Journal Watch: Blood Products vs. Crystalloid for Resuscitation

Reviewed This Month

Prehospital Blood Product and Crystalloid Resuscitation in the Severely Injured Patient: A Secondary Analysis of the Prehospital Air Medical Plasma Trial

Authors: Guyette FX, Sperry JL, Peitzman AB, et al.      

Published in: Ann Surg, 2021 Feb 1; 273(2): 358–64 

Hemorrhage is the leading cause of preventable death following injury. Historically, crystalloid fluids were our primary method of treating patients suffering significant loss of blood. However, the Prehospital Air Medical Plasma (PAMPer) trial found that prehospital administration of plasma during resuscitation reduced 30-day mortality of severely injured patients by almost 10%. Data from other studies have shown packed red blood cells (PRBCs), plasma, and platelets also can improve outcomes for patients at risk for hemorrhagic shock.

The authors of this month’s study point out the optimal prehospital blood product for resuscitation of hemorrhagic shock from trauma is unknown. Their study objective was to determine whether prehospital blood product resuscitation reduced 30-day mortality in patients at risk for hemorrhagic shock compared with crystalloid-only resuscitation. Their hypothesis was that either PRBCs or plasma would be associated with lower mortality than crystalloid resuscitation alone. Their secondary hypothesis was that the combination of PRBCs and plasma would be associated with the greatest reduction.

PAMPer Trial

To test their hypotheses, the authors performed a secondary analysis of data collected during the PAMPer trial. This was a pragmatic, multicenter cluster-randomized trial involving patients at risk for hemorrhagic shock who were transported by civilian helicopter emergency medical services (HEMS). Each of the 27 HEMS bases that participated in the trial was randomly selected to carry plasma for one-month blocks. During control months, patients cared for by HEMS from half the bases would receive crystalloid only, and patients at the remaining bases would receive PRBCs with or without additional crystalloid. During intervention months, patients cared for by HEMS from half the bases would receive plasma with or without additional crystalloid, and patients at the remaining bases would receive plasma and PRBCs with or without additional crystalloid.

Inclusion criteria required that patients be transported to one of nine participating trauma centers. Further, patients were only enrolled if they had an SBP of 70–80 mm Hg and a heart rate of 108 bpm or more. Patients were also enrolled if they had an SBP less than 70 mm Hg, regardless of heart rate, at any time before arrival at the trauma center.

Filling in the Blanks

Missing data is an obstacle most studies deal with in some way. The authors noted that 30-day mortality, injury severity scores, and head-abbreviated injury scores were missing in 6%, 2%, and 2% of cases, respectively. To address this missing data, the authors chose to perform multiple imputation. Imputation is when a missing value is replaced with a substituted value. 

Now, rather than simply impute an average or make up a number (both of which can introduce bias), the authors created multiple data sets, each with different imputed values based on available data. They then analyzed each of these data sets and pooled the results for analysis. This technique has been validated repeatedly in scientific literature. 

In addition, the authors performed a complete-case analysis for comparison. In other words, they removed any cases with missing data prior to performing their analysis. They indicated that similar results were obtained between the complete-case and imputed data sets and therefore only reported the results obtained following multiple imputation. 

They also performed Cox proportional hazards regression analyses. This is a method for investigating the effect several variables have on the time an event of interest (in this study 30-day mortality) takes to happen. The statistic produced is called a hazard ratio. A hazard ratio of 1 means at any time during follow-up, the rates of the event of interest were the same in both groups. A hazard ratio less than 1 means at any time during follow-up, fewer patients in the treatment group experienced the event of interest than in the control group. Finally, a hazard ratio greater than 1 means more patients in the treatment group experienced the event of interest than in the control group. 

Results

There were 407 patients included in the analysis; 34% received crystalloid only, 20% received PRBCs, 36% received plasma, and 10% received PRBCs and plasma. Regression analyses showed all prehospital blood product groups performed significantly better with respect to 30-day mortality compared to crystalloid-only resuscitation. The PRBC-plus-plasma group had the greatest mortality benefit, with a 62% reduction, followed by the plasma group with 43% and the PRBC group with 32%. There was no significant difference in cause of death across groups (p=0.180).

There was a significant reduction in the hazard of 30-day mortality per unit of prehospital PRBCs (hazard ratio=0.69; 95% CI 0.52–0.92; p=0.009) and per unit of prehospital plasma (HR=0.68; 95% CI 0.54–0.88; p=0.003) transfused. There was not a significant difference between the dose of crystalloid and mortality (HR 1.20; 95% CI 0.81–1.77, p=0.372).

When evaluating only patients who received a prehospital blood product, each liter of crystalloid was associated with a 65% increase in the hazard of 30-day mortality (HR=1.65; 95% CI 1.17– 2.32; p=0.004). However, among patients who received crystalloid only, the volume of crystalloid was not associated with 30-day mortality (HR=0.92; 95% CI 0.57–1.50; p=0.748).

Every study has limitations. This study was a secondary analysis, had some missing data, and some of the treatment groups had small numbers of patients. Further, the prehospital blood product groups all received some crystalloid. Nevertheless, these data show that any blood product resuscitation was associated with lower mortality than crystalloid-only resuscitation. Compared with crystalloid-only resuscitation, patients who received prehospital PRBCs and plasma had the greatest mortality benefit, followed by plasma, then PRBCs.  

Antonio R. Fernandez, PhD, NRP, FAHA, is a research scientist at ESO and serves on the board of advisors of the Prehospital Care Research Forum at UCLA.