Thromboelastography for Hemorrhagic Resuscitation

Wikkelsø A, Wetterslev J, Møller AM, Afshari A. Thromboelastography (Teg) or thromboelastometry (Rotem) to monitor haemostatic treatment versus usual care in adults or children with bleeding. Cochrane Database Syst Rev. 2016 Aug 22;2016(8):CD007871.

Study Population: 1493 total elective surgery patients included those with bleeding from cardiac surgery (1435 patients), excision of burn wounds (30 patients) and liver transplantation (28 patients)

All-cause mortality, requiring transfusion (PRBC, FFP, platelets), surgical re-intervention for bleeding control

Any adverse events

Severe bleeding and coagulopathy contribute to high morbidity and mortality in patients undergoing surgery. While conventional coagulation testing is commonly used to guide transfusion, these assays do not reveal the complexities of hemostatic derangements and were not designed to predict bleeding or guide coagulation management. Point of care thromboelastography (TEG) and rotational thromboelastometry (ROTEM) assays measure fibrinolysis and were developed partly for the purpose of guiding hemostatic resuscitation.1 TEG/ROTEM are currently used in many trauma centers to guide acute resuscitation, within minutes of a patient's arrival to the emergency department (ED). We examined the evidence pertaining to the benefit and safety of these assays for ED patients with acute hemorrhage.

The systematic review and meta-analysis2 included 17 randomized controlled trials with 1493 subjects, most of whom were adults undergoing elective cardiac surgery. The sample size of the 16 trials ranged from 28 to 224 per trial (median of 96 subjects per trial). The trials compared TEG- or ROTEM-guided transfusion strategies to care guided by clinical judgment or conventional testing. The systematic review’s primary outcome was overall mortality; secondary outcomes included the proportion of patients requiring transfusion of packed red blood cells (PRBC), platelets, and plasma, the amount of blood products transfused, rate of surgical re-intervention, hospital length of stay, and complications from bleeding or transfusion.2

All-cause mortality was lower with TEG/ROTEM (7.4% control vs 3.9% with Accepted Article TEG/ROTEM, Relative Risk [RR]: 0.5; 95%CI, CI 0.3 to 0.9; Absolute Risk Difference [ARD]: 3.6%; Number-needed-to-treat [NNT]: 28; low quality evidence, 8 trials, n=717). The proportion of patients receiving transfusion of packed red blood cells (PRBC) was lower in patients assigned to TEG/ROTEM (RR 0.9; CI 0.8 to 0.9; ARD: 10%; NNT: 10; low quality evidence, 10 trials, n=832). Similarly, the need for plasma (RR 0.6; CI 0.3 to 0.9; ARD: 20%; NNT: 5; low quality evidence, 8 trials, n=761), and platelets (RR 0.7; CI 0.6 to 0.9; ARD: 9%; NNT: 11; low quality evidence, 10 trials, n=832) was also lower. The systematic review found no difference in rate of surgical re-intervention to control bleeding. Some trials reported the amount of blood products used in each group as a continuous variable. However, because of the variability of the blood product measures (some trials reported them in units, others in milliliters, or milliliters/kg) and methods of summarizing data (medians and quartiles versus means and standard deviation), as well as the significant heterogeneity between these trials, the systematic review did not summarize these data.2

Adverse outcomes and complications were inconsistently reported, thereby making estimates unreliable.

Given the many limitations of the data, the authors of this systematic review offer guarded support for their findings and rate the quality of evidence as low. In addition, due to the wide array of transfusion algorithms and varied thresholds for administering blood products in each trial, there was significant clinical and statistical heterogeneity among the trials.2 Furthermore, no trial used a 1:1:1 hemorrhage transfusion strategy of red cells, plasma, and platelets; an approach currently favored by many experts.11^, 12

Mortality data was missing or incompletely reported in many trials. In addition, statistical significance was lost for mortality when comparing fixed-effect and random-effect models, thereby suggesting the results were unreliable. Therefore, we have not reported mortality outcomes in the summary table.

Methodological weaknesses were rife throughout the included trials. Only two of the 17 trials (12%) were appropriately blinded and had low risk of bias. Additionally, the dataset was too heterogenous and biased to allow for meaningful appraisal of most outcomes.

Lastly, 91% of study participants were adults undergoing elective cardiac surgery. Data from these participants are not generalizable to other clinical settings of hemorrhage management, such as the ED or intensive care units. It may be premature to extrapolate the findings to acute trauma resuscitations, or spontaneously bleeding medical patients, until trials are conducted in these specific populations. Severely injured patients or those with gastrointestinal bleeding, for example, may arrive in the ED already coagulopathic, with ongoing blood loss that continues until source control of bleeding is accomplished. In the two randomized trials comparing TEG/ROTEM-guided strategies to conventional management in acute care environments, no meaningful benefits were seen.9^, 10

In summary, TEG/ROTEM-guided hemostatic resuscitation requires further study. While early, low quality data from elective surgical settings suggests a possible benefit, more high-grade evidence from ED-relevant resuscitations are needed. Therefore, we have assigned a color recommendation of Yellow (Unclear if benefits, more data needed).

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

October 12, 2021