The authors declare no conflicts of interests.
Systematic evaluation of evidence on veterinary viscoelastic testing Part 2: Sample acquisition and handling
Article first published online: 28 JAN 2014
© Veterinary Emergency and Critical Care Society 2014
Journal of Veterinary Emergency and Critical Care
Volume 24, Issue 1, pages 30–36, January/February 2014
How to Cite
Flatland, B., Koenigshof, A. M., Rozanski, E. A., Goggs, R. and Wiinberg, B. (2014), Systematic evaluation of evidence on veterinary viscoelastic testing Part 2: Sample acquisition and handling. Journal of Veterinary Emergency and Critical Care, 24: 30–36. doi: 10.1111/vec.12142
- Issue published online: 28 JAN 2014
- Article first published online: 28 JAN 2014
- Manuscript Accepted: 15 NOV 2013
- Manuscript Received: 8 NOV 2013
To examine systematically the evidence on sample acquisition and handling for the thrombo elastography (TEG) and rotational thromboelastometry (ROTEM) viscoelastic point of care instruments and to identify knowledge gaps.
Six questions were considered, addressing sampling site, collection system, anticoagulant, collection procedure, and sample storage. Standardized, systematic evaluation of the literature was performed. Relevant articles were categorized according to level of evidence (LOE). Consensus was developed regarding conclusions for application of concepts to clinical practice.
Academic and referral veterinary medical centers.
PubMed and CAB abstracts were searched. Eighteen papers were initially chosen; 5 of these papers applied to > 1 domain question. Three papers were used to address 2 questions each, and 2 papers were used to address 3 questions each. Most papers were judged LOE 3 (Good or Fair). Two of 5 papers were judged to be the same LOE each time they were used; 2 papers were judged to be LOE 3, Fair for 1 question and 3, Good for a second question; 1 paper used to address 3 questions was judged LOE 3, Good twice and 3, Fair once. Fourteen additional papers were evaluated post hoc during manuscript preparation.
Jugular venipuncture is recommended, but samples from IV catheters can be used. Consistent technique is important for serial sampling, and standardized sampling protocols are recommended for individual centers performing TEG/ROTEM. There is insufficient evidence to recommend use of a specific blood collection system, although use of evacuated blood tubes and 21-Ga or larger needles is suggested. Use of 3.2% buffered sodium citrate in a strict 1:9 ratio of citrate to blood is suggested. Suggested tube draw order is discard/serum, followed by citrate, EDTA, and then heparin. Samples should be held at room temperature for 30 minutes prior to analysis.
activated thromboplastin time
global clot strength (shear elastic modulus) (TEG variable)
Kappa value (TEG variable)
level of evidence
maximum amplitude (TEG variable)
maximum clot firmness (ROTEM variable)
platelet function analyzer-100
reaction time (TEG variable)
Viscoelastic coagulationtesting was originally developed as a point-of-care test designed for use with fresh whole blood analyzed within minutes of collection. This approach is not practical if thromboelastography (TEG)/rotational thromboelastometry (ROTEM) testing is performed distant from patients, a limitation that led to use of citrate-anticoagulated whole blood. Use of citrated whole blood for TEG/ROTEM requires recalcification, with calcium being added as a reagent at the beginning of the assay. Generation of factor XIIa (contact activation) is not calcium-dependent and continues during storage of citrated whole blood; use of citrated blood samples for TEG/ROTEM means that thrombin generation is affected by the degree of factor XIIa generation during storage (as a result of contact with surfaces) and prior to recalcification. Accordingly, preanalytical factors such as sample storage time and temperature have potential to impact TEG/ROTEM results. Additionally, as with other coagulation tests, platelet activation, and tissue factor contamination of the blood sample resulting from traumatic venipuncture have potential to affect results. Optimal sample collection and handling for TEG/ROTEM are the main focus of this domain.
Databases and Search Terms Used
PubMed1 and CAB Abstracts2 were searched using these terms as “topic” (CAB abstracts) or “all fields” (PubMed): ((dog OR canine) OR (cat OR feline) OR (horse OR equine)) AND ((TEG OR thromboelastograph*) OR (ROTEM OR thromboelastom*)). Additional terms used to narrow initial search results are presented in Table 1. Fifty-eight papers were identified using CAB and 57 using PubMed; there was much overlap in results from the 2 databases. Results from both databases were reviewed, and the most relevant papers (n = 18) were chosen for inclusion in the sample domain worksheets and systematic review. During manuscript preparation, additional relevant papers (n = 14, 2 of which were not published until after completion of the initial systematic review) were added. All papers were evaluated for level of evidence (LOE) relative to the question at hand as previously described.
|Additional search terms applied|
|Database||Search field||to results from initial search|
|PubMed||All fields||• Preanalytical OR preanalytical|
|CAB abstracts*||Topic||• Hemolysis OR hemolyzed|
|• Icterus OR icteric|
|• Lipemia OR lipemic|
|CAB abstracts*||Description||• Factor OR variable|
|• Technique OR method|
2.1 2.1 PICO question
In companion animals with hemostatic abnormalities (P), does the use of one blood vessel for sample collection (I), compared with another (C), alter testing results (O)?
Jugular venipuncture is recommended in horses and suggested for dogs and cats. Consistent sampling technique should be used when obtaining serial samples in the same patient.
Summary of the evidence
Most papers reviewed did not investigate the specific vessel chosen for venipuncture. Jugular venipuncture was used to collect blood samples from cats in one TEG study (LOE 3, Fair). Studies in horses (LOE 2–5, Good-Fair) also report sampling the jugular vein.[5-11] A recent TEG/ROTEM review article (LOE 5, Fair) supports clean venipuncture with limited tissue trauma or excessive aspiration but did not mention specific blood vessels and was considered neutral to the question of sampling site.
Opposing the question of whether sample site affects results is a recent paper investigating kaolin-activated TEG in clinically healthy dogs (LOE 3, Good). This study compared samples obtained via jugular and lateral saphenous vein sampling and additionally compared different sampling methods for each vein (see below). Sampling site (jugular versus saphenous) had no significant effect on R, K, angle, maximum amplitude (MA), or G.
2.2 2.2 PICO question
In companion animals with hemostatic abnormalities (P), does the use of one blood collection system (I), compared with another (C), alter testing results (O)?
There is insufficient evidence to recommend universal use of a particular blood collection system; however, use of evacuated tubes and 21-Ga or larger needles is suggested. Blood samples obtained from IV catheters may be suitable for TEG/ROTEM testing; use of heparinase cups to prevent exogenous heparin from impacting results is suggested in that case. Individual centers should use standardized sample acquisition protocols for all patients undergoing TEG/ROTEM testing. TEG/ROTEM results should be interpreted in light of method-specific reference intervals.
Summary of the evidence
Two recent studies support that sample collection technique impacts TEG results. One (LOE 3, Good) investigated effects of delayed anticoagulation and use of vacuum-assisted blood collection tubes on results of nonactivated TEG in clinically healthy adult dogs. Sampling methods delaying anticoagulation or using evacuated tubes resulted in more coagulable-appearing samples. Authors concluded that sample methods impact results and should be considered when developing reference intervals and interpreting patient results. A second study (LOE 3, Good) investigating kaolin-activated TEG in clinically healthy dogs compared samples obtained from the jugular and lateral saphenous veins, respectively, using direct venipuncture with syringe aspiration and sampling using a winged catheter with an evacuated tube. For the jugular vein, samples obtained via central venous catheter with syringe aspiration were additionally compared. For saphenous vein samples, sampling method had no significant effect. Jugular vein samples collected via direct venipuncture/syringe aspiration had significantly lower R and K and higher angle than those collected via winged catheter/evacuated tube. No significant differences were observed when jugular samples collected via central venous catheter were compared to jugular samples collected via direct venipuncture/syringe aspiration and winged catheter/evacuated tube, respectively. Authors concluded that all sampling methods are clinically acceptable and that sampling methods should be standardized for research purposes and in clinical patients undergoing serial sampling.
Two studies were judged neutral to the question of sampling technique. One study of 15 adult dogs (LOE 3, Fair) investigated whether jugular venipuncture quality and use of a discard tube affected TEG results. Two venipuncture techniques, each used with and without discard tubes, were evaluated. Authors concluded that mild to moderate vein trauma predominantly and significantly affected R. Clot formation was more rapidly initiated in samples obtained with poorer quality technique, but the effect was mild and mitigated by use of discard tubes. The second (LOE 3, Good) investigated TEG results in clinically healthy dogs using jugular vein blood samples obtained using a butterfly catheter and vacuum tube (discard tube used) with those obtained using direct venipuncture and syringe aspiration (no discard tube). Syringe-drawn samples had significantly shorter R time; however, no other significant differences were observed.
Five studies were judged to oppose the need for specific sample collection or handling techniques; however, 3 of these investigated coagulation testing but did not examine viscoelastic coagulation testing per se. An older study (LOE 3, Fair) investigated coagulation testing (prothrombin time [PT], activated partial thromboplastin time [aPTT], fibrin(ogen) degradation product, and fibrinogen concentration) of samples obtained using direct venipuncture and those obtained through heparinized jugular catheters (using a 2-syringe technique involving a discard sample) in clinically healthy dogs and concluded that sampling method did not significantly impact results. A study of healthy human volunteers (LOE 6, Poor) compared platelet function testing (optical aggregometry and PFA-100 [platelet function analyzer-100]) in samples obtained using direct venipuncture and those obtained through a butterfly cannula and concluded that the maximum aggregation response was not significantly different between collection methods. A study of coagulation testing (PT, aPTT, and fibrinogen concentration) in dogs (LOE 3, Fair) admitted to an intensive care unit compared samples obtained using direct venipuncture and through an intravenous catheter at the time of catheter placement and 24 hours later. Authors concluded that results agreement between the sampling techniques was clinically acceptable, and that sampling through IV catheters has potential to minimize venous trauma and patient discomfort in critically ill patients. More recently, a study investigating normal human volunteers (LOE 6, Fair) concluded that delivery of blood samples to the laboratory through a pneumatic tube system did not significantly affect ROTEM results. Another recent study investigated kaolin-activated TEG in clinically healthy dogs (LOE 3, Good) concluded that sample collection through variously sized IV catheters (20- Ga, 18-Ga, 14-Ga, and 13-Ga) did not affect results.
2.3 2.3 PICO question
In companion animals with hemostatic abnormalities (P), does the use of one citrate concentration (I) compared with another (C) alter testing results?
There is insufficient evidence to recommend a universal citrate concentration for TEG/ROTEM assays. Collection of blood into 3.2% buffered sodium citrate in a strict 1:9 ratio (final concentration 10.8 mM citrate) is suggested.
Summary of the evidence
No study was identified that specifically addressed impact of citrate concentration on TEG or ROTEM specifically, in either humans or animals. Two studies were judged to support that citrate concentration affects TEG/ROTEM results. The first (LOE 6, Good) evaluated the effect of 3.2% and 3.8% citrate on PT and aPTT testing (using 2 sets of reagents each) in 5 groups of humans (clinically healthy, those receiving stable oral anticoagulant treatment, and 3 groups of hospitalized patients). Authors concluded that 3.2% citrate was preferable. The second study (LOE 6, Good) investigated effects of varying total sample volumes and citrate concentration (3.2% and 3.8%) on PT and aPTT testing in clinically healthy humans and those receiving oral anticoagulant treatment. Under-filling sample tubes significantly prolonged results of both assays, and this effect was more pronounced using 3.8% citrate tubes. Authors concluded that 3.2% citrate is the preferred concentration. Studies in humans (all LOE 6, Fair) investigating PFA-100 closure times have shown that citrate concentration affects closure time and coefficient of variation.[24-26] International Society of Thrombosis and Hemostasis Special Scientific Committee (ISTH-SSC) recommendations for TEG/ROTEM in humans (LOE 6, Fair) recommend use of 3.2% citrate.
Three studies were judged to oppose that citrate concentration alters TEG/ROTEM results. One study in humans (LOE 6, Good) investigated various coagulation tests (but not viscoelastic coagulation assays) in 3 types of 3.2% buffered sodium citrate tubes and concluded that, although statistically significant differences were identified, these were not clinically significant. A second study (LOE 3, Good) investigated effects of 3.2% and 3.8% citrate on coagulation testing (PT, aPTT, fibrinogen concentration, factor VIII activity, and factor IX activity) in healthy dogs and dogs having hereditary hemostatic disorders. A significant prolongation in aPTT was observed using one mechanical test system for 3.2% citrate samples, as compared to 3.8% citrate samples; however, for the other tests and systems evaluated, no significant differences were observed. A third study (LOE 3, Good) investigated effects of 3.2% and 3.8% citrate on coagulation testing (PT, aPTT, fibrinogen concentration, platelet concentration, von Willebrand factor antigen testing, and platelet function (as assessed by the PFA-100) in 20 healthy dogs. Except platelet function, authors concluded that there was no significant difference in results between the 2 sample types. PFA-100 closure times were significantly shorter in 3.2% citrate samples compared to 3.8% citrate samples.
2.4 2.4 PICO question
In companion animals with hemostatic abnormalities (P), does use of a discard tube (I), compared with no discard tube (C), alter TEG/ROTEM results?
Use of a discard volume of blood is not required if “clean” (ie, atraumatic) venipuncture is achieved; however, use of a discard blood volume can mitigate the effects of suboptimal venipuncture on TEG/ROTEM results and, given the difficulty of achieving perfectly atraumatic venipuncture in many veterinary patients, routine use of a discard blood volume is suggested. As stated above in question 2.2, each center performing TEG/ROTEM should establish its own blood sampling protocol to be used for all patients undergoing this type of testing. The following blood tube sampling order is recommended (given in order of draw): discard/serum, citrate, EDTA, and heparin.
Summary of the evidence
One study (LOE 3, Good) directly supported the use of a discard blood volume and is summarized above under question 2.2. Five additional studies were judged neutral to the question of whether discard blood volume use is necessary. No studies were judged opposed to this question.
The 5 studies judged neutral were published prior to the supporting study, and the issue of discard blood volume was not directly addressed or evaluated in these studies; however, blood samples for TEG/ROTEM were collected after samples for serum biochemical or complete blood count analysis. Often, a statement was made that the order of tube collection was selected to minimize the impact of any excess tissue factor on TEG/ROTEM results. One study (LOE 3, Fair) validated a tissue factor (TF)-activated TEG assay in dogs; blood for TEG was collected last. A second study (LOE 3, Fair) validated a ROTEM assay in horses; blood for ROTEM was collected last. A third study (LOE 3, Fair) focused on establishing reference intervals for kaolin-activated TEG in clinically healthy dogs; blood for TEG analysis was collected after tubes for a serum biochemistry panel and CBC. A fourth study (LOE 3, Fair) investigated effects of sample storage conditions and time on TF-activated TEG in horses; blood for TEG was collected last. A fifth study (LOE 3, Fair) investigated impact of preanalytical factors on ROTEM in dogs; blood for ROTEM was collected last or following a discard blood volume.
2.5 2.5 PICO question
In companion animals with hemostatic abnormalities (P), does prolonged sample storage (I), compared with short-storage time (C), alter TEG/ROTEM results (O)?
Blood samples for TEG/ROTEM should be held at room temperature for 30 minutes prior to analysis.
Summary of the evidence
Four studies (2 in dogs, 2 in horses) have directly investigated the effect of storage time on TEG/ROTEM results. Two studies were judged to support that sample storage time affects test results, and 2 were considered neutral to this question. No studies were considered opposed.
Of 2 supporting studies, 1 (LOE 3, Good) validating a TF-activated TEG assay in dogs (1:50,000 final tissue factor dilution) 30 and 120 minutes following collection found that samples rested for 120 minutes appeared hypercoagulable compared to those rested 30 minutes. A second study (LOE 3, Fair) investigated the effect of sample storage time (30, 60, and 120 minutes) on TF-activated TEG in horses (1:3,600 final tissue factor dilution) and concluded that samples rested for longer appeared hypercoagulable.
Of 2 studies considered neutral to the question, 1 validating a ROTEM assay in horses (LOE 3, Fair) found that CT was significantly lower in samples rested for 20 hours, compared with those rested for 2 hours, and that maximum clot firmness (MCF) was significantly lower in samples rested for 4 hours compared to 2 hours. Otherwise, no significant differences were detected. A second study investigating preanalytical factors influencing ROTEM in dogs (LOE 3, Good) evaluated sample holding times ranging from 0 to 30 minutes and concluded that results are affected if a weak clotting activator (diluted TF) is used, but not if a strong clotting activator (ROTEM Ex-TEM or In-TEM reagents) is used.
2.6 2.6 PICO question
In companion animals with hemostatic abnormalities (P), does sample storage at 37°C (I), compared with sample storage at room temperature (20°C) (C), alter TEG/ROTEM results?
Blood samples for TEG/ROTEM should be held at room temperature prior to analysis and should not be refrigerated.
Summary of the evidence
Two studies directly investigated effects of sample storage temperature on TEG/ROTEM. One study supporting that sample storage temperature affects ROTEM results (LOE 3, Fair) validated a ROTEM assay in horses and compared samples stored at 4°C to those stored at room temperature (approximately 20°C). Authors found that CT and MCF were significantly higher and angle significantly lower in refrigerated samples, compared to room temperature samples. A second study also supporting that sample storage temperature affects TEG results (LOE 3, Good) compared samples held at 37°C for 30 minutes, those prewarmed to 37°C immediately prior to analysis, and samples held at room temperature for 30 minutes. Samples rested at 37°C for 30 minutes had significantly greater angle and MA than samples rested at room temperature, and immediately prewarmed samples had significantly greater MA than those rested at room temperature; all differences were judged to be minor.
A third study (LOE 3, Good), judged neutral to the question of temperature, investigated effects of preanalytical factors on ROTEM in dogs. Authors performed ROTEM following 0, 10, 20, and 30 minutes of sample rest at 37°C and following 20 minutes of sample rest at room temperature (22.5°C). Focus of this investigation was storage time rather than temperature, but authors concluded that results from samples rested at room temperature (which were rewarmed to 37°C for 10 minutes prior to ROTEM, for a total rest period of 30 minutes) were not significantly different from those rested at 37°C for 30 minutes.
All studies evaluated support atraumatic acquisition of free-flowing blood for TEG/ROTEM analysis, although evidence is lacking to support universal use of a specific blood vessel. Contamination of the whole blood samples with TF during the collection process (eg, as the needle is redirected into perivascular tissue) is presumed to influence TEG/ROTEM results, although to date no studies have investigated this specifically, eg, by comparing TF concentration in traumatically and atraumatically acquired samples. Venipuncture technique predominantly affects R time, which may have limited clinical significance. Further, use of a discard blood volume can mitigate this effect. Any larger vessel from which free-flowing blood may be collected can be considered acceptable for TEG/ROTEM sampling; realistically, in most companion animals, this equates to the jugular vein. In large dogs, large veins other than the jugular potentially can be used. In horses, cats, and small dogs, use of the jugular vein seems most likely to facilitate atraumatic collection of a free-flowing blood sample.
Several studies investigated sampling technique for traditional coagulation tests but did not investigate viscoelastic testing per se. It seems likely that sampling technique has greater potential to affect TEG/ROTEM results (whole blood sample) than PT/aPTT results (plasma sample), given that platelet activation by shear stress may occur during suboptimal venipuncture. On the basis of the available literature, evidence for a specific sample collection system for TEG/ROTEM is lacking, but suggestions include using evacuated blood sample tubes (to minimize delay in anticoagulation), 21-Ga or greater needles sizes (to facilitate free-flowing blood), and a discard blood volume (to minimize tissue factor contamination of the blood sample). There is no need to use a discard tube if a serum sample is to be obtained first (the serum sample tube serves the same function as the discard volume). EDTA and heparin tubes should be drawn following citrate tubes to minimize effects of EDTA and heparin contamination, respectively, on all coagulation testing. Although not addressed specifically by Bauer et al, if heparinized catheters are used to collect blood samples for TEG/ROTEM, it seems prudent to use heparinase sample cups during TEG/ROTEM analysis to prevent exogenous heparin from impacting the results.
The population-intervention-comparison-outcome (PICO) question regarding citrate concentration did not address the issue of blood to citrate ratio. As summarized in question 2.3, studies have shown that under- or overfilling citrate tubes (regardless of citrate concentration) affects results of coagulation testing, although this has not been investigated in TEG/ROTEM specifically. Recommendations for TEG/ROTEM citrate concentration and citrate to blood ratio are extrapolated from other coagulation test studies. Pending studies in viscoelastic testing specifically, use of 3.2% buffered sodium citrate in a strict 1:9 ratio (final concentration 10.8 mM citrate) is suggested.
Based on studies to date, TEG/ROTEM results become more coagulable with time. Equine blood may be more stable than canine blood for TEG/ROTEM analysis on the basis of one study. TEG/ROTEM results for citrated blood samples utilizing no coagulation activator or weak activators are more variable, likely due to the variability in factor XIIa activity at the time of recalcification. Strong coagulation activators can minimize or eliminate the effects of prolonged sample storage. A universal sample resting time of 30 minutes is reasonable to allow sample transportation to the laboratory while minimizing effects of contact activation of coagulation. Resting samples at room temperature is more practical for most testing sites than resting at 37°C, particularly if transport to a laboratory is involved. If serial TEG/ROTEM analyses are to be performed in one patient, samples should be rested for the same amount of time and at the same temperature for each analysis.
All centers performing TEG/ROTEM should establish sample acquisition and handling protocols to be used with all patients undergoing this type of testing, including standardized venipuncture technique, blood collection system, sample storage time, and coagulation activator used. Patient data should be interpreted in light of method-specific (meaning both sampling method and TEG/ROTEM assay method) reference intervals. Papers reporting results of TEG/ROTEM studies should provide details of sample collection (including vein and collection system used) and handling (including sample rest time and temperature).
PubMed, U.S. National Library of Medicine, Bethesda, MD.
CAB Abstracts, CABI, Oxfordshire, UK.
- 3RECOVER evidence and knowledge gap analysis on veterinary CPR. Part 1: evidence analysis and consensus process: collaborative path toward small animal CPR guidelines.” J Vet Emerg Crit Care 2012; 22 (Suppl 1):S4–S12., . “
- 13Comparison of venous sampling methods for thromboelastography in clinically normal dogs. Am J Vet Res 2012; 73(12):1864–1870., , , et al.
- 14Effects of delayed anticoagulation and use of evacuated tubes on non-activated thrombelastography in dogs. Vet Clin Pathol 2012; 41(1):63–70., , .
- 27Recommendations for performing thromboelastrography/thromboelastometry in hemophilia: communication from the SSC of the ISTH. J Thromb Haemost 2013; DOI 10.111/jth.12458., , , et al.
- 33Establishment of reference intervals for kaolin-activated thromboelastography in dogs including an assessment of the effects of sex and anticoagulant use. J Vet Diagn Invest 2009; 21(5):641–648., , .
- 35Clot formation in canine whole blood as measured by rotational thromboelastometry is influenced by sample handling and coagulation activator. Blood Coagul Fibrinolysis 2010; 21(7):692–702., , , et al.