Prehospital factors associated with mortality in injured air medical patients.

BACKGROUND: Air medical transport provides rapid transport to definitive care. Overtriage and the expense and risk of transport may offset survival benefits.

OBJECTIVE: We assessed the ability of prehospital factors to predict resource need for helicopter-transported patients.

METHODS: We performed a prospective, observational cohort analysis of injured scene patients taken to one of two level I trauma centers from October 2009 to September 2010. Variables analyzed included patient demographics, diagnoses, and clinical outcomes (in-hospital mortality, emergent surgery within 24 hours, blood transfusion within 24 hours, and intensive care unit [ICU] admission ≥24 hours, as well as a combined outcome of all clinical outcomes). Prehospital variables were prospectively obtained from air medical providers at the time of transport and included past medical history, mechanism of injury, and clinical factors. We compared those variables with and without the outcomes of interest via χ(2) analysis and the Kruskal-Wallis test, where appropriate. Multivariate logistic regression identified factors associated with outcomes of interest with the intent of developing a clinical prediction tool.

RESULTS: Five hundred fifty-seven patients were transported during the study period. The majority of the patients were male (67%) and white (95%) and had an injury that occurred in a rural location (58%). Most injuries were blunt (97%), and patients had a median Injury Severity Score (ISS) of 9. The overall mortality was 4%; 48% of the patients had one of the four outcomes. The most common reasons for requesting air transport were motor vehicle collision (MVC) with high-risk mechanism (18%), MVC at a speed greater than 20 mph (18%), Glasgow Coma Scale score (GCS) less than 14 (15%), and loss of consciousness (LOC) greater than 5 minutes (15%). Factors associated with mortality were age greater than 44 years, GCS less than 14, systolic blood pressure (SBP) less than 90 mmHg, and flail chest. This model had 100% sensitivity and 50% specificity and missed no deaths. The combined endpoint of all four outcomes (death, receipt of blood, surgery, ICU admission) included intubation by emergency medical services, two or more fractures of the humerus/femur, presence of a neurovascular injury, a crush injury to the head, failure to localize to pain on examination, GCS less than 14, or the presence of a penetrating head injury. This model had a sensitivity of 57% (53%-61%) and a specificity of 78% (75%-87%).

CONCLUSIONS: Very few prehospital criteria were associated with clinically important outcomes in helicopter-transported patients. Evidence-based guidelines for the most appropriate utilization of air medical transport need to be further evaluated and developed for injured patients.