Risk factors for post-traumatic pneumonia in patients with retained haemothorax: results of a prospective, observational AAST study

Introduction: Retained haemothorax (RH) is a problematic sequela of thoracic trauma, reported in up to 20% of patients following chest injury. RH is associated with a higher severity of thoracic trauma and may portend the onset of other serious post-traumatic complications, including pneumonia. The development of pneumonia has previously been reported to be as high as 19.5% in the setting of traumatic RH. The purpose of this study was to identify risk factors for the development of pneumonia as a complication in RH.

Methods: We utilized the American Association for the Surgery of Trauma Post-Traumatic Retained Haemothorax database. Patients with post-traumatic RH were prospectively enrolled from 2009 to 2011. Inclusion criteria were placement of a thoracostomy tube within 24h of admission for the evacuation of pneumothorax or haemothorax and subsequent chest computed tomography scan chest showing RH. Patients treated with thoracotomy before placement of tube thoracostomy were excluded. For univariate analysis, the Chi-square test with Yates correction was used for comparison of categorical risk factors and the Student's t-test or the Mann-Whitney test for comparison of continuous risk factors. To identify independent risk factors for the development of pneumonia, variables from the univariate analysis significant at p<0.2 were entered into a forward logistic regression model. Adjusted odds ratio and 95% confidence intervals (CI) were derived.

Results: 328 patients with post-traumatic RH from 20 United States centres were enrolled. After stepwise regression analysis, ISS>25 (adjusted OR: 7.1; 95% CI: 3.1, 16.4; p<0.001), blunt mechanism of injury (adjusted OR: 3.5; 95% CI: 1.7, 7.2; p=0.001), and failure to administer peri-procedural antibiotics on the initial thoracostomy tube placement (adjusted OR: 2.6; 95% CI: 1.30, 5.4; p=0.01) were found to be independent predictors of the pneumonia in patients with post-traumatic RH.

Conclusions: To our knowledge, our current study is the largest attempt to identify the independent predictors for pneumonia in this population. Our data show that elevated ISS, blunt thoracic trauma, and failure to administer peri-procedural antibiotics on tube thoracostomy placement are the statistically significant independent risk factors.