Prediction of blunt trauma outcomes with pulmonary contusion quantification

To assess for associations between pulmonary contusion volume and complications of thoracic trauma.

Consecutive patients were retrospectively identified as sustaining thoracic trauma in a single tertiary level 1 trauma center from January 1, 2013 to October 31, 2017. Individuals ≥ 16 years old presenting with blunt thoracic trauma who underwent computed tomography (CT) within 24 hours of admission and Injury Severity Score (ISS) ≥ 15 were included. Exclusion criteria were penetrating injury, death before CT, intracranial injury, or unavailable imaging. A body fellowship-trained radiologist with more than 12 years of clinical experience manually measured pulmonary contusion, blinded to clinical data using the three largest orthogonal dimensions in axial and sagittal planes. Outcomes included mortality, severe hypoxemia, and ventilator-associated pneumonia (VAP). Logistic regression was used to evaluate the association between outcome variables and pulmonary contusion volume.

Of the 1,116 patients, 153 were included in the study. The number of pulmonary lobes injured on CT was associated with severe hypoxemia (p= 0.0047). For every 100 mL increase in contusion volume, the odds of death increased by 20% (p = 0.04), and the odds of severe hypoxemia increased by 21% prior to adjustment for ISS (p = 0.015).

Pulmonary contusion volume was associated with severe hypoxemia prior to adjustment for ISS and mortality. Flail chest, the number of pulmonary lobes injured, and the number of ribs fractured were associated with mortality, severe hypoxemia, and VAP, respectively.