Objective: A retrospective study was performed to assess CT sensitivity for diagnosing tracheal rupture. Intubated cadaver tracheas were examined to assess endotracheal tube balloon overdistention and deformity and to evaluate the relationship of balloon pressures to tracheal injury.
Materials and methods: Neck or chest CT scans of 14 patients with tracheal rupture and 41 control trauma patients with pneumomediastinum but without tracheal injury were reviewed and compared to assess the presence and location of extrapulmonary air, whether direct visualization of tracheal wall disruption was possible, the size and shape of endotracheal tube balloon, signs of transtracheal balloon herniation in intubated patients, and the location of the extratracheal endotracheal tube. Intact and experimentally injured cadaver tracheas were used to evaluate tube balloon pressure and configuration.
Results: All 14 patients with tracheal rupture had deep cervical air and pneumomediastinum. Overdistention of the tube balloon occurred in 71% (5/7) of the intubated patients, and balloon herniation occurred in 29% (2/7). Direct tracheal injury was seen in 71% (10/14) of the patients as a wall defect (n = 8) or deformity (n = 2). Overall, CT was 85% sensitive for detecting tracheal injury. Patients with tracheal injury had a significantly lower incidence of pneumothorax (p = 0.01) than did the control group. The CT appearance of balloon herniation through defects in the cadaver tracheas closely mimicked those of patients with tracheal injury. The amount of balloon pressure required to rupture the intubated trachea was extremely high and rupture was difficult to obtain.
Conclusion: CT can reveal tracheal injury and can be used to select trauma patients with pneumomediastinum for bronchoscopy, leading to early confirmation and treatment.