Abstract
Background:
Prior research suggests that plasma may improve outcomes in traumatic brain injury (TBI). We examined the association between plasma administration and mortality in moderate-severe TBI in a large retrospective cohort, hypothesizing plasma is associated with decreased mortality after accounting for confounding covariates.
Materials and Methods:
Patients from the 2017–2020 Trauma Quality Improvement Program (TQIP) dataset ≥18 years with moderate-severe TBI were included. We excluded patients with comorbidities associated with bleeding or sensitivity to volume (antiplatelet or anticoagulation medications, bleeding disorders, cirrhosis, congestive heart failure, chronic obstructive pulmonary disease). Multivariable logistic regression examined the association between plasma volume transfused in the first four hours and hospital mortality, adjusting for sociodemographics, severity of injury/illness, neurologic status, and volume of other blood products. We also adjusted for and included interactions with hemorrhage markers (shock; need for hemorrhage control).
Results:
Of 63,918 patients included, hospital mortality was 37.0 %. 82.8 % received no plasma. Each unit of plasma was associated with greater unadjusted mortality, with odds ratio (OR): 1.13 (95 % confidence interval: 1.12–1.14), but after confounder adjustment, plasma units were not associated with greater mortality, with OR: 1.01 (0.99–1.03). While the overall adjusted effect of plasma was not significant, significant interactions between hemorrhage markers and plasma were present (p < 0.001).
Conclusions:
Administration of plasma within the first four hours after hospital presentation was not associated with decreased or increased mortality in adult patients with moderate to severe TBI after confounder adjustment. Interaction analysis suggests the presence of hemorrhage improves the effect of plasma on mortality in TBI. This important clinical question should be answered with a prospective randomized study of plasma for nonbleeding patients with TBI.
Keywords: Traumatic brain injury, Plasma, Mortality, Resuscitation, Trauma quality improvement program
Introduction
Traumatic injury is a leading cause of death[1] and disability[2] with traumatic brain injury (TBI) accounting for a significant portion of that death and disability [3]. While processes and systems of TBI care have advanced in the past decades [4], specific therapies for TBI remain limited.
Over the past decade, multiple high-quality trials have demonstrated the importance of plasma in balanced resuscitation after major traumatic hemorrhage [5,6], with balanced resuscitation including plasma becoming the standard of care [7]. There has also been emerging evidence suggesting improved outcomes, both survival and neurologic, after administration of plasma in TBI, from animal models [8,9], retrospective cohorts [10–12], and a secondary analysis of a randomized controlled trial (RCT) [13]. While the prevention of secondary brain injury from hypotension is thought to be a main benefit of plasma [8,12, 13], there is also evidence for an independent effect outside of treatment of hemorrhagic shock, which is likely due to better brain vasodilation with subsequent improved cerebral perfusion pressure and oxygenation in addition to improved repair of the blood-brain barrier endothelium [9–11].
Given increased plasma administration and balanced massive transfusion protocols in recent clinical practice [14,15], we aimed to examine the association between plasma administration and mortality in patients with moderate and severe TBI in a large retrospective cohort from 2017 to 2020 based on a quality improvement database. We hypothesized that plasma would be associated with decreased mortality after controlling for variables that confound the relationship between plasma administration and mortality, especially hemorrhage which may confound this association by indication.
Patients and methods
Study design and population
This is a retrospective cohort study using data from the Trauma Quality Improvement Program (TQIP), a national quality monitoring and improvement program for the care of the injured patient established by the American College of Surgeons Committee on Trauma [16]. Criteria for inclusion into TQIP are established by the National Trauma Data Standard (NTDS) [17], but briefly, it includes patients presenting to participating centers with an injury from a traumatic mechanism. This study was considered exempt from review by the Vanderbilt Institutional Review Board given its analysis of already deidentified data collected for quality improvement.
This study includes patients recorded in TQIP presenting to participating trauma centers from 2017 to 2020, age 18 and above, with moderate to severe TBI, defined as having a Glasgow Coma Scale (GCS) [18] between 3 and 12 on the calendar day after hospital admission and an Abbreviated Injury Scale-Head (AIS-Head)[19] greater than 3. Patients with known use of antiplatelet or anticoagulation medications, bleeding disorders, cirrhosis, congestive heart failure, chronic obstructive pulmonary disease, and chronic renal failure were excluded as the benefits and harms of plasma are known to be different for patients with these comorbidities as opposed to the general population. Patients transferred from another hospital were excluded as plasma administration at a referring facility is not known or recorded in this database.
Exposures, outcomes, and covariates
Our primary exposure was the volume of plasma transfused within the first four hours after presentation to the emergency department, measured in blood transfusion units. Our primary outcome was hospital mortality, defined as an emergency department or hospital disposition of “deceased/expired.”
Our covariates are shown in Table 1. They include patient sociodemographics: age, biologic sex (currently what is reported in TQIP), race/ethnicity, and insurance/payment status; neurologic injury severity: severity of TBI by GCS on calendar day after hospital admission and Abbreviated Injury Scale (AIS) Head, pupillary response at presentation, and midline shift at presentation; and other measures of injury severity: total volume of other blood products (packed red blood cells, platelets, cryoprecipitate, and whole blood) transfused within the first four hours after presentation (units), modified Injury Severity Score (ISS)[20] excluding AIS-Head (to avoid collinearity as AIS-Head is already in the model), hypoxia at presentation (oxygen saturation < 92 %), and total ventilatory days. Briefly, this modified ISS was calculated by taking the three highest AIS scores across every body region except Head and then squaring their sums [21,22]. Blood products other than plasma were summed as the main indications for receiving blood products in the first four hours after trauma were as part of a balanced resuscitation in the event of severe hemorrhagic shock.
Table 1.
TQIP cohort characteristics of patients with moderate-severe TBI by plasma received.
| Plasma transfused in first 4 h (units) | 0 (N = 52,935) | 0–2 (N = 4391) | 2–6 (N = 4070) | 6–12 (N = 1607) | >12 (N = 915) | Overall (N = 63,918) |
|---|---|---|---|---|---|---|
|
| ||||||
| Age | 43 (28–61) | 38 (27–56) | 37 (25–54) | 35 (25–54) | 36 (25–50) | 42 (28–60) |
| Male sex, N (%) | 12,782 (75.8) | 3287 (74.9) | 3050 (74.9) | 1179 (73.4) | 679 (74.2) | 48,321 (75.6) |
| Race/ethnicity, N (%) | ||||||
| American Indian | 365 (0.7 %) | 41 (0.9 %) | 21 (0.5 %) | 6 (0.4 %) | 4 (0.4 %) | 437 (0.7 %) |
| Asian | 1547 (2.9 %) | 107 (2.4 %) | 125 (3.1 %) | 39 (2.4 %) | 21 (2.3 %) | 1839 (2.9 %) |
| Black | 8555 (16.2 %) | 794 (18.1 %) | 782 (19.2 %) | 315 (19.6 %) | 187 (20.4 %) | 10,633 (16.6 %) |
| Hispanic | 7760 (14.7 %) | 608 (13.8 %) | 616 (15.1 %) | 246 (15.3 %) | 152 (16.6 %) | 9382 (14.7 %) |
| Other/Unknown | 3542 (6.7 %) | 301 (6.9 %) | 301 (7.4 %) | 115 (7.2 %) | 81 (8.9 %) | 4340 (6.8 %) |
| Pacific Islander | 132 (0.2 %) | 6 (0.1 %) | 8 (0.2 %) | 3 (0.2 %) | 2 (0.2 %) | 151 (0.2 %) |
| White | 31,034 (58.6 %) | 2534 (57.7 %) | 2217 (54.5 %) | 883 (54.9 %) | 468 (51.1 %) | 37,136 (58.1 %) |
| Primary method of payment, N (%) | ||||||
| Medicaid | 11,003 (20.8 %) | 913 (20.8 %) | 840 (20.6 %) | 315 (19.6 %) | 191 (20.9 %) | 13,262 (20.7 %) |
| Self-Pay | 7957 (15.0 %) | 783 (17.8 %) | 770 (18.9 %) | 342 (21.3 %) | 221 (24.2 %) | 10,073 (15.8 %) |
| Private/Commercial | 19,938 (37.7 %) | 1844 (42.0 %) | 1720 (42.3 %) | 667 (41.5 %) | 364 (39.8 %) | 24,533 (38.4 %) |
| Medicare | 9187 (17.4 %) | 456 (10.4 %) | 377 (9.3 %) | 146 (9.1 %) | 58 (6.3 %) | 10,224 (16.0 %) |
| Other | 3487 (6.6 %) | 300 (6.8 %) | 254 (6.2 %) | 111 (6.9 %) | 62 (6.8 %) | 4214 (6.6 %) |
| ISS | 26 (20–33) | 34 (26–43) | 34 (27–43) | 41 (29.5–50) | 43 (34–54) | 26 (21–34) |
| ISS excluding AIS Head | 8 (1–17) | 17 (9–29) | 22 (13–33) | 27 (17–38) | 34 (22–43) | 9 (2–18) |
| Hypoxia, N (%) | 6031 (11.4 %) | 929 (21.2 %) | 947 (23.3 %) | 421 (26.2 %) | 225 (24.6 %) | 8553 (13.4 %) |
| Total ventilatory days * | 4 (2–11) | 5 (2–12) | 5 (2–12) | 4 (1.5–11) | 2 (1–9) | 5 (2–12) |
| Severe TBI by GCS, N (%) | 35,655 (67.4) | 3469 (79.0) | 3187 (78.3) | 1310 (81.5) | 751 (82.1) | 44,372 (69.4) |
| Severe TBI by AIS Head, N (%) | 23,543 (44.5) | 2143 (48.8) | 1950 (47.9) | 738 (45.9) | 370 (40.4) | 28,744 (45.0) |
| Pupillary response at presentation | ||||||
| Both reactive | 32,689 (61.8 %) | 2057 (46.8 %) | 1839 (45.2 %) | 666 (41.4 %) | 383 (41.9 %) | 37,634 (58.9 %) |
| One reactive | 3599 (6.8 %) | 370 (8.4 %) | 299 (7.3 %) | 111 (6.9 %) | 54 (5.9 %) | 4433 (6.9 %) |
| Neither reactive | 14,870 (28.1 %) | 1808 (41.2 %) | 1766 (43.4 %) | 735 (45.7 %) | 417 (45.6 %) | 19,596 (30.7 %) |
| Midline shift on admission CT, N (%) | 16,120 (30.5 %) | 1284 (29.2 %) | 1019 (25.0 %) | 385 (24.0 %) | 164 (17.9 %) | 18,972 (29.7 %) |
| Shock, N (%) | 2220 (4.2) | 762 (17.4) | 997 (24.5) | 473 (29.4) | 300 (32.8) | 4752 (7.4) |
| Need for hemorrhage control procedure, N (%) | 2386 (4.5) | 1522 (34.7) | 1964 (48.3) | 1125 (70) | 811 (88.6) | 7808 (12.2) |
| Hospital mortality, N (%) | 17,521 (33.1) | 2257 (51.4) | 2190 (53.8) | 1008 (62.7) | 674 (73.7) | 23,650 (37.0) |
| Total units other blood products transfused in 4 h, N (%) | ||||||
| 0 | 47,282 (89.3 %) | 62 (1.4 %) | 29 (0.7 %) | 0 (0 %) | 0 (0 %) | 47,373 (74.1 %) |
| ≤2 | 3961 (7.5 %) | 2238 (51.0 %) | 289 (7.1 %) | 7 (0.4 %) | 1 (0.1 %) | 6496 (10.2 %) |
| 2–7 | 1554 (2.9 %) | 1807 (41.2 %) | 2425 (59.6 %) | 165 (10.3 %) | 10 (1.1 %) | 5961 (9.3 %) |
| 7–14 | 115 (0.2 %) | 258 (5.9 %) | 1157 (28.4 %) | 865 (53.8 %) | 63 (6.9 %) | 2458 (3.8 %) |
| >14 | 23 (0.0 %) | 26 (0.6 %) | 170 (4.2 %) | 570 (35.5 %) | 841 (91.9 %) | 1630 (2.6 %) |
AIS: Abbreviated Injury Scale; CT: computerized tomography; GCS: Glasgow Coma Scale; SD: standard deviation; TBI: traumatic brain injury; TQIP: Trauma Quality Improvement Program.
All values median (interquartile range) unless otherwise specified.
58,600 of 63,918 patients on mechanical ventilator.
Additionally, two covariates were markers of hemorrhage: shock at presentation (defined as systolic blood pressure (SBP) <90 mmHg)[23] and need for a hemorrhage control procedure. Hemorrhage control procedures are defined by the NTDS according to the data fields “Surgery for Hemorrhage Control,” which includes laparotomy, thoracotomy, sternotomy, extremity exploration or amputation, neck exploration, soft tissue exploration, or extraperitoneal pelvic packing; and “Angiography,” which includes angiogram only, angiogram with embolization, or angiogram with stenting [17]. These covariates were also specifically used to analyze for interactions between plasma administration and hemorrhage in the effect on mortality (explained in more detail below). Laboratory hemoglobin and hematocrit values are not captured in TQIP and not used here as a marker of hemorrhage, as the utility of these laboratory markers is thought to be limited in acute hemorrhage prior to plasma hemoglobin equilibration.
These covariates were informed by our a priori hypothesis that hemorrhage may confound by indication, by prior evidence of the association of pre-injury socioeconomic status on outcome [24], and by an analysis of optimal risk adjustment in TQIP cohorts using the mortality endpoint, which included total ventilatory days for optimal adjustment [25]. Collinearity was assessed with variance inflation factor analysis comparing models with and without variables within the same category as listed above (pre-injury sociodemographics, neurologic injury severity, and other markers of injury severity). No signficant collinearity was found.
Statistical analysis
A locally estimated scatterplot smoothing (LOESS) function was used to examine the relationship between volume of plasma transfused and hospital mortality to categorization of volume of plasma transfused and was found to be linear.
We used simple and multiple logistic regression to determine the association between volume of plasma transfused and hospital mortality, adjusting for all covariates specified above in the multiple regression (including the two markers of hemorrhage). This was followed by multiple logistic regression that included interaction terms between volume of plasma transfused and the two markers of hemorrhage: shock at presentation and need for hemorrhage control procedure. To determine if these interactions were significant, a likelihood ratio test was used to compare the model with and without interaction terms.
Because of the potential for confounding by indication of plasma for hemorrhage, we performed a sensitivity analysis excluding patients with shock or need for hemorrhage control procedure. Only complete cases were considered for analysis. Collinearity between covariates was assessed with groups of covariates suspected to be colinear with a variance inflation factor diagnostic. All data cleaning and statistical analysis was performed in R version 4.2.1 [26].
Results
Cohort characteristics
From 2017 to 2020, there were 4273,914 patients recorded in TQIP, representing over 850 hospitals. Of those, 63,918 met inclusion and exclusion criteria, as shown in Fig. 1. Characteristics of this cohort are shown in Table 1. In terms of sociodemographics, the cohort was 75.6 % male, median age of 43 years old, 58.1 % white, and 38.4 % privately insured. With regards to neurologic severity of injury, 69.4 % of the cohort had severe TBI by GCS criteria (post-resuscitation GCS 3–8 on the calendar day after hospital admission) but only 45.0 % were severe by AIS-Head criteria (AIS-Head 5–6). 37.6 % had at least one unreactive pupil and 29.7 % had midline shift.
Fig. 1.
Study population, inclusion, exclusion, and final cohort.
For markers of hemorrhage, 10.8 % presented in shock (SBP <90 mmHg), 12.2 % underwent a procedure for hemorrhage control, and 25.9 % were administered blood products other than plasma within the first 4 h of admission. Of patients in shock, 23.1 % received plasma transfusion, and of patients undergoing a procedure for hemorrhage control, 49.4 % received plasma transfusion. Finally, the median ISS was 26 but median ISS excluding Head AIS component was 9. For respiratory function, 13.4 % were hypoxic (oxygen saturation < 92 % at presentation) and included patients had a median 4 days of intubation.
Primary analysis: adjusted effect of plasma on mortality
82.8 % of the cohort received no plasma and overall hospital mortality was 37.0 %, as shown in Table 2. In the unadjusted analysis, the administration of each unit of plasma was associated with a significantly higher odds of hospital mortality, with an odds ratio (OR) of 1.13 (95 % confidence interval (CI): 1.12–1.14). As shown in Fig. 2, administration of each unit of plasma is no longer associated with increased mortality after adjustment for covariates, with OR 1.01 (95 %CI: 0.99–1.03). All covariates were significantly associated with hospital mortality except for race/ethnicity.
Table 2.
TQIP cohort hospital mortality by plasma received.
| Plasma transfused in first 4 h (units) | 0 (N = 52,935) | 0–2 (N = 4391) | 2–6 (N = 4070) | 6–12 (N = 1607) | >12 (N = 915) | Overall (N = 63,918) |
|---|---|---|---|---|---|---|
|
| ||||||
| Hospital mortality, N (%) | 17,521 (33.1) | 2257 (51.4) | 2190 (53.8) | 1008 (62.7) | 674 (73.7) | 23,650 (37.0) |
Fig. 2.
Cohort unadjusted and adjusted odds ratios of hospital mortality with each additional unit of plasma administered.
Sensitivity analysis excluding hemorrhage and effect modification by hemorrhage
Sensitivity analyses, which excluded patients with shock and undergoing hemorrhage control procedures, demonstrated similar results to the above analysis and are shown Supplementary Figure 1.
We introduced interaction terms between plasma administration in the first four hours and markers of hemorrhage, shock and need for hemorrhage control procedure. While the overall adjusted effect of plasma on mortality in TBI was not significant, this interaction analysis demonstrated that these hemorrhage markers significantly (p < 0.001) modified the relationship between plasma and mortality. Specifically, the interaction effect sizes (which can be thought of as “ratios of odds ratios”) are significantly <1 (shock: 0.92 (95 %CI: 0.90–0.94), need for hemorrhage control procedure: 0.91 (95 %CI: 0.88–0.93)). This suggests the presence of hemorrhage increases the need for plasma transfusion, indicating a population (those bleeding) in which plasma improves mortality in TBI.
Discussion
In this national retrospective cohort study from 2017 to 2020, administration of plasma within the first four hours after hospital presentation was not associated with decreased or increased mortality in adult patients with moderate to severe TBI after adjustment for confounders. While there is an association between plasma and mortality in unadjusted analysis, this is likely due to confounding by indication, specifically that of plasma administration being indicated for major hemorrhage. In turn, hemorrhage causes mortality. Of all the adjusted covariates, the total volume of other blood products infused and the two markers of hemorrhage – shock and need for hemorrhage control procedure – are most likely the drivers of this effect being present in unadjusted analysis but not present in adjusted analyses. This was a finding replicated in sensitivity analyses excluding patients in shock or needing a hemorrhage control. Notably, our hypothesis was disproved, as we did not find an association between plasma administration and decreased mortality among patients with moderate to severe TBI.
Prior retrospective cohort studies have shown a benefit of plasma in TBI with regard both to survival[10,11] and neurologic outcomes [12]. Two of those studies compared different transfusion strategies [10,12], which meant comparison groups were both experiencing hemorrhage. Animal studies[8,9] and a secondary analysis of a RCT[13] similarly compared homogenous groups experiencing both hemorrhage and TBI. Only one study, performed at a single center in patients with isolated TBI, compared patients with and without major hemorrhage. [11] In contrast, this study includes both isolated TBI and polytrauma patients, ranging from patients with no indicators of hemorrhagic shock all the way to patients undergoing massive transfusion.
Shock and the need for a hemorrhage control procedure both modified the effect of plasma administration on mortality, as indicated by the significant interaction term. Specifically, these interaction effects were <1, suggesting that while there is no overall association between plasma and mortality, the presence of hemorrhage improves the effect of plasma on TBI. Put more simply, plasma administration was associated with improved survival among bleeding polytrauma patients with TBI. This may suggest that the benefits of plasma administration in this mixed population of moderate to severe TBI that includes both isolated head injury and polytrauma, may be due to its role in hemorrhage resuscitation after hemorrhage[5,6,13] rather than an intrinsic effect on the brain through better local brain oxygenation and repair of the blood-brain endothelium [9–11]. Teasing out this difference remains extremely difficult in retrospective cohorts given that blood products like plasma are limited resources and thus are reserved primarily, at least in the trauma population, for bleeding patients. It should be noted that this analysis is difficult to interpret for extremes of blood transfusion: the patients receiving greater than 12 units of plasma represent extreme outliers for which there is also a significant survival bias.
Strengths of this study include its size and generalizability but moreover the diversity of practice captured. TQIP data during this time period represents over 4 million patients from 850 centers, of which over 60,000 met inclusion criteria. With this large sample size, we have a robust ability to adjust for many known and recorded confounders. However, weaknesses include the retrospective nature of this cohort and the purpose of the data source. TQIP includes a wide variety of centers, from small community programs that are not ACS verified to large ACS verified Level I centers, and so the data is not captured with the granularity nor verification of a research dataset. The most accurately recorded data points for these patients with TBI are related to neurotrauma processes of care (e.g., midline shift and intracranial monitor placement) rather than hemorrhage processes of care (e.g. blood transfusion, where further limitations remain as only the first four hours of transfusion are collected). Additionally, we are unable to control for unmeasured confounders in an administrative quality improvement dataset that is not constructed with research questions in mind. Nevertheless, given its enormous size compared to most research datasets, it may be able to provide clinically meaningful analyses. A significant limitation is the inability to delineate the effects of different hospitals, including transfer between them, to adhere institutional and patient data privacy standards. Furthermore, given our large sample size, there are potential for detecting statistically significant but not clinically relevant results.
While this study aimed to leverage its large sample size to adjust for numerous variables, at present, few clinicians give plasma for patients not suspected to be hemorrhaging. It is thus unlikely that sufficient real-world variation exists in plasma administration practices for large retrospective cohort studies to assess the utility of plasma specifically for TBI, especially for primary neurologic protection (rather than prevention of secondary injury from hypotension). Furthermore, hemorrhage is a clinical diagnosis, and while shock and need for hemorrhage control procedure may be specific for patients with hemorrhage, they are not sensitive, a further limitation of this study. We did not assess whether plasma administration was associated with mortality regardless of volume transfused. A prospective RCT, the upcoming multi-center FFP In Traumatic BRAin INjury (FIT-BRAIN) Trial (NCT06062888) [27], comparing plasma to usual care for patients with TBI without major hemorrhage may better answer the important clinical question of whether plasma improves both survival and neurologic outcomes after TBI.
Conclusions
Administration of plasma within the first four hours after hospital presentation was not associated with decreased or increased adjusted odds of mortality in adult patients with moderate to severe TBI in this large, retrospective quality improvement cohort. Interaction analysis demonstrated effect modification, suggesting plasma administration was more associated with improved survival among bleeding polytrauma patients with TBI than nonbleeding patients with TBI. Our analyses highlight the difficulty of examining this question with these types of data and demonstrates the need for a high-quality randomized controlled trial to provide better answers.
Supplementary Material
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.injury.2024.112040.
Acknowledgements and disclosure
We would like to thank all course faculty and teaching assistants for Biostatistics 140.624: Statistical Methods in Public Health IV, a course at the Johns Hopkins Bloomberg School of Public Health. with special acknowledgement to Leah Jager, PhD. Additionally, we would like to thank Frank Harrell, PhD for statistical consultation. This research was presented at the 19th Annual Academic Surgical Congress held February 2024 in Washington, D.C.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. The authors are supported by NIH T32GM135094 (SR, ELR, MBP), K23GM150110 (AWM), R01GM120484, R01AG058639, and I01RX002992 (MBP). The authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article.
Footnotes
Declaration of competing interest
None.
Ethics statement
This study was considered exempt from review by the Vanderbilt Institutional Review Board given its analysis of already deidentified data collected for quality improvement.
CRediT authorship contribution statement
Shayan Rakhit: Writing – review & editing, Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Conceptualization. Areg Grigorian: Writing – review & editing, Data curation. Erika L Rivera: Writing – review & editing, Conceptualization. Francisco A Alvarado: Writing – review & editing, Conceptualization. Mayur B Patel: Writing – review & editing, Supervision, Funding acquisition, Conceptualization. Amelia W Maiga: Writing – review & editing, Supervision, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.
References
- [1].Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2018;392(10159):1736–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 2020;396(10258):1204–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Callcut RA, Kornblith LZ, Conroy AS, et al. The why and how our trauma patients die: a prospective Multicenter Western Trauma Association study. J Trauma Acute Care Surg 2019;86(5):864–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Rakhit S, Nordness MF, Lombardo SR, Cook M, Smith L, Patel MB. Management and Challenges of Severe Traumatic Brain Injury. Semin Respir Crit Care Med 2021;42(1):127–44. [DOI] [PubMed] [Google Scholar]
- [5].Holcomb JB, Tilley BC, Baraniuk S, et al. Transfusion of plasma, platelets, and red blood cells in a 1:1:1 vs a 1:1:2 ratio and mortality in patients with severe trauma: the PROPPR randomized clinical trial. Jama 2015;313(5):471–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Sperry JL, Guyette FX, Brown JB, et al. Prehospital plasma during air medical transport in trauma patients at risk for hemorrhagic shock. N Engl J Med 2018;379(4):315–26. [DOI] [PubMed] [Google Scholar]
- [7].Committee on Trauma. Advanced trauma life support: student course manual. 10th ed. Chicago, IL: American College of Surgeons; 2018. [Google Scholar]
- [8].Halaweish I, Bambakidis T, He W, et al. Early resuscitation with fresh frozen plasma for traumatic brain injury combined with hemorrhagic shock improves neurologic recovery. J Am Coll Surg 2015;220(5):809–19. [DOI] [PubMed] [Google Scholar]
- [9].Imam A, Jin G, Sillesen M, et al. Fresh frozen plasma resuscitation provides neuroprotection compared to normal saline in a large animal model of traumatic brain injury and polytrauma. J Neurotrauma 2015;32(5):307–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Jokar TO, Khalil M, Rhee P, et al. Ratio-based Resuscitation in Trauma Patients with Traumatic Brain Injury: is There a Similar Effect? Am Surg 2016;82(3):271–7. [PubMed] [Google Scholar]
- [11].Chang R, Folkerson LE, Sloan D, et al. Early plasma transfusion is associated with improved survival after isolated traumatic brain injury in patients with multifocal intracranial hemorrhage. Surgery 2017;161(2):538–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Hernandez MC, Thiels CA, Aho JM, et al. Prehospital plasma resuscitation associated with improved neurologic outcomes after traumatic brain injury. J Trauma Acute Care Surg 2017;83(3):398–405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Gruen DS, Guyette FX, Brown JB, et al. Association of Prehospital Plasma With Survival in Patients With Traumatic Brain Injury: a Secondary Analysis of the PAMPer Cluster Randomized Clinical Trial. JAMA Netw Open 2020;3(10):e2016869. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Kutcher ME, Kornblith LZ, Narayan R, et al. A paradigm shift in trauma resuscitation: evaluation of evolving massive transfusion practices. JAMA Surg 2013;148(9):834–40. [DOI] [PubMed] [Google Scholar]
- [15].Ohbe H, Tagami T, Endo A, et al. Trends in massive transfusion practice for trauma in Japan from 2011 to 2020: a nationwide inpatient database study. J Intensive Care 2023;11(1):46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Shafi S, Nathens AB, Cryer HG, et al. The Trauma Quality Improvement Program of the American College of Surgeons Committee on Trauma. J Am Coll Surg 2009;209(4):521–30. e1. [DOI] [PubMed] [Google Scholar]
- [17].Hashmi ZG, Kaji AH, Nathens AB. Practical Guide to Surgical Data Sets: national Trauma Data Bank (NTDB). JAMA Surg 2018;153(9):852–3. [DOI] [PubMed] [Google Scholar]
- [18].Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale. Lancet 1974;2(7872):81–4. [DOI] [PubMed] [Google Scholar]
- [19].Rating the severity of tissue damage. I. The abbreviated scale. Jama 1971;215(2):277–80. [DOI] [PubMed] [Google Scholar]
- [20].Baker SP, O’Neill B, Haddon W Jr, Long WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974;14(3):187–96. [PubMed] [Google Scholar]
- [21].Lingsma H, Andriessen TM, Haitsema I, et al. Prognosis in moderate and severe traumatic brain injury: external validation of the IMPACT models and the role of extracranial injuries. J Trauma Acute Care Surg 2013;74(2):639–46. [DOI] [PubMed] [Google Scholar]
- [22].Kesinger MR, Juengst SB, Bertisch H, et al. Acute Trauma Factor Associations With Suicidality Across the First 5 Years After Traumatic Brain Injury. Arch Phys Med Rehabil 2016;97(8):1301–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Blackmore AR, Leonard J, Madayag R, Bourg PW. Using the Trauma Quality Improvement Program Metrics Data to Enhance Clinical Practice. J Trauma Nurs 2019;26(3):121–7. [DOI] [PubMed] [Google Scholar]
- [24].Haider AH, Weygandt PL, Bentley JM, et al. Disparities in trauma care and outcomes in the United States: a systematic review and meta-analysis. J Trauma Acute Care Surg 2013;74(5):1195–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Haider AH, Hashmi ZG, Zafar SN, et al. Developing best practices to study trauma outcomes in large databases: an evidence-based approach to determine the best mortality risk adjustment model. J Trauma Acute Care Surg 2014;76(4):1061–9. [DOI] [PubMed] [Google Scholar]
- [26].R Core Team. R: a language and environment for statistical computin. Vienna, Austria: R Foundation for Statistical Computing; 2023. [Google Scholar]
- [27].Alam HB FFP In Traumatic BRAin INjury (FIT-BRAIN) Trial. 2024. https://classic.clinicaltrials.gov/ct2/show/NCT060628882024). [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.