Pre-analytical Variables for Coagulation Testing

- This is Lab Medicine Rounds, a curated podcast for physicians, laboratory professionals and students. I'm your host, Justin Kreuter, a transfusion medicine pathologist and assistant professor of laboratory medicine and pathology at Mayo Clinic. And today we're rounding with Dr. Jansen Seheult, Assistant professor of laboratory medicine and pathology in the division of Hematopathology at Mayo Clinic in Rochester, Minnesota, to talk about pre-analytic variables for coagulation testing.

Thanks for joining us today, Dr. Seheult. - Thanks for having me, Dr. Kreuter. - Hey, so why don't we kick off with, you know,

for our podcast audience, why is it important for them to appreciate kind of the role of pre-analytical variables in coagulation testing? - That's a good question. So I think 20 years ago, laboratory community and patient facing clinicians probably did not truly understand the impact of pre-analytic variables. It took some seminal papers with the Mario Banani in 1997 and then 2006 to show how prevalent these errors were.

At that time, they were done in stat lab testing, and I think he, he quoted a figure like 4,000 parts per million. If you do 1 million in laboratory tests, 4,000 of them could have errors associated with pre-analytic phase of testing. And then when he followed that study up, it decreased slightly, but it never went away.

So what we've learned over the last two decades is that we can hone in the analytical phase of testing quite well through standardization of our practices, advances in instrumentation and methodologies. But the pre analytical phase, and to some extent, the post analytical phase remain a challenge to tackle. And it is because it encompasses so many human factors in the process that we cannot fully automate.

Some of the more recent literature and coagulation testing in particular have cited figures as high as 5% of all coagulation testing being affected by pre variables. It is a bit challenging to capture the full scale of this. We can identify patterns in the lab that we know are definitely, or we are confident are associated with pre variables. My lasting concern has always been what is the true extent of what we do not detect?

And those are the subtle patterns that might be associated with a true clinical phenomenon or it might not. And that is all the more impactful or problematic in a reference laboratory test setting like we have at Mayo Clinic, where we do not necessarily get information on the clinical history, the medications that the patient is on. We do not know the provenance of the specimen from the time it was ordered through the time that it arrives at us, most likely as an aliquot that is frozen.

So that's why it, it takes a lot of education collaboration to, to address or try to begin to address this problem of pre-analytic variables. - Well, that's awesome. I think you got my attention with the prevalence.

And I love, I think I, I really like this idea that, you know, what is it that we don't know that we don't know? And your point's very well taken that, you know, these may be artifacts or there may be things yet to be discovered in, in clinical coagulation. Maybe can we kind of take a little bit of a dive into maybe a few of those, you know, are there a few of the common pre-analytic variables that you're, that you've talked about recently that are podcast audience might, should be aware of?

- Yeah, I, I like to categorize or classify pre-analytic variables into five pin buckets. I call it the PC PST, Penta patient Collection. And then the last three, processing storage and transportation don't necessarily happen in a single order. And they can be duplicative process. We could store a sample multiple times along the chain. So they're almost like a cycle when it comes to the most prevalent, I think in the laboratory, we like to think about things that we can test for.

Hmm. So in the laboratory, if you asked a trainee what are some pre-analytic variables, you're probably gonna hear things like hemolysis, ictaurs, lipemia, those are testable problems. We have pre an check modules that we can test for those using different light wavelengths. In the grand scheme of things, they probably are not the most prevalent. I think patient related factors are probably the most prevalent throughout all laboratory disciplines.

But in coagulation in particular, when I say patient related, I'm talking about medications. Many patients who are being tested for a bleeding or thrombosing disorder might be on an anticoagulant or a procoagulant drug. And those drugs can actually influence the results what we report from the laboratory. There are also patient diseases that might have an impact on laboratory assays testing for a different disease. So, or, or even the same disorder.

So let's say a patient comes in with acute venous thromboembolism. Many professional organizations have said that that is not the right time to be testing for thrombophilias or prothrombotic states. And that is because the clotting cascade is disarrayed at that point when the patient has an acute clot and you can have a, a decrease in natural procoagulant and anticoagulant factors.

So that's one case where if you see a borderline abnormal result, you don't know if it is related to clot consumption or if it's a true reflector of, of the patient's underlying disease state. If patients get thrombolytic therapy, a clot buster drug, you can see influences on procoagulant factors like fibrinogen factor five and factor viii. This is not an artifact, this is a natural phenomenon that is a consequence or sequela of giving a treatment.

But we have to be cognizant that if we request laboratory assays for procoagulant factors, they might be influenced by some of these therapies or medications. So I, I think the patient related factors are probably the most prevalent. And those are related to the patient condition or the, the medication. One area that is, is kind of pervasive throughout laboratory testing is did you order the right test and also patient misidentification problems.

I think those are big, no-no from the point of view of blood bankers and transfusion medicine staff, but it's, it's also a problem from our point of view, we don't have direct knowledge of when those occur in coagulation testing. It's not like it's a reportable event to the FDA as the analogy with blood banking or transfusion medicine, but I, I presume that it happens just as frequently.

So I think we need to have robust systems deployed at the point of collection to make sure those errors aren't happening as well. - Wow. I I really like how you're, you know, one of the things for our audience to appreciate the perspective that you're adding to us, right? This idea of there's a lot of the things that are maybe top of mind. You're talking about lipemia, hemolysis, these things

that are testable as you, as you put it, right. You know, these things that we're aware of it and that's why they're top of mind. But highlighting that they're probably not the most prevalent. And then highlighting that there's several talking about diseases, medications, I certainly appreciate this challenge of when somebody has an acute DVT, right? That's when somebody wants to do coag testing, and I'm totally on board with, that's probably not the ideal time.

That kind of maybe leads me into my next question, which is, you know, as you're saying, these are things that are a challenge to automate. How do you, how could you envision our medical community kind of better coming together to together to collaborate to kind of mitigate some of these patient related issues?

- Yeah, the, the medical community probably needs to be very, an expansive view of what the medical community is from allied health staff that do phlebotomy or specimen collection all the way through laboratory staff that can identify errors as they happen. But I, I would premise that the medical community probably includes industry partners as well.

So I, I think if we were to tackle Pre-analytical variables, we have to tackle it at each step of process along the way where an error can happen if you look at the failure modes of what a pre variable can cause. So we, we are developing better systems for patient positive patient identification. I think we have trained our collection staff very well.

And, and, and that's a testament to decades of efforts and initiatives that went into education and making sure people are aware that positive patient identification is critical to providing an accurate laboratory test result for a patient. We also have to facilitate some of that. So making sure that phlebotomy stations are set up in a way that, that are compatible or commensurate with the safety critical tests that they're performing.

Sometimes we don't think of phlebotomy as safety critical, but if you're gonna act on a laboratory test result, I, I think that it is a safety critical task that they're performing. One shortcoming, I believe is that, and this gets into physician burnout slightly and, and allied health staff burnout, we would like more information about the patient when we receive a specimen to perform testing.

There is always a trade off between the physician or allied health staff effort that goes into populating what our requirements might be because they're dealing with such a high volume practice, they're collecting multiple specimens. The person that is collecting the specimen may not have knowledge of the condition that is important to the laboratory and when interpreting the results. So that is a gap. I don't think it is an easy gap to address.

Maybe computerized physician order entry has gotten us one step further along the way to, to addressing that. But that has created more administrative overhead, I think, for physicians and allied health staff as well. So I think it probably is contingent on pathologists and laboratorians to identify maybe the assays or collection practices that where pre variables may be most critical and ask for key pieces of information only for those, or for a small set of assays.

I think you get buy in a couple ways from that. You, you are providing education by telling people these are things that might influence the result that and, and the interpretation of those results that we provide. But you probably also encourage people to start thinking about how these might affect other assays as well. And then I, I talked about industry partnerships. So I think some of that is, is creating technology instrumentation assays that are less susceptible to pre-analytic variables.

We have some examples of that in coagulation testing. There have been a lot of efforts to make ddi er assays less susceptible to heterophile antibody interferences. The, the, one of the challenges in coagulation testing is a lot of, of what we do are functional assays and they rely on a complex interaction of multiple proteins and enzymes along with the methodology that we use to detect an endpoint.

So it is, it will remain a challenge to address all the pre-analytic variables, primarily by optimizing or advancing the science of coagulation assays or instrumentation itself without addressing all of the steps that go before. - Hmm. And with, in your practice in this kind of building, this collaboration, I really like that you're kind of broadening this out to who is in the medical community and really highlighting that.

'cause I think that probably accurately ref reflects our listenership, you know, are there ways or things that you have worked on doing and, you know, is there a, a story that you can kind of share about in an intervention or something that was tried and how that worked out? - Specifically addressing a pre-analytic variable? Yeah. - Yeah. - I, I can think of a couple, I'll follow on the example I just gave about d dimer testing.

So a lot of those new methodologies are not available yet for use in the United States because they're not FD approved or cleared, but they are in use in, in Europe. We have had to come up with workarounds to addressing this d dimer, heterophile antibody interference issue. And it was a laboratory partnership between our special coagulation lab and the core clinical chemistry lab because the chemists have a long history of dealing with heterophile antibodies. They are the experts in this area.

I, I think it starts with identifying that there might be a problem. And many times in DDI er testing that is a patient who has had hundreds of thousands of dollars spent on complex radiographic workups to explain a significantly elevated DDI er without an explanation. And then it takes an astute clinician to identify, well, could this be a pre analytical variable

that is in interfering with the D dimer assay? And that triggers, now we have a, a, a stepwise algorithm that we follow that includes things like dilution studies, testing with a second method, try to test with hetero vial blocking reagents. So we've created a process now for doing that, but it starts with the astute clinician identifying that there's a problem here and that that process we, we can't replace.

I, I think we, we need to train more clinicians to try to identify these issues when they come up. But that's where I think partnership with patient facing clinicians and laboratorians and even within the Department of Pathology and lab medicine collaborations among work units or sections can, can provide a lot of value.

- Yeah, I I love that you're showing that example, you know, that there are experts within our community and how we can go to the clinical chemist and they can really inform and help the coagulation laboratory out. Right. The answers can sometimes come within from within - And you would be surprised.

Well, we were surprised to know that they had very detailed standard operating procedures to deal with these issues, but we sometimes live in a silo where we don't know what exists outside of that silo. - Exactly. Wow. So what do you think the, the future of coagulation testing looks like? You've kind of preempted a little bit of this, but maybe to kind of wrap up this interview, what do you think the future looks like?

- I think hemostasis testing is heading towards genomic testing and also proteomic mass spectrometric based testing. Whether the assays that we use today will become antiquated. I cannot predict that it is likely that for the next 10 to 15 years, we will add to the activity menu of assays that we perform, but they would supplement one another. So you would start screening with the functional coagulation tests that we used today and then reflex on to more esoteric, potentially more costly assays.

Eventually, when the cost of doing genomic testing is less than the cost of doing a panel of coagulation assays, the calculus might change.

I don't know if that will happen in my field. It potentially could, but one of the challenges to recognize is that as we broaden that activity menu, we are now dealing with a even more di diverse array of pre-analytic variables that we have to think about. We work actively on developing mass spectrometric assays for antigen levels or coagulation factors, or even for activity assays.

And we're learning every day that the same pre-analytic variables that might influence our conventional coagulation assays are not what you consider when you think about a mass spec assay. But there, there are unique considerations for a mass spectrometric method. And the same can be said of genomic testing. I think there are additional issues that you have to think about. Patient identification I think is gonna be critical for all of this and ordering the right test.

Those are two commonalities that apply across all of laboratory testing. And when it comes to things like, does a medication interfere with a functional coagulation test, that may not be an issue anymore. If we are doing a mass spec asay that's using a substrate and the coagulation enzyme is cleaving that substrate, so the field is gonna change.

I don't know what adoption will look like, and I don't know what the rate of that change will look like, but it's, it's actually an exciting time to be in coagulation right now. And, and we're trying to implement procedures and perform the correct validation verification activity so that we fully understand upfront I priori what these pre-analytic variant might be before we implement these assays. - We're rounding with Dr. Seheult talking about pre-analytical variables for coagulation testing.

I, I always appreciate your insights and, and the perspective you bring to this topic. Thank you so much, Dr. Seheult. Thanks - Very much for the opportunity. - And to all of our listeners, thank you for joining us today. We invite you to share your thoughts and suggestions via email to MCL [email protected].

If you've enjoyed this podcast, please subscribe and until our next rounds together, we encourage you to continue to connect lab medicine and the clinical practice through educational conversations.