 or more -- 2.5 Coag-Sense vs 3.5 Coaguchek XS. THAT kind of 1. Tenths are within normal variations. I consider values within 20% to be acceptable.

If the lab results are the true 'gold standard,' then comparing the results of the two meters to the lab results would be the way to determine the margins of error.

The differences, at INRs of around 1.0, are small -- just a few tenths of a point. When the INR gets higher, the difference between the two meters increases pretty rapidly.

What I was asking about was a comparison of the two meters, and comparisons to the labs. I wasn't looking for any statistically relevant conclusions - I just wanted to see which meter gets closest to the lab results.

 

[Pete81]

Big differences indeed. I just had my Coagucheck XS check last Thursday. Of course when I go to clinic for this (2x a year) I test against another, clinic calibrated, XS. it was spot on so I would say the XSs can be precise but because of the fact there was no lab test I have no idea about the accuracy indeed. The lady also looked at my history (I report into an app) and noticed my 100% in range.....Your 1.0 differences put that in a whole different perspective.
Would the clinic have you cross-check your machine with lab test a few times? Trouble is though that because you don’t know what causes these differences there’s also no guarantee the best match then will still be best at a later point. I also don’t know if the biochemistry used in both machines is intrinsically different.

 

[Protimenow]

Pete, there are major differences in the way that the two meters work:

The Coag-Sense uses a small wheel that is built into each strip and is spun when a test is run. An optical system detects when a clot forms, and the wheel stops moving. When the blood is mixed with the reagent, the timer starts. An actual clot is formed. Based on the value of the reagent, the INR can be calculated.

The CoaguChek XS uses impedance to measure clotting time. As the blood clots, the impedance on the strip changes. At some point, the meter 'detects' that a clot has (or would have) formed. The INR is then computed, based on a reagent value.

In theory, being able to detect an actual, physical clot (Coag-Sense) should assure a more accurate result.

On the Roche site for professionals, I think that there IS some way to calibrate the meter. Most of the time, the meters - used in professional settings, and by home testers, usually aren't calibrated. Instead, quality control that is built into the strip takes care of checking the strip (and machine) to see that the test will be accurate (or something).

The Coag-Sense also does quality control and a High INR and Low INR test strip is included with each box of strips.

I can probably bring one or both meters to the 'clinic' and test them there - rather than test at home before I go to the 'clinic.' They have a hard enough time getting their heads around the fact that I self-test and manage my dosing. Bringing my own meters may just throw them over the top. So will showing them how I'm able to create a 6.75 mg dose. (7.0 seemed a bit high, 6.5 seemed a bit low, so I'll try the average of the two). In their world, if you can't get anything other than 5 or 7.5 (or smaller), that MUST be what you're taking.

If I can control myself, I'll see how they do their testing and dosing, then, perhaps, let them know what (if anything, other than letting me self-test) they've done wrong.

I would not be too surprised if my CoaguChek XS results are close to the one at the 'clinic.'

FWIW - like you, when I had to go to a clinic years ago, they called me Mr. Consistent. It was no surprise - I managed my dosing and was sure to be in range when I went to the clinic.

 

[Sheenas7]

I read about the differences and the meters and also the comparisons to blood draws and it makes me feel uneasy. It was really hard for me to start testing myself. I always thought it was a huge responsibility and I was afraid of doing it wrong. But with Covid19 I took the step to self testing. Everyone in this group has been a huge help. I am feeling more confident now and thankful that I don’t have to venture out every 2 weeks to get tested. I guess it makes me wonder how long can I trust this little machine? I have the Vantus model from Roche. Does accuracy change with time (years) and use?
Hope all of you are well and that we stay safe.

 

[ScribeWithALancet]

Anyone who reads this thread knows that I've been a strong advocate of the Coag-Sense. Years ago, when I compared many meters to blood draws, my focus went to the Coag-Sense and CoaguChek XS. At the time, the Coag-Sense was usually slightly below the lab, and the CoaguChek XS was slightly higher. If my INR somehow rose above 2.5 or so, the difference between Coag-Sense and Labs from the CoaguChek XS was larger, with the XS being increasingly higher than the labs.

I chose the Coag-Sense as the meter that I would trust with my life. I would occasionally have blood draws.

Over the last few years, though, I've questioned the blood draw results, which were often 1.0 higher than the Coag-Sense. Coagusense (the manufacturer) swore that their meters were accurate. They still do.

It was a bit troubling that the labs - even the hospital labs - were always considerably higher than the Coag-Sense.

I recently starated using the CoaguChek XS (I was able to get some strips) and I began testing both against each other. As the INR increased, so did the difference between the two meters. The XS results seemed to be closer to the labs than the Coag-Sense (not SEEMED - WERE).

So, although Coagusense insists that their rsults are accurate, that they carefully test the accuracy of their strips, etc., etc., I get a strong feeling that the Coag-Sense is consistently reporting values that are too low - and that the difference from the lab results and the Coaguchek XS increases as the INR increases.

(I'm stuck going to a 'Coumadin Clinic' on Wednesday. I'll check their CoaguChek XS to my own).

Here are a couple questions:

If you use a Coag-Sense meter, and have gone to a lab or hospital for INR testing, how closely did the results compare to each other?

If you use a Coag-Sense meter and also test (or have tested) with a CoaguChek XS, how close were the results to each other?

I'm not only trying to verify that there IS a significant difference from the Coag-Sense and other methods, and also curious about whether there's something in my blood that somehow confuses the Coag-Sense (my guess is that maybe their values for the reagents are inaccurate - because the method detects an actual clot).

Any feedback would be appreciated.

Protime,
I only have seven data pairs to date. I have not been able to collect any new data pairs since I last wrote to you. I am still under orders from my internist NOT to go to the Health Plan Lab for a new blood draw until well after the epidemic is over. At one point I had Pulmonary Sarcoidosis which is an immune system dysfunction of the lungs and makes me catch airborne infections with the greatest of ease.

My seven data points yielded a regression equation of y=0.87X+0.05
for a prediction of my Lab Draw value based on my Coag-Sense Value.
The largest difference were two time differences of 0.5 with one difference of 0.4 and the other four differences of 0.2.

The Coag-Sense was always higher then the Lab Draw value. This is the opposite of your experience. All the blood draws / finger pricks were made within 30 minutes of each other.

With the regression equation, 6 of the 7 predicted values were within 0.1 of the Lab Draw. The remaining data pair was within 0.2. I found a WHO standard that said for a six hour delay, the Lab Draw was allowed to drop by up to 10%. Applying the correction to the Lab Draw INR value showed as close a match between the values as did the original regression equation. The revised regression equation would be y=0.97X+ 0.06.
For me, either of these two equations say that the Coag-Sense plus its thromboplastin is marching in exact drumbeat to the monster machine that my health plan uses in conjunction with the thromboplastin they use. I cannot begin to explain how your results are so different. I will resume my testing in a couple months after the Northern Virginia area has been at zero hospitalizations for a considerable period and let you know if the Coag-Sense continues to march with the Lab Draw results.

My health plan still refuses to talk with me about my meter results but they now say that I should be self-managing with the meter since my results have been so good. I send them a report sheet following the pattern of London Andy's once a monthe. With weekly Coag-Sense testing and tiny (0.25) "mid-course" corrections, this works fine.

Walk in His Peace,
Scribe With a Lancet

 

[Pete81]

ProTimeNow, you sure have done some research....Interesting how, from what I get from your description, the CoagSense seems to contain a Tiny mechanical part, I would argue that the electrochemical solution of Coagusense would be more robust but that’s only slightly educated and more of a gut feeling.

it also sounds as if your clinic does not have a strong background and experience with large patient groups that are self testing and/or self monitoring. That’s unfortunate because what I experience with clinics across the country here, is the acknowledgment of the fact that a well trained individual is in principle always in better control of their INR than one with a bi-weekly lab test and a prescribed dosing scheme. It would be great if your clinic could somehow support you in getting good numbers so you can continue to self monitor with good confidence.

Something that just came to mind, I have seen people use the small glass pipettes to draw up some blood and transfer it to the strip instead of directly dripping it on there. It wouldn’t surprise me if even though this was also probably tested, the small differences in ‘operating’ are a potential cause for errors.

 

[pellicle]

As the INR increased, so did the difference between the two meters. The XS results seemed to be closer to the labs than the Coag-Sense (not SEEMED - WERE).

this is a well documented thing

... but it would be helpful if you provided something more than qualitative results, because by how much and at what INR you're calling High is pretty crucial to evaluating its significance.

Also in the past when you published this there was sometimes 24 hours difference in samples (IIRC, I'll see if I can search out those posts to check I'm not wrong there)

 

[pellicle]

was unable to find the discussion about sample dates being different but did quicly uncover this discussion

https://www.valvereplacement.org/threads/a-different-meter-coagusense.42343/post-776166

 

[ScribeWithALancet]

this is a well documented thing

... but it would be helpful if you provided something more than qualitative results, because by how much and at what INR you're calling High is pretty crucial to evaluating its significance.

Also in the past when you published this there was sometimes 24 hours difference in samples (IIRC, I'll see if I can search out those posts to check I'm not wrong there)

pellicle

In my 7 data pair sample, after I corrected with the regression analysis the values were within 0.2 for one value and within 0.1 for the other 6. (range of INRs being 2.0 to 3.2 for Blood Draw lab values and 2.4 to 3.7 for the Coag-Sense Meter values) Similarly, after correcting the Blood Draw Labs INR for a 10% reduction, 3 values within 0.2 - scattered top, middle, bottom. The remaining 4 within 0.1. Regression and correlation can work to minimize dispersion. My absolute values differed between the Lab and Meter much more then the corrected values did. Unfortunately, with the dispersion in values in the Roche test of different Labs (different machines) being as high as the dispersion between meters and a particular lab, who knows - "What is Truth?". I suspect you can only get regression analysis and correlation between one lab with one machine and reagent to work with one meter with one reagent.
I am looking forward to the ending of this epidemic and the gathering of a larger cloud of data both by Protime and myself. I suspect you are right that the divergence will be higher at the two extremes of INR beyond the safe range of 2.5-3.5 and the moderate range of 2.0-4.0. However, I have not collected data pairs in those areas.
Walk in His Peace,
Scribe With a Lancet

 

[pellicle]

Hi
nice work

I suspect you are right that the divergence will be higher at the two extremes of INR

don't quote me on that, that's established in the literature and even by Roche ... let me see if I can fish something out

 

[pellicle]

Scribe With a Lancet

Accuracy and clinical utility of the CoaguChek XS portable international normalised ratio monitor in a pilot study of warfarin home-monitoring | Journal of Clinical Pathology

Accuracy of the Point-Of-Care Coagulometer CoaguChek XS in the Hands of Patients - PubMed (may need access to journal)


Comparison and Evaluation of Two Point-of-Care Coagulation Analyzers for Monitoring Warfarin Oral Anticoagulant Therapy

 

[Protimenow]

Yes, the variance between the labs and the CoaguChek XS are well documented.

According to the former CEO at Coagusense, their meter was being used to verify the CoaguChek XS results if it reports an INR of 4 or above. Because the Coag-Sense meter uses a physical device to detect the actual clot formation, it's supposed to be accurate for higher INRs.

The Vantus should be fine for tests. I haven't heard of any XS or Vantus meters losing their accuracy as they age. I can't see how age would matter much -- a code chip that is included with each batch of strips provides a value that is used to calculate prothrombin time and INR. It wouldn't be the number of tests run, either - some clinics probably do more testing in a month or two than some users would do in many years (figuring that some clinics may do 5 tests a day, 5 days a week, means that this clinic does 100 tests a month. If a person does one test a week, it would take nearly two years to match what a clinic can do in a month). These meters are designed to keep working, on and on and on -- the cost of the meters isn't factored into the marketing plan -- it's the sales of strips that are where the real money seems to be.

I haven't run any regression analysis on any of my tests. I take the results exactly as I get them. A 2.4 on my Coag-Sense is compared to a 3.4 on my CoaguChek XS, with tests made on different fingers just a few minutes apart. That's enough to show that there's a difference in values reported by the two meters. At times, I'll repeat the tests just to make sure that the difference is repeatable, and not the result of operator error or some strange artifact that yielded an incorrect result.

Comparisons between the Coag-Sense and the labs has, for the past year or more (I have to check my spreadsheet), often as much as 1-1.5 below the lab results. Years ago, all three - lab, CoaguChek XS, and Coag-Sense were usually very close (with the exception of the XS being increasingly higher the higher the INR went).

I'm faced with some decisions that I'm not yet able to make:

The Coag-Sense strips, for the past two years (or so) are wrong. Perhaps they have an inaccurate reagent value. I've tested on two different Coag-Sense meters, and the strips were in nearly absolute agreement. I've tested with different batches of strips, and the values were again in near perfect agreement.

There's something in my blood that somehow confounds the meter. Something has changed, and the strips can't detect the correct values. Perhaps the reagent used reacts more rapidly to my blood than it does for most other testers.

I don't know which is correct. That's why I'm reaching out for user experiences.

 

[Sheenas7]

The other day I received a letter from Roche. They announced that they sold the it’s Coaguchek Patient Services to Bio Tel Heart. Effective June 23,2020. New service name is Remote INR by BioTel Heart. They said no action was required from my part. I know not everyone used them but I was sorry to hear. Hope no changes are coming.

 

[Protimenow]

The other day I received a letter from Roche. They announced that they sold the it’s Coaguchek Patient Services to Bio Tel Heart. Effective June 23,2020. New service name is Remote INR by BioTel Heart. They said no action was required from my part. I know not everyone used them but I was sorry to hear. Hope no changes are coming.

There probably won't be many changes - at least, not right away. They probably took over contracts to provide services - and you (or your insurance) may have some kind of contractual arrangement for fees that won't be renegotiated for a while.
It could just be a 'sale' that's just a paper shuffle. Maybe Roche wanted to create the appearance that they were separating Coaguchek Patient Services from Roche, the manufacturer and drug maker, and spun off BioTel Heart as a separate entity that is still closely related to, or owned by, Roche.
Perhaps not.

CoaguChek Patient Services, from what I understand, owns the meters. If you stop using their services, you have to return the meter. We'll see if Remote INR does the same thing.

I wouldn't worry about any major changes. Maybe they'll change a phone number to call in results.

 

[Sheenas7]

There probably won't be many changes - at least, not right away. They probably took over contracts to provide services - and you (or your insurance) may have some kind of contractual arrangement for fees that won't be renegotiated for a while.
It could just be a 'sale' that's just a paper shuffle. Maybe Roche wanted to create the appearance that they were separating Coaguchek Patient Services from Roche, the manufacturer and drug maker, and spun off BioTel Heart as a separate entity that is still closely related to, or owned by, Roche.
Perhaps not.

CoaguChek Patient Services, from what I understand, owns the meters. If you stop using their services, you have to return the meter. We'll see if Remote INR does the same thing.

I wouldn't worry about any major changes. Maybe they'll change a phone number to call in results.

Thank you for your comments. I guess time will tell.

 

[tom in MO]

...CoaguChek Patient Services, from what I understand, owns the meters. If you stop using their services, you have to return the meter. We'll see if Remote INR does the same thing..

I've had three different providers and each one told me the meter is free, I pay for supplies and service. This was true of my mother-in-law's meter as well. This is often true per my coumadin clinic nurse.

 

[Protimenow]

Yes. The meter is free--as long as you use their services.

I spoke to Roche last year when the Vantus became available. They told me that it was NOT for sale. They told me that it was sent to subscribers to the service, and that it would have to be returned if the person stopped using the service.

In other words - they'll let you keep the razor as long as they send you the blades. Once you stop getting blades, they want the razor back.

I don't know WHAT the other services do about their meters. I suspect that, if you only use the service for a month or two, they'll also want the meter back.

 

[MdaPA]

If you use a Coag-Sense meter and also test (or have tested) with a CoaguChek XS, how close were the results to each other?

My darling wife has been self-testing using CoaguChek XS for about 2 1/2 years now and during the times compared to the lab (Quest), the INR reported by the lab has mostly been lower by .2 or .3. More recently, however, we noticed the past few comparisons with the lab have been lower upwards of .5 (e.g. last lab comparison came in at 2.5 where CoaguChek XS reported 3.0). Upon noticing this pattern, I called Coaguchek Patient Services and spoke with someone in Technical Support. The representative said a 20% deviation from the lab or even against another machine was the upper limit due to variations in testing procedures, reagents used, and other limitations which are included in the manufactures literature (usually accompanies the PT test strips). The representative confirmed that we were testing according to procedure (e.g. putting sample on test strip within 15 secs, storing test strips are appropriate temp, expiration date, etc) and that none of the documented limitations applied. One possible factor the representative did mention that might interact with the reagent used with the XS PT test strips was vitamin C. Not sure regarding the truth to this as this is something that I did not see mentioned in the literature. Coincidentally, however, my wife had recently increased her vitamin C intake (from 1000mg to 1500mg per day) since the Covid-19 out-break. Deviations upwards of .5 were also noted prior to this up-take.

Yes. The meter is free--as long as you use their services.

This is my understanding as well after several calls with Roche looking to drop their service but keep/buy their machine.