Amy Carpenter
August 2025—Anticoagulant interference, INR calculation, and D-dimer reporting are among the coagulation testing analytical and postanalytical variables that merit caution. Andrew Goodwin, MD, and Eric Salazar, MD, PhD, in a CAP24 session dug into why and what the requirements are, including for viscoelastic testing.
Second of two parts
Part one in the July issue
Protocols are necessary for the weekly, if not daily, anticoagulant interferences laboratories encounter, said Dr. Goodwin, professor, Department of Pathology and Laboratory Medicine, University of Vermont Larner College of Medicine, and medical director of the thrombosis and hemostasis laboratory, University of Vermont Medical Center.
“The heparins, the vitamin K antagonists like warfarin, the direct oral anticoagulants—which include the direct Xa inhibitors and the direct thrombin inhibitors—can be very problematic, particularly if you think about the lupus anticoagulant that has been known to cause false-positives or false-negative results as well” (Favaloro EJ, et al. Res Pract Thromb Haemost. 2024;8[3]:102385).
The sixth edition of “CLSI H21: Collection, Transport, and Processing of Blood Specimens for Testing Plasma-Based Coagulation Assays,” published last year, and other publications provide recommendations on how to address samples in which an anticoagulant is identified (Favaloro EJ, et al. Res Pract Thromb Haemost. 2022;6[2]:e12676).
“The key is, Do you identify that anticoagulant?” said Dr. Goodwin, former chair and now advisor to the CAP Hemostasis and Thrombosis Committee. “Do you have a process in your laboratory to look for potential interfering substances due to anticoagulants? Do you check a thrombin time? An anti-Xa? Do you check the INR to detect the potential of an anticoagulant?”
In Dr. Goodwin’s laboratory, the lupus anticoagulant starts the first step in the cascade in an anti-Xa assay. If there is a measurable anti-Xa, a heparin neutralization will be done, “and if it doesn’t neutralize, we stop.”
“If we identify anti-Xa,” he said, “we tell them that we have anti-Xa activity that’s unaffected by heparinase,” which suggests the presence of a direct Xa inhibitor or other novel anticoagulant that could lead to false-positive or false-negative results, and that repeat testing once the patient is free of anticoagulant therapy is recommended, if clinically indicated. “The guidelines say if you’re on a DOAC, you should interrupt for 48 hours, maybe longer if they have renal impairment, transition them to a low-molecular-weight heparin if necessary, and then draw them at a trough level.”
Dr. Goodwin cites a table in a 2022 article that lists the effect of anticoagulants on tests used to investigate lupus anticoagulant and other routine coagulation tests (Favaloro EJ, et al. Res Pract Thromb Haemost. 2022;6[2]:e12676). The dilute Russell’s viper venom time test in his laboratory will neutralize up to one unit of unfractionated and low-molecular-weight heparin. “So our protocol is, if the patient has an anti-Xa greater than 1.0 [U/mL], we stop. We don’t care if it’s heparin or a DOAC; we stop because we don’t believe our system can neutralize that degree of heparin.”
The package insert for the laboratory’s silica clot assay used to list what concentrations it would neutralize, but it now says to follow the CLSI H60 guideline, which says samples containing heparin may exhibit falsely prolonged clotting times. “We still have the 1.0 U/mL cutoff for any anti-Xa activity,” he said.
Thrombin time is very sensitive to detecting not only unfractionated heparin but also a direct thrombin inhibitor, Dr. Goodwin said.
In the CAP hematology and coagulation checklist, HEM.37165 Coagulation Testing and Therapeutic Anticoagulant Recommendations requires that laboratories provide to clinicians information on the laboratory tests used for monitoring heparin, low-molecular-weight heparin, direct thrombin inhibitors, and/or oral anticoagulant therapy and about potential interferences of anticoagulant medications on coagulant testing.
The novel antihemophilic medications on the market warrant attention, too, he said. He asked the medical center pharmacy for a list of drugs on formulary for patients with hemophilia, read the package inserts, and identified some drugs that his laboratory cannot monitor. “So I communicate this to our clinicians.” Additionally, there is a growing amount of evidence-based literature describing how various assay methodologies, activators, reagents, and so on measure novel antihemophilic medications.
Among the postanalytical variables that can affect coagulation testing is the INR calculation, and the verification of these calculations is cited relatively frequently in CAP accreditation inspections, Dr. Goodwin said.
HEM.37830 INR Calculation Adjustment for ISI requires the calculation of the INR to be adjusted using the appropriate ISI value for every new lot of PT reagent, changes in types of reagent, or change in instrumentation.
GEN.43450 Verification of Calculations Producing Patient Results requires calculations that use patient results to produce other reported results be reviewed every two years or when a system change is made that may affect the calculations. The requirement note says more frequent checks may be required for certain specific calculations, as delineated elsewhere in the checklists (e.g. INR). “So INR is required annually, not every two years,” Dr. Goodwin said.
HEM.37860 Report Verification Criteria requires that patient reports be checked for correct INR calculations, patient values, and reference intervals in these circumstances: a change in lot or type of PT reagent, or instrument, establishment of a new PT reference interval, and a change in INR calculation. In addition, in the absence of the above changes, reports should be checked at defined intervals. A note suggests that the calculations be checked at the INR values of 2.0 and 3.0, “just to make sure the calculation is working,” Dr. Goodwin said.
Units and magnitude of measurement for D-dimer are another postanalytical reporting variable.
“Unfortunately, D-dimer reporting has evolved such that there are two units you can use to report,” said Dr. Salazar, associate professor, Department of Pathology and Laboratory Medicine, University of Texas Health San Antonio; medical director of clinical laboratories, UT Health Multispecialty and Research Hospital; and vice chair, CAP Hemostasis and Thrombosis Committee.
They are fibrinogen equivalent units (FEU) and D-dimer units (DDU). “They’re not the same, and different reagents will give you either DDU or FEU,” he said. FEU equals DDU times 1.74, but a multiplication of two is commonly used (FEU = 2 × DDU). (DDU = .57 × FEU and .5 is commonly used.)
In the CAP checklist, HEM.37924 D-dimer Unit Results requires that the unit type (FEU or DDU) and unit of magnitude (e.g. ng/mL) reported with the patient results be the same units as generated directly by the D-dimer method (following the manufacturer’s product insert). If different units are reported, the laboratory must verify the correct conversion of the units annually. “Mostly, the [Hematology and Thrombosis] Committee and CAP would prefer that you report the package insert units,” Dr. Salazar said, noting that some laboratories are converting units.
HEM.37930 D-dimer Reporting says that if a D-dimer test is used for the evaluation of venous thromboembolism, the laboratory must report the VTE exclusion cutoff value as stated by the manufacturer. If the D-dimer test is intended for other purposes (e.g. DIC evaluation), a reference interval is required.
And HEM.37925 D-dimer—Evaluation of VTE requires that if a quantitative D-dimer method is used in evaluating venous thromboembolism, the method must be valid for this purpose. “Not all D-dimer reagents are approved or labeled for exclusion purposes,” Dr. Salazar noted.
Members of the Hematology and Thrombosis Committee are studying whether, in proficiency testing data, there are problems in how D-dimer units are being reported. As part of their work, they reviewed various D-dimer reagent package inserts and found variability in the unit types. “Some are DDU; some are FEU. Some provide a threshold for exclusion of venous thromboembolism. Some do not,” Dr. Salazar said. “Some are FDA approved or cleared for venous thromboembolism evaluation, and some are not.” Some assays are approved for exclusion and others are labeled as an aid in the exclusion of DVT/pulmonary embolism or in the assessment and evaluation of DIC/PE. “Some are only labeled for the quantitative measurement of D-dimer. It’s important to understand whether your reagent is labeled for exclusion purposes,” Dr. Salazar said.
In one proficiency testing participant summary report for D-dimer reagent results, 32 laboratories reported Hemosil D-dimer HS 500 results in ng/mL or µg/L DDU (mean: 1,700.219 DDU), and 657 laboratories reported Hemosil D-dimer HS 500 results in ng/mL or µg/L FEU (mean: 1,972.682 FEU).
“The package insert says to report in FEU,” Dr. Salazar said, adding that for the 32 laboratories that reported in DDU, “it looks like it’s an FEU unit,” judging by the mean value they reported.
“Whether this is filling in the wrong bubble, reporting in the wrong units, converting incorrectly, we don’t know. But there’s clearly some kind of problem.”
In another example from the participant summary report, 66 laboratories reported Quidel Triage D-dimer reagent results with a mean that looks like an FEU value, 1,715.152, but the package insert says to report in DDU, Dr. Salazar said. “It’s very confusing.”
Members of the committee harmonized all units in the proficiency testing D-dimer data to ng/mL FEU, he said, and plotted histograms of all data per reagent instrument combination. They found that for most of the data, there was a dominant population and a clear mean, but they also found pockets of distinct clusters of data that differed from the dominant population, with some reporting twice as high and others half as high. “Clearly these are participants that are reporting results that suggest the wrong units,” Dr. Salazar said, noting the issue is present for every manufacturer. (A paper written by committee members is under review at Seminars in Thrombosis and Hemostasis.)
The committee also employed additional analyses to evaluate results further. “Mostly when you report the package insert units, you seem to get it right if we consider the mean of the peer group correct,” Dr. Salazar said, “but if you report in units that are different from the package insert, it often appears wrong. So labs that report D-dimer proficiency testing in units that differ from the package insert tend to have different results.”
To use the age-adjusted D-dimer for elderly patients who tend to have a higher baseline D-dimer, “you have to use the right conversion, the right multiplier, depending on the units you’re using.” If using DDU, multiply patient age by five. If using FEU, multiply patient age by 10.
The problem of reporting D-dimer in units that differ from the package inserts is one that International Society on Thrombosis and Haemostasis 2024 Congress participants discussed at length, with the aim of bringing manufacturers together to harmonize units and “land on one,” Dr. Salazar said. For many other coagulation tests, there is an international standard managed by ISTH. “There is no D-dimer standard,” he said.
Traditional viscoelastic testing is a cup-and-pin mechanism; the whole blood and pin are in the cup, and either the cup or pin rotates back and forth. Clot time, stability, and development are measured.
The Haemonetics TEG 6s is an example of a newer mechanism at work. The whole blood is put into a microfluidics cartridge and the sample is subjected to a vibration signal. “What you’re measuring is the shear elastic modulus,” Dr. Salazar said, which is directly proportional to the force applied across a gel with a height inversely proportional to the degree of displacement, delta x. “If you had a piece of Jello on a table and then put your hand on top of it and tried to move it, the top of the Jello would shift but not the bottom. That displacement is the shear elastic modulus that a TEG or a ROTEM is measuring,” he explained.
Ultrasound is another newer mechanism used for viscoelastic testing.
Traditional testing provides clot time in seconds. Viscoelastic testing provides the same in addition to “clot formation, the angle of the curve, the clot firmness or the maximum amplitude of the curve, and the stability—does it stay or does it disappear?” Dr. Salazar said.
Viscoelastic testing is essentially started with reagents similar to those for traditional tests—for example, kaolin for TEG, ellagic acid for the INTEM, and tissue factor for the EXTEM.
“Modern viscoelastic assays can now be viewed in real time, so providers can log in remotely and watch the curves develop over time,” Dr. Salazar said.
This was the only test in Dr. Salazar’s laboratory, and likely in others, for which a provider was able to review the results that developed on the curve over time before the lab could ensure they’re valid results. “The providers are already acting on it—they’re in the OR, they’re transfusing, they’re changing care. And we had multiple instances of test error. Maybe we added the wrong reagent, maybe we mixed up a sample or something. And providers may have already acted on those results,” he said.
The CAP understood this risk, Dr. Salazar said, and a new requirement has been added to the hematology and coagulation checklist for viscoelastic testing. HEM.38700 Viscoelastic Testing—Error Communication requires that if viscoelastic testing for hemostasis analysis is performed in the laboratory and the results are viewable remotely by clinical personnel in real time, the laboratory must promptly communicate analytic errors to the responsible clinical personnel and the communication must be recorded.
“It’s an example of a postanalytic issue that the committee is tackling,” he said, “with a new checklist standard.”
Amy Carpenter is CAP TODAY senior editor. Whole Blood Viscoelastic Assays in Clinical Diagnosis: An Illustrated Case-Based Guide, by Oksana Volod, MD, is available from the CAP at https://bit.ly/CT_0825_coag.