Background: Recent literature supports the notion that bronchoalveolar lavage (BAL) in ventilated trauma patients may improve our ability to diagnose and treat ventilator-associated pneumonia (VAP). We hypothesized that BAL would decrease the number of cases of VAP diagnosed and impact our antibiotic use and ventilator days.
Methods: Prospective data on all infectious complications were collected for patients admitted to the trauma-burn service for the year 2001. All VAPs between January 1, 2001, through June 30, 2001, were diagnosed without BAL (No BAL group) using clinical signs of fever, sputum production, leukocytosis, chest radiographs, and sputum culture. After July 1, 2001, VAP was diagnosed with the use of BAL.
Results: There were 37 cases of VAP in the No BAL group (11%) and 29 cases of VAP (8%) in the BAL group. There were no statistical differences in Injury Severity Score, hospital length of stay, ventilator days, or mortality between the two groups. The time to initial treatment of VAP was shorter for the BAL group, but did not reach significance. The number of patients who had their VAP pathogens correctly treated with empiric antibiotics was also the same between the two groups. There was no difference in the rate of recurrent pneumonias. The antibiotic costs and respiratory therapy/ventilator costs were not statistically different between the groups for trauma patients, although antibiotic costs were higher for burn patients.
Conclusion: The routine use of BAL to diagnose VAP in our mixed trauma-burn population did not impact on clinical outcomes or antibiotic use. Our results do not justify the additional costs and potential risks of BAL for all patients. The means of VAP diagnosis may not be as important as choosing the appropriate antibiotics for common VAP organisms in any given intensive care unit.