Jason L. Sperry, Francis X. Guyette, Bryan A. Cotton, James F. Luther, Richard B. Utarnachitt, Matthew E. Kutcher, Brian J. Daley, Allan B. Peetz, Mayur B. Patel, Michael D. Goodman, Jeffrey A. Claridge, Nimitt Patel, Brian G. Harbrecht, Zain G. Hashmi, Ryan Zarychanski, Matthew D. Neal, Mark H. Yazer, Christian Martin‑Gill, Laura E. Vincent, Ashley M. Harner, David E. Meyer, Andrew J. Latimer, Bryce R. Robinson, Catherine L. McKnight, William R. Hinckley, Keith R. Miller, Jan O. Jansen, Douglas Martin, Erin E. Fox, Bedda L. Rosario‑Rivera
8 min
Traumatic hemorrhage remains a major cause of preventable death. While prehospital blood transfusion is associated with improved survival, it is unclear whether whole blood offers superior outcomes over traditional component-based resuscitation (red cells, plasma, or both). Furthermore, the impact of blood storage duration on patient outcomes in the prehospital setting remains poorly characterized.
The Type O Whole Blood and Assessment of Age during Prehospital Resuscitation (TOWAR) trial was a pragmatic, multicenter, phase 3, cluster-randomized trial. Researchers compared the efficacy and safety of up to 2 units of low-titer group O whole blood against standard-of-care blood components during air medical transport for trauma patients experiencing hemorrhagic shock. The primary endpoint was all-cause mortality within 30 days. Additionally, an observational substudy evaluated whether the storage age of whole blood (1 to 14 days vs. 15 to 21 days) influenced mortality.
The study enrolled 1020 patients. Analysis revealed that 30-day mortality was 25.9% in the whole-blood group compared to 20.5% in the component group, showing no statistically significant benefit for whole blood (adjusted odds ratio, 1.24; P = 0.24). Secondary outcomes, including the quantity of blood products transfused within 24 hours, organ failure, and infection rates, did not differ significantly between the groups. In the observational substudy, the storage age of whole blood was not associated with differences in 30-day mortality.
This trial provides high-level evidence regarding the role of whole blood in prehospital resuscitation. Although whole blood is operationally simpler to deploy in austere settings—such as military or far-forward environments—this study suggests that it does not confer a survival advantage over standard blood component therapy in civilian trauma scenarios. These findings are important for health systems refining their prehospital transfusion protocols.
BACKGROUND: Blood transfusion before arrival at a hospital reduces mortality from traumatic hemorrhage and shock. Whether transfusion with whole blood is more beneficial than transfusion with blood components is uncertain, as are the effects of the length of time that blood products are in storage between donation and transfusion. METHODS: In this pragmatic, multicenter, phase 3, cluster-randomized trial, we assigned 44 air medical bases in a 2:1 ratio to the use of up to 2 units of whole blood or as-indicated blood components (plasma, red cells, or both) for prehospital transfusion in trauma patients during 1-month blocks. The primary outcome was death from any cause within 30 days after randomization. An observational substudy assessed outcomes according to the storage age of whole blood. RESULTS: Of 1020 eligible patients transported to hospitals by the air bases, 715 were assigned to receive whole blood and 305 to receive blood components; 695 and 298, respectively, were included in the primary analysis. Mortality at 30 days was 25.9% in the whole-blood group and 20.5% in the component group (adjusted odds ratio, 1.24; 95% confidence interval [CI], 0.87 to 1.76; P = 0.24). No substantial between-group differences in adverse events were observed. In the observational substudy, 30-day mortality was 27.1% among 210 patients who received whole blood with a storage age of 15 to 21 days and 26.4% among 443 patients who received whole blood with a storage age of 1 to 14 days (adjusted odds ratio, 0.99; 95% CI, 0.74 to 1.32). CONCLUSIONS: In injured patients with hemorrhagic shock, the use of whole blood for prehospital transfusion did not result in lower 30-day mortality than the use of blood components. (Funded by the Defense Health Agency Research Technology Portfolio Management, Combat Casualty Portfolio; TOWAR ClinicalTrials.gov number, NCT04684719.).
Alex: Whole blood can mean different products. What exactly were they carrying?
Sam: Low-titer group O whole blood. That means group O blood with low levels of anti-A and anti-B antibodies. Apart from the assigned blood product, the trial did not change care before or after hospital arrival.
Alex: And who qualified? Someone bleeding at the scene could be quite different from someone transferred after initial treatment elsewhere.
Sam: Both routes were included, although most patients came directly from the injury scene. Eligibility required low blood pressure together with a fast heart rate, or very low blood pressure alone. Most injuries were blunt, and traumatic brain injury was common.
Alex: These were emergencies. How did the researchers handle consent and the possibility that knowing the treatment could affect assessment?
Sam: Enrollment used an approved exception from informed consent, with community consultation and later notification. Patients could withdraw their data. Crews knew which product they gave, but the personnel assessing outcomes did not know treatment assignments.
Alex: Let's get to the result. How large was the difference in deaths?
Sam: Deaths by thirty days occurred in about twenty-six percent of the whole-blood group, compared with about twenty-one percent of the component group. That numerical difference favored components, but the adjusted analysis did not establish a statistically significant difference.
Alex: The direction is opposite to the hypothesis. Should listeners take that as evidence that whole blood is worse?
Sam: Not as a demonstrated harm. The confidence interval allowed effects on either side of no difference. The supported conclusion is narrower: whole blood did not demonstrate lower mortality. Calling the treatments equivalent would also go beyond the evidence.
Alex: What makes that result credible, beyond the trial's size?
Sam: It was randomized across multiple centers, and survival status was available for nearly everyone. The analysis accounted for the base assignments and imbalances in brain injury and prehospital intubation. Different approaches to missing survival data gave similar results.
Alex: Did another outcome suggest a benefit that the main mortality comparison missed?
Sam: Earlier mortality and in-hospital mortality appeared similar. The study found no apparent differences in subsequent blood use, organ failure, hospital-acquired infection, or acute respiratory distress syndrome, a severe lung condition. Adverse events also appeared similar, without establishing identical safety.
Alex: Now the difficult part: were the groups actually receiving distinct treatments, or did real-world care blur the comparison?
Sam: That is the limitation I'd put first. Some patients assigned to whole blood also received components. Some assigned to components received whole blood. Transfusion before randomization was allowed, and the prehospital blood volume was relatively low compared with what patients received in hospital.
Alex: So this tested a practical early transfusion policy, not two completely separate treatments maintained throughout resuscitation.
Sam: That's how I'd read it. Other care continued normally, and treatments overlapped. The authors say those features could increase the chance of missing a real difference. They also note that many component patients received both plasma and red cells, making the approaches more alike.
Alex: The title also asks about blood storage. Was that another randomized comparison?
Sam: No. The storage-age substudy was observational and limited to patients assigned whole blood. Researchers compared older with fresher blood and used weighting to address measured differences between recipients. Storage through twenty-one days was not associated with apparent differences in mortality or other outcomes.
Alex: That sounds useful for supply planning, but weaker than assigning patients to different storage ages.
Sam: It is weaker evidence for causation. Adjustment cannot rule out unmeasured differences. The result offers reassurance within the storage period studied, not proof that storage age never matters. It also says nothing about blood stored beyond that period.
Alex: Where does that leave services for which separate components are difficult to carry?
Sam: The logistical case remains. The authors note that component-based resuscitation may not be feasible in austere, forward military settings. Whole blood reduces the need to manage multiple products outside a blood bank. Those advantages don't require demonstrated survival superiority to be operationally relevant.
Alex: Who should read the full study, and where should they start?
Sam: Trauma researchers and people choosing prehospital blood supplies should read the intervention assignment and protocol adherence sections first. They show what was intended and what patients actually received. Then read the primary outcome and the Discussion's limitations; military blood planners should also examine the storage-age substudy.
Alex: And what should everyone else carry away from this comparison?
Sam: Whole blood may simplify emergency transfusion, but this trial did not show that it saves more lives than separate components. Keep that distinction between practical convenience and proven survival benefit.
Alex: Thanks for listening.