An overview of the potential sources of diagnostic errors in (classic) thromboelastography curve interpretation and preventive measures

Thromboelastography (TEG), a hemostatic point-of-care assay, provides global information about fibrin formation, platelet activation, and clot retraction in real-time. As it is an operator-dependent technique, error in any phase of the testing process can result in the misinterpretation of the thromboelastogram, and subsequently lead to mismanagement of the patient, wastage of blood products besides increasing the financial burden on the hospital and the patient. The present paper describes the possible errors leading to wrong thromboelastogram interpretation, and the respective preventive measure. In the light of limited resources available for operational challenges in TEG, this review paper can prove to be helpful.


Introduction
Thromboelastography (TEG), invented in 1948, is an assay that detects the contribution of both cellular and plasma components of hemostasis [1]. It provides global information about fibrin formation, platelet activation, and clot retraction in real-time. The ability of TEG to assess the hemostasis in whole blood makes it an ideal point-of-care diagnostic modality for identifying patients with abnormal clot dynamics [2]. It is used bedside or in central laboratories. TEG is widely used in the management of liver transplant patients, in obstetrics, in cardiac surgeries, and for trauma care [3,4]. Multiple studies have found that using a TEG based transfusion algorithm, significantly decreases the number of blood components transfused [5][6][7].

Working principle of thromboelastography
Thromboelastography measures the viscoelastic properties of whole blood clot formation under low shear stress. A pin connected to a torsion wire linked to a mechanical-electrical transducer is suspended from above in a cup containing whole blood and calcium chloride, maintained at 37 C as shown in Fig. 1. As the physical properties like the elasticity and strength of the developing clot changes, the rotation of the pin also gets affected and the movement of the pin is converted into electrical signals to create a graphical and a numerical output on the computer. A normal thromboelastogram is schematically represented in Fig. 2. The modern instrumentation can be used to analyze both native whole blood (without any additive) and citrated whole blood by using different initiators [1,8].
The test and the result interpretation are complex, and requires the correlation of medical and treatment history of the patient, and other laboratory parameters like the platelet count, prothrombin time and activated partial thrombolpastin time along with due considerations to the pre-analytical and analytical factors that influence the thromboelastogram. The misinterpretation of the thromboelastogram may lead to mismanagement of the patient, wastage of blood products besides increasing the financial burden on the hospital and the patient [9]. The present paper describes the possible errors leading to wrong thromboelastogram interpretation, based on our 10-year-experience of TEG in a trauma care set up.

Errors in the pre-analytical phase
In general, incorrect laboratory test results are most commonly due to factors occurring outside the laboratory, i.e. in the preanalytical phase and account for about 70% of all errors [10,11]. These are mostly preventable but are beyond the control of a laboratory personnel. On an average, pre-analytical errors account for 0.23-1.2% of the total hospital expenditure [12]. The common pre-analytical errors affecting a TEG tracing can occur either during sample collection or during transport.

Sources of error during sample collection
Blood samples to be tested could be either run directly after collection without any additive (native blood), often performed at bedside or could be adequately filled in a citrate vial and sent to a centralised laboratory. Although a mislabelled sample is one of the frequent pre-analytical errors but it obviously does not cause any peculiar change in the TEG tracing [13]. Variables that can lead to changes in the TEG tracings are: 1.2.1.1. Use of a wrong vial for blood collection. Use of a vial with a stronger anticoagulant like Ethylenediaminetetraacetic acid (EDTA) produces a straight line (Fig. 3) in TEG even when the patient has normal clot dynamics or is in a hypercoagulable state. This is due to inadequate reversal of calcium chelation in the blood sample by the specified amount of calcium added in vitro.

Preventive measure.
To collect whole blood up to the 'mark' in blue-capped blood collection tube. Nine volume of whole blood is collected in a one volume of 3.2% sodium citrated (0.105-0.109 mol/L) ensured by a 'mark' on the blue-capped plastic vial [14]. This ratio of 9:1 is critical to ensure minimal osmotic effects and changes in the free calcium concentration in the whole blood [15]. The  sample must be inverted several times to mix well after filling the vial to ensure proper mixing. Heparinised blood should be avoided as heparin has enzyme inhibiting property and interferes with end point determination [16]. Sample acceptance and rejection criteria must be defined for the diagnostic test in advance.

Incorrect blood withdrawal technique.
An already clotted blood sample is invariably rejected. Blood tested from an under filled citrated vial produces a graph representing a hypocoagulable state (Fig. 4) due to the dilution of clotting factors present in the blood with the excess anticoagulant in the vial.

Preventive measure.
To follow good phlebotomy practice. The importance of the quality of the representative sample received in the laboratory for testing should not be undervalued. The choice of needle size is based on better accessibility for either an arterial or venous sample and should be age appropriate. Blood should be collected keeping in mind the 'order of draw' recommended by CSLI and EFLM. [17,18] If blood is drawn too slowly, it may lead to some coagulation before it is tested. Regular training of the phlebotomists and the health care workers regarding the steps of sample collection is recommended.

Sources of error during transport
The two factors that commonly affect a TEG tracing during transport are as follows:

Delay in transit.
A delay in transport leads to delayed testing and hence loss of clotting factors with shorter plasma half-life [19]. This can give an appearance of false hypocoagulable state on TEG.

Preventive measure.
To follow the recommended testing time. A non-coagulated whole blood has to be run within 4 min of collection. It is reliable and preferred where the instrument is placed bed side for immediate results [20]. A citrated sample, is preferred in a set up where the sample is sent to a central laboratory for analysis. Sample is considered stable and useable for up to 2 h at room temperature [21]. Any delay may lead to the loss of factors like V, VII and VIII with shorter plasma half-life and may produce a graph representing poor clot dynamics. Time of collection of the sample along with the time of receiving it in the laboratory must be noted. Interpreting a stored sample is not advisable. Although, in a study that used kaolin as an activator, found that the storage of citrated whole blood for up to 30 min affected r-time only and not the other parameters [22].

Inappropriate handling of blood sample.
Excess shaking of the sample during transportation which may cause hemolysis or partial coagulation is unacceptable. In addition, the clotting factors that are mostly proteins, may be lost due to extreme temperature changes during transportation. This may result in absence of clot formation or a hypocoagulable state on TEG ( Fig. 3 or Fig. 4 respectively).

Preventive measure.
To standardise sample handling during transportation. In order to maintain the integrity of the representative sample, shipping the sample in appropriate container and regulate and maintain the optimum temperature during sample transport is preferable [23].

Errors in the analytical phase
The errors occurring in the analytical phase of the testing process are usually the hardest to identify. They may be related to the operator, to the instrument, to the quality of reagents or to the environmental. The entire in vitro clot dynamics occur in a reaction cup as described above and thus the end result depends a lot on how the cup has been handled. If the cup is not pushed down precisely by the operator into the specified socket in the TEG machine, then it produces a typical graph with a beak like opening. (Fig. 5). Incorrect placement of the reaction cup results in reduction of the free space of 1 cm between the cup and the pin thus influencing the sheer force that is measured for the formation of the TEG curve.
To appoint a trained personnel dedicated for operating TEG. Like for any other technique or equipment used in the laboratory, one must follow the same stringent quality process for TEG also. A trained personnel, dedicated for operating TEG should be appointed for the daily quality assessment program and maintenance of the instrument [24,25]. Competency assessment of laboratory personnel for operating the machine can be already defined by international bodies can be implemented [26]. To place the reaction cup correctly, gently push the cup down into the machine socket in the cup holder while supporting the cup holder from down. In most machines, final 'click' sound denotes that the cup has been placed accurately. Mentioning step-by-step instructions for its start-up, operation and shutdown along with the steps of sample preparation described in manufacturer's or laboratory's Standard Operating Procedure (SOP) manuals should be readily available.   (Fig. 4).

Preventive measure.
To use freshly prepared calcium chloride solution for the test. Calcium has a pivotal role in the coagulation cascade [28]. Citrate being a calcium chelator prevents the blood from clotting before the beginning of the test by binding to the calcium in blood. The use of freshly prepared 0.2 M calcium chloride (not older than 24 h) stored at room temperature is advised [29].

1.3.2.3.
Repeated re-use of the reaction cup. Repeated cleaning of cups with hypochlorite solution or any other cleaning agent leads to shrinking of the internal diameter. Reusing such cups may lead to accidental gripping of the pin during a run. This can lead to a jerk like movement in the pin and thus an abrupt opening of the TEG tracing (Fig. 6).

Preventive measure.
To use only disposable reaction cups. Cups used to hold the blood sample are made of cyrolite or medical grade polyvinyl chloride [30]. Multiple washing and re-use affects the internal diameter of the well due to deposition of salts and makes it prone to pin gripping. Single use of the disposable cups for each testing process will help reduce errors due to pin gripping.
1.3.2.5. Instrument error due to failed auto-calibration. The failure to auto-calibrate before testing a sample leads to the formation of a typical widely open graph starting at the origin (Fig. 7). A very short r-time and k-time, a large alpha angle and a high maximum amplitude is seen in these cases. An inexperienced person may misinterpret it as a hypercoagulable state as it appears to be a case of rapid clot formation.
1.3.2.6. Preventive measure. To perform frequent instrument maintenance. The calibration of an instrument is a critical activity. Calibration and instrument validation is recommended if the instrument has been displaced from one position to the other. Worn out, loose or faulty machinery parts, incorrect procedure and mishandling of the instrument by an untrained operator are common reasons for a failed calibration. Performance checks must be carried out at frequent intervals and recorded. Daily, weekly and monthly maintenance protocols for servicing the instrument should be formulated with the help of the manufacturer's guideline. The contact number Fig. 7. Widely open graph close to the origin appearing to have for increased rate of clot formation and high clot strength suggestive of a hypercoagulable state is actually due to failed auto-calibration before the sample run.  (Fig. 8).  Vibrations, also referred to as 'noise' which occur at any point during the sample analysis usually reflect as continuous or sudden spikes in the TEG tracing. It is due to the instability of either the pin or the reaction cup or both. It is associated with abnormal values of various amplitudes like amplitude at 10 min (A10) or at 20 min (A20), or maximum amplitude (Fig. 9).

Preventive measure.
To maintain a vibration-free environment. TEG is a sensitive equipment. Avoiding loud noises, mechanical disturbances or shocks to the instrument and to the platform on which it is placed, are essential to obtain a smooth curve. The displacement of the instrument during or in-between runs should be avoided. The location of equipment in use should be specified in the laboratory equipment inventory. The manufacturer's guidelines must be followed for transporting the instrument. Workbench space must be sufficient to prevent disturbances [34].

Evaporation of the sample from the reaction cup.
Evaporation of the plasma from the sample cup during the sample run alters the plasma to cell (platelets) ratio. The proportion of platelets and its activity become higher than actual, and thus are reflected as higher clot strength or abnormally high value of maximum amplitude in the TEG graph (Fig. 10).

Preventive measure.
To optimise the humidity and room temperature of the laboratory. Errors due to evaporation of the plasma are generally unpredictable and inevitable. The room and equipment temperatures should be monitored regularly and adjusted whenever required. Depending on the geographical area, a relative humidity of 35-50% and a temperature range of 20-25 C should be maintained for a comfortable working environment [35]. In case it still happens, repeat testing with a fresh sample and allowing it to run up to 30-35 min may prevent abnormal results. The laboratory facilities must meet the required environmental conditions, including any needed separation of work areas.  account for 13-20% of all errors [36]. Even with the advancement of technology, flagging a potentially spurious result is still lacking in the current instrument and thus needs meticulous review of the test results.

Preventive measure
To increase awareness amongst medical professionals. The physician finally interpreting the TEG tracing must be aware and competent to identify the common sources of pre-analytical and analytical errors, and should know the ways to troubleshoot the problems. In addition, the interpretation must be based on the statistically established reference interval for the local population, and it is suggested that each laboratory performing the test should refer to its own list of reference interval [8,37].
The above discussed errors in TEG curve interpretation have been categorised according to the nature of preventability in Table 1. The role of the laboratory physician in the context of emerging technologies is elaborated in an opinion paper [38]. It is the responsibility of the laboratory and the laboratory physician to identify and establish appropriate corrective actions and preventive measures for these errors. Sometimes, it may not be possible to judge the exact error in real-time, especially, if one is not aware of the technical operations involved to obtain the TEG graph. One has to correlate with the patient's clinical profile and other laboratory parameters while keeping a high index of suspicion for any possible pre-analytical or analytical error. In the light of limited resources available for operational challenges in TEG, this review paper can prove to be helpful.

Conclusion
Strengthening the knowledge of the technical staff and the concerned laboratory physicians on the working principle of thromboelastography, methodological loopholes, and gaining clarity on each step of the testing process will help to anticipate the potential sources of errors and to plan strategies accordingly to overcome them.

Conflicting interest
Nil.
CRediT authorship contribution statement