Rapid laboratory testing for trauma patients: where a perfect result may not be in the best interests of the patient

The French writer and philosopher Voltaire famously stated “Le mieux est l'ennemi du bien”[commonly translated as “The perfect is the enemy of the good”]. This principle is what Chandler and coworkers1 in this issue of TRANSFUSION appear to have hit upon for coagulation testing in massively bleeding patients. Clinicians appear to be happier with a less precise result in 15 minutes, rather than perfect result in 30 to 60 minutes, as long as most of the time the result they receive is close to the “true” value or at least sufficiently accurate to prevent the inappropriate transfusion of blood products. For the traumatologist, this principle is already well established. When a severely injured patient is brought to the emergency department, the physician does not follow the standard clinical process of obtaining a thorough history and physical exam, then ordering laboratory tests and diagnostic imaging, and then finally embarking on a course of treatment. Rather, they start with Advance Trauma Life Support (ATLS) guidelines and administer intravenous fluids and red blood cells (RBCs), intubate and ventilate the patient, and then get straight to the operating room. If they followed the normal, slow, and methodical medical approach in this situation, the patient would be dead by the time the surgeon had obtained the patient's past medical history. In transfusion medicine, fast turnaround time of pretransfusion testing has long been a desired characteristic. Blood transfusion began as a surgical technique, and it was clinicians who created the first hospital blood banks. Over time, the number of atypical antibodies grew, the serologic techniques required to identify them became more complex, and gradually, clinical pathologists laid claim to the field.2 Considerations of precision and accuracy increased: only with the most stringent compatibility testing could a patient be protected from a hemolytic transfusion reaction. Subsequently, transfusion specialists learned to scale back pretransfusion compatibility testing for the sake of speed. Long gone are the days of the minor antiglobulin crossmatch and the routine use of elaborate enhancement additives and room temperature or 37°C readings are fading from use; for more and more patients the AHG crossmatch has been replaced by the immediate spin or electronic crossmatch.3 Undoubtedly, some clinically important antibodies go undetected,4,5 but these occasions are rare and are surely outweighed by the benefit of providing blood products quickly without having to forgo compatibility testing altogether. Now, as blood transfusion laboratories continue to accelerate their turnaround time, pressure has begun to shift to the coagulation laboratory to either speed up or risk being bypassed altogether. The creation of STAT laboratories in close proximity to the patient's location may shave some time off the preanalytical phase of testing by shortening specimen transport time and eliminating competition with routine samples.6 However, even the analytic phase of standard coagulation tests (30-60 min) is often too long an interval to wait for reliable guidance on the use of blood products for a massively bleeding patient. As a result, many clinicians have turned their attention toward point-of-care coagulation testing. Although progress is being made in obtaining accurate prothrombin time (PT) and/or international normalized ratio (INR) with these devices,7,8 there has been less success to date with fibrinogen measurements.9 Thromboelastography (TEG), another point-of-care or near-care test, can provide a more nuanced depiction of the patient's coagulation status10-12 and has accordingly been adopted into the resuscitation protocols of some centers.13 However, while the turnaround time for rapid TEG (19 + 3.1 min10) may seem advantageous compared to routine laboratory tests, in most centers this time is spent by the clinician performing laboratory testing, instead of direct treatment for the patient. Given the limitations of both standard laboratory testing and point-of-care testing, many clinicians have opted to simply forgo laboratory testing altogether in the setting of massive blood loss and attempt instead to resuscitate the patient with their best approximation of whole blood, administering RBCs, platelets (PLTs), and plasma in equal unit ratios. While there is evidence to suggest that massively bleeding trauma patients who receive blood products in these proportions have better outcomes,14 these results must be interpreted cautiously. All data currently available on 1:1:1 ratio-based transfusion are retrospective in nature and therefore prone to various sources of bias and confounding. In addition, it is not clear whether the same benefit can be extrapolated to nontrauma patients experiencing massive blood loss or to patients who bleed less than 10 units in a 24-hour period. Indeed, in this latter group there is evidence that aggressive and early use of plasma and PLTs may actually worsen outcomes.15 Most importantly, there is as of yet no evidence per se that a strategy of transfusing based on ratios leads to better outcomes than transfusing based on laboratory values; indeed, it is likely that most physicians would prefer to have their transfusion decisions guided by objective markers of coagulation, if only they were available when they needed them. The study by Chandler and coworkers1 published in this issue does much to address this need. Recognizing that standard coagulation testing is often too slow to usefully guide transfusion decisions in the setting of emergent blood loss, they embarked on a campaign to minimize testing time while still providing reasonably accurate results. While they are not the first to make such an attempt,16-18 they appear to have set a “world record,” cutting turnaround time from 35 ± 37 to 14 ± 3 minutes for their emergency hemorrhage panel (complete blood count, PT, partial thromboplastin time [PTT], and fibrinogen), without a clinically important impact on the accuracy of their test results. Their description of their action plan provides a useful roadmap for other laboratories who may wish to follow in their footsteps. The 10 modifications made to standard practice were as follows: Emergency hemorrhage panels prioritized over all other STAT and routine samples. PTT removed from their hemorrhage panel. Thrombin time (TT) removed from the hemorrhage panel. Samples moved directly to testing area upon arrival to the laboratory, with computer entry and accessioning labels performed while the sample was commencing centrifugation. Centrifugation time shortened from 8 minutes at 2000 × g to 2 minutes at 4440 × g. Checks for clots eliminated after determining that clots in samples did not cause a clinically significant impact on the PT result and would only result in one erroneous fibrinogen result (normal to falsely low reading) in 2300 samples. Checks for hemolysis eliminated after determining that no clinically significant impact on the PT or the fibrinogen result was noted. Critical results not repeated before release. All results, critical or not, called to the clinical team immediately. Calibration curve for fibrinogen extended down to 53 mg/dL. If the result fell below this level, the result was reported as less than 53 mg/dL instead of doing the standard dilutions and repeat testing to get an “exact” result. While all of the above modifications are sensible compromises in the name of speed, several require additional comment. Elimination of the PTT and TT seem very reasonable, as in most cases neither is relevant to transfusion decisions: plasma is usually administered in response to an elevated PT or INR, while cryoprecipitate transfusions are generally triggered by low fibrinogen levels. However, in situations where a patient has been administered recombinant activated Factor VII, the PT may be artificially reduced to near normal range, despite persisting and severe underlying coagulation factor deficiencies; in these cases, ongoing need for plasma transfusion may only be detectable by an elevated PTT. On the other hand, while the TT generally provides no additional information once a functional fibrinogen assay has been obtained, it does have the advantage of being a faster test. As it turns out, the authors were obliged to do a fair amount of additional work modifying their fibrinogen assay to meet their target turnaround time of 20 minutes for more than 95% of samples tested. Considering that many clinicians have a somewhat binary approach to fibrinogen levels when deciding whether or not to administer cryoprecipitate or antifibrinolytics (e.g., yes if below 100 mg/dL, no if above), one wonders whether basing decisions on the TT instead might have been a simpler alternative. The authors also found that turnaround time could be shortened significantly by eliminating a number of routine quality assurance processes. Validation studies were reported for some of these changes (e.g., eliminating checks for clots or hemolysis) but not for others (e.g., change in centrifuge settings, elimination of retesting critical results). It should be noted as well that care must be taken to avoid identification errors when registering a sample into the laboratory computer system without actually having the sample in hand, as the authors propose for the sake of speed. If only an occasional sample is processed this way, the likelihood of either affixing the laboratory label to the wrong tube or sending the result to the wrong patient is probably low. However, if several samples are processed this way simultaneously then the risk may increase.19 The frequency with which clinicians order the rapid emergency hemorrhage panel is in fact a key determinant of its success. As the authors acknowledge, excess use of this panel would likely result in an increase in overall turnaround time, and thus clinicians must be careful not to abuse it. A rapid hemorrhage panel may prove to be a difficult tendency to resist. Some studies have found that up to one-third of laboratory tests ordered as “STAT” are not in response to actual medical emergencies but represent attempts to compensate for delays in pre- and postanalytic phases of testing or just general impatience on the part of the clinician.20 The fact that Chandler and colleagues1 found that, at baseline, two-thirds of all samples received by their laboratory were ordered as STAT suggests their institution may be prone to the same phenomenon. Laboratories implementing the rapid emergency hemorrhage panel will therefore need to keep a close eye on the appropriateness of the requests they receive lest they end up simply contributing to the steady inflation of ordering urgency. Even if used appropriately, the panel may prove victim to its own success—with results obtained fast enough to guide therapy, it would not be unreasonable to order a rapid emergency hemorrhage panel every 30 to 60 minutes on a massively bleeding patient. However, it should be possible to meet this increased demand by redistributing rather than increasing laboratory resources: a recent study from the United Kingdom reported a remarkable frequency of INR testing on intensive care patients—1.01 tests per day, of which 83% of the results were less than 1.521; presumably many of these tests did not need to be ordered in the first place. Despite these reservations, many clinicians and laboratory physicians (ourselves included) will be very interested in implementing a similar rapid emergency hemorrhage panel in their own institutions, and as real-world experience with abbreviated coagulation testing grows there may be opportunities to address other questions which the study by Chandler and colleagues1 was not designed to answer. For example, it may eventually prove possible to identify preanalytical variables, such as sample volume or transit time, that unduly influence the precision of the results obtained from the rapid panel and which would therefore indicate reinclusion of routine quality assurance checks. Even more importantly, it may be possible to show that use of the panel, by providing timely and objective information regarding the patient's coagulation status, results in more appropriate use of blood products and better patient outcomes than blindly giving plasma and PLTs to all bleeding patients. Undoubtedly, we will find that not all massively bleeding patients are the same and that one approach does not work for all. For some, waiting a little longer for a more precise characterization of their coagulation status will be worthwhile, although for others a quick and “close enough” approximation will be more appropriate. Knowing which route to go will require effective communication between the laboratory and the bedside, with consideration of the severity of the injury, the rate of hemorrhage, the speed with which surgical hemostasis will be reached, the severity of the patient's coagulopathy, and the existing hemostatic impairment. When conditions are changing rapidly at the bedside and every minute counts, laboratory physicians and technologists, both in the blood bank and the coagulation laboratory, must be able to keep up if they hope to influence clinical decision-making in the best interests of the patient. None.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.