Background: Ventilator-associated events (VAE), using objective diagnostic criteria, are the preferred quality indicator for patients requiring mechanical ventilation (MV) for greater than 48 hours. We aim to identify the occurrence of VAE in our trauma population, the impact on survival, and length of stay, as compared to the traditional definition of ventilator-associated pneumonia (VAP).
Methods: This retrospective review included adult trauma patients, who were Washington residents, admitted between 2012 and 2017, and required at least 3 days of MV. Exclusions included patients with Abbreviated Injury Scale head score greater than 4 and burn related mechanisms of injury. We matched trauma registry data with our institutional, physician-adjudicated, and culture-confirmed ventilator event database. We compared the clinical outcomes of ventilator-free days, intensive care unit length of stay, hospital length of stay, and likelihood of death between VAE and VAP.
Results: One thousand five hundred thirty-three trauma patients met criteria; 124 (8.1%) patients developed VAE, 114 (7.4%) patients developed VAP, and 63 (4.1%) patients met criteria for both VAE and VAP. After adjusted analyses, patients with VAE were more likely to die (hazard ratio [HR], 2.86; 95% confidence interval [CI], 1.44-5.68), than those with VAP, as well those patients with neither diagnosis (HR, 2.83; 95% CI, 1.83-4.38). Patients with VAP were no more likely to die (HR, 1.55; 95% CI, 0.91-2.68) than those with neither diagnosis. Patients with VAE had fewer ventilator-free days than those with VAP (HR, -2.71; 95% CI, -4.74 to -0.68).
Conclusion: Critically injured trauma patients who develop VAE are three times more likely to die and utilize almost 3 days more MV than those that develop VAP. The objective criteria of VAE make it a promising indicator on which quality indicator efforts should be focused. Future studies should be aimed at identification of modifiable risk factors for VAE and their impact on outcome, as these patients are at high risk for death.
Level of evidence: Retrospective cohort study, level III.