Functional and survival outcomes in traumatic blunt thoracic aortic injuries: An analysis of the National Trauma Databank

Abstract

Objective: Blunt thoracic aortic injury (BAI) remains a leading cause of trauma deaths, and off-label use of endovascular devices has been increasingly utilized in an effort to reduce the morbidity and mortality in this population. Utilizing a nationwide database, we determined the incidence of BAI, and analyzed both functional and survival outcomes at discharge compared with matched controls.

Methods: Patients with BAI were identified by International Classification of Disease-9 codes from the National Trauma Data Bank (Version 6.2), 2000-2005. Patients were analyzed based on aortic repair, associated physiologic burden, and coexisting injuries. Control groups were matched by age, mechanism, major thoracic Abbreviated Injury Scale score (AIS >/= 3), major head AIS, and major abdominal AIS. Outcomes were assessed using the functional independence measure (FIM) score and overall mortality. FIM scores were scored from 1 (full assistance required) to 4 (fully independent) for three categories: feeding, locomotion, and expression.

Results: During the study period, 3,114 patients with BAI were identified among 1.1 million trauma admissions for an overall incidence of 0.3%. One hundred thirteen (4%) were dead on arrival, and 599 (19%) died during triage. Of the patients surviving transport and triage (n = 2402), 29% had a concomitant major abdominal injury and 31% had a major head injury. Sixty-eight percent (1,642) underwent no repair, 28% (665) open aortic repair, and 4% (95) endovascular repair with associated mortality rates of 65%, 19%, and 18%, respectively (P < .05). Aortic repair independently improved survival when controlling for associated injuries and physiologic burden (odds ratio (OR) = 0.36; 95% confidence interval (CI), 0.24-0.54, P < .05). Compared with matched controls, BAI resulted in a higher mortality (55% vs. 15%, P < .05), and independently contributed to mortality (OR = 4.04; 95% CI, 3.53-4.63, P < .05). In addition, BAI patients were less likely to be fully independent for feeding (72% vs. 82%, P < .05), locomotion (33% vs. 55%, P < .05), and expression (80% vs 88%, P < .05).

Conclusion: This manuscript is the first to define the incidence of BAI utilizing the NTDB. Remarkably, two-thirds of patients are unable to undergo attempts at aortic repair, which portends a poor prognosis. When controlling for associated injuries, blunt aortic injury independently impacts survival and results in poor function in those surviving to discharge.