Blunt pulmonary contusion: admission computed tomography scan predicts mechanical ventilation.

Background: Blunt pulmonary contusion (BPC) evolves over 12 to 24 hours, and the initial plain radiographs fail to reliably identify patients at risk of clinical deterioration. Admission computed tomography (CT) may offer accurate characterization of BPC and early prediction of the need for mechanical ventilation (MV). Methods: This was a combination of a 5-year retrospective study (January 2002 to April 2007) and a 6-month prospective study (September 2007 to February 2008) of adult blunt trauma (BT) patients with thoracic injuries and a chest CT upon hospital arrival. The primary outcome was MV due to thoracic trauma. To ensure that MV was required for BPC and not for associated injuries, all patients with significant extrathoracic injuries (Abbreviated Injury Scale score >2) were excluded. The extent of BPC was measured by two scoring systems. Results: Of 392 patients (67% males; age, 48 years +/- 21 years; Abbreviated Injury Scale score chest, 3 +/- 1; and Injury Severity Score [ISS], 13 +/- 6), 243 (62%) had BPC. Twenty-five (6%) patients required MV and two (0.5%) died. The combination of Glasgow Coma Scale (GCS) score <14, BPC score >2, and >4 ribs fractured predicted MV in 100% of the cases, and the absence of all factors precluded MV in 100%. In the prospective period of 6 months, 55 patients had BPC and we confirmed our finding that the absence of the three factors precludes MV. Conclusion: A simple score derived by the initial chest CT, in combination with GCS and the number of fractured ribs, can predict the need for MV early. In the presence of these predictors, patients should be admitted to a high level of monitoring.