EDITORIAL

Prehospital Whole Blood for Traumatic Hemorrhage — Consistent Evidence from Two Randomized Trial

Susan E. Rowell, Selwyn O. Rogers, Jr.

N Engl J Med 2026;394:2372-2373

DOI: 10.1056/NEJMe2605735

Hemorrhagic shock remains the leading cause of preventable death after injury.¹ Over the past two decades, early administration of blood products has transformed care for patients with bleeding. The Pragmatic, Randomized Optimal Platelet and Plasma Ratios (PROPPR) trial²established that among patients with severe trauma and bleeding, the use of balanced blood-component therapy (i.e., transfusing plasma, platelets, and red cells in equal ratios) early after hospital arrival improved hemostasis and reduced the risk of death from early hemorrhage. Subsequently, prehospital transfusion of plasma showed a survival benefit in a multicenter trial of air medical transport services but not in a single-center trial of rapid ground transportation.^3,4

The use of whole blood has gained interest as a simpler way to deliver balanced hemostatic resuscitation. In this issue of the Journal, Smith et al.⁵ report the results of the Study of Whole Blood in Frontline Trauma (SWiFT) trial, and Sperry et al.⁶ report the findings of the Type O Whole Blood and Assessment of Age during Prehospital Resuscitation (TOWAR) trial. These two phase 3, randomized trials provide the most rigorous evidence to date with regard to the use of prehospital whole blood in trauma patients with bleeding. Neither trial showed a difference in survival.

Both trials involved trauma patients with bleeding who underwent prehospital transfusion of up to 2 units of leukoreduced whole blood or component therapy (plasma, red cells, or both). The SWiFT trial, which was conducted across air ambulance services in England, showed that the incidence of death and massive transfusion at 24 hours appeared to be similar with whole blood and component therapy. The TOWAR trial, which was conducted across air medical bases in the United States and Canada, similarly showed that 30-day mortality did not differ significantly between the whole-blood and component groups. It is important to note that an observational substudy of the TOWAR trial showed that the storage age of whole blood through 21 days from donation was not associated with apparent differences in outcomes.

These results should not be misread as evidence that the use of whole blood is a failed resuscitation strategy. Rather, they reflect how dramatically prehospital care of patients with hemorrhage has improved. The 24 to 28% 30-day mortality observed in both trials is meaningfully lower than that reported in earlier trials conducted before widespread tourniquet use, training for hemorrhage control, damage-control resuscitation, and adoption of prehospital blood-transfusion programs. Because the baseline mortality in both trials was lower than that observed in earlier prehospital trials, including the PAMPer trial,³ any true mortality difference between the treatment groups was most likely smaller than the trials were powered to detect. Neither the SWiFT trial nor the TOWAR trial was designed as an equivalence trial, and their confidence intervals do not exclude clinically meaningful differences in either direction.

The similarity of the outcomes in the two trials is notable because their designs were so different. They tested different whole-blood products against different comparison groups using different randomization strategies, and they measured different primary outcomes in patients enrolled on two continents. However, the investigators in both trials reached the same conclusion — that the administration of 1 or 2 units of cold-stored, leukoreduced whole blood in the prehospital setting produced outcomes indistinguishable from those with component therapy. This consistency suggests that cold-stored whole blood does not produce a measurable survival benefit at the doses tested.

To better understand this finding, it is important to understand what cold-stored, leukoreduced whole blood is. It is not the warm fresh whole blood that was associated with survival benefits in military settings, which has platelets that are fully functional and coagulation factors that have full potency.⁷ Leukoreduction reduces the platelet count and impairs platelet function. Storage degrades coagulation factor activity and platelet hemostatic capacity over time.⁸ However, cold-stored, leukoreduced whole blood is not a single product. The SWiFT trial used a platelet-sparing filter and produced uniform, platelet-replete units across all sites, whereas the TOWAR trial deferred to local blood-banking practices and tested a heterogeneous mix of conventionally leukoreduced and nonleukoreduced units. Nonetheless, the results of both trials reflect the real-world performance of a stored, processed product, and the nonsuperiority observed reflects practical clinical utility.

Two other trial characteristics — how much whole blood was given and to whom — are noteworthy. Protocol-indicated transfusion was capped at 2 units, a volume that is possibly too small to show benefit. A secondary analysis of the PROPPR trial showed that early delivery of platelets was associated with improved survival.⁹ Thus, whether 1 or 2 units of cold-stored whole blood delivered enough platelet activity is unclear. Whole blood is fundamentally a hemostatic therapy, and its benefit should be greatest in patients dying from hemorrhage. In the TOWAR trial, hemorrhage was reported as a cause of death in 9.0% of the patients in the whole-blood group and 7.8% of those in the component group, and brain injury was reported as a cause of death in 9.0% and 7.8%, respectively. In the SWiFT trial (in which causes of death were reported as percentages of deaths rather than as percentages of the total number of patients in each group), 26.4% of the deaths in the whole-blood group and 37.8% of those in the standard-care group were due to traumatic brain injury, and 11.5% and 10.8%, respectively, were due to uncontrolled bleeding. Thus, it is plausible that neither trial included the patients most likely to benefit from the use of whole blood. Whether higher doses or a more bleeding-enriched cohort would change the outcome remains unknown.

The SWiFT and TOWAR trials show us that whole blood and component therapy appear to perform similarly at the doses studied, so blood-product selection should be guided by availability, capacity, cost, and operational realities. Neither trial identified a safety concern with either approach, so prehospital systems with established blood programs can safely continue to use either whole blood or component therapy. These trials also show us that up to 2 units of whole blood does not outperform component therapy. Whether one is better than the other in specific patient groups remains unknown.