Is Mechanism a Biological Variable?: A Secondary Analysis of the PROPPR Trial

The purpose of this study was to evaluate for differences in the baseline mortality rates of patients injured by different mechanisms, in the Pragmatic Randomized Optimal Platelet and Plasma Ratios (PROPPR) trial, and compare their responses to 2 resuscitation paradigms. Our hypothesis was there are differences between the blunt and penetrating trauma patients, with regard to baseline and effect size.

Previous research including clinical trials and basic science research suggests differences in mortality among patients injured by blunt or penetrating mechanisms, although differences between these 2 mechanisms-both baseline and effect size-are rarely considered explicitly. The objective of this analysis was to compare mortality and other clinical outcomes of trauma patients stratified specifically by injury mechanism and resuscitation strategy.

We performed a retrospective review of the PROPPR trial to assess for differences in mortality outcomes in patients with blunt or penetrating injuries who received a 1:1:1 or 1:1:2 resuscitation strategy. Our primary outcome was 24-hour mortality with additional endpoints at proximate (ie, 1 hour, 3 hours, and 6 hours) times post-arrival. A logistic regression model utilizing general estimating equations and adjusted for age, Injury Severity Score (ISS), and first documented pulse and Glasgow Coma Scale (GCS) score were used to assess the interaction of mortality outcomes by resuscitation type and injury mechanism. Secondary outcomes evaluated include acute kidney injury, ventilator-associated pneumonia, cardiac arrest, symptomatic and asymptomatic pulmonary embolism, deep vein thrombosis, acute respiratory distress syndrome, and stroke. Additional nonmortality outcomes of interest included total hospital and ventilator- and ICU-free days, time to hemostasis, time to exsanguination, and time to death.

The original trial enrolled 680 patients, 338 (49.7%) received 1:1:1 and 342 (50.3%) 1:1:2 resuscitation. 8 patients had combined blunt and penetrating injuries and were excluded from this analysis, leaving 672 patients with blunt (350, 52.1%) and penetrating (322, 47.9%) injuries. Compared to penetrating injuries, patients with blunt injuries were older, more likely to be white, had a higher rate of air transfers, longer transport time and longer time to hemostasis, lower GCS Score, and higher ISS and R time on thromboelastography (P < 0.001). Overall mortality between blunt and penetrating injuries was similar at 1 hour (2.6% vs 4.0%, P = 0.286) and 3 hours (7.4% and 8.1%, P = 0.754). However, mortality in both groups steadily increased overtime, and more markedly at 24 hours for patients with blunt compared to penetrating injuries (16.9% and 11.8%, P = 0.063). When comparing resuscitation strategies, receipt of a 1:1:1 resuscitation significantly decreased the odds of mortality among blunt-injured patients who received a 1:1:1 transfusion (relative to blunt-injured patients receiving a 1:1:2 transfusion) at 3 hours (odds ratio [OR]: 0.26, 95% confidence interval [CI]: 0.10-0.66, P = 0.005) and 6 hours (OR: 0.38, 95% CI: 0.19-0.77, P = 0.0007). We observed no statistically significant differences in patients with penetrating injuries or at any time other points.

We found a significant difference in adjusted mortality at 3 hours in patients with blunt injuries when comparing resuscitation strategies, which was consistent with previous studies. Responses to resuscitation may differ depending on the mechanism of injury, and some interventions may be more beneficial depending on injury type.