Prehospital whole blood transfusion in hemorrhagic shock: A systematic review

Hemorrhagic shock is a life-threatening form of circulatory failure resulting from acute blood loss that causes inadequate tissue perfusion, cellular hypoxia, and metabolic derangement.1, 2 It remains a major global cause of morbidity and mortality across both traumatic and non-traumatic etiologies. Traumatic hemorrhage is the leading cause of preventable death in adults under 45 years of age and accounts for approximately 1.5 million deaths annually worldwide.1, 3, 4 Non-traumatic causes, including gastrointestinal (GI) bleeding, obstetric hemorrhage, and vascular rupture, also contribute substantially to the global burden of shock, particularly in settings where access to blood products and rapid resuscitation is limited.4-6 Despite advances in trauma systems and critical care, the temporal distribution of hemorrhage-related deaths has changed little over time, with most fatalities occurring within hours of injury or onset.7-9 Early mortality is frequently the result of uncontrolled bleeding before definitive hemostasis can be achieved, emphasizing the critical importance of timely recognition and early initiation of hemostatic resuscitation.3, 10 Efforts to improve survival from hemorrhagic shock have increasingly focused on early, balanced transfusion strategies that restore circulating volume and coagulation function simultaneously.11 Historically, whole blood (WB) was the standard resuscitative fluid during much of the twentieth century but was replaced by component therapy (CT) to facilitate inventory management and targeted transfusion.12 Modern balanced resuscitation approaches, such as the 1:1:1 ratio of red blood cells, plasma, and platelets, were developed to approximate WB and have demonstrated improved hemostatic control and reduced exsanguination in trauma.13 WB has since re-emerged as a practical and physiologic alternative, offering red cells, plasma, and platelets in a single product with preserved hemostatic properties.14-16 In the prehospital environment, where transport time, storage capacity, and logistics limit access to multiple components, a single balanced product may simplify resuscitation and expedite delivery of hemostatic therapy.17-19 Several foundational trials have evaluated prehospital transfusion of blood products for patients with hemorrhagic shock.20 The Prehospital Air Medical Plasma (PAMPer) and Control of Major Bleeding After Trauma (COMBAT) trials evaluated prehospital plasma compared with crystalloid resuscitation and reported mixed results.21, 22 These studies were conducted between 2009 and 2013 and primarily involved helicopter emergency medical services (HEMS). The RePHILL trial reported no survival difference between prehospital red blood cell and plasma transfusion versus crystalloid.23 Differences in patient acuity, injury mechanism, transport timelines, and prehospital team composition limit direct comparison with contemporary civilian paramedic-based emergency medical services (EMS).21-23 Observational studies suggest that earlier transfusion, rather than product type alone, is associated with improved short-term survival.24-26 Civilian and military systems have increasingly adopted WB for early resuscitation, reporting favorable feasibility and safety profiles in both trauma and non-traumatic hemorrhage.27-32 However, the existing evidence remains limited, and uncertainty persists regarding the specific effectiveness and safety of WB when administered in the prehospital setting. The literature evaluating prehospital transfusion demonstrates substantial clinical and methodological heterogeneity.30, 31 Studies differ in patient populations, including traumatic and non-traumatic hemorrhage, transfusion strategies, EMS system structure, transport times, and outcome definitions. Most available data are derived from observational designs with moderate to serious risk of bias, which limits causal inference and generalizability. The objective of this systematic review was to evaluate the safety and effectiveness of prehospital WB transfusion compared with CT or crystalloid-based resuscitation in adults with hemorrhagic shock from traumatic and non-traumatic causes. This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines33, 34 and was prospectively registered with PROSPERO (CRD42024604327). Study eligibility criteria were defined using the Population/Intervention/Comparison/Outcomes/Study Design/Timeframe framework.35 The population included adults aged 16 years or older with hemorrhagic shock from either traumatic or non-traumatic causes. Both civilian and military populations were eligible. The intervention was prehospital WB transfusion administered before hospital arrival as part of early resuscitation. Studies describing in-hospital transfusion were included only when WB administration was initiated during the prehospital phase as part of early resuscitation. Out-of-hospital surgical teams, such as forward surgical units, were excluded because their capacity for definitive surgical hemostasis precludes direct comparison with conventional prehospital care. Comparators included CT or crystalloid-based resuscitation. The primary outcome was 24-h all-cause mortality.36 Secondary outcomes included mortality at 1 h, 3–6 h, and 28–30 days; 24-h transfusion requirements (volume and units of blood products); and transfusion-related adverse events, including febrile non-hemolytic transfusion reactions (FNHTR), allergic or anaphylactic reactions, transfusion-associated circulatory overload (TACO), transfusion-related acute lung injury (TRALI), and transfusion-transmitted infection.37 Eligible study designs included randomized controlled trials, non-randomized controlled trials, and observational studies (cohort or case–control). While included as safety outcomes, complications such as TACO, TRALI, and transfusion-transmitted infection typically manifest beyond the prehospital timeframe. Accordingly, these complications are unlikely to be detected during prehospital care and are primarily relevant to in-hospital monitoring. The electronic search strategy was developed with an experienced information specialist (R.S.) and peer-reviewed by a second specialist following Peer Review of Electronic Search Strategies (PRESS) guidelines.38 The following databases were searched: EMBASE (Ovid), MEDLINE (Ovid), Scopus, Web of Science, Cumulative Index to Nursing and Allied Health Literature, and the Cochrane Central Register of Controlled Trials (CENTRAL). Searches were conducted independently by two information specialists using Medical Subject Headings (MeSH) and free-text terms combined with Boolean operators. Searches were performed from database inception to August 27, 2025. All publication years were considered, but only studies published in English or French were included. Non-English or French language papers, conference abstracts, and unpublished data were excluded to ensure all included studies had undergone peer review. Reference lists were screened for additional studies, and subject matter experts reviewed the final selection for completeness and relevance. Complete search strategies for each database, including all search terms, filters, and limits, are provided in Table S1: Electronic Search Strategies. The strategy incorporated both controlled vocabulary and free-text keywords related to “whole blood,” “prehospital transfusion,” and “hemorrhagic shock,” adapted to each database's indexing system. All records were imported into Covidence systematic review software (version 2.0; Veritas Health Innovation, Melbourne, Australia), and duplicates were removed. Two reviewers (P.-M.D. and A.G.) independently screened titles and abstracts according to the eligibility criteria. Discrepancies were resolved by discussion, and unresolved disagreements were adjudicated by a third reviewer (B.N.). No automation tools were used. Two reviewers (P.-M.D. and A.G.) independently extracted data in duplicate using a standardized, piloted form in Covidence. Extracted data were compared for accuracy, with discrepancies resolved by consensus or third-party adjudication (B.N.). No automation or machine-assisted tools were used. When information was incomplete or unclear, study authors were contacted, and data were extracted as reported in the published article or updated upon author response. Data were extracted for all reported measures of mortality (1-h, 3-6-h, 24-h, and 28- to 30-day), transfusion requirements (volume and number of blood units), and transfusion-related adverse events, including febrile non-hemolytic, allergic, or anaphylactic reactions, as well as TACO, TRALI, and transfusion-transmitted infection. Additional data included publication details (author, year, and country), study characteristics (design, setting, and inclusion criteria), population characteristics (age, Injury Severity Score [ISS]), and intervention details (type, volume, and timing of transfusion). Assumptions regarding missing or unclear data were documented, and no imputation was performed. Risk of bias was assessed independently by two reviewers (P.-M.D. and A.G.). Randomized controlled trials were evaluated using the Cochrane Risk of Bias tool for randomized trials (Risk of Bias 2 tool [RoB 2])39 and, when applicable, the version with additional considerations for cluster-randomized trials.40 Non-randomized studies were assessed using the Risk of Bias in Non-Randomized Studies of Interventions (ROBINS-I) tool.41 Disagreements were resolved through discussion or by a third reviewer (B.N.). For mortality outcomes, risk ratios or odds ratios (ORs) with corresponding 95% confidence intervals (CIs) were extracted as reported. For continuous variables, mean differences or medians with interquartile ranges (IQR) were collected. Data were tabulated and analyzed descriptively using Microsoft Excel 365 (Version 16.101.3; Microsoft Corporation, Redmond, WA, USA). A quantitative meta-analysis was planned using a random-effects model when studies were sufficiently homogeneous in design, population, intervention, and outcome reporting. Pooled effect estimates were to include risk ratios, ORs, or mean differences with 95% CIs. Statistical significance was defined as a two-tailed p-value <.05. When meta-analysis was not feasible, a structured narrative synthesis was conducted following the Cochrane Handbook for Systematic Reviews of Interventions42 and Synthesis Without Meta-Analysis (SWiM) reporting guidelines.43 Findings were summarized by outcome domain (mortality, transfusion requirements, and safety) and time interval (≤6 h, 24 h, and 28–30 days). The direction and magnitude of effects were described, and consistency and certainty of evidence were evaluated. Publication bias was assessed using funnel plot asymmetry and Egger's test when at least 10 studies with comparable outcomes were available.44, 45 The certainty of evidence for each outcome was evaluated using the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework.46 Each outcome was rated as high, moderate, low, or very low certainty based on risk of bias, inconsistency, indirectness, imprecision, and publication bias. Database searches identified 5482 records. After removing 2910 duplicates, 2572 unique records were screened, and 98 full-text articles were assessed for eligibility. A total of seven studies met inclusion criteria, comprising 6954 patients.47-53 The study selection process is illustrated in Figure 1. Characteristics of included studies are summarized in Table 1. All included studies were conducted in the United States and evaluated low-titer group O whole blood (LTOWB) use in civilian prehospital programs or EMS-linked trauma systems (2020–2025). The included designs consisted of a pragmatic cluster-randomized pilot trial50 and six observational cohort studies.47-49, 51-53 Comparator strategies included CT47, 49, 50, 52, 53 or crystalloid-based resuscitation approaches.48, 51 Across included studies, transfusion strategies varied. Some programs administered WB only during the prehospital phase, while others initiated WB prehospital with continuation after emergency department (ED) arrival, or evaluated early WB administered as part of a combined prehospital and ED pathway. In one multicenter study, WB was administered prehospital at a single site and initiated early in-hospital at other participating centers.52 Given heterogeneity in study designs, populations, and outcome measures, a meta-analysis was not performed; instead, a structured narrative synthesis was conducted in accordance with SWiM reporting guidelines.42, 43 Risk-of-bias assessments are presented in Tables 2 and 3. Among observational studies, overall risk ranged from moderate to serious, largely due to confounding, selection bias, and inconsistent covariate adjustment.47-49, 51-53 The cluster-randomized trial showed some concerns regarding randomization and participant enrollment but demonstrated low risk for missing data and outcome assessment.50 Mortality and survival outcomes are summarized in Table 4. The study by Williams et al. analyzed 350 patients (198 LTOWB vs. 152 CT); 30-day survival was 73% versus 74% (p = .81), with an adjusted odds ratio (aOR) 2.19 (95% CI 1.01–4.76; p = .047); the authors reported this adjusted association favoring LTOWB despite similar unadjusted survival.47 Braverman et al. included 538 patients, 214 of whom were propensity-matched (58 prehospital WB vs. 156 non-WB); unadjusted mortality was 0% versus 7.1% in the ED (p = .04), 5.3% versus 14.1% at 6 h (p = .08), 17.2% versus 23.1% at 24 h (p = .36), and 29.0% versus 34.8% at discharge (p = .25); none reached statistical significance after adjustment.48 Brill et al. evaluated 1377 patients (840 LTOWB vs. 537 CT) and reported 30-day survival of 75% versus 76%; unweighted OR 4.10 (95% CI 2.22–7.45; p < .001) and weighted OR 1.59 (95% CI 1.28–1.98; p < .001) both indicated improved adjusted survival with LTOWB.49 Guyette et al. studied 86 patients (40 LTOWB vs. 46 CT); mortality was 10.0% versus 10.9% at 3 h (p = .90), 12.5% versus 13.0% at 6 h (p = .94), 15.0% versus 17.4% at 24 h (p = .76), and 25.0% versus 26.1% at 28 days (p = .91), with no significant differences.50 Braverman et al. analyzed 1562 patients (171 prehospital WB vs. 1391 non-WB) and found mortality of 2.3% versus 4.5% in the ED (p = .18), 12.9% versus 16.1% at 24 h (p = .27), and 18.7% versus 23.7% in hospital (p = .14), with no adjusted survival difference.51 Sperry et al. conducted a multicenter prospective cohort of 1051 patients (624 LTOWB vs. 427 CT). After propensity adjustment, relative risks (RR) for mortality were 0.90 (95% CI 0.59–1.39) at 4 h, 1.08 (95% CI 0.77–1.52) at 24 h, and 1.10 (95% CI 0.83–1.47) at 28 days. In prespecified high-risk subgroups, LTOWB was associated with lower mortality: ≥5% predicted mortality, 4 h RR 0.52 (0.32–0.87); ≥10% predicted mortality, 24 h RR 0.67 (0.47–0.97); ≥20% predicted mortality, 28 days RR 0.70 (0.51–0.96); and ≥50% predicted mortality, 28 days RR 0.64 (0.45–0.93).52 Kostolni et al. analyzed 1990 prehospital transfusions (1515 LTOWB vs. 475 CT); adjusted aOR 1.68 (95% CI 0.58–4.89) for mortality showed no significant difference. Unadjusted mortality was versus 7.1% (p = in as reported by the study trauma mortality was versus (p = and medical mortality versus (p = and physiologic outcomes are summarized in Table 4. Findings by Guyette et al. indicated lower with LTOWB vs. 2 units at 6 h and vs. 2 units at 24 p < Brill et al. reduced 24-h component transfusion 95% CI p < .001) and lower total transfusion volume (p < Braverman et al. reported transfusion vs. p = and shock vs. p = Braverman et al. lower transfusion vs. p = Williams et al. found products vs. 3 p = .001) with aOR (95% CI p = Sperry et al. comparable total units at 24 h vs. 10 p = but total transfusion volume vs. p < Kostolni et al. reported differences in vs. p = and vs. p = outcomes are summarized in Table 4. No study reported an of transfusion reactions with Williams et al. two reactions in the component group and none LTOWB (p = Brill et al. no transfusion reactions and a lower of death in the LTOWB with in versus (p = Sperry et al. found no transfusion reactions and similar of infection vs. p = vs. p = vs. p and multiple failure vs. p = Kostolni et al. reported no transfusion reactions in the ED and LTOWB all Across all studies, no evidence of TRALI, TACO, or transfusion-transmitted infection associated with LTOWB was assessments are presented in Tables and low certainty for mortality outcomes and moderate certainty for transfusion and safety This systematic review found that prehospital administration of WB in adults with hemorrhagic shock is and demonstrates a safety across included Across studies, WB transfusion was associated with lower transfusion requirements compared with CT or crystalloid-based resuscitation. These suggest that early administration of WB and balanced resuscitation during the phase of shock reactions most to during prehospital transport intervals include allergic, and febrile non-hemolytic These were across included No included study demonstrated an risk of transfusion-related complications such as TACO, or acute lung that prehospital WB can be under structured The risk of of the and associated with LTOWB is low and be in the of the mortality risk associated with uncontrolled The in use the physiologic of which and platelets Early of and of effects are in the management of hemorrhagic 2 WB transfusion logistics by multiple with a single balanced time and during 16 Mortality outcomes across studies, with some reporting improved early survival following prehospital WB and others no significant difference compared with 53 Differences in strategies may have to in reported Studies WB with crystalloid-based resuscitation or no prehospital transfusion frequently reported physiologic or early outcome while with balanced CT demonstrated This may differences in study design, inclusion criteria, and such as transport of the seven included studies evaluated traumatic hemorrhage study included both traumatic and non-traumatic hemorrhage and the of Mortality following hemorrhagic shock is by multiple These include use of such as and effectiveness of hemorrhage trauma time to definitive surgical intervention, and ED to These were reported across included the use of WB in non-traumatic hemorrhagic shock remains very limited, and In this non-traumatic hemorrhage included and other medical were not reported with to and were not to differences between traumatic and non-traumatic non-traumatic hemorrhage medical with and resuscitation with definitive direct comparison with traumatic hemorrhage and may the of prehospital studies have demonstrated between transfusion and mortality, while earlier initiation of hemostatic resuscitation has associated with improved physiologic and early survival.24-26 In the transport time was reported across included studies and not be assessed based on transport is derived primarily from studies evaluating prehospital plasma transfusion and rather than from the between transport time and the effectiveness of prehospital WB transfusion be from the available data that prehospital WB transfusion is and when within systems of and effectiveness remains and is by that were across studies evaluating prehospital plasma or red blood cell transfusion reported inconsistent survival The RePHILL trial patients with preserved blood of initiation of transfusion, and prehospital The primary outcome included which may be by such as strategies, and These characteristics limit with paramedic-based EMS systems patients with hemorrhagic These with the of this which that system structure, and transport time may the of across This review to methodological and was prospectively It a peer-reviewed search strategy across multiple databases and included and data Risk of bias and certainty of evidence were evaluated using methodological and Several be Most included studies were related to injury and The number of studies and in design, population, and outcome reporting the of one study, all hemorrhage, to non-traumatic causes of The of data from the United States may limit to other prehospital were but none of the included studies identified significant safety prospective civilian evaluating LTOWB as the primary prehospital resuscitative fluid Most published evidence of observational studies, pilot trials, or prehospital and ED data for prehospital transfusion are not in the United in electronic of shock and transfusion to trauma and completeness of to inconsistent reporting across The available evidence the of prehospital WB programs within systems to and WB resuscitation by a balanced transfusion product and may facilitate earlier initiation of hemostatic resuscitation when compared with or crystalloid The timing and feasibility of early administration on and between prehospital and While timing of administration relevant to transport was reported in existing studies, and effect be outcomes following prehospital WB transfusion are within the of including early recognition of hemorrhagic shock, between EMS and trauma and timely access to definitive hemorrhage to which patient populations the from prehospital WB transfusion, as well as the timing of administration and that randomized controlled trials, including the Study of in Trauma and the O and of Prehospital are to evidence on clinical and to include non-traumatic causes of hemorrhage, such as obstetric and bleeding, the of WB transfusion in prehospital care. Prehospital WB transfusion is and demonstrates a safety across the included Mortality outcomes were and similar to CT or crystalloid-based resuscitation, while transfusion requirements were lower and no in adverse was reported. the certainty of evidence remains limited, the available data that WB may be a practical for early, balanced resuscitation in the prehospital when within systems of and The as the during the search strategy The authors have no of The data that the of this study are available from the corresponding author upon Data The is not for the or of information by the than missing be to the corresponding author for the