Is ventilator-associated pneumonia in trauma patients an epiphenomenon or a cause of death?

BACKGROUND: Ventilator-associated pneumonia (VAP) is a common infection among patients in trauma intensive care units (ICUs). It has been suggested by different investigators that VAP is an indicator of injury severity and not necessarily associated with mortality. Crude mortality rates approximating 20% have been reported for trauma patients with VAP. Most studies have involved the most severely injured patients, making it difficult to determine the relative contribution of either VAP or injury severity to death. If VAP is independently associated with mortality, this relationship should be most evident in less severely injured patients. We studied patients with less severe injuries (Injury Severity Score, ISS < 25) to determine the impact of VAP on outcomes.

METHODS: Patients admitted to the trauma ICU with ISS < 25 were identified from the trauma registry of a level I trauma center. Patients with penetrating injuries and those who died within 48 h of injury were excluded. Pneumonia was diagnosed using quantitative cultures of bronchoalveolar lavage effluent (>/= 10(5) colony forming units/mL). Risk factors for VAP, including age, transfusions with 24 h of admission, brain injury, and chest injury severity were analyzed. Logistic regression analysis was then performed to determine independent factors for death.Results: There were 15,492 blunt admissions over a 5.5 year study period who survived >48 h. Of these, 5,860 (38%) were admitted to the ICU, and 4,111 (70% of ICU admissions) had ISS < 25. The incidence of VAP in this group was 8%. Patients with VAP were older (47 vs 39 years), had more transfusions within 24 h (2.5 vs 0.9 units of red blood cell concentrates) and had greater injury severity by ISS (16.7 vs 12.6 points), GCS (Glasgow Coma Scale) score (11.8 vs. 13.7 points) and chest AIS (Abbreviated Injury Scale) (1.7 vs 0.9 points; all p < 0.001). Overall mortality was 4%. Mortality was 16% in patients with VAP compared to 3% in those without VAP (p < 0.0001). Logistic regression analysis identified transfusions, age, and VAP as independent predictors of mortality. Other descriptors of injury severity (ISS, GCS, or chest AIS) were not associated with death.

RESULTS: There were 15,492 blunt admissions over a 5.5 year study period who survived . 48 h. Of these, 5,860 (38%) were admitted to the ICU, and 4,111 (70% of ICU admissions) had ISS , 25. The incidence of VAP in this group was 8%. Patients with VAP were older (47 vs 39 years), had more transfusions within 24 h (2.5 vs 0.9 units of red blood cell concentrates) and had greater injury severity by ISS (16.7 vs 12.6 points), GCS (Glasgow Coma Scale) score (11.8 vs. 13.7 points) and chest AIS (Abbreviated Injury Scale) (1.7 vs 0.9 points; all p , 0.001). Overall mortality was 4%. Mortality was 16% in patients with VAP compared to 3% in those without VAP (p , 0.0001). Logistic regression analysis identified transfusions, age, and VAP as independent predictors of mortality. Other descriptors of injury severity (ISS, GCS, or chest AIS) were not associated with death.

CONCLUSIONS: Ventilator-associated pneumonia is independently associated with death in less severely injured trauma patients. This demonstrates the need for effective diagnostic techniques so that adequate therapy may be initiated. Prevention of VAP in less severely injured trauma patients should increase survival.