Model to Predict Duration of Ventilation and 30-Day Mortality in Patients with Traumatic Injuries.

摘要

Learning Objectives: Scoring systems are widely used in civilian practice to predict intensive care patient outcomes. Tools to predict outcomes from combat trauma are not available; specifically prediction tools for aeromedical evacuation from theater. The objective was to identify pre-flight variables associated with an increased requirement for ventilation time and 30-day mortality. Methods: This IRB-approved, retrospective cohort study included all patients evacuated from combat theaters by the U.S. Air Force Critical Care Air Transport Team (CCATT) between 2007 and 2011. Pre-flight physiological variables and treatments were assessed for associations and examined in stepwise regression models. The primary outcome was total ventilator time (72 hours); the secondary outcome was 30-day mortality. Receiver operating characteristic (ROC) curves were produced for both outcomes. Data are presented as percentages, median [IQR], and odds ratio (OR [95% CI]). Results: 1308 combat trauma patients (24 years, 98% male) were included: 72% blast, 17% penetrating, 9% blunt, and 2% burns. Pre-flight systolic blood pressure was 121 [109-143] mmHg, pulse 100 [84-116] bpm, and base deficit 0 [-2-2]. The median number of blood products administered pre-flight were 4 [0-13] units packed red blood cells (PRBC), and 3 [0-12] units fresh frozen plasma. When modeling for ventilator time, injury severity score (ISS) (OR 1.04 [1.03-1.06]), pre-flight PRBC units transfused (OR 1.05 [1.04-1.07]), and pre-flight intubated status (OR 11.9 [8.53-16.89]) were independently associated with increased ventilator days. A composite of the variables produced an AUC of 0.85 with 86% sensitivity and 56% specificity. Using mortality as the outcome, ISS (OR 1.06 [1.03-1.09]), prothrombin time (OR 2.13 [1.18-4.47]), pre-flight intubated status (OR 9.2 [1.88-166.11]), and whole blood (OR 3.18 [1.38-7.04]) were associated with death.