Blind protected specimen brush and bronchoalveolar lavage in ventilated children

Objective: To determine whether nonbronchoscopic protected specimen brush (PSB) and bronchoalveolar lavage (BAL) are contributive for diagnosing ventilator-associated pneumonia in mechanically ventilated children, Design: Prospective study. Setting: Fifteen-bed pediatric intensive care unit in a university hospital. Patients: A total of 103 mechanically ventilated children, ranging in age from 7 days to 8.8 yrs, most with a high clinical suspicion for bacterial pneumonia. Interventions: All the children underwent nonbronchoscopic PSB and BAL. Nonbronchoscopic PSB was performed with a plugged double-sheathed brush and BAL with a double-lumen plugged catheter. Endotracheal secretions and blood cultures were also collected, Open-lung biopsy was performed for any child who died within 7 days after the inclusion in the study, according to the parental consent. Measurements and Main Results: The PSB specimens were submitted for bacteriologic quantitative culture (positive threshold, 103 colony-forming units [cfu]/mL), The BAL samples were processed for microscopic quantification of the polymorphonuclear cells containing intracellular bacteria (positive threshold, 1%) and quantitative culture (positive threshold, 10(4) cfu/mL). According to diagnostic categories based on clinical, biological, radiologic, and pathologic criteria, 29 children had bacterial pneumonia and 64 did not. Ten children were classified as having an uncertain status. Of the 29 children with bacterial pneumonia, 26 (90%) met one of the following three criteria: a) PSB specimen culture, greater than or equal to 10(3) cfu/mL; b) intracellular bacteria in cells retrieved by BAL, greater than or equal to 1%; and c) BAL fluid culture, greater than or equal to 10(4) cfu/mL. In contrast, 56 (88%) of the 64 patients without pneumonia did not. Conclusion: The results of this study indicate the following: a) nonbronchoscopic PSB and BAL were feasible in a large population of mechanically ventilated children; b) nonbronchoscopic techniques were contributive for diagnosing ventilator-associated pneumonia in children; and c) a combined diagnostic approach, using nonbronchoscopic PSB and BAL, was superior to using either test alone.