Pre-analytical Variables in Coagulation Testing Associated With Diagnostic Errors in Hemostasis

labmedicine.com February 2012 ■ Volume 43 Number 2 ■ LABMEDICINE 1

CE Update

Modern instrumentation is generally capable of providing highly accurate test results. Utilized with appropriate internal quality control and external quality assurance measures, ana-lytical errors within hemostasis testing are generally minimal. Nevertheless, incorrect or inappropriate test results will on occa-sion be reported to clinicians, most often due to circumstances beyond the control of the laboratories performing these tests. Overall, a significant impact on patient care arising from

diagnostic errors has been estimated to arise in around 9% to 15% of errors, while the likelihood of inappropriate care has been described in 2% to 7% of such cases.1 Many of these er-rors will originate due to the inappropriate collection, handling, or processing of samples referred for testing and sometimes because testing has been initiated on the wrong patient or at the wrong timepoint. In these instances, test results will accurately reflect the status of the test sample, but conversely they will not accurately reflect the clinical status of the patient being investi-gated. These issues are referred to as pre-analytical variables.

Consequences of Pre-analytical IssuesIt is inherently challenging to establish a direct relation-

ship between spurious test results and patient outcomes since laboratory errors do not always and necessarily translate into serious harm for the patient as may occur in cases of mis-handled surgery or inappropriate drug therapy. Nevertheless, the consequences of incorrect test results might still be clinically meaningful and lead to several unwanted clinical outcomes or adverse economical consequences, and place laboratories at risk.2 The seriousness of the potential consequences relates to the test being performed, the extent of the difference between the reported result and the true result, as well as the ability of laboratory personnel and clinicians to recognize the issues.3,4 Consequences may be particularly serious for errors related to spe-cialized hemostasis tests because these assays are often considered

Pre-analytical Variables in Coagulation Testing Associated With Diagnostic Errors in Hemostasis Emmanuel J. Favaloro, PhD, MAIMS, FFSc (RCPA),1 Dorothy M. (Adcock) Funk, MD,2 Giuseppe Lippi, MD3

(1Department of Haematology, ICPMR, Westmead Hospital, Westmead, NSW, Australia, 2Esoterix Inc., Englewood, CO, 3Clinical Chemistry and Hematology Laboratory, Academic Hospital of Parma, Parma, Italy)DOI: 10.1309/LM749BQETKYPYPVM

Abstract The use of modern laboratory instrumentation with high levels of test reliability and appropriate quality assurance measures will lead to very few analytical errors within hemostasis testing. Nevertheless, incorrect or inappropriate test results are still reported, often due to events outside the control of the laboratories performing the tests. This is due primarily to pre-analytical events associated

with sample collection and processing, as well as post-analytical events related to the reporting and interpretation of test results. This review focuses on the pre-analytical phase, highlighting contributory elements and providing suggestions on how problems can be minimized or prevented, thereby improving the likelihood that reported test results actually represent the true clinical status of the patient rather than that of an inappropriate

sample. This review should be of value to both laboratory personnel and clinicians because an appreciation of these issues will enable the optimal clinical management of patients. Keywords: pre-analytical variables, extra-analytical variables, diagnostic errors, hemostasis, coagulation

After reviewing this article, readers should be able to describe pre-analytical variables affecting various coagulation tests and discuss how these problems can be avoided in order to ensure the accurate reporting of patient results.

Coagulation exam 51202 questions and corresponding answer form are located after this CE Update on page 59.

Corresponding AuthorEmmanuel J. Favaloro, PhD, MAIMS, FFSc (RCPA)[email protected]

Submitted 6.3.11 | Revision Received 7.4.11 | Accepted 7.25.11

AbbreviationsLA, lupus anticoagulant; APS, antiphospholipid (antibody) syn-drome; VWD, von Willebrand disease; VWF, von Willebrand factor; PT, prothrombin time; INR, international normalized ratio; APTT, activated partial thromboplastin time; TT, thrombin time; DIC, dis-seminated intravascular coagulation; F, factors; CLSI, Clinical and Laboratory Standards Institute; ISI, international sensitivity index; EDTA, ethylenediaminetetraacetic acid; aPL, antiphospholipid; aCL, anticardiolipin; aB2GPI, anti-beta-2-glycoprotein I; NRR, normal reference range; MNPT, mean normal PT; APCR, activated protein C resistance

CE Update

2 LABMEDICINE ■ Volume 43 Number 2 ■ February 2012 labmedicine.com

“diagnostic.” Thus, a patient might be diagnosed with a par-ticular disorder, when he/she does not have it (ie, false-positive test result obtained) or a patient with a true disorder might be missed (ie, false-negative test result obtained). Both situations will cause adverse consequences for both patients and the health care system. As examples, 1) a false-negative antiphospholipid (aPL) antibody or lupus anticoagulant (LA) test result in a pa-tient with the aPL antibody syndrome (APS) may lead to lack of appropriate treatment with anticoagulant therapy to prevent a future thrombosis; and 2) a false diagnosis of von Willebrand disease (VWD) may lead to inappropriate treatment with fac-tor concentrates or to a lifelong label of a congenital disorder affecting quality of life. Serious consequences may also result due to errors in routine coagulation testing. For example, 1) a falsely prolonged “screening” coagulation test might influence a clinical decision to undertake further costly and time consum-ing (eg, “specific diagnostic”) investigations, unnecessarily delay invasive procedures, and raise unnecessary anxiety in the patient being investigated; 2) a false-normal screening test result might prevent further evaluation of factor assays, thus incorrectly dis-counting hemophilia and possibly placing a patient at an unjus-tified risk of bleeding with invasive procedures (ie, surgery, den-tal extraction, biopsies); and 3) a false low or high coagulation test time in a patient being monitored for anticoagulant therapy may lead to subsequent incorrect dosing of anticoagulant therapy with a risk of thrombosis or bleeding depending on the direction of the error.

Overview of Hemostasis, Laboratory Testing, and Pre-analytical Issues

Hemostasis is commonly thought of in terms of “coagula-tion pathways” or as a surrogate of “coagulation.” However, hemostasis is far more complex than “coagulation,” which in essence reflects clot formation, as it incorporates several components unrelated to the “coagulation” process. The com-ponents of hemostasis can be considered within the context

of Virchow’s Triad, or as a composite of primary hemostasis (platelet plug formation, von Willebrand factor [VWF]/plate-lets/subendothelial components), secondary hemostasis (fibrin clot formation, procoagulant “clotting” factors, and the natural anticoagulants) and fibrinolytic pathways.5,6

The modern hemostasis laboratory performs a large num-ber of distinct tests, often using a variety of methodologies (Table 1). All hemostasis laboratories perform routine coagula-tion tests comprising the prothrombin time (PT)/international normalized ratio (INR) and the activated partial thromboplas-tin time (APTT), sometimes supplemented by specific fibrino-gen assays, and occasionally thrombin time (TT) assays. Most routine test laboratories also perform D-dimer assays. These tests are variably performed to investigate hemostasis in patients suspected of having a potential dysfunction in the secondary hemostasis pathway, either congenital (eg, hemophilia) or ac-quired (eg, disseminated intravascular coagulation [DIC]).5,7-10 This is because PT/INR, APTT, and TT are sensitive to deficiencies or defects in various procoagulant factors. Thus, the PT/INR is sensitive to factors (F) I, II, VII, V, and X, and the APTT to F I, II, V, VIII, IX, X, XI, and XII. The single or compound deficiency or absence of most of these factors will lead to an increased tendency to bleeding and will occasion-ally define hemophilia (eg, deficiency in FVIII, hemophilia A; deficiency in FIX, hemophilia B). In contrast, an excess of some procoagulant factors (eg, FVIII, FIX, and FXI) may lead to thrombophilia.11 Although the PT/INR and APTT are not highly sensitive to elevations in procoagulant factors, a short APTT is sometimes indicative of this, and hence may reflect an increased risk of thrombosis.8,12,13 The PT/INR is also used to monitor vitamin K antagonists such as warfarin,7,14 and the APTT to monitor unfractionated heparin.8 Indeed, these tests are more often performed for monitoring anticoagulant therapy than for assessing secondary hemostasis. Other tests performed by hemostasis laboratories in general comprise a battery of spe-cific “diagnostic” assays (largely listed in Table 1).

The large number of distinct tests involving a variety of methodologies may result in significant problems when unsuitable

Table 1_Summary of Hemostasis Tests and Sample Requirements

Comprise Usually Performed Via Sample Type

A. Routine coagulation tests Citrate anticoagulated plasma post single centrifugationPT/INR, APTT, TT, fibrinogen Clot-based tests, automated instrument, primary collection tube (sometimes separated plasma) D-Dimer (D-D) ELISA or ELFA or agglutination (primary or secondary tube) B. Specialized Hemostasis Tests Factor assays (ie, II, V, VII, VIII, IX, X, XI, XII), factor Clot-based tests, automated instrument Separated citrate anticoagulated plasma, post single inhibitor assessments, protein S, protein C centrifugation (usually post freezing)VWF tests ELISA, immunoassay, or agglutination Protein C, protein S, antithrombin ELISA, immunoassay, clot based, or chromogenic assays Heparin (anti-Xa) assay Chromogenic assays Separated citrate anticoagulated plasma, post single (or preferably double) centrifugation (usually post freezing)APCR Clot-based tests, automated instrument LA Clot-based tests, automated instrument Separated citrate anticoagulated plasma, post double centrifugation (usually post freezing)Solid phase aPL tests including aCL ELISA or immunoflourescent assay Separated serum preferred; separated citrate anticoagulated and ab2GPI plasma post single centrifugation sometimes acceptable. Usually post freezing.Platelet function tests Specialized instrumentation Citrate anticoagulated whole blood or special processing required.Genetic thrombophilia tests Specialized instrumentation EDTA or citrate anticoagulated whole blood or special processing required

CE Update


samples are submitted for testing. Although guidelines are avail-able for how to manage and when to reject unsuitable speci-mens,15,16 it is not always clear when unsuitable samples have been received. Pre-analytical problems can arise at any point prior to sample testing, including (but not limited to) sample collection, handling, transportation, processing, and storage. Whereas analytical errors are largely avoided or intercepted by using appropriate test methodologies and by incorporation of appropriate control measures, pre-analytical issues present a more difficult scenario for laboratories as they are often outside the control of the laboratory performing the tests, and often the laboratory is unaware that the adverse pre-analytical event has occurred. Thus, the laboratory may issue a test result with the best of intentions as reflecting an accurate patient-related result, but this may not be the case. Clinicians would be even less aware of the issue of pre-analytical variables than the labo-ratory, and would base their clinical actions on the test result received (as reflecting a true and correct result). For this reason, guidelines for specimen collection and handling must be strictly followed and deviations avoided unless their impact, or lack thereof, on coagulation testing is known.

Appropriate Sample Collection, Processing, and Storage

These are critical to the attainment of appropriate test results but are often neglected, overlooked, or poorly applied.

Positive Patient and Sample Identification The importance of proper patient identification cannot

be overemphasized. In an outpatient setting, the principle of “double identifiers” should be used, specifically; the conscious patients should be asked to identify themselves and also pro-duce some form of identification. Within the hospital, positive patient identification should follow institutional rules and will typically entail electronic or bar-code methods to reduce the risk of patient misidentification. Other guidelines to ensure positive patient and sample identification include those related to printing tube labels; matching patient identification with the patient’s full name, and an additional identifier, such as date of birth or medical record number; and identification of collection date and time.16,17

Sample CollectionAll tests have specific collection requirements. Sample col-

lection issues might arise because of inexperience or time pres-sures when collectors are faced with a busy clinic and multiple collection requirements. Most samples referred for coagulation testing must be drawn into citrate-based anticoagulant tubes (generally 105-109 mM or 129 mM sodium citrate, also re-ferred to as 3.2% or 3.8%, respectively). The current Clinical and Laboratory Standards Institute (CLSI) guidelines16 favor the use of the lower citrate concentration, except for specific applica-tions.4,16 Specimens collected in 129 mM (3.8%) buffered so-dium citrate may overestimate the PT and APTT and underes-timate fibrinogen if the normal range is based on 3.2% citrated samples.18 Conversely, samples collected into 129 mM (3.8%) citrate may provide a more stable sample for assessing antiplate-let (eg, aspirin) therapy response using the PFA-100. Sometimes there is no apparent difference in relation to testing (eg, anti-Xa [heparin] testing) based on citrate concentration. The major recommendation therefore is that laboratories standardize to 1

citrate concentration and develop normal ranges appropriate for that concentration. This standardization should include all components of the assay (eg, including determination of patient PT, mean normal PT [NMPT], and international sensitivity index [ISI] for the INR).

Coagulation samples should preferably be collected before other test samples are drawn, if these contain stronger antico-agulant agents such as ethylenediaminetetraacetic acid (EDTA) (for a complete blood count), lithium-heparin (for clinical chemistry testing), as well as clot activators (ie, thrombin), since these materials may contaminate a subsequent coagulation test sample. A specific sequence of tube collections (so-called “order of draw”) is provided by the CLSI.19 The old dogma that the first collection tube should be discarded may not generally be required, as evidence for differential effects on coagulation as-says are lacking.20 Nevertheless, a discard tube is needed if the sample is drawn using a winged collection with variable tub-ing length so air in the tubing is not introduced into blood collection tubes leading to under-filling.16,21 Tubes should be adequately filled (to the mark noted on the tube if provided) or to no less than 90% of this total volume. Under-filling may cause significant sample dilution and may also provide falsely prolonged clotting times due to the excess calcium-binding ci-trate present. This effect depends on the citrate concentration, the tube size, and the test performed being more pronounced with 3.8% citrate tubes and small volume (pediatric) collection tubes.22,23 Sample dilution will also lead to under-estimation of quantitative test results (eg, clotting factor levels).

Blood should never be transferred from 1 collection tube to another in an effort to provide the required complete fill vol-ume. This is true even if 2 sodium citrate tubes are combined, as this may lead to doubling up of anticoagulant citrate levels and further dilution of the plasma sample. The introduction of stronger anticoagulants (eg, EDTA or lithium-heparin) or clot activators (eg, thrombin) must also be avoided, and this will occur if blood from non-citrate collection tubes is added to citrate tubes.

Samples should be mixed thoroughly (but gently) by 3 to 6 end-over-end tube inversions to ensure adequate mixing of test sample with anticoagulant19 and to prevent sample clot-ting. Insufficient mixing may have a greater effect on special-ized hemostasis assays performed some time after collection than on basic coagulation tests performed soon after collection. Conversely, too vigorous mixing (eg, by shaking of tubes) might lead to in vitro hemolysis or spurious factor activation resulting in false shortening of test clotting times and even pos-sible false elevation of clotting factor activity (eg, FVII).

Some tests referred to hemostasis laboratories may require sample matrices other than sodium citrate anticoagulated plasma, leading to additional scope for pre-analytical error. For example, while the test sample for LA testing must be citrate anticoagulated plasma, the preferred test sample for solid phase testing of aPL antibodies, such as anticardiolipin (aCL) anti-body and anti-beta-2-glycoprotein I (aB2GPI), in serum. As all of these different tests might be requested for a patient being investigated for APS, problems may arise should the laboratory inadvertently perform LA testing using the serum sample.

Other issues arising from blood collections include difficult collections, or those derived from central venous lines, leading to partially clotted, hemolyzed, or activated samples, or samples diluted by saline or contaminated with heparin. Collections from venous lines should include a process for flushing and/or discarding the initial collection volumes. Size and type of

CE Update


needle used may also influence results and too large (less than 16 gauge) or too small a needle bore (greater than 25 gauge) should be avoided, and heparinized needles (sometimes used for blood gas collection) not used.24

Sample TransportSamples should be transported as per current guidelines,16

non-refrigerated at ambient temperature (15-22°C) in as short a time as possible. Ideally, testing for routine coagulation tests like the PT and the APTT should be accomplished within 4 hours of collection, although allowable tolerances may be greater than this.25,26 However, APTT testing for unfraction-ated heparin monitoring should preferably be processed within 1 hour due to the potential for heparin neutralization by plate-let releasates.16,27,28 Extremes of temperature (ie, both refriger-ated or high) should be avoided. Delays in transport may affect in particular the labile factors (FV, FVIII), leading to prolonged clotting times and in vitro loss of factor activity.29 In such cases, local centrifugation and separation of plasma followed by freez-ing and frozen transport of the plasma should be considered.

Sample Processing and StorageThis should also in general proceed as per current CLSI

guidelines,16 noting limitations according to which test is being performed. Most coagulation-based tests, including PT, APTT, and clotting factor assays, are performed on plasma derived from once-centrifuged samples (Table 1). Some samples, such as those for LA testing, should be double centrifuged to ensure platelet-free preparations prior to freezing.30 Centrifugation should essentially be at an ambient temperature (15°C-22°C), but this is sometimes difficult to control. Non-refrigerated centrifuges are adequate, providing they do not overheat. Alternatively, refrigerated centrifuges may be used but should be set to maintain ambient temperatures, rather than low tem-peratures, which can lead to platelet activation and adverse ef-fects. Nevertheless, refrigerated centrifugation does not appear to affect routine coagulation tests when testing is performed soon after centrifugation. Centrifugation should ideally be at 1500 g for a minimum of 10-15 minutes.16 Shorter centri-fuge times might be acceptable for routine coagulation tests performed immediately post-centrifugation when there are no subsequent test requirements (ie, plasma not to be frozen or processed for additional assays). Using centrifugal forces greater than 1500 g are not recommended as this may induce plate-let activation and lysis of RBCs. The use of centrifuge breaks should also be avoided or monitored to avoid remixing of test samples, particularly if plasma is to be frozen, since there is a potential for hemolysis and platelet contamination, which may subsequently affect most hemostasis assays.

Testing generally proceeds using the once-centrifuged test sample in the primary collection tube or on the once-centrifuged separated plasma sample before or after freezing (Table 1). For some assays, samples should be double centrifuged (“double-spun”), which entails the re-centrifugation of the separated plasma, and re-separation of this double-spun plasma from any residual cellular pellet prior to freezing. Since all plasma-based hemostasis tests can safely be performed on double-spun mate-rial, it might be prudent to institute this process as a general lab-oratory policy for any plasma that will be frozen prior to testing. Use of filtered plasma is no longer recommended for LA testing, since this might produce spurious test results with some assays, as highlighted later.30 Lastly, some tests require additional special differential processing (eg, platelet function testing).31

The stability of coagulation samples varies depending on a number of variables such as the blood collection system, whether the samples are stored as whole blood or centrifuged, the temperature at which samples are maintained during stor-age, the reagent/instrument system used for analysis, and the test parameter to be analyzed. For example, whole blood stored up to 24-48 hours prior to centrifugation has been reported as acceptable for many hemostasis tests (although not for FV, FVIII, and protein S),25,32 but other studies have reported sig-nificant changes in some test results over such time periods.4 Moreover, storage of refrigerated whole blood is now actively discouraged and leads to activation events affecting FVII, FVIII, VWF, and possibly others.16,33

In general, to afford the greatest sample integrity, samples should be processed as quickly as possible (ideally within 1 hour of collection) and testing performed within 4 hours of procure-ment (or else be processed by centrifugation and plasma frozen). During this short-term storage, whole blood samples should be kept capped and maintained at room temperature. If testing is not to be performed within about 4 hours for the APTT and 24 hours for the PT, the plasma should be separated from the cellu-lar fraction of the once or twice-centrifuged sample, without dis-turbing the cell pellet. For many tests of hemostasis, the separated plasma can be safely frozen for later testing. Separated plasma can generally be maintained at room temperature or refrigerated for a few hours without an adverse effect on coagulation. Otherwise, separated plasma samples should be frozen. Frost-free freezers with automatic defrost cycles are generally unsuitable, since they cycle freeze-thaw events to maintain the frost-free environment, and this adversely affects subsequent coagulation tests. However, the use of frost-free freezers for patient samples is acceptable where freezers are monitored by a continuous-monitoring tem-perature recording device, or a minimum-maximum thermom-eter, enabling the laboratory to show the acceptable temperature range is never exceeded. When storing plasma, the lower the freezer temperature, the longer the specimens can be maintained for future testing. As a general rule of thumb, testing for samples maintained at around -20°C should be finalized within 2-4 weeks of storage, whereas testing for samples maintained at around -80°C can occur several months and sometimes years later (useful for research studies and prospective trials).34

Controlled Thawing of Frozen Plasma SamplesPreviously frozen samples should be rapidly thawed in a

37°C water bath for 5-10 minutes or until completely thawed.4,16 Close monitoring during this time is necessary to avoid inad-equate or excessive incubation at 37°C. Sample integrity may be compromised if samples are either not completely thawed or if maintained too long at 37°C. Furthermore, water baths must be properly maintained to make certain they are not inadvertently maintained at a higher temperature because this may lead to deterioration of coagulation factor activities and spurious coagula-tion test results. Once samples are thawed, it is imperative they are thoroughly and adequately mixed prior to testing.

Typical Issues Related to Inappropriate Sample Collection, Processing, and Storage

Incorrect Patient Collected or Wrong Label AttachedPatient misidentification errors are potentially associated

with the worst clinical outcome due to the potential for misdi-agnosis and inappropriate therapy. Whenever misidentification

CE Update


is suspected, the laboratory might be able to identify this as being the case by investigation, but the safest approach is gener-ally recollection and retesting.

Incorrect Anticoagulant Matrix Collected or Provided to the Laboratory

Although serum, heparin, or EDTA samples provided as a primary collection tube can be quickly identified as unsuitable by the laboratory, collection into the wrong anticoagulant may be missed when samples are provided in secondary tubes, or if samples have been mixed or transferred or added to a primary citrate tube.4,16 The potential consequences will differ accord-ing to the sample type received and the tests being performed as will the ability of laboratory personnel to recognize an incorrect sample. For example, testing for routine coagulation tests such as the PT and APTT will result in no clot or prolonged clot-ting times, but effects on other test results may be more subtle. Thus, testing of normal serum for VWF tests will not provide extreme changes but might instead give rise to patterns consis-tent with Type 2 VWD. Similarly, testing of heparin or EDTA

plasma may derive “plasma-like” test results for D-dimer and some VWF tests but a false impression of absent FVIII activity by 1-stage clotting assays.

Serum or Clotted SamplesSamples in which the blood is slow to fill the collection

container, where there is prolonged use of a tourniquet, or considerable manipulation of the vein by the needle may be prone to develop a clot in vitro. Clots may also develop when samples are incompletely mixed immediately following col-lection or in under-filled tubes. Although modern laboratory instrumentation is increasingly being equipped with various additional sensors (eg, bubble/volume/clot), samples yielding long clotting times should routinely be checked for the presence of a clot, either visually or preferably by inserting 2 wooden applicator sticks into a whole blood sample. The presence of a clot is a cause for rejection of the specimen. Serum will lead to loss of fibrinogen and many other coagulation factors (notably FII, FV, and FVIII) as well as differential loss of high molecular weight VWF (Tables 2 and 3).35,36 Serum may also yield high

Table 3_Summary of Effects of Inappropriate Sample Processing Issues on Select Hemostasis Tests

Issue Effect on Hemostasis Tests

Whole blood refrigerated Platelet activation and loss of FVIII and VWF; can lead to false diagnosis of hemophilia or VWD prior to centrifugation Filtered plasma Loss of fibrinogen, FVIII, and VWF; can lead to false diagnosis of dysfibrinogenemia, hypofibrinogenemia, hemophilia, or VWD; prolongs routine coagulation test times (PT, APTT, and TT); false LA feasibleDelayed transport, delayed 1. Loss of labile factors (especially FV and FVIII); can lead to false impression of hemophilia; prolongs routine coagulation test testing, poor storage, several times (PT, APTT); 2. Samples with unfractionated heparin can yield lower than expected APTTs and lower anti-FXa (heparin) freeze-thaw events; storage test levels; 3. Potential activation of FVII in frost-free freezer Poor centrifugation, heavy braking, Platelet contamination, hemolysis, and platelet disruption post freezing; activation, false low APTT, false low heparin levels, sample remixing prior to freezing false negative LA, false high factor levels

Table has been adapted and updated from reference 4.

Table 2_Summary of Differential Effects of Testing Different Sample Types on Select Hemostasis Tests

Potential Consequences Potential Consequences On Sample Type Routine Coagulation Tests On Factor Assays Other Hemostasis Tests

EDTA plasma Prolongs PT and APTT, and occasionally TT. False low levels (especially False impression of inhibitors to FV and Might influence fibrinogen and FV and FVIII) FVIII, and may show time dependence D-dimer assays (ie, enhanced with incubation); false LA feasibleSerum or fully clotted No fibrinogen, so no clot in PT, APTT, or TT. False low levels (especially FII, FV, False impression of factor inhibitors or VWD; coagulation sample False impression of afibrinogenemia. and FVIII); false high FVII false LA feasible D-dimer assays can be affected especially if testing delayedPartially clotted Depending on relative extent of platelet False low factor levels or false Flow obstructions in PFA-100 testing coagulation sample activation, hemolysis and loss of fibrinogen high factor VII might lead to false prolongation of PT, APTT, and TT, or false shortening of APTTUnderfilled primary citrate Will typically prolong PT, APTT, and TT. May False low factor levels likely False low levels of most hemostasis tests anticoagulant tube underestimate fibrinogen and D-dimer likelyVitamin K-deficient plasma, Prolongs PT and APTT (PT raised >APTT False low factors (especially False low protein C (potentially different patient on vitamin K antagonist raised) FII, FVII, FIX, FX) effect with clot-based assays vs therapy, liver disease sample chromogenic assays); false low protein S; false APCR; false LA feasible Heparin ‘contamination’ (either Prolongs PT, APTT, and TT (usually TT raised Reduced factors (especially False low Antithrombin; false LA feasible ex-vivo or due to collection >APTT raised >PT raised), false low FVIII, FIX, FXI, FXII) tube error) fibrinogen False impression of factor inhibitors


CE Update


values for some factors (eg, FVII) due to activation. Testing of serum will therefore lead to non-coagulation in tests such as the PT, APTT, and TT, possible diagnosis of coagulation factor deficiencies, false diagnosis of certain subtypes of VWD, and problems with LA identification. Alternatively, test results using serum might be normal with some other tests. Testing of par-tially clotted blood may lead to prolongation or shortening of coagulation tests depending on the extent of fibrinogen/factor loss vs activation events, and is often harder to identify.

To determine if the sample is serum, a TT can be per-formed. Non-clotting will suggest either serum or heparin con-tamination, which can then be differentiated by mixing studies (ie, TT performed on sample mixed 1:1 with normal plasma; if the mixed plasma clots, then serum is confirmed, whereas if the mixed plasma does not clot, this would suggest heparin con-tamination). The presence of heparin can also be determined by use of a heparin anti-FXa assay or by measuring a TT or APTT before and after the addition of a heparin-neutralizer.37

EDTA PlasmaThis will raise coagulation test times such as PT and

APTT and reduce FV and FVIII, leading to the potential false identification of factor deficiencies and/or factor inhibitors.35,36 If normal plasma mixing studies are performed on EDTA plasma, lack of correction is seen, suggesting the presence of an inhibitor. In some cases it may also lead to false identifica-tion of weak LA. As before, some test results might be normal (eg, VWF:Ag, D-dimer), and so, the ability of a laboratory to recognize an EDTA plasma sample will depend on the tests performed. Assessment of potassium (extremely increased) or calcium (very low to absent) will usually identify the presence of an inappropriate EDTA collection.37

HeparinWithin a hospital setting heparin contamination is much

more common than hemostasis assays incorrectly collected in either EDTA or submitted as serum. Heparin is used as therapy for treating patients (eg, post thrombosis), in “heparin flushes” to maintain flow in central lines, within “heparinized needles” (eg, blood gas collection), and for many surgical applications. Effects on hemostasis tests depend on the heparin concentration (or level of contamination) and the test performed. In general, clotting times (APTT and especially the TT) are prolonged, and fibrinogen and clotting factors (especially APTT based, viz FVIII, FIX, FXI, FXII) reduced.35,36 This might lead to false identification of dysfibrinogenemia/hypofibrinogenemia and certain factor deficiencies. There is also a potential for false identification of LA and factor inhibitors and reduced antithrombin. Sometimes, test reagents (eg, for PT or LA detec-tion) include heparin neutralizers, at levels sufficient to neutral-ize about 1 U/mL unfractionated heparin. Although useful, this can lead to complex patterns of test results and laboratories are sometimes falsely reassured that these tests are not influenced by heparin. Thus, a normal PT but abnormal APTT, or an abnormal PT and APTT, can both arise, depending on the contaminating level of heparin and whether the PT reagent contains heparin neutralizers. Results might be normal with some other tests (eg, D-dimer, VWF:Ag), and so whether the laboratory recognizes a heparin-contaminated sample as such will again depend on the tests performed.

Heparin contamination can be provisionally identified by testing of select clot-based assays (especially APTT and TT), and then by mixing studies (see end of serum section above),

and confirmed by using an anti-FXa assay or by repeat APTT or TT testing after addition of a heparin neutralizer.37

Processing IssuesBadly processing samples can lead to hemolysis or plate-

let activation, falsely prolonging or shortened clotting times, depending on the extent of hemolysis vs platelet activation. Alternatively, inadequate mixing may lead to clotting or partial clotting and prolongation or shortening of clotting times and elevated or diminished clotting levels. Freezing of plasma con-taminated with cellular material may also lead to hemolysis or activation events, as well as the potential for false-negative LA.

HemolysisThis results from cellular destruction within whole blood

and the release of cellular lysis products including hemoglo-bin into the plasma. Although in vitro hemolysis might be a byproduct of a problematic collection or the result of poor handling of blood post collection, hemolysis can also derive from in vivo blood cell lysis (eg, from hereditary, acquired, and iatrogenic conditions such as autoimmune hemolytic anemia, severe infections, intravascular disseminated coagulation, or transfusion reactions).38 Hemolysis increases the spectrometric absorbance of the plasma sample and leads to high background absorbance readings, which may compromise clot detection by some instruments and thus affect the accuracy of test times. Instruments utilizing mechanical means of clot detection are not affected by this interference, but the test result may still be compromised since cell lysis products include tissue factors that may activate coagulation. The net effect is that detected fibrino-gen levels may fall with increasing hemolysis, whereas D-dimer levels may increase. Prothrombin time values may fall in line with decreasing fibrinogen, whereas APTTs may increase or decrease depending on the net effect of activation vs the loss of fibrinogen. Hemolysis may also influence other test results (eg, decrease antithrombin levels).

If possible, grossly hemolyzed specimens should be re-jected. If testing must be pursued (eg, if in vivo hemolysis is present), testing using a mechanical end point detection system is recommended, although the potential effect of activation should also be noted. Samples appearing hemolyzed due to the presence of a hemoglobin substitute are not a cause of specimen rejection, and these samples should be evaluated using a me-chanical or electromechanical method for clot detection.

HematocritThe presence of significant anemia has not been shown to

influence test results.39 Too high a hematocrit will influence the anticoagulant to plasma ratio and thus test results.4,16 An ad-justment in the ratio of anticoagulant solution/volume of blood at different packed cell volume when hematocrit values are above 55% may be undertaken using CLSI recommendations, although a simplified method is to remove 0.1 mL of sodium citrate from a 5 mL 3.2% sodium citrate evacuated tube prior to collection.40

LipemiaIt is not easy to dichotomize the biological and analytical

effect of lipemia on coagulation tests.4,16 Acute elevation of the coagulant activity of FVII is observed after consumption of high-fat meals, mostly due to an increase in the concen-tration of activated FVII (FVIIa). High-fat meals also have a substantial, acute effect on platelet function and may also

CE Update


induce a lowering of some clotting factor activities (eg, FII, FIX, FX, FVII, FVIIa, FXIIa). Analytical interferences in some laboratory assays (especially those based on optical clot detec-tion) also occur but are minimized using mechanical or electro-mechanical-based procedures or using analyzers comparing the absorption of samples at 2 wavelengths or performing coagula-tion assays at alternative wavelengths.3,4 Nevertheless, regardless of the potential source of interference (biological or analytical), the best approach might be recollection of blood samples at fasting, provided that metabolic problems (ie, dyslipidemia) are absent.

Freeze-thawing EventsThese result in the loss of some labile factors, notably

FV and FVIII. Since it is not always clear how many times a sample has been thawed and refrozen prior to testing, retesting using a fresh sample is always indicated should an unexpected low factor result be obtained.

Under-recognized Pre-analytical Issues

Normal Reference Range Derivations and Related Issues

Laboratories use normal reference ranges (NRRs) to iden-tify whether a test result is within the normal range or outside this range (and to thus identify an abnormal result). Use of an inappropriate NRR may mean some normal individuals will yield apparently abnormal test results. However, even the use of a typical and potentially appropriate NRR, generated as the mean +/- 2 standard deviations, will identify 5% of the normal population as outside this range, simply based on the statistical model used (ie, to capture 95% of the normal population). Another way to consider this is to recognize that a standard laboratory NRR will correctly identify only 95 out of every 100 normal test results. Put into clinical context, 5 in every 100 (or 1 in every 20) tests a clinician orders (using such NRR estimates) will likely reflect a false abnormal test result, again simply based on the statistical model used to generate the NRR. The relative false positive to true positive rate increases substantially for rare disorders and is a particular problem with congenital disorders such as protein C, protein S, and antithrombin, especially when patient cohorts are inap-propriately selected for testing.41,42

Miscellaneous VariablesAge, gender, ethnicity, and blood group might influence

reference values for certain parameters of laboratory hemostasis, and/or generate variable test results for some tests.4,43 For ex-ample, FVIII and VWF and platelet function tests are generally influenced by such factors. Thus, interpretation of test results should consider these issues to prevent misdiagnosis.

International Normalized Ratio (INR) The INR is the most common test performed by coagula-

tion laboratories. The INR derives as a mathematical calcula-tion, viz: INR=(patient PT/MNPT)ISI where MNPT=mean normal PT, and ISI=international sensitivity index. The patient’s PT is an analytical event and is derived from the in-strument. However, the ISI and MNPT are derived separately and might be considered as pre-analytical variables within the context of inaccurate INRs.4,7

Filtered Plasma

Plasma for LA testing must be essentially platelet free (<10 × 109/L) if frozen prior to testing.4,16,30 Platelet-free prepara-tions can be achieved by a process of double centrifugation, high-speed (ultra-) centrifugation, microfiltration, or combina-tions thereof; however, microfiltration, commonly used in the past because of its ease, is no longer recommended. This is because the process leads to loss of other plasma components, including FVIII and VWF, and may cause potential problems in LA detection because it artificially elevates the baseline clot-ting times observed using some LA clot-based assays, and may also lead to a false conclusion of an elevated APTT. In extreme cases, microfiltration may even generate false (weak) positive LA findings. Moreover, in many routine hemostasis laboratories, LA testing is requested not only for specific investigation of APS but also for investigation of unexplained prolongation of APTT test times, a common incidental finding in a laboratory practice. The most common explanations are low levels of FXII or (unless the laboratory uses an LA-insensitive reagent) the presence of (usually asymptomatic) LA. However, since an el-evated APTT may also define a clinically significant event, such as hemophilia or VWD, prolonged APTTs should be further investigated to determine the underlying cause. It is therefore not uncommon to receive requests including test combinations for LA, and FVIII and/or VWF to exclude LA, hemophilia or VWD, respectively. The dilemma is that should the laboratory process the sample for LA testing by filtration, and then unwit-tingly test that sample for FVIII and VWF, a false diagnosis of hemophilia or VWD is then quite feasible. Accordingly, the double centrifugation approach for preparation of hemostasis samples prior to freezing is now strongly favored.

Physical Activity, Illness, and StressExcess physical activity in patients immediately prior to

collection leads to certain in vivo events (eg, plasma volume expansion and increased basal metabolism), which may in turn lead to significant effects on hemostasis. However, perhaps the best-known acute effects are related to acute phase reactants, which may rise due to physical activity, illness or stress, and include fibrinogen, VWF, and FVIII. In the worse case sce-nario, these elevations may result in a misdiagnosis of (mild) hemophilia A or VWD Type 1 patients as a non-hemophilia or non-VWD (false negatives). Blood collection may sometimes be stressful for some patients (particularly children) leading to acute phase changes in proteins secondary to the phlebotomy itself.

Circadian and Diurnal RhythmsLevels of some hemostasis components follow a circadian

or diurnal rhythm, with differential levels detectable at different times of the day.4,44 For example, fibrinogen and plasminogen activator inhibitor-1 levels tend to be higher in the early morn-ing hours. PFA-100 closure times and possibly VWF may also provide different values throughout a 24-hour period. Although most changes tend to be fairly subtle, in the worse case scenario this might also lead to some clinically significant differences.

Patients on Anticoagulant TherapyTesting for thrombophilia is often performed in patients

who have recently suffered a thrombotic event. Patients are placed on anticoagulant therapy after a thrombosis. Testing while on anticoagulant therapy will affect (both biologically and analytically) many of the tests undertaken in this context,

CE Update


including LA, activated protein C resistance (APCR), anti-thrombin, protein C, and protein S. Thus, false-positive and false-negative diagnoses can both occur, depending on the ex-tent of the anticoagulant effect, and the test performed.41,42

Other Medications That May Interfere With Coagulation Testing

A variety of therapeutic agents may cause spurious coagu-lation results due to variable mechanisms. This effect is not always intuitive based on the pharmaceutical product.4

Clinical Ordering and Inappropriate Requests as a Pre-analytical Issue

The concept of clinical ordering patterns as a pre-analytical issue is also worth mentioning, in particular for the case of inappropriate clinical orders.4 There is heightened concern cur-rently with respect to thrombophilia tests, which comprise an area of investigation that is growing rapidly within hemostasis, and perhaps leading to “over” or “inappropriate” ordering.41,42 The proper timing of test orders is an important but poorly recognized issue. Following a thrombotic event, some loss (con-sumption) of the natural anticoagulants might arise; hence, test-ing too soon after a thrombosis might lead to false conclusion of a deficiency. FVIII may also be elevated post thrombosis, leading to a missed LA diagnosis if only APTT-based screening tests are used. Alternatively, anticoagulant therapy will affect the detected levels of the natural anticoagulants, as mentioned previously (viz, heparin therapy may influence antithrombin detection, warfarin therapy may influence protein C and pro-tein S levels, and heparin and warfarin therapy may influence APCR testing). Heparin and warfarin therapy may also influ-ence the appropriate identification of LA. Recent audits of clinical practice indicate that up to 1/3 of samples destined for thrombophilia investigations are from patients on warfarin and/or heparin therapy, or the sample is otherwise heparin contami-nated, and thus representing high potential for diagnostic error. Considered another way, upwards of 80% of abnormal throm-bophilia test results may be a reflection of inappropriate testing while on anticoagulant therapy.42

Test Methodology and Test Panel SelectionWhile test results and methodologies comprise analytical

issues, the choice of which particular methodologies or test pan-els to use might best be considered as pre-analytical variables. The presence of LA or APCR, seen in about 2%-5% of the

general Caucasian population, may interfere with some clot-based protein C and protein S assays, and lead to false identifi-cation of such deficiencies.4 Insufficient laboratory test panels may also miss significant disease. For example, VWD may be misdiagnosed or missed if the test panel does not include tests for VWF activity, such as a collagen-binding assay.45 Some methodologies are also poor at identifying low levels of VWF, so type 3 VWD may be misidentified as type 1. In another example, different laboratories and even experts use different tests (or methodologies) and test panels for the identification (or exclusion) of APS.46 Moreover, there are wide variations in the detection of solid phase aPL by different commercial assays, and different perceptions will arise among practitioners regard-ing general sensitivities and specificities of different tests and panels for APS, and different perceptions of positive or negative aPL for any given patient will arise among clinicians, depending on both the methodologies, as well as the test panels, used to identify APS.

ConclusionPre-analytical issues in hemostasis testing are an important

cause of diagnostic error (summarized in Tables 2 and 4) and can lead to significant adverse clinical events. However, the bur-den of laboratory errors is estimated to remain globally modest (ie, 1 in every 900-2074 patients or every 214-8316 laboratory results).47 Notably, a large number of errors are likely to be in-tercepted before they are released to the clinician and, thereby, before they translate into real harm for the patient. The ulti-mate aim of laboratory practice would be to have no errors or to at least detect and correct all errors before the test result is released. Accordingly, several tools might assist in their identi-fication, including a comprehensive education of all personnel regarding types and sources of errors, the accurate evaluation of sample quality (ie, volume, blood to anticoagulant ratio, pres-ence of potential interferents, or contaminants), and the sys-tematic recording of suspect results along with pertinent clinical information.48 When data are considered clinically question-able, the original test request should be checked and the speci-men inspected and retested, sometimes with different assays and instruments. The most reliable approach to deal with laboratory errors is to establish a total quality management system.49-51 This would entail the elimination or strict supervision of the most vulnerable activities, the implementation of customized

Table 4_Summary of Misdiagnosis and/or Misidentification in Hemostasis Possibly Arising From Inappropriate Sample Types

Misdiagnosis/Misidentification Can Arise From Testing Of

False-positive LA Normal EDTA plasma, normal serum, vitamin K deficiency patient, anticoagulated patient, heparin-contaminated sample, plasma containing factor inhibitorsFalse diagnosis of VWD: Filtered normal plasma, normal serum, normal plasma derived from refrigerated whole blood sampleFalse subtype identification (Type 2 Type 1 VWD plasma derived from refrigerated whole blood sample, testing of filtered plasma or serum diagnosis in Type 1 VWD patient) False diagnosis of hemophilia A Filtered normal plasma, normal serum, normal plasma derived from refrigerated whole blood sample, aged sample, sample post several freeze-thaw events, heparin-contaminated sample, EDTA sample, normal serum sample, underfilled primary citrate collection tubeFalse identification of factor inhibitors Heparinized normal sample, EDTA sample, normal serum sample, lupus anticoagulant


CE Update


informatics systems for error identification and recording, and for highlighting collection requirements as related to specifically ordered tests, and facilitate the continuous education of opera-tors (both inside and outside the laboratory) by dissemination of best practice recommendations.49-51 This would include pro-viding adequate training and guidance to blood collectors.

Additional useful advice to laboratories and clinicians is be vigilant of all these issues; select/order the best tests and test panels available, undertake testing only when necessary, at the correct point in time for the condition under investigation; incorporate as much clinical information as possible into the diagnostic approach; follow the recommendations of local labo-ratory experts/specialists; repeat tests when not in keeping with clinical expectations or when an abnormal finding is reported; implement restrictive specimen acceptance policies and toler-ance criteria for inappropriate specimens; put quality practices into place where possible to aid in identifying problem samples; and establish a mutually beneficial clinical-laboratory interface where both parties discuss the problems within meetings or teaching moments as well as actively collaborate to achieve the best possible patient outcome (Table 5). We also strongly rec-ommend that laboratories use appropriate post-test guidance to assist clinicians in the interpretation of test results, as well as to guide when repeat, confirmatory, and follow-up testing may be required.4,52

There have been several new oral anticoagulants recently released onto the market (most notably Dabigatran and Rivaroxaban) for a variety of clinical indications including pre-vention of venous thromboembolism after major orthopaedic surgery or secondary prevention in atrial fibrillation. These agents will variously affect coagulation and hemostasis tests as described elsewhere,53,54 but should now also be considered within the context of preanalytical problems associated with hemostasis testing. LM

1. Plebani M. Errors in clinical laboratories or errors in laboratory medicine? Clin Chem Lab Med. 2006;44:750-759.

2. Lippi G, Guidi GC. Risk management in the preanalytical phase of laboratory testing. Clin Chem Lab Med. 2007;45:720-727.

3. Lippi G, Guidi GC, Mattiuzzi C, et al. Preanalytical variability: The dark side of the moon in laboratory testing. Clin Chem Lab Med. 2006;44:358-365.

4. Favaloro EJ, Lippi G, Adcock DM. Preanalytical and postanalytical variables: The leading causes of diagnostic error in hemostasis? Semin Thromb Hemost. 2008;34:612-634.

5. Lippi G, Favaloro EJ, Franchini M, et al. Milestones and perspectives in coagulation and hemostasis. Semin Thromb Hemost. 2009;35:9-22.

6. Favaloro EJ, Lippi G. Coagulation update: What’s new in hemostasis testing? Thromb Res. 2011;127(Suppl 2):S13-S16.

7. Favaloro EJ, McVicker W, Hamdam S, et al. Improving the harmonisation of the International Normalized Ratio (INR): Time to think outside the box? Clin Chem Lab Med. 2010;48:1079-1090.

8. Lippi G, Favaloro EJ. Activated partial thromboplastin time: New tricks for an old dogma. Semin Thromb Hemost. 2008;34:604-611.

9. Lippi G, Franchini M, Targher G, et al. Help me, Doctor! My D-dimer is raised. Ann Med. 2008;40:594-605.

10. Favaloro EJ. Laboratory testing in disseminated intravascular coagulation. Semin Thromb Hemost. 2010;36:458-467.

11. Coppola A, Tufano A, Cerbone AM, et al. Inherited thrombophilia: Implications for prevention and treatment of venous thromboembolism. Semin Thromb Hemost. 2009;35:683-694.

12. Mina A, Favaloro EJ, Mohammed S, et al. A laboratory evaluation into the short activated partial thromboplastin time. Blood Coagul Fibrinolysis. 2010;21:152-157.

13. Lippi G, Salvagno GL, Ippolito L, et al. Shortened activated partial thromboplastin time: Causes and management. Blood Coagul Fibrinolysis. 2010;21:459-463.

14. Lippi G, Franchini M, Favaloro EJ. Pharmacogenetics of vitamin K antagonists: Useful or hype? Clin Chem Lab Med. 2009;47:503-515.

15. Lippi G, Banfi G, Buttarello M, et al. Recommendations for detection and management of unsuitable samples in clinical laboratories. Clin Chem Lab Med. 2007;45:728-736.

16. CLSI. Collection, Transport, and Processing of Blood Specimens for Testing Plasma-Based Coagulation Assays and Molecular Hemostasis Assays: Approved Guideline. 5th ed. CLSI document H21-A5. Wayne, PA: Clinical and Laboratory Standards Institute; 2008.

17. Kiechle FL, Adcock DM, Calam RR, et al. So You’re Going to Collect a Blood Specimen. An Introduction to Phlebotomy. College of American Pathologists. 12th ed. Northfield, IL; 2007.

Table 5_Important Issues for Laboratories and Clinicians to Consider Within the Context of Pre-analytical Issues in Hemostasis Testing, and Some Recommendations

Issue Consideration/Recommendation

Test selection Select/order the best tests/test processes/test panels for the condition being investigatedPopulation to be tested and Select the appropriate population/methodology to determine the normal reference range clinical condition/medication Only order the test(s) when clinically appropriate and in the right patient at the right time at time of testing Sample collection Proper patient and sample identification Atraumatic phlebotomy with minimal tourniquet use Draw blue stopper tube (citrate anticoagulant) first or only after a non-additive tube Fill tube adequately (no less than 90% fill) Adequately and thoroughly mix with tube anticoagulant Sample transport Transport promptly at room temperature Sample processing Ideally centrifuge within 1 hour of phlebotomy to obtain platelet-poor plasma (most tests) Double centrifuge plasma for some tests, namely LA, APCR, and heparin (anti-Xa) assays Aliquot (in a non-activating secondary tube) immediately following centrifugation for those tests to be performed later Special requirements for some tests such as platelet function and PFA-100Sample storage Test plasma within appropriate timeframe; store as required samples to be tested subsequentlySample testing Select the best test/methodology/test panel for the analyte/parameter being tested Perform test in timely manner and according to best practiceResult interpretation Laboratory: Provide clinician with appropriate guidance/test interpretation Clinician: Recognize test limitations/extra-analytical issues that may influence test results and follow local expert laboratory advice


CE Update


18. Adcock DM, Kressin DC, Marlar RA. Effect of 3.2% vs 3.8% sodium citrate concentration on routine coagulation testing. Am J Clin Pathol. 1997;107:105-110.

19. CLSI. Procedures for the Collection of Diagnostic Blood Specimens by Venipuncture. Approved Standard. 6th ed. CLSI document H3-A6. Wayne PA: Clinical and Laboratory Standards Institute; 2007.

20. Raijmakers MT, Menting CH, Vader HL, et al. Collection of blood specimens by venipuncture for plasma-based coagulation assays: Necessity of a discard tube. Am J Clin Pathol. 2010;133:331-335.

21. Favaloro EJ, Lippi G, Raijmakers MT, et al. Discard tubes are sometimes necessary when drawing samples for hemostasis. Am J Clin Pathol. 2010;134:851.

22. Adcock DM, Kressin DC, Marlar RA. Minimum specimen volume requirements for routine coagulation testing: Dependence on citrate concentration. Am J Clin Pathol. 1998;109:595-599.

23. Chuang J, Sadler MA, Witt DM. Impact of evacuated collection tube fill volume and mixing on routine coagulation testing using 2.5 mL (pediatric) tubes. Chest. 2004;126:1262-1266.

24. Sharp MK, Mohammad SF. Scaling of hemolysis in needles and catheters. Ann Biomed Eng. 1998;26:788-797.

25. Zürcher M, Sulzer I, Barizzi G, et al. Stability of coagulation assays performed in plasma from citrated whole blood transported at ambient temperature. Thromb Haemost. 2008;99:416-426.

26. Awad MA, Selim TE, Al-Sabbagh FA. Influence of storage time and temperature on international normalized ratio (INR) levels and plasma activities of vitamin K dependent clotting factors. Hematology. 2004;9:333-337.

27. van den Besselaar AM, Meeuwisse-Braun J, Jansen-Grüter R, et al. Monitoring heparin by the activated partial thromboplastin time—the effect of pre-analytical conditions. Thromb Haemost. 1987;57:226-231.

28. Adcock DA, Kressin DC, Marlar RA. The effect of time and temperature variables on routine coagulation tests. Blood Coagul Fibrinolysis. 1998;9:463-470.

29. O’Neill EM, Rowley J, Hansson-Wicher H, et al. Effect of 24-hour whole-blood storage on plasma clotting factors. Transfusion. 1999;39:488-491.

30. Pengo V, Tripodi A, Reber G, et al. Subcommittee on Lupus Anticoagulant/Antiphospholipid Antibody of the Scientific and Standardisation Committee of the International Society on Thrombosis and Haemostasis. Update of the guidelines for lupus anticoagulant detection. J Thromb Haemost. 2009;7:1737-1740.

31. Favaloro EJ, Lippi G, Franchini M. Contemporary platelet function testing. Clin Chem Lab Med. 2010;48:579-598.

32. Heil W, Grunewald R, Amend M, et al. Influence of time and temperature on coagulation analytes in stored plasma. Clin Chem Lab Med. 1998;36:459-462.

33. Refaai MA, van Cott EM, Lukoszyk M, et al. Loss of factor VIII and von Willebrand activities during cold storage of whole blood is reversed by rewarming. Lab Hematol. 2006;12:99-102.

34. Woodhams B, Girardot O, Blanco MJ, et al. Stability of coagulation proteins in frozen plasma. Blood Coagul Fibrinolysis. 2001;12:229-236.

35. Favaloro EJ, Bonar R, Duncan E, et al. Identification of factor inhibitors by diagnostic haemostasis laboratories: A large multi-centre evaluation. Thromb Haemost. 2006;96:73-78.

36. Favaloro EJ, Bonar R, Duncan E, et al. Mis-identification of factor inhibitors by diagnostic haemostasis laboratories: Recognition of pitfalls and elucidation of strategies. A follow up to a large multicentre evaluation. Pathology. 2007;39:504-511.

37. Lippi G, Salvagno GL, Adcock DM, et al. Right or wrong sample received for coagulation testing? Tentative algorithms for detection of an incorrect type of sample. Int J Lab Hematol. 2010;32(1 Pt 2):132-138.

38. Lippi G, Blanckaert N, Bonini P, et al. Haemolysis: An overview of the leading cause of unsuitable specimens in clinical laboratories. Clin Chem Lab Med. 2008;46:764-772.

39. Siegel JE, Swami VK, Glenn P, et al. Effect (or lack of it) of severe anemia on PT and aPTT results. Am J Clin Pathol. 1998;110:106-110.

40. Marlar RA, Potts RM, Marlar AA. Effect on routine and special coagulation testing values of citrate anticoagulant adjustment in patients with high hematocrit values. Am J Clin Pathol. 2006;126:400-405.

41. Favaloro EJ, McDonald D, Lippi G. Laboratory investigation of thrombophilia: The good, the bad, and the ugly. Semin Thromb Hemost. 2009;35:695-710.

42. Favaloro EJ, Mohammed S, Pati N, et al. A clinical audit of congenital thrombophilia investigation in tertiary practice. Pathology. 2011;43:266-272.

43. Montagnana M, Favaloro EJ, Franchini M, et al. The role of ethnicity, age and gender in venous thromboembolism. J Thromb Thrombolysis. 2010;29:489-496.

44. Banfi G, Del Fabbro M. Biological variation in tests of hemostasis. Semin Thromb Hemost. 2009;35:119-126.

45. Favaloro EJ. Toward a new paradigm for the identification and functional characterization of von Willebrand disease. Semin Thromb Hemost. 2009;35:60-75.

46. Favaloro EJ, Wong RC. Laboratory testing for the antiphospholipid syndrome: Making sense of antiphospholipid antibody assays. Clin Chem Lab Med. 2011;49:447-461.

47. Plebani M, Lippi G. To err is human. To misdiagnose might be deadly. Clin Biochem. 2010;43:1-3.

48. Lippi G. Governance of preanalytical variability: Travelling the right path to the bright side of the moon? Clin Chim Acta. 2009;404:32-36.

49. Lippi G, Chance JJ, Church S, et al. Preanalytical quality improvement: From dream to reality. Clin Chem Lab Med. 2011;49:1113-1126.

50. NCCLS. Continuous Quality Improvement: Integrating Five Key Quality Systems; Approved Guideline. 2nd ed. NCCLS document GP22-A2. Wayne, PA: NCCLS; 2004.

51. CLSI. Quality Management System: A Model for Laboratory Services; Approved Guideline. CLSI document GP26-A4. Wayne, PA: Clinical and Laboratory Standards Institute; 2011.

52. Favaloro EJ, Lippi G. Laboratory reporting of haemostasis assays: The final post-analytical opportunity to reduce errors of clinical diagnosis in hemostasis? Clin Chem Lab Med. 2010;48:309-321.

53. Favaloro EJ, Lippi G. Laboratory testing and/or monitoring of the new oral anticoagulants/antithrombotics: For and against? Clin Chem Lab Med. 2011;49:755-757.

54. Favaloro EJ, Lippi G, Koutts J. Laboratory testing of anticoagulants - the present and the future. Pathology. 2011;43:682–692.

Documents

Pre-analytical Variables in Coagulation Testing Associated With Diagnostic Errors in Hemostasis