Prehospital fresh frozen plasma: Universal life saver or treatment in search of a target population?

Recently within 2 days of each other, two of the leading medical journals published trials on the use of prehospital fresh frozen plasma (FFP) for trauma patients, reaching apparently different conclusions. In the New England Journal of Medicine article reporting on the Prehospital Air Medical Plasma (PAMPer) trial, Sperry et al1 found that two units of prehospital FFP was associated with an almost 10% survival advantage. In the Control of Major Bleeding After Trauma Trial (COMBAT) reported in the Lancet, Moore et al2 found that the same volume of plasma had no survival advantage. The concept of prehospital plasma is in theory attractive. You have an injured individual who is likely to need plasma in the next few hours, so why not preemptively reduce the bleeding by administering plasma first? An important issue to consider, however, is that at the injury scene the patient is not likely to be deficient in clotting factors yet. The prehospital transfusion of blood products has been shown to improve survival in US military combat casualties injured in Afghanistan.3 The normal range for clotting factors measured as percent of normal is approximately 50150 with a mean of 100. It follows, that the average individual can lose half their plasma volume and still have clotting factor levels in the normal range even after reconstitution of blood volume. A major reduction in clotting factors can occur in two relevant settings, firstly when there is marked hemodilution such as that following major fluid resuscitation and secondly due to consumption in the presence of disseminated intravascular coagulation (DIC). With the exception of obstetric DIC at the time of delivery, this consumption is not usually so rapid to become a major issue in the prehospital management of trauma. Fresh frozen plasma is obtained from whole blood donation or plasmapheresis. While in the COMBAT trial they specify that they gave two units of approximately 250 mL each of FFP, the PAMPer investigators do not specify the volume of the two units they administered. The PAMPer FFP was prethawed and could be up to 5 days since thawing, while the COMBAT product was collected by plasmapheresis, frozen within 24 hours and thawed rapidly in special equipment before administration. In a study one of us was involved in, measuring clotting factor levels before and after administration of four units of optimally thawed FFP just before infusion, the increase in clotting factor levels was 9%14%.4 Based on the volume of FFP infused in the PAMPer and COMBAT studies, the increase in factor levels would be 7% at best. This suggests that if prehospital FFP provides benefit, it is unlikely that this is through the substitution of coagulation factors. Other modes of FFP benefit include protecting the endothelial glycocalyx and reducing vascular permeability and inflammation.5,6 We compared the PAMPer and COMBAT trials looking for differences in their design and conduct to shed light on the true effect of FFP, if any (Table 1). Of course, the simplest explanation for the difference in results could well be a combination of chance and low power for the COMBAT study, due to low event rate and small sample size. The contribution of the COMBAT trial under a fixed effect metaanalysis approach would have been <10%, and the pooled estimates, driven by the larger PAMPer study, would have shown a significant benefit for plasma administration at both 24 hours (OR = 0.64, 95% CI 0.420.98) and 1 month (OR = 0.67, 95% CI 0.460.98). Even under the more conservative randomeffect approach, the visual inspection of the forest plot (Figure 1) would support both the hypothesis the two trials observing a different effect (as their effect sizes lie on opposite sides of the identity line) and instead representing random variation of the same effect (as the confidence intervals do overlap). However, there are clear differences between the trials, starting from the choice of results to report and the modality used to report them. The entry criteria for the two trials were the same in terms of blood pressure and heart rate measurements, but