This cohort study assesses the association between whole blood as an adjunct to component therapy–based massive transfusion protocol and survival among patients presenting with severe hemorrhage at trauma centers.
Key Points
Question
Is whole blood as an adjunct to component therapy–based massive transfusion protocol (MTP) compared with MTP alone associated with improved survival among adult trauma patients presenting with severe hemorrhage?
Findings
In this cohort study of 2785 patients who presented with severe traumatic hemorrhage, whole blood as an adjunct to MTP compared with MTP alone was associated with lower mortality at 24 hours and 30 days, with a survival benefit found as early as 5 hours after emergency department arrival.
Meaning
The findings suggest that whole-blood resuscitation as an adjunct to component–based MTP is associated with improved survival among adult patients presenting to trauma centers with severe hemorrhage, with a benefit found early after administration.
Abstract
Importance
Whole-blood (WB) resuscitation has gained renewed interest among civilian trauma centers. However, there remains insufficient evidence that WB as an adjunct to component therapy–based massive transfusion protocol (WB-MTP) is associated with a survival advantage over MTP alone in adult civilian trauma patients presenting with severe hemorrhage.
Objective
To assess whether WB-MTP compared with MTP alone is associated with improved survival at 24 hours and 30 days among adult trauma patients presenting with severe hemorrhage.
Design, Setting, and Participants
This retrospective cohort study using the American College of Surgeons Trauma Quality Improvement Program databank from January 1, 2017, and December 31, 2018, included adult trauma patients with a systolic blood pressure less than 90 mm Hg and a shock index greater than 1 who received at least 4 units of red blood cells within the first hour of emergency department (ED) arrival at level I and level II US and Canadian adult civilian trauma centers. Patients with burns, death within 1 hour of ED arrival, and interfacility transfers were excluded. Data were analyzed from February 2022 to September 2022.
Exposures
Resuscitation with WB-MTP compared with MTP alone within 24 hours of ED presentation.
Main Outcomes and Measures
Primary outcomes were survival at 24 hours and 30 days. Secondary outcomes selected a priori included major complications, hospital length of stay, and intensive care unit length of stay.
Results
A total of 2785 patients met inclusion criteria: 432 (15.5%) in the WB-MTP group (335 male [78%]; median age, 38 years [IQR, 27-57 years]) and 2353 (84.5%) in the MTP-only group (1822 male [77%]; median age, 38 years [IQR, 27-56 years]). Both groups included severely injured patients (median injury severity score, 28 [IQR, 17-34]; median difference, 1.29 [95% CI, −0.05 to 2.64]). A survival curve demonstrated separation within 5 hours of ED presentation. WB-MTP was associated with improved survival at 24 hours, demonstrating a 37% lower risk of mortality (hazard ratio, 0.63; 95% CI, 0.41-0.96; P = .03). Similarly, the survival benefit associated with WB-MTP remained consistent at 30 days (HR, 0.53; 95% CI, 0.31-0.93; P = .02).
Conclusions and Relevance
In this cohort study, receipt of WB-MTP was associated with improved survival in trauma patients presenting with severe hemorrhage, with a survival benefit found early after transfusion. The findings from this study are clinically important as this is an essential first step in prioritizing the selection of WB-MTP for trauma patients presenting with severe hemorrhage.
Introduction
Hemorrhage remains the leading cause of preventable in-hospital deaths after injury in US adult civilian trauma centers, accounting for 40% of deaths within the first 24 hours.1 Trauma-induced coagulopathy (TIC) plays a substantial role in bleeding-related deaths, with an associated mortality of up to 50%.2
Current trauma-resuscitative strategies underscore the importance of a balanced transfusion approach of separated blood components to mitigate the adverse effects of TIC on patient outcomes. Massive transfusion protocol (MTP), an evidence-based, high-ratio component transfusion (CT) strategy, attempts to mimic the original composition of whole blood (WB) and is associated with improved hemostasis and reduced death from bleeding in patients with severe hemorrhage.3 Logically, transfusion of WB would confer the same benefits.
The use of WB for treating hemorrhage after injury is a century-old practice with its original foundation in military medicine since World War I.4 However, shortly after the Vietnam War, civilian blood banks gradually shifted away from WB inventories, gravitating toward the routine use of massive crystalloid replacement and fractionated blood components for hemorrhagic shock.5 However, over the years, much has been learned about the detriments of crystalloid replacement and the limitations of CT in patients with hemorrhage.
The modern military experience has been credited with the resurgence of WB use, primarily predicated on addressing the logistical challenges of providing a limited or, at times, unavailable inventory of stored components for MTP in the time and space of active military conflict.6,7 Retrospective studies have demonstrated the practicality of the military application of WB with the added benefits of an enhanced hemostatic profile and improved survivability.8 As a result, WB has gained traction in the civilian setting as a rediscovered treatment for injured patients with severe hemorrhage.9
The adoption of WB as the initial transfusion product among US civilian trauma centers has gradually increased, replacing conventional CT as the initial transfusion product among 40 to 70 adult civilian trauma centers across the US.10,11 This trend currently lacks robust and consistent evidence on patient outcomes. It is unknown whether WB improves survival in patients with severe hemorrhage. Given this uncertainty, the American College of Surgeons Trauma Quality Improvement Program (ACS-TQIP) management guidelines for massive transfusion in trauma currently do not include WB as part of the MTP.12
The evidence is lacking on whether WB is associated with reduced mortality in patients presenting with severe hemorrhage and on the timing in which patients receive the most substantial survival benefit. Therefore, the objective of this study was to analyze survival associated with WB as an adjunct to MTP (WB-MTP) compared with MTP alone in patients presenting with severe hemorrhage in US and Canadian adult civilian trauma centers over a 2-year period. We hypothesized, a priori, that WB-MTP would be associated with improved survival at 24 hours and 30 days without an increase in major complications.
Methods
Study Design, Setting, and Data Source
In this retrospective cohort study, we performed a survival and secondary analysis of adult patients treated at level I and level II US and Canadian civilian trauma centers participating in the ACS-TQIP between January 1, 2017, and December 31, 2018. TQIP is a voluntary performance improvement program containing deidentified, validated, risk-adjusted patient and hospital data collected by trained abstractors. The study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. The Johns Hopkins Hospital institutional review board approved the study as an exemption and waived informed consent because the data were publicly accessible, retrospectively obtained, and deidentified.
Study Participants
The study included adult (aged ≥18 years) civilian trauma patients presenting with severe hemorrhage who received MTP within the first 24 hours of emergency department (ED) presentation. Massive transfusion protocol was defined as receiving a 1:1:1 ratio of packed red blood cells (pRBCs), plasma, and platelets after at least 4 units of transfused pRBCs within 1 hour of ED presentation or 10 units within 24 hours of ED arrival. Severe hemorrhage was defined as systolic blood pressure less than 90 mm Hg, shock index greater than 1, and receipt of at least 4 units of pRBCs within 1 hour of admission. Hypotension at ED presentation in the context of a shock index greater than 1 has been identified as a major risk factor and predictor for TIC and the potential for severe hemorrhage necessitating massive transfusion.13,14,15 Similar definitions have been used in the trauma literature.16,17 We excluded patients with burns, those who died within 1 hour of ED arrival, and those who had interfacility transfers.
Exposure
Patient exposure was separated into 2 groups: those who received WB-MTP and those who received MTP alone within 24 hours of ED presentation. Data on WB use from Trauma Quality Programs Participant Use File 2017 to 2018 level I and level II trauma centers were abstracted using International Statistical Classification of Diseases and Related Health Problems, Tenth Revision procedure codes. The exact type of WB is not specified in this data set. Data on the volume of WB transfused within the first 4 hours of hospital arrival had not been collected by TQIP until the beginning of the first quarter of 2020.
Outcomes
The primary outcomes measured were survival times at 24 hours and 30 days. Secondary outcomes selected a priori were hospital length of stay (LOS), intensive care unit (ICU) LOS, and in-hospital major complications. Major complications were defined as deep venous thrombosis, pulmonary embolism, acute respiratory distress syndrome, cerebrovascular accident, and acute kidney injury.
Potential Confounders
Patient baseline characteristics included age, sex, race (Asian, Black, and White; self-reported or identified by a family member), body mass index, and comorbidities (hypertension, diabetes, chronic obstructive pulmonary disease, stroke, and myocardial infarction). Injury characteristics were the mechanism of injury, injury severity score category, and ED vital signs (systolic blood pressure, heart rate, and Glasgow Coma Scale score). Hospital characteristics included ACS trauma center verification level and hospital site clustering. Interventions for hemorrhage, time to hemorrhage control intervention, time to first blood product transfusion, and time to WB transfusion were also included.
Statistical Analysis
Data were analyzed from February 2022 to September 2022. Parametric and nonparametric continuous variables were summarized with medians and IQRs. Categorical variables were described as counts and proportions. Independent-sample Mann-Whitney U tests and t tests were used as appropriate. Pearson χ2 tests were used to assess differences among categorical variables; all statistical tests were 2-sided.18 A 2-tailed P value of .05 was considered statistically significant. We performed multiple imputation to address missing data that were missing completely at random (eMethods in the Supplement).19,20,21
We performed univariate Kaplan-Meier survival analysis for the primary outcome. Log-rank statistics were calculated. Survival curves were generated using Stata SE, version 17 (StataCorp, LLC). Next, multivariable models were created by first selecting potential confounding variables a priori using the 10% change-in-estimate approach described by Mickey and Greenland22 and based on existing trauma literature that was clinically relevant (eMethods in the Supplement). We then conducted a multivariable, hierarchical, mixed-effects Cox proportional hazards regression model with a random effect to evaluate treatment with adjustment for possible confounders associated with survival while accounting for site clustering (eMethods and eTable 1 in the Supplement). The model met the Cox proportional hazards regression assumption based on complementary log-log plots and Schoenfeld residuals. Final model selection was based on the Akaike information criterion. The model’s goodness of fit was assessed using the Grønnesby and Borgan test. The Harrell C statistic was performed to determine the model’s discrimination. Results from the Cox proportional hazards regression model were presented as hazard ratios (HRs) and 95% CIs.
We then performed univariate analysis followed by a hierarchical mixed-effects multivariable logistic regression model to evaluate in-hospital complications. Next, the Hosmer-Lemeshow goodness-of-fit test was performed to assess the model fit. We then performed residual checks of covariates using locally weighted scatterplot smoothing of Pearson residuals. Likelihood ratio tests were conducted to assist with model building.
To address the significant differences in unbalanced comparison groups, propensity score matching was performed and outcomes were analyzed (eMethods, eTable 2, and eFigure 1 in the Supplement). Finally, an independent parametric statistical test was performed to determine hospital LOS and total ICU LOS.
Results
From January 1, 2017, through December 31, 2018, we identified 432 adult trauma patients who had received WB-MTP. There were 370 ACS-verified trauma centers (178 level I and 192 level II centers) identified in the study, with 45 centers using WB transfusion specific to the study population. A total of 4698 patients met criteria for severe hemorrhage (height and weight were missing for 903 patients [19%]). Of those patients, 1913 (41%) were excluded based on prespecified exclusion criteria. In total, 2785 patients were identified; 432 (15.5%) received WB-MTP (97 female [22%]; 335 male [78%]; median age, 38 years [IQR, 27-57 years]), and 2353 (84.5%) received MTP alone (531 female [23%]; 1822 male [77%]; median age, 38 years [IQR, 27-56 years]) (Table 1 and eFigure 2 in the Supplement). The overall 30-day mortality rate was 22%. Within the study cohort, demographics and comorbidities did not differ across resuscitation groups. Patients in both groups were profoundly injured, with a median injury severity score of 28 (IQR, 17-34; median difference 1.29 [95% CI, −0.05 to 2.64]). The MTP-only group was more likely to undergo surgery for hemorrhage control during hospitalization than was the WB-MTP group (1905 [81%] vs 324 [75%]; P = .004; risk difference, −0.06 [95% CI, −0.11 to −0.02]). Transfusion requirements during the first 4 hours and 24 hours were higher in the MTP-only group. For the WB-MTP group, the median WB units transfused was 1 (IQR, 1-1 unit), with 22 patients (5%) receiving more than 2 units of WB within 24 hours. After propensity score matching, 526 individuals were paired (263 in each group) (Table 1 and eTable 3 in the Supplement). There was a greater distribution of patients at level II trauma centers compared with level I centers (median, 1540 [IQR, 842-2102] vs 1320 [IQR, 628-2083]; median difference, −106.00 [95% CI, −175.05 to −37.81]) (eMethods and eTable 4 in the Supplement).
Table 1. Baseline Characteristics of the Study Population and Trauma Centers by Transfusion Group.
| Characteristic | Patientsa | |||
|---|---|---|---|---|
| Unmatched | Propensity score matched | |||
| WB-MTP (n = 432) | MTP (n = 2353) | WB-MTP (n = 263) | MTP (n = 263) | |
| Demographics | ||||
| Age, median (IQR), y | 38 (27-57) | 38 (27-56) | 37 (27-51) | 38 (27-53) |
| Race | ||||
| Asian | 6 (1) | 67 (3) | 4 (2) | 5 (2) |
| Black | 120 (28) | 700 (30) | 72 (27) | 68 (26) |
| White | 242 (56) | 1237 (53) | 149 (57) | 145 (55) |
| BMI, median (IQR) | 28 (24-32) | 28 (24-32) | 28 (24-31) | 26 (22-30) |
| Sex | ||||
| Female | 97 (22) | 531 (23) | 62 (24) | 59 (22) |
| Male | 335 (78) | 1822 (77) | 201 (76) | 204 (78) |
| ED vital sings, median (IQR) | ||||
| Systolic blood pressure, mm Hg | 70 (60-79) | 69 (59-77) | 70 (60-79) | 70 (59-79) |
| Heart rate, beats/min | 112 (97-129) | 117 (99-133) | 117 (100-129) | 117 (100-134) |
| Glasgow Coma Scale scoreb | 14 (13-15) | 8 (3-15) | 14 (13-15) | 14 (13-15) |
| Injury | ||||
| Penetrating | 150 (35) | 908 (39) | 87 (33) | 102 (39) |
| ISSc | ||||
| Median (IQR) | 26 (17-35) | 27 (19-36) | 26 (17-34) | 27 (19-34) |
| 1-8 | 14 (3) | 46 (2) | 6 (2) | 8 (3) |
| 9-15 | 56 (13) | 243 (10) | 33 (13) | 23 (9) |
| 16-24 | 112 (26) | 616 (26) | 75 (29) | 84 (32) |
| 25-75 | 250 (58) | 1448 (62) | 149 (57) | 148 (56) |
| AIS, median (IQR)d | ||||
| Head | 0 (0-1) | 0 (0-3) | 0 (0-1) | 0 (0-1) |
| Chest | 3 (1-4) | 3 (1-4) | 3 (2-4) | 3 (0-4) |
| Abdomen | 3 (0-4) | 3 (0-4) | 3 (0-4) | 3 (0-4) |
| Spine | 0 (0-2) | 0 (0-2) | 0 (0-2) | 0 (0-2) |
| Comorbidities | ||||
| Hypertension | 70 (16) | 333 (14) | 47 (18) | 43 (16) |
| Diabetes | 32 (7) | 149 (6) | 24 (9) | 14 (5) |
| COPD | 14 (3) | 57 (2) | 11 (4) | 11 (4) |
| Myocardial infarction | 2 (1) | 16 (1) | 1 (1) | 1 (1) |
| Stroke | 7 (2) | 17 (1) | 3 (1) | 1 (1) |
| Trauma center | ||||
| ACS trauma center | ||||
| Level I | 301 (70) | 1758 (75) | 171 (65) | 171 (65) |
| Level II | 131 (30) | 595 (25) | 92 (35) | 92 (35) |
| Intervention for hemorrhage control | 324 (75) | 1905 (81) | 197 (75) | 197 (75) |
| Time to intervention, median (IQR), min | 62 (41-115) | 60 (37-109) | 66 (43-115) | 59 (38-109) |
| Time to first blood product transfusion, median (IQR), min | 37 (17-76) | 37 (17-72) | 37 (17-76) | 37 (17-76) |
| Transfusion amount, median (IQR), U | ||||
| 24 h | ||||
| pRBCs | 8 (4-14) | 14 (9-22) | 9 (5-14) | 9 (5-14) |
| Plasma | 6 (3-10) | 10 (6-17) | 6 (3-10) | 8 (5-12) |
| Platelets (pooled pack) | 1 (0-3) | 4 (2-7) | 3 (1-5) | 4 (2-6) |
| Cryoprecipitate | 0 (0-1) | 0 (0.3-2) | 0 (0-1) | 0 (0-1) |
| WB | 1 (1-1) | NA | NA | NA |
| 4 h | ||||
| pRBCs | 7 (4-11) | 11 (6-18) | 7 (4-11) | 7 (5-11) |
| Plasma | 4 (2-8) | 8 (4-14) | 5 (2-8) | 6 (4-9) |
| Platelets (pooled pack) | 1 (0-2) | 3 (2-5) | 1 (0-3) | 2 (1-4) |
| Cryoprecipitate | 0 (0-0.5) | 0 (0-1) | 0 (0-0) | 0 (0-0) |
| WB | 1 (1-1) | NA | NA | NA |
Abbreviations: ACS, American College of Surgeons; AIS, Abbreviated Injury Score; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); COPD, chronic obstructive pulmonary disease; ED, emergency department; ISS, Injury Severity Score; MTP, massive transfusion protocol; NA, not applicable; pRBCs, packed red blood cells; WB, whole blood; WB-MTP, WB as an adjunct to component therapy–based MTP.
Data are presented as number (percentage) of patients unless otherwise indicated.
Scores range from 3 to 15, with higher scores indicating improved responsiveness.
Scores range from 1 to 75, with higher scores indicating a currently untreatable injury.
Scores range from 1 to 6, with higher scores indicating a currently untreatable injury.
The log-rank statistic provided evidence for a difference in overall time to death at 30 days between the 2 groups. On univariate survival analysis of death at 24 hours and 30 days, patients who received WB-MTP had significantly improved survival at 24 hours (HR, 0.72; 95% CI, 0.52-0.99; P = .049) and at 30 days (HR, 0.74; 95% CI, 0.59-0.95; P = .02) (Figure 1).
Figure 1. Unadjusted Kaplan-Meier Survival Estimates by Transfusion Group.
MTP indicates massive transfusion protocol and WB-MTP, whole blood as an adjunct to component therapy–based MTP.
We then performed a survival analysis at 24 hours, adjusting for confounders using the Cox proportional hazards regression model. There was an association between improved survival at 24 hours and WB-MTP, with a 37% lower risk of mortality (HR, 0.63; 95% CI, 0.41-0.96; P = .03). The 24-hour survival curve of both transfusion groups demonstrated early separation, before 5 hours, following ED arrival and initial blood product transfusion (Figure 2A). Similarly, adjusted Cox proportional hazards regression demonstrated improved survival benefit associated with WB-MTP at 30 days, with a reduction in the hazard of mortality (HR, 0.53; 95% CI, 0.31-0.93; P = .02) (Figure 2B). After propensity score matching, the survival benefit associated with WB-MTP persisted at 24 hours (HR, 0.76; 95% CI, 0.62-0.95; P = .02) and 30 days (HR, 0.48; 95% CI, 0.25-0.91; P = .03) (Table 2). Among patients who received WB-MTP, there was no significant difference in the adjusted odds ratio (OR) for overall major complications (OR, 0.82; 95% CI, 0.37-1.81; P = .63) (Table 3) (eMethods and eFigure 3 in the Supplement).
Figure 2. Adjusted Kaplan-Meier Survival Estimates by Transfusion Group.
MTP indicates massive transfusion protocol and WB-MTP, whole blood as an adjunct to component therapy–based MTP.
Table 2. Adjusted Cox Proportional Hazards Regression Treatment Effect Estimates by Transfusion Group for Unmatched and Matched Results.
| Treatment characteristic | Unmatched | Propensity score matched | ||||||
|---|---|---|---|---|---|---|---|---|
| Mortality at 24 h | Mortality at 30 d | Mortality at 24 h | Mortality at 30 d | |||||
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| MTP alone | 1 [Reference] | NA | NA | NA | NA | NA | NA | NA |
| WB-MTP | 0.63 (0.41-0.96) | .03 | 0.53 (0.31-0.93) | .02 | 0.76 (0.62-0.95) | .02 | 0.48 (0.25-0.91) | .03 |
| ISS, each category increase | 1.02 (1.02-1.03) | <.001 | 1.02 (1.01-1.02) | <.001 | 1.00 (0.99-1.06) | .10 | 1.03 (1.01-1.05) | .001 |
| Total GCS score, 1-point increase | 0.90 (0.85-0.97) | .001 | 0.93 (0.88-0.97) | .002 | 0.86 (0.75-1.01) | .06 | 0.91 (0.82-0.99) | .04 |
| Penetrating injury | 1.67 (1.02-2.76) | .04 | 1.08 (0.77-1.51) | .65 | 0.99 (0.30-3.38) | .10 | 1.11 (0.48-2.60) | .80 |
| Time to bleeding control, per min increase | 0.99 (0.99-1.01) | .07 | 0.99 (0.99-1.01) | .53 | 0.99 (0.98-1.01) | .14 | 0.99 (0.99-1.01) | .36 |
| Intervention for bleeding control | 0.87 (0.78-0.97) | .01 | 0.80 (0.69-0.92) | .002 | 0.74 (0.49-1.10) | .14 | 0.61 (0.42-0.88) | .009 |
| Trauma center | ||||||||
| Level I | 0.91 (0.51-1.67) | .78 | 0.73 (0.49-1.49) | .12 | 0.81 (0.33-2.01) | .66 | 0.96 (0.86-1.08) | .55 |
| Level II | 1.29 (1.10-1.51) | .002 | 1.46 (0.99-2.12) | .051 | 1.07 (0.39-2.90) | .14 | 1.84 (1.00-3.37) | .048 |
| Time to first blood product transfusion | 1.00 (1.00-1.00) | .003 | 1.00 (1.00-1.00) | <.001 | 0.99 (0.99-1.01) | .19 | 1.00 (1.00-1.00) | .008 |
| Delay WB transfusion ≥2 h | 1.00 (1.00-1.01) | .007 | 2.10 (0.77-5.73) | .14 | 1.00 (1.00-1.01) | .008 | 2.16 (0.76-6.15) | .15 |
| Age, 10-y increase | 1.26 (1.16-1.39) | <.001 | 1.45 (1.30-1.63) | <.001 | 1.80 (1.28-2.55) | .001 | 1.90 (1.46-2.47) | <.001 |
| Male | 1.38 (1.01-1.90) | .04 | 1.58 (1.06-2.34) | .02 | 2.36 (0.87-6.43) | .09 | 1.36 (0.68-2.73) | .39 |
| Systolic blood pressure, 1-mm Hg increase | 0.97 (0.96-0.97) | <.001 | 0.99 (0.99-1.01) | .60 | 0.99 (0.98-1.01) | .49 | 0.99 (0.98-0.99) | .04 |
| Pulse, 10-point increase, beats per min | 1.01 (1.01-1.02) | <.001 | 1.06 (0.97-1.16) | .18 | 1.01 (0.99-1.03) | .10 | 1.11 (0.92-1.34) | .28 |
Abbreviations: GCS, Glasgow Coma Scale; HR, hazard ratio; ISS, Injury Severity Score; MTP, massive transfusion protocol; NA, not applicable; WB, whole blood; WB-MTP, WB as an adjunct to component therapy–based MTP.
Table 3. Adjusted Odds Ratios for Major Complications in the WB-MTP Group Compared With the MTP-Only Group as Reference.
| Outcome | Odds ratio (95% CI) | P value |
|---|---|---|
| Acute kidney injury | 0.47 (0.22-1.01) | .055 |
| Pulmonary embolism | 0.84 (0.32-2.19) | .73 |
| Deep vein thrombosis | 2.11 (0.99-4.45) | .06 |
| ARDS | 1.58 (0.72-3.51) | .25 |
| Stroke | 0.61 (0.17-2.16) | .44 |
| Overall | 0.82 (0.37-1.81) | .63 |
Abbreviations: ARDS, acute respiratory distress syndrome; MTP, massive transfusion protocol; WB-MTP, WB as an adjunct to component therapy–based MTP.
When considering hospital LOS and total ICU LOS, there was a statistically significant greater hospital LOS in the WB-MTP group (median, 15 days [IQR, 6-26 days] vs 11 days [IQR, 2-28 days]; mean difference, −1.67 days [95% CI, −2.80 to −0.53 days]; P = .004). Conversely, there was no significant difference in ICU LOS between the groups (median, 6 days [IQR, 3-14 days] vs 5 days [IQR, 2-14 days]; mean difference, 0.15 days [95% CI, −1.11 to 1.42 days]; P = .81) (eMethods and eTable 5 in the Supplement).
Discussion
The results from this analysis showed a survival benefit at 24 hours and 30 days associated with WB-MTP compared with MTP alone among patients presenting with or at risk of severe hemorrhage in adult civilian trauma centers in the US and Canada. Furthermore, the survival curves for the 2 groups separated within 5 hours of ED arrival and initial blood product transfusion, suggesting that WB is associated with an early beneficial effect in blunting the pathophysiology of TIC. The early favorable results associated with WB indicate the importance of timely administration in patients presenting with severe hemorrhage shortly after ED arrival.
Hemostatic resuscitation strategies, which are associated with improved survival among patients with severe trauma, aim to counter TIC. These strategies involve replacing lost and defective blood components by transfusing CT to mirror the original composition of WB. Nonetheless, transfusing separated blood components has consistently failed to restore the original hemostatic profile and robustness of native WB.23 In a previous study,24 investigators found a disadvantaged biologic profile of CT compared with WB, evidenced by a reduction in hematocrit concentration (29% vs 38%-50%), factor activity (65% vs 100%), platelet count (88 000 vs 150 000-400 000 cells/mm3), and fibrinogen (750 vs 1000 mg/dL). Furthermore, the decreased hemostatic capacity observed after the processing and collection of separated blood components is further hindered by the accumulation of storage lesions and the anticoagulant effects of additive storage solutions.24
Recognizing the constraints of CT, the military sought to revisit the use of WB resuscitation throughout the Iran and Afghanistan conflicts. As a result, several retrospective studies from the military experience have demonstrated improved outcomes associated with WB-MTP compared with component therapy–based MTP, prompting the Committee on Tactical Combat Casualty Care to recommend fresh WB (FWB) as the initial product for hemostatic resuscitation.25,26,27 More recently, Gurney et al28 described survival at 30 days for FWB compared with CT in a retrospective cohort of individuals critically injured during combat in Afghanistan. After adjusting for confounders, the study demonstrated a survival benefit associated with FWB, with an increased hazard of mortality in the cohort with no FWB (HR, 2.8; 95% CI, 1.2-6.4). However, there were several notable differences between that study’s findings and the current study’s findings. First, the WB product used in the military was FWB, reflecting the military’s austere environment. In contrast, modified, cold-stored WB was used in this study’s civilian setting. The initial concerns for cold-stored WB in civilian trauma centers included the theoretical reduced effectiveness related to temperature-related platelet dysfunction and hemolytic reactions. Prior studies have suggested that cold-stored platelets demonstrate improved hemostatic function in the acute setting with a more extended shelf half-life.29,30 Second, a substantial number of patients in both transfusion groups in the study conducted by Gurney et al28 had not received any platelets due to the lack of inventory. In our study, both groups received a median of 1 pooled pack of platelets. Despite the differences between the study by Gurney et al28 and the current study, our findings of a statistically significant reduction in time to death were consistent with the military experience.
Realizing the potential benefits of WB in the civilian setting from the military experience, Cotton et al31 conducted a randomized clinical pilot trial of 107 patients that compared WB-MTP with MTP alone in severely injured trauma patients requiring large-volume transfusions. The study showed no difference in secondary outcomes of 24-hour or 30-day mortality rates. However, there are several plausible reasons for the discordant survival outcomes compared with our study. The study by Cotton et al31 was conducted in 2013, prior to the US Food & Drug Administration’s and the Association for the Advancement of Blood & Biotherapies’ approval of WB in the civilian setting. The absence of approval limited the authors’ use of WB by excluding patients in the study with group B and group AB blood types. Additionally, the process of generating ABO blood type–compatible WB rendered platelets ineffective. Furthermore, there was a substantial delay in time to WB transfusion; the need for blood typing restricted the efficiency and, at times, the ability to use WB altogether.
As more trauma centers began to emphasize the practice of WB, uncertainties remained about patient outcomes due to the lack of consistent evidence. Given this knowledge gap, Hanna et al32 conducted a nationwide retrospective analysis comparing WB-MTP with MTP only in trauma patients from 2015 to 2016. The authors found a reduction in 24-hour mortality (OR, 0.78; 95% CI, 0.59–0.89; P = .01) and hospital mortality (OR, 0.88; 95% CI, 0.81–0.90; P = .01) associated with WB-MTP. The significant survival benefits associated with WB-MTP observed in that study align with the results of our study. However, the current study was distinct in many aspects. We chose to conduct a survival analysis as opposed to logistic regression. Survival analysis has the advantage of recognizing outcome events at different time points, which can detect different treatment effects among groups, thereby identifying where along the timeline patients may benefit the most from intervention. Additionally, the survival analysis considered patients who were censored (discharged or transferred) before the end of the evaluation period of 30 days. Lastly, we included time to WB and blood product transfusion in our analysis, a limitation not considered in the study by Hanna et al.32
More recently, preliminary results from a multicenter, prospective, observational study conducted by Hazelton et al33 that compared WB transfusion with MTP without WB showed that WB transfusion was associated with a 48% reduction in in-hospital mortality. That study’s results are consistent with our study’s results, demonstrating a statistically significant improvement in survival associated with WB transfusion. However, several critical differences exist between the present study and the preliminary work of Hazelton and colleagues.33 First, the study by Hazelton et al33 included patients of all ages, from pediatric patients to older patients, with a median age of 40 years (IQR, 11-30 years). Although the authors adjusted for age, there remain several biologic, anatomic, and practical differences between pediatric and adult patients to consider when interpreting the results. Second, patients who received any blood transfusion during the initial resuscitation were included in the study by Hazelton et al33 without any additional physiologic criteria. In contrast, we chose patients who received massive blood transfusion in the setting of shock, with a design to evaluate severely injured trauma patients most likely to benefit from an intervention.
Limitations
This study has several limitations. The study is a retrospective analysis from a national database; the improved survival and secondary benefits associated with WB transfusion can only be interpreted as an association, not causation. As an observational study, the lack of randomization places the study at an inherent risk of confounding by indication and other unmeasured biases. However, trauma centers that use WB typically have it stored in the trauma bay, providing readily accessible hemostatic resuscitation to severely injured patients. Given the proximity and immediate access to WB, severely injured patients may be more likely to receive WB than to wait for the release and arrival of CT from a blood bank. Confounding by indication in this manner may underestimate the beneficial effects of WB. Conversely, the beneficial effects of WB may be overestimated, as patients who received WB early after arrival may have done well otherwise. However, we performed a propensity score analysis to mitigate this limitation. Another limitation of our study is the lack of laboratory data (eg, coagulation parameters, lactate), practitioner-level data, and data on tranexamic acid administration. These variables are not captured in the TQIP database, which allows for unmeasured biases.
The total amount of WB given to each patient was low (median, 1 unit; IQR, 1-1 unit), which leaves an ambiguous interpretation of the survival benefit associated with a single intervention. However, as a single transfusion product, the underlying mechanism of survival benefit could be related to its faster and more efficient correction of coagulopathy. In addition, recipients received fewer donor immunologic constituents as a single product compared with multiple immune-provoking products of CT derived from various donor sources. In addition, the close proximity of WB to the trauma bay may have also contributed to the survival benefit observed. Furthermore, details on the type of WB used at varying institutions are not specified in the TQIP data.
Our findings did not detect a statistically significant difference in major complications between the comparison groups. This limitation could be explained by our study not being powered enough to detect smaller differences (eMethods and eFigure 4 in the Supplement). Nevertheless, these pertinent patient-centered outcomes warrant further scrutiny in future studies.
In addition, prehospital administration of blood products is not specified in the TQIP data set. The study by Sperry et al34 demonstrated survival benefit associated with prehospital blood product transfusion in patients with severe hemorrhage. However, the studies by Moore et al35 and Crombie et al17 did not demonstrate a survival advantage associated with prehospital blood product transfusion, which in part may be due to the shorter transport times seen in these studies.
Conclusions
In this cohort study, receipt of WB-MTP was associated with improved survival in trauma patients presenting with severe hemorrhage, with a survival benefit found early after transfusion. Therefore, initial WB resuscitation may offer marked benefits in this patient population. The findings from this study are clinically important as this is an essential first step in prioritizing the selection of WB resuscitation for trauma patients presenting with severe hemorrhage. Further studies are warranted to identify which subgroup of patients will have the most benefit and to incorporate these findings into MTPs.
eMethods.
eTable 1. Adjusted Cox Proportional Hazard Regression Multivariable Model. Treatment Effect Estimates by Transfusion Group for 30-Day Mortality
eTable 2. Measures of Balance After Propensity Score Matching
eFigure 1. Distributions of the Estimated Propensity Scores Between Comparison Groups
eFigure 2. Study Selection
eTable 3. Baseline Characteristics, Expanded Detail Including Risk Differences and Median Differences
eTable 4. Distribution of Patients by Trauma Center Level
eFigure 3. Forest Plot Displaying Adjusted OR and 95% CI for Major Complications by Transfusion Group
eTable 5. Unadjusted Hospital and ICU Length of Stay by Transfusion Group
eFigure 4. Sample Size and Power Plot
eReferences
References
- 1.Callcut RA, Kornblith LZ, Conroy AS, et al. ; Western Trauma Association Multicenter Study Group . The why and how our trauma patients die: a prospective multicenter Western Trauma Association study. J Trauma Acute Care Surg. 2019;86(5):864-870. doi: 10.1097/TA.0000000000002205 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Brohi K, Singh J, Heron M, Coats T. Acute traumatic coagulopathy. J Trauma. 2003;54(6):1127-1130. doi: 10.1097/01.TA.0000069184.82147.06 [DOI] [PubMed] [Google Scholar]
- 3.Cannon JW, Khan MA, Raja AS, et al. Damage control resuscitation in patients with severe traumatic hemorrhage: a practice management guideline from the Eastern Association for the Surgery of Trauma. J Trauma Acute Care Surg. 2017;82(3):605-617. doi: 10.1097/TA.0000000000001333 [DOI] [PubMed] [Google Scholar]
- 4.Jenkins DH, Winchell RJ, Coimbra R, et al. Position statement of the American College of Surgeons Committee on Trauma on the National Academies of Sciences, Engineering and Medicine Report, A National Trauma Care System: integrating military and civilian trauma systems to achieve zero preventable deaths after injury. J Trauma Acute Care Surg. 2016;81(5):819-823. doi: 10.1097/TA.0000000000001217 [DOI] [PubMed] [Google Scholar]
- 5.Spinella PC, Pidcoke HF, Strandenes G, et al. Whole blood for hemostatic resuscitation of major bleeding. Transfusion. 2016;56(suppl 2):S190-S202. doi: 10.1111/trf.13491 [DOI] [PubMed] [Google Scholar]
- 6.Cordova CB, Cap AP, Spinella PC. Fresh whole blood transfusion for a combat casualty in austere combat environment. J Spec Oper Med. 2014;14(1):9-12. doi: 10.55460/6WR8-NER8 [DOI] [PubMed] [Google Scholar]
- 7.Spinella PC, Dunne J, Beilman GJ, et al. Constant challenges and evolution of US military transfusion medicine and blood operations in combat. Transfusion. 2012;52(5):1146-1153. [DOI] [PubMed]
- 8.Pidcoke HF, Aden JK, Mora AG, et al. Ten-year analysis of transfusion in Operation Iraqi Freedom and Operation Enduring Freedom: increased plasma and platelet use correlates with improved survival. J Trauma Acute Care Surg. 2012;73(6)(suppl 5):S445-S452. doi: 10.1097/TA.0b013e3182754796 [DOI] [PubMed] [Google Scholar]
- 9.Bahr MP, Yazer MH, Triulzi DJ, Collins RA. Whole blood for the acutely haemorrhaging civilian trauma patient: a novel idea or rediscovery? Transfus Med. 2016;26(6):406-414. doi: 10.1111/tme.12329 [DOI] [PubMed] [Google Scholar]
- 10.Yazer MH, Spinella PC, Anto V, Dunbar NM. Survey of group A plasma and low-titer group O whole blood use in trauma resuscitation at adult civilian level 1 trauma centers in the US. Transfusion. 2021;61(6):1757-1763. doi: 10.1111/trf.16394 [DOI] [PubMed] [Google Scholar]
- 11.Hashmi ZG, Chehab M, Nathens AB, et al. Whole truths but half the blood: addressing the gap between the evidence and practice of pre-hospital and in-hospital blood product use for trauma resuscitation. Transfusion. 2021;61(S1)(suppl 1):S348-S353. doi: 10.1111/trf.16515 [DOI] [PubMed] [Google Scholar]
- 12.American College of Surgeons . ACS TQIP massive transfusion in trauma guidelines. Published October 2014. Accessed December 16, 2021. https://docslib.org/doc/9119686/acs-tqip-massive-transfusion-in-trauma-guidelines
- 13.Kushimoto S, Kudo D, Kawazoe Y. Acute traumatic coagulopathy and trauma-induced coagulopathy: an overview. J Intensive Care. 2017;5(1):6. doi: 10.1186/s40560-016-0196-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Schroll R, Swift D, Tatum D, et al. Accuracy of shock index versus ABC score to predict need for massive transfusion in trauma patients. Injury. 2018;49(1):15-19. doi: 10.1016/j.injury.2017.09.015 [DOI] [PubMed] [Google Scholar]
- 15.Mutschler M, Nienaber U, Münzberg M, et al. ; TraumaRegister DGU . The shock index revisited—a fast guide to transfusion requirement? a retrospective analysis on 21,853 patients derived from the TraumaRegister DGU. Crit Care. 2013;17(4):R172. doi: 10.1186/cc12851 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Guyette FX, Brown JB, Zenati MS, et al. ; STAAMP Study Group . Tranexamic acid during prehospital transport in patients at risk for hemorrhage after injury: a double-blind, placebo-controlled, randomized clinical trial. JAMA Surg. 2020;156(1):11-20. doi: 10.1001/jamasurg.2020.4350 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Crombie N, Doughty HA, Bishop JRB, et al. ; RePHILL collaborative group . Resuscitation with blood products in patients with trauma-related haemorrhagic shock receiving prehospital care (RePHILL): a multicentre, open-label, randomised, controlled, phase 3 trial. Lancet Haematol. 2022;9(4):e250-e261. doi: 10.1016/S2352-3026(22)00040-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Austin PC. A tutorial on multilevel survival analysis: methods, models and applications. Int Stat Rev. 2017;85(2):185-203. doi: 10.1111/insr.12214 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Haider AH, Saleem T, Leow JJ, et al. Influence of the National Trauma Data Bank on the study of trauma outcomes: is it time to set research best practices to further enhance its impact? J Am Coll Surg. 2012;214(5):756-768. doi: 10.1016/j.jamcollsurg.2011.12.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Oyetunji TA, Crompton JG, Ehanire ID, et al. Multiple imputation in trauma disparity research. J Surg Res. 2011;165(1):e37-e41. doi: 10.1016/j.jss.2010.09.025 [DOI] [PubMed] [Google Scholar]
- 21.White IR, Royston P, Wood AM. Multiple imputation using chained equations: issues and guidance for practice. Stat Med. 2011;30(4):377-399. doi: 10.1002/sim.4067 [DOI] [PubMed] [Google Scholar]
- 22.Mickey RM, Greenland S. The impact of confounder selection criteria on effect estimation. Am J Epidemiol. 1989;129(1):125-137. doi: 10.1093/oxfordjournals.aje.a115101 [DOI] [PubMed] [Google Scholar]
- 23.Ponschab M, Schöchl H, Gabriel C, et al. Haemostatic profile of reconstituted blood in a proposed 1:1:1 ratio of packed red blood cells, platelet concentrate and four different plasma preparations. Anaesthesia. 2015;70(5):528-536. doi: 10.1111/anae.13067 [DOI] [PubMed] [Google Scholar]
- 24.Mays JA, Hess JR. Modelling the effects of blood component storage lesions on the quality of haemostatic resuscitation in massive transfusion for trauma. Blood Transfus. 2017;15(2):153-157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Nessen SC, Eastridge BJ, Cronk D, et al. Fresh whole blood use by forward surgical teams in Afghanistan is associated with improved survival compared to component therapy without platelets. Transfusion. 2013;53(suppl 1):107S-113S. doi: 10.1111/trf.12044 [DOI] [PubMed] [Google Scholar]
- 26.Spinella PC, Perkins JG, Grathwohl KW, Beekley AC, Holcomb JB. Warm fresh whole blood is independently associated with improved survival for patients with combat-related traumatic injuries. J Trauma. 2009;66(4)(suppl):S69-S76. doi: 10.1097/TA.0b013e31819d85fb [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Butler FK Jr, Holcomb JB, Schreiber MA, et al. Fluid resuscitation for hemorrhagic shock in tactical combat casualty care: TCCC guidelines change 14-01—2 June 2014. J Spec Oper Med. 2014;14(3):13-38. doi: 10.55460/DPOC-JWIY [DOI] [PubMed] [Google Scholar]
- 28.Gurney J, Staudt A, Cap A, et al. Improved survival in critically injured combat casualties treated with fresh whole blood by forward surgical teams in Afghanistan. Transfusion. 2020;60(S3)(suppl 3):S180-S188. doi: 10.1111/trf.15767 [DOI] [PubMed] [Google Scholar]
- 29.Zhu CS, Pokorny DM, Eastridge BJ, et al. Give the trauma patient what they bleed, when and where they need it: establishing a comprehensive regional system of resuscitation based on patient need utilizing cold-stored, low-titer O+ whole blood. Transfusion. 2019;59(S2):1429-1438. doi: 10.1111/trf.15264 [DOI] [PubMed] [Google Scholar]
- 30.Pidcoke HF, McFaul SJ, Ramasubramanian AK, et al. Primary hemostatic capacity of whole blood: a comprehensive analysis of pathogen reduction and refrigeration effects over time. Transfusion. 2013;53(0 1)(suppl 1):137S-149S. doi: 10.1111/trf.12048 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Cotton BA, Podbielski J, Camp E, et al. ; Early Whole Blood Investigators . A randomized controlled pilot trial of modified whole blood versus component therapy in severely injured patients requiring large volume transfusions. Ann Surg. 2013;258(4):527-532. doi: 10.1097/SLA.0b013e3182a4ffa0 [DOI] [PubMed] [Google Scholar]
- 32.Hanna K, Bible L, Chehab M, et al. Nationwide analysis of whole blood hemostatic resuscitation in civilian trauma. J Trauma Acute Care Surg. 2020;89(2):329-335. doi: 10.1097/TA.0000000000002753 [DOI] [PubMed] [Google Scholar]
- 33.Hazelton JP, Ssentongo AE, Oh JS, et al. Whole blood vs blood component therapy for transfusion in trauma patients: a prospective, multicenter study. Abstract presented at: American Surgical Association Annual Meeting; April 7, 2022; Chicago, IL. Scientific Session II.
- 34.Sperry JL, Guyette FX, Brown JB, et al. ; PAMPer Study Group . Prehospital plasma during air medical transport in trauma patients at risk for hemorrhagic shock. N Engl J Med. 2018;379(4):315-326. doi: 10.1056/NEJMoa1802345 [DOI] [PubMed] [Google Scholar]
- 35.Moore HB, Moore EE, Chapman MP, et al. Plasma-first resuscitation to treat haemorrhagic shock during emergency ground transportation in an urban area: a randomised trial. Lancet. 2018;392(10144):283-291. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods.
eTable 1. Adjusted Cox Proportional Hazard Regression Multivariable Model. Treatment Effect Estimates by Transfusion Group for 30-Day Mortality
eTable 2. Measures of Balance After Propensity Score Matching
eFigure 1. Distributions of the Estimated Propensity Scores Between Comparison Groups
eFigure 2. Study Selection
eTable 3. Baseline Characteristics, Expanded Detail Including Risk Differences and Median Differences
eTable 4. Distribution of Patients by Trauma Center Level
eFigure 3. Forest Plot Displaying Adjusted OR and 95% CI for Major Complications by Transfusion Group
eTable 5. Unadjusted Hospital and ICU Length of Stay by Transfusion Group
eFigure 4. Sample Size and Power Plot
eReferences