A machine learning-based Coagulation Risk Index predicts acute traumatic coagulopathy in bleeding trauma patients.

BACKGROUND: Acute traumatic coagulopathy (ATC) is a well-described phenomenon known to begin shortly after injury. This has profound implications for resuscitation from hemorrhagic shock, as ATC is associated with increased risk for massive transfusion (MT) and mortality. We describe a large-data machine learning-based Coagulation Risk Index (CRI) to test the early prediction of ATC in bleeding trauma patients. METHODS: Coagulation Risk Index was developed using continuous vital signs (VSs) available during the first 15 minutes after admission at a single trauma center over 4 years. Data to compute the CRI were derived from continuous features of photoplethymographic and electrocardiographic waveforms, oximetry values, and blood pressure trends. Two groups of patients at risk for ATC were evaluated: critical administration threshold and patients who received an MT. Acute traumatic coagulopathy was evaluated in separate models and defined as an international normalized ratio (INR) >1.2 and >1.5 upon arrival. The CRI was developed using 2 years of cases for training and 2 years for testing. The accuracy of the models is described by area under the receiver operator curve with 95% confidence intervals. RESULTS: A total of 17,567 patients were available for analysis with continuous VS data, 52.8% sustained blunt injury, 30.2% were female, and the mean age was 44.6 years. The ability of CRI to predict ATC in critical administration threshold patients was excellent. The true positive and true negative rates were 95.6% and 88.3%, and 94.9% and 89.2% for INR >1.2 and INR >1.5, respectively. The CRI also demonstrated excellent accuracy in patients receiving MT; true positive and true negative rates were 92.8% and 91.3%, and 100% and 88.1% for INR >1.2 and INR >1.5, respectively. CONCLUSION: Using continuous VSs and large-data machine learning capabilities, the CRI accurately predicts early ATC in bleeding patients. Clinical application may guide early hemostatic resuscitation. Extension of this technology into the prehospital setting could provide earlier treatment of ATC. LEVEL OF EVIDENCE: Retrospective, Prognostic Study; Level III.

Out with the old, in with the new? The revised AAST grading schema better predicts splenic salvage but not splenectomy.

BACKGROUND: The revised American Association for the Surgery of Trauma (AAST) organ injury scale (OIS) for splenic injury incorporates radiologic features but the implications of this are unknown. We hypothesized that the revised AAST-OIS would better predict outcomes. METHODS: Patients with a blunt splenic injury admitted to a Level I trauma center were reviewed from 2016 to 2021. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for splenectomy were calculated for high-grade injuries (AAST-OIS grades IV-V) using both schemas. RESULTS: Of the 852 patients analyzed, 48.5% were observed, 24.6% were embolized, and the remaining underwent operative intervention. The median AAST-OIS increased from II to III (p â€‹< â€‹0.01). Sensitivity (38.0% vs. 73.7%) and NPV (80.9% vs. 88.2%) for splenectomy increased for high-grade injuries but specificity (93.5% vs 70.1%) and PPV (67.5% vs 46.7%) decreased. CONCLUSION: The revised AAST-OIS better predicted splenic salvage but is less accurate at predicting need for splenectomy.

INHIBITORS OF INORGANIC POLYPHOSPHATE AND NUCLEIC ACIDS ATTENUATE IN VITRO THROMBIN GENERATION IN PLASMA FROM TRAUMA PATIENTS.

ABSTRACT: Background: Inorganic polyphosphate (polyP) is a procoagulant polyanion. We assessed the impact of polyP inhibition on thrombin generation after trauma using the novel polyP antagonists, macromolecular polyanion inhibitor 8 (MPI 8), and universal heparin reversal agent 8 (UHRA-8). Methods: Plasma thrombin generation (calibrated automated thrombogram, CAT), in 56 trauma patients and 39 controls +/- MPI 8 and UHRA-8 (50 µg/mL), was expressed as lag time (LT, minutes), peak height (PH, nM), and time to peak (ttPeak, minutes), with change in LT (ΔLT) and change in ttPeak (ΔttPeak) quantified. Results expressed in median and quartiles [Q1, Q3], Wilcoxon matched-pairs testing, P < 0.05 significant. Results: Trauma patients had greater baseline PH than controls (182.9 [121.0, 255.2]; 120.5 [62.1, 174.8], P < 0.001). MPI 8 treatment prolonged LT and ttPeak in trauma (7.20 [5.88, 8.75]; 6.46 [5.45, 8.93], P = 0.020; 11.28 [8.96, 13.14]; 11.00 [8.95, 12.94], P = 0.029) and controls (7.67 [6.67, 10.50]; 6.33 [5.33, 8.00], P < 0.001; 13.33 [11.67, 15.33]; 11.67 [10.33, 13.33], P < 0.001). UHRA-8 treatment prolonged LT and ttPeak and decreased PH in trauma (9.09 [7.45, 11.33]; 6.46 [5.45, 8.93]; 14.02 [11.78, 17.08]; 11.00 [8.95, 12.94]; 117.4 [74.5, 178.6]; 182.9 [121.0, 255.2]) and controls (9.83 [8.00, 12.33]; 6.33 [5.33, 8.00]; 16.67 [14.33, 20.00]; 11.67 [10.33, 13.33]; 55.3 [30.2, 95.9]; 120.5 [62.1, 174.8]), all P < 0.001. Inhibitor effects were greater for controls (greater ΔLT and ΔttPeak for both inhibitors, P < 0.001). Conclusion: PolyP inhibition attenuates thrombin generation, though to a lesser degree in trauma than in controls, suggesting that polyP contributes to accelerated thrombin generation after trauma.

A pseudo-dilemma: Are we over-diagnosing and over-treating traumatic splenic intraparenchymal pseudoaneurysms?

BACKGROUND: Splenic embolization for traumatic vascular abnormalities in stable patients is a common practice. We hypothesize that modern contrast-enhanced computed tomography (CT) over diagnoses posttraumatic splenic vascular lesions, such as intraparenchymal pseudoaneurysms (PSA) that may not require embolization. METHODS: We reviewed the experience at our high-volume center with endovascular management of blunt splenic injuries from January 2016 to December 2021. Multidisciplinary review was used to compared initial CT findings with subsequent angiography, analyzing management and outcomes of identified vascular lesions. RESULTS: Of 853 splenic injuries managed overall during the study period, 255 (29.9%) underwent angiography of the spleen at any point during hospitalization. Vascular lesions were identified on 58% of initial CTs; extravasation (12.2%) and PSA (51.0%). Angiography was performed a mean of 22 hours after admission, with 38% done within 6 hours. Embolization was performed for 90.5% (231) of patients. Among the 130 patients with PSA on initial CT, 36 (27.7%) had no visible lesion on subsequent angiogram. From the 125 individuals who did not have a PSA identified on their initial CT, 67 (54%) had a PSA seen on subsequent angiography. On postembolization CT at 48 hours to 72 hours, persistently perfused splenic PSAs were seen in 41.0% (48/117) of those with and 22.2% (2/9) without embolization. Only one of 24 (4.1%) patients with PSA on angiography observed without embolization required delayed splenectomy, whereas 6.9% (16/231) in the embolized group had splenectomy at a mean of 5.5 ± 4 days after admission. CONCLUSION: There is a high rate of discordance between CT and angiographic identification of splenic PSAs. Even when identified at angiogram and embolized, close to half will remain perfused on follow-up imaging. These findings question the use of routine angioembolization for all splenic PSAs. LEVEL OF EVIDENCE: Therapeutic/Care Management; Level IV.

A murine multiple-injury model for the study of thromboinflammation.

INTRODUCTION: Neutrophil extracellular traps (NETs) contribute to trauma-induced coagulopathy. We aimed to develop a murine multiple-injury model that induces thrombo-inflammatory response, that is, NETosis and accelerated thrombin generation. METHODS: Wild-type male mice (n = 10, aged 8-12 weeks) underwent multiple injuries (gastrocnemius crush, femur fracture, and laparotomy) and were compared with an uninjured control group (n = 10). Mice were euthanized by cardiac puncture performed 3 hours after injury. Whole blood samples were immediately processed to platelet poor plasma for thrombin generation kinetics (calibrated automated thrombogram), myeloperoxidase (MPO), and von Willebrand factor quantification. Immunohistochemistry of lung tissue was performed to assess for citrullinated histone 3 (CitH3) and MPO. A NETosis cluster was defined as 3+ neutrophils staining for CitH3 at 400× magnification (CitH3 cluster). Data were presented either as mean (SD) or median (interquartile range) with p < 0.05 significant. Sham and trauma treated animals were compared by the two-sample Wilcoxon rank-sum test. RESULTS: Animals subjected to multiple injuries had accelerated thrombin generation compared with controls with greater peak height (61.3 [41.2-73.2] vs. 28.4 [19.5-37.5] nM, p = 0.035) and shorter time to peak (3.37 [2.81-3.81] vs. 4.5 [4.08-4.75] minutes, p = 0.046). Markers of neutrophil activation were greater following multiple injuries than in controls (MPO, 961.1 [858.1-1116.8] vs. 481.3 [438.0-648.9] ng/mL; p = 0.004). NETosis, as evidenced by the aforementioned defined number of CitH3 clusters in the lung, was greater in multiple-injury animals than in controls (mean [SD], 3 [2.9] vs. 0.2 [0.7]; p = 0.009). CONCLUSION: This is the first study to demonstrate that NETosis and accelerated thrombin generation can be induced using a murine multiple-injury model, as early as 3 hours following injury.

Early lyophilized cryoprecipitate enhances the ADAMTS13/VWF ratio to reduce systemic endotheliopathy and lessen lung injury in a mouse multiple-trauma hemorrhage model.

BACKGROUND: Recent studies in severely injured patients suggest an important role of von Willebrand Factor (VWF) and ADAMTS13 in the endotheliopathy of trauma (EoT). We hypothesized that the early use of cryoprecipitate would be effective as an endothelial protector by supplementing physiologic VWF and ADAMTS13 to reverse the EoT. We tested a pathogen-reduced lyophilized cryoprecipitate (LPRC) that could expedite the early administration of cryoprecipitate in the battlefield. METHODS: A mouse multiple-trauma model with uncontrolled hemorrhage (UCH) from liver injury was utilized followed by hypotensive resuscitation (mean arterial pressure, 55-60) × 3 hours with lactated Ringer's (LR), fresh frozen plasma (FFP), conventional pathogen-reduced cryoprecipitate (CC), and LPRC. Blood was collected for measurement of syndecan-1, VWF, and ADAMTS13 by ELISA. Lungs were stained for histopathologic injury and syndecan-1 and bronchial alveolar lavage (BAL) fluid harvested for protein as an indicator of permeability. Statistical analysis was by ANOVA followed by Bonferroni correction. RESULTS: Following multiple trauma and UCH, blood loss was similar across groups. Mean volume of resuscitation was higher in the LR group compared with the other resuscitation groups. Lung histopathologic injury, syndecan-1 immunostaining and BAL protein were higher with LR compared with resuscitation with FFP and CC, while LPRC further reduced BAL compared with FFP and CC. The ADAMTS13/VWF ratio was significantly lower in LR but improved with FFP and CC, comparable to shams while LPRC further increased this ratio. CONCLUSION: The protective effects of CC and LPRC were comparable to FFP in ameliorating the EoT in our murine multiple trauma and UCH model. Lyophilized cryoprecipitate may also provide additional benefit by enhancing the ADAMTS13/VWF ratio. These data provide evidence of the safety and efficacy of LPRC and warrants further investigation for its potential application in military settings once approved for human administration.

Contemporary management and outcomes of penetrating colon injuries: Validation of the 2020 AAST Colon Organ Injury Scale.

INTRODUCTION: The American Association for the Surgery of Trauma Colon Organ Injury Scale (OIS) was updated in 2020 to include a separate OIS for penetrating colon injuries and included imaging criteria. In this multicenter study, we describe the contemporary management and outcomes of penetrating colon injuries and hypothesize that the 2020 OIS system correlates with operative management, complications, and outcomes. METHODS: This was a retrospective study of patients presenting to 12 Level 1 trauma centers between 2016 and 2020 with penetrating colon injuries and Abbreviated Injury Scale score of <3 in other body regions. We assessed the association of the new OIS with surgical management and clinical outcomes and the association of OIS imaging criteria with operative criteria. Bivariate analysis was done with χ 2 , analysis of variance, and Kruskal-Wallis, where appropriate. Multivariable models were constructed in a stepwise selection fashion. RESULTS: We identified 573 patients with penetrating colon injuries. Patients were young and predominantly male; 79% suffered a gunshot injury, 11% had a grade V destructive injury, 19% required ≥6 U of transfusion, 24% had an Injury Severity Score of >15, and 42% had moderate-to-large contamination. Higher OIS was independently associated with a lower likelihood of primary repair, higher likelihood of resection with anastomosis and/or diversion, need for damage-control laparotomy, and higher incidence of abscess, wound infection, extra-abdominal infections, acute kidney injury, and lung injury. Damage control was independently associated with diversion and intra-abdominal and extra-abdominal infections. Preoperative imaging in 152 (27%) cases had a low correlation with operative findings ( κ coefficient, 0.13). CONCLUSION: This is the largest study to date of penetrating colon injuries and the first multicenter validation of the new OIS specific to these injuries. While imaging criteria alone lacked strong predictive value, operative American Association for the Surgery of Trauma OIS colon grade strongly predicted type of interventions and outcomes, supporting use of this grading scale for research and clinical practice. LEVEL OF EVIDENCE: Prognostic and Epidemiological; Level III.

Time to Splenic Angioembolization Does Not Impact Splenic Salvage Rates.

We aimed to determine whether early (<6 hours) vs delayed (≥6 hours) splenic angioembolization (SAE) after blunt splenic trauma (grades II-V) impacted splenic salvage rates at a level I trauma center (2016-2021). The primary outcome was delayed splenectomy by timing of SAE. Mean time of SAE was determined for those who failed vs those who had successful splenic salvage. We retrospectively identified 226 individuals, from which 76 (33.6%) were in the early group and 150 (66.4%) were in the delayed group. The early group had higher AAST grade, greater amount of hemoperitoneum on CT, and 3.9x greater odds of undergoing delayed splenectomy (P = .046). Time to embolization was shorter in the group that failed splenic salvage (5 vs 10 hours, P = .051). On multivariate analysis, timing of SAE had no effect on splenic salvage. This study supports performing SAE on an urgent rather than emergent basis in stable patients after blunt splenic injury.

Surveillance Imaging Associated With Delayed Splenectomy in High-Grade Blunt Splenic Trauma.

This retrospective, single-site study at a level I trauma center (2016-2021) sought to determine whether repeat CT had an impact on clinical decision making after splenic angioembolization following blunt splenic trauma (grades II-V). The primary outcome was need for intervention after subsequent imaging (defined as angioembolization and/or splenectomy) by high- or low-grade injury. Of the 400 individuals examined, 78 (19.5%) underwent intervention after repeat CT, from which 17% were in the low-grade group (grades II and III) and 22% were in the high-grade group (grades IV and V). Individuals in the high-grade group were 3.6 times more likely to undergo delayed splenectomy than those in the low-grade group (P = .006). Delayed intervention after surveillance imaging in blunt splenic injury is driven mostly by the identification of new vascular lesions and leads to greater rates of splenectomy in high-grade injuries. Surveillance imaging should be considered for all AAST injury grades II or higher.

Surgical Science and the Evolution of Critical Care Medicine.

Surgical science has driven innovation and inquiry across adult and pediatric disciplines that provide critical care regardless of location. Surgically originated but broadly applicable knowledge has been globally shared within the pages Critical Care Medicine over the last 50 years.

Contemporary Management and Outcomes of Injuries to the Inferior Vena Cava: A Prospective Multicenter Trial From PROspective Observational Vascular Injury Treatment.

INTRODUCTION: Injuries to the inferior vena cava (IVC), while uncommon, have a high mortality despite modern advances. The goal of this study is to describe the diagnosis and management in the largest available prospective data set of vascular injuries across anatomic levels of IVC injury. METHODS: The American Association for the Surgery of Trauma PROspective Observational Vascular Injury Treatment (PROOVIT) registry was queried from November 2013 to January 2019. Demographics, diagnostic modalities, injury patterns, and management strategies were recorded and analyzed. Comparisons between anatomic levels were made using non-parametric Wilcoxon rank-sum statistics. RESULTS: 140 patients from 19 institutions were identified; median age was 30 years old (IQR 23-41), 75% were male, and 62% had penetrating mechanism. The suprarenal IVC group was associated with blunt mechanism (53% vs 32%, P = .02), had lower admission systolic blood pressure, pH, Glasgow Coma Scale (GCS), and higher ISS and thorax and abdomen AIS than the infrarenal injury group. Injuries were managed with open repair (70%) and ligation (30% overall; infrarenal 37% vs suprarenal 13%, P = .01). Endovascular therapy was used in 2% of cases. Overall mortality was 42% (infrarenal 33% vs suprarenal 66%, P<.001). Among survivors, there was no difference in first 24-hour PRBC transfusion requirement, or hospital or ICU length of stay. CONCLUSIONS: Current PROOVIT registry data demonstrate continued use of ligation extending to the suprarenal IVC, limited adoption of endovascular management, and no dramatic increase in overall survival compared to previously published studies. Survival is likely related to IVC injury location and total injury burden.

TRAUMA-DERIVED EXTRACELLULAR VESICLES ARE SUFFICIENT TO INDUCE ENDOTHELIAL DYSFUNCTION AND COAGULOPATHY.

ABSTRACTINTRODUCTION: Although a number of studies have demonstrated increased release of extracellular vesicles (EVs) and changes in their origin differentials after trauma, the biologic significance of EVs is not well understood. We hypothesized that EVs released after trauma/hemorrhagic shock (HS) contribute to endotheliopathy and coagulopathy. To test this hypothesis, adoptive transfer experiments were performed to determine whether EVs derived from severely injured patients in shock were sufficient to induce endothelial dysfunction and coagulopathy. Methods: Total EVs were enriched from plasma of severely injured trauma/HS patients or minimally injured patients by ultracentrifugation and characterized for size and numbers. Under isoflurane anesthesia, noninjured naive C57BL/6J mice were administered EVs at varying concentrations and compared with mice receiving equal volume vehicle (phosphate-buffered saline (PBS)) or to mice receiving EVs from minimally injured patients. Thirty minutes after injection, mice were sacrificed, and blood was collected for thrombin generation (thrombin-antithrombin, thrombin-antithrombin complex [TAT] assay) and syndecan-1 by enzyme-linked immunoabsorbent assay (ELISA). Lungs were harvested for examination of histopathologic injury and costained with von Willebrand factor and fibrin to identify intravascular coagulation. Bronchial alveolar lavage fluid was aspirated from lungs for protein measurement as an indicator of the endothelial permeability. Data are presented as mean ± SD, P < 0.05 was considered significant, and t test was used. Results: An initial proof-of-concept experiment was performed in naive mice receiving EVs purified from severely injured trauma/HS patients (Injury Severity Score [ISS], 34 ± 7) at different concentrations (5 × 106 to 3.1 × 109/100 µL/mouse) and compared with PBS (control) mice. Neither TAT nor syndecan-1 levels were significantly different between groups at 30 minutes after EV infusion. However, lung vascular permeability and histopathologic injury were significantly higher in the EV group, and lung tissues demonstrated intravascular fibrin deposition. Based on these data, EVs from severely injured trauma/HS patients (ISS, 32 ± 6) or EVs from minimally injured patients (ISS, 8 ± 3) were administered to naive mice at higher concentrations (1 × 109 to 1 × 1010 EV/100 µL/mouse). Compared with mice receiving EVs from minimally injured patients, plasma TAT and syndecan-1 levels were significantly higher in the trauma/HS EV group. Similarly, bronchial alveolar lavage protein and lung histopathologic injury were higher in the trauma/HS EV group, and lung tissues demonstrated enhanced intravascular fibrin deposition. Conclusion: These data demonstrate that trauma/HS results in the systemic release of EVs, which are capable of inducing endotheliopathy as demonstrated by elevated syndecan-1 and increased permeability and coagulopathy as demonstrated by increased TAT and intravascular fibrin deposition. Targeting trauma-induced EVs may represent a novel therapeutic strategy.

Biomarkers of endothelial cell dysfunction persist beyond resuscitation in patients with hemorrhagic shock.

BACKGROUND: It has been shown that microRNA-19b (miR-19b) binds to and degrades syndecan-1 after hemorrhagic shock (HS) and contributes to endothelial dysfunction in vitro and in vivo. The objective of the current study was to assess longitudinal changes in miR-19b and syndecan-1 in HS patients. METHODS: Blood samples from HS patients (blood pressure <90 mm Hg and ≥2 U blood) were collected upon admission, completion of hemostasis, and after 24 hours for miR-19b (quantitative reverse transcription PCR) and syndecan-1 (enzyme-linked immunosorbent assay) and compared with controls and minimally injured (Injury Severity Score, ≤9). Inflammatory cytokines were measured (Luminex [Thermo Fisher, Waltham, MA]). Correlations between syndecan-1, miR-19b, inflammatory markers, and patient outcomes were performed. Logistic regression models were developed for outcomes. RESULTS: Thirty-four HS patients were studied: age, 46 (19-89) years; male, 82%; penetrating, 35%; Injury Severity Score, 24 ± 10; and blood products at 24 hours, 21 ± 19 U. MicroRNA-19b was increased upon arrival and further increased over time: 4.6 → 6.7 → 24.1-fold change compared with 0.1 and 1.2 for minimally injured patients and controls, respectively. Syndecan-1 was increased to 42.6 → 50 → 51.5 ng/mL over time compared with 14.7 and 23.5 for minimally injured and controls, respectively. Values for both biomarkers remained significantly increased through 24 hours and were associated with a persistent increase in inflammatory cytokines. Admission syndecan-1 significantly predicted mortality, coagulopathy, and massive transfusion. CONCLUSION: We have shown for the first time that miR-19b and syndecan-1 were biomarkers for endothelial dysfunction independent of resuscitation. MicroRNA-19b did not demonstrate a strong correlation with syndecan-1 nor outcomes. Admission syndecan-1, however, remains a strong prognostic marker, but its elevation over time suggests a versatile role following HS that requires further investigation. LEVEL OF EVIDENCE: Prognostic/Epidemiological; Level II.

Validation of the American Association for the Surgery of Trauma Organ Injury Scale for Penetrating Colon Injuries.

In 2020, the American Association for the Surgery of Trauma (AAST) published a revision of the organ injury scale (OIS) for bowel injuries. The update included for the first time a separate OIS for penetrating colon injuries as well as imaging criteria. To validate the new OIS and its correlation with outcomes, we performed a retrospective review of patients with penetrating colon injuries (AIS<3 in other body regions) between 2016 and 2020 at a single institution. Sixty-six patients met inclusion criteria. Most were young (29 years median) and male (90%). All underwent operative intervention and 23 (34%) had pre-operative imaging. Imaging grade was higher than operative grade in 11 patients (48%). Higher AAST operative grade was associated with a higher likelihood of resection and anastomosis or colostomy, need for damage control laparotomy, and development of intra-abdominal abscess and acute kidney injury. A multicenter study is underway to confirm these findings.

Near Disappearance of Splenorrhaphy as an Operative Strategy for Splenic Preservation After Trauma.

BACKGROUND: Splenorrhaphy was once used to achieve splenic preservation in up to 40% of splenic injuries. With increasing use of nonoperative management and angioembolization, operative therapy is less common and splenic injuries treated operatively are usually high grade. Patients are often unstable, making splenic salvage unwise. Modern surgeons may no longer possess the knowledge to perform splenorrhaphy. METHODS: The records of adult trauma patients with splenic injuries from September 2014 to November 2018 at an urban level I trauma center were reviewed retrospectively. Data including American Association for the Surgery of Trauma splenic organ injury scale, type of intervention, splenorrhaphy technique, and need for delayed splenectomy were collected. This contemporary cohort (CC) was compared to a historical cohort (HC) of splenic injuries at a single center from 1980 to 1989 (Ann Surg 1990; 211: 369). RESULTS: From 2014 to 2018, 717 adult patients had splenic injuries. Initial management included 157 (21.9%) emergent splenectomy, 158 (22.0%) angiogram ± embolization, 371 (51.7%) observation, and only 10 (1.4%) splenorrhaphy. The HC included a total of 553 splenic injuries, of which 313 (56.6%) underwent splenectomy, while splenorrhaphy was performed in 240 (43.4%). Those who underwent splenorrhaphy in each cohort (CC vs HC) were compared. CONCLUSION: The success rate of splenorrhaphy has not changed. However, splenorrhaphy now involves only electrocautery with topical hemostatic agents and is used primarily in low-grade injuries. Suture repair and partial splenectomy seem to be "lost arts" in modern trauma care.

Outcomes after ultramassive transfusion in the modern era: An Eastern Association for the Surgery of Trauma multicenter study.

BACKGROUND: Despite the widespread institution of modern massive transfusion protocols with balanced blood product ratios, survival for patients with traumatic hemorrhage receiving ultramassive transfusion (UMT) (defined as ≥20 U of packed red blood cells [RBCs]) in 24 hours) remains low and resource consumption remains high. Therefore, we aimed to identify factors associated with mortality in trauma patients receiving UMT in the modern resuscitation era. METHODS: An Eastern Association for the Surgery of Trauma multicenter retrospective study of 461 trauma patients from 17 trauma centers who received ≥20 U of RBCs in 24 hours was performed (2014-2019). Multivariable logistic regression and Classification and Regression Tree analysis were used to identify clinical characteristics associated with mortality. RESULTS: The 461 patients were young (median age, 35 years), male (82%), severely injured (median Injury Severity Score, 33), in shock (median shock index, 1.2; base excess, -9), and transfused a median of 29 U of RBCs, 22 U of fresh frozen plasma (FFP), and 24 U of platelets (PLT). Mortality was 46% at 24 hours and 65% at discharge. Transfusion of RBC/FFP ≥1.5:1 or RBC/PLT ≥1.5:1 was significantly associated with mortality, most pronounced for the 18% of patients who received both RBC/PLT and RBC/FFP ≥1.5:1 (odds ratios, 3.11 and 2.81 for mortality at 24 hours and discharge; both p < 0.01). Classification and Regression Tree identified that age older than 50 years, low initial Glasgow Coma Scale, thrombocytopenia, and resuscitative thoracotomy were associated with low likelihood of survival (14-26%), while absence of these factors was associated with the highest survival (71%). CONCLUSION: Despite modern massive transfusion protocols, one half of trauma patients receiving UMT are transfused with either RBC/FFP or RBC/PLT in unbalanced ratios ≥1.5:1, with increased associated mortality. Maintaining focus on balanced ratios during UMT is critical, and consideration of advanced age, poor initial mental status, thrombocytopenia, and resuscitative thoracotomy can aid in prognostication. LEVEL OF EVIDENCE: Prognostic, level III.