Background: More than one half of lower respiratory cultures are negative for ventilator-associated pneumonia (VAP) and final reporting requires 72 hours to 96 hours. A previous retrospective study concluded that preliminary bronchoalveolar lavage (BAL) culture results (pBAL), reported at approximately 24 hours, accurately predicted final BAL culture results (fBAL). Our objective was to verify the predictive value of pBALs for fBALs, and evaluate the use of insignificant (1-99,999 cfu/mL) pBALs for rapid discontinuation of empirical antibiotics.
Methods: All BALs performed in the intensive care unit (ICU) of the Presley Regional Trauma Center were used to compare pBALs with fBALs. Trauma intensive care unit patients whose antibiotics were discontinued using pBALs (study group), were compared with a historical control group who had their antibiotics discontinued using fBALs.
Results: Preliminary and final results of 474 prospectively collected BALs were compared. Significant pBALs had a positive predictive value of 100%, whereas the negative predictive values of "no growth to date" and insignificant pBALs were 99% and 95%, respectively. Study patients (n = 176) were similar to control patients (n = 112) except for a higher median injury severity score (34 vs. 29; p < 0.001). The median (interquartile range) duration of empirical antibiotic therapy in patients without VAP was shorter for study patients [1.5 (1.25, 2) vs. 3 (3, 4) days, p < 0.001]. Empirical antibiotics were temporarily interrupted in four patients with VAP because of falsely negative pBALs without adverse clinical sequelae.
Conclusions: Preliminary BALs were highly predictive for final results. Using insignificant pBALs appears to be a safe strategy for promptly discontinuing empirical antibiotics in trauma patients with suspected VAP.