Background: We evaluated the role of serial catheter-directed bronchoalveolar lavage (CDBAL) in the diagnosis and management of pneumonia in ventilated surgical intensive care unit patients.
Methods: Intubated surgical intensive care unit patients were prospectively evaluated with serial CDBALs from September 1, 2012, to May 31, 2013. Initial CDBALs were performed if patients developed the following signs of pneumonia: white blood cell count greater than 11 or less than 4, temperature greater than 38.5°C or less than 36°C, qualitative purulent sputum, worsening oxygenation, or new infiltrate on plain chest x-ray. Subsequent CDBALs were performed every 4 days. Pneumonia was diagnosed using a Clinical Pulmonary Infection Score of greater than 6 and CDBAL cultures with greater than or equal to 10 colony-forming units of pathogenic organisms. Patients were also evaluated for sustained (≥48 hours) respiratory deterioration (increased FIO2 or positive end-expiratory pressure) corresponding to the National Healthcare Safety Network definition of ventilator-associated event (VAE).
Results: A total of 159 patients were intubated for 5 days or longer, of whom 80 patients were diagnosed with clinical pneumonia. Of these patients, 67 had serial CDBALs performed, and 81 ventilator-associated pneumonias (VAPs) were diagnosed in these patients. Of the patients with VAP, 16 also met the National Healthcare Safety Network criteria for VAE. Patients with VAP that had sustained respiratory deterioration demonstrated resolution of their compromise 60 hours (interquartile range [IQR], 41-107 hours) after starting antibiotics. Of the patients with pneumonia, 66 (81%) had resolution of the pathogenic bacteria in subsequent CDBAL cultures or were extubated within 4 days (IQR, 4-5 days) after starting antibiotics. The duration of antibiotic therapy in this group was 8 days (IQR, 7-9 days). The remaining 15 patients had multiple positive serial CDBAL cultures that isolated the same organism despite antibiotic treatment. The duration of antibiotic therapy was 14 days (IQR, 10-19 days) in these patients. The culture results were used to adjust antibiotic regimens a median of one time (IQR, 1-2 times) in 13 (87%) and two or more times in 6 (40%) of these patients.
Conclusion: Serial CDBALs help guide antibiotic treatment duration in patients with pneumonia and VAE. Patients with sustained hypoxia or persistent bacterial growth may require prolonged therapy.
Level of evidence: Diagnostic test, level III. Therapeutic study, level IV.