Objective: Nonbronchoscopic bronchoalveolar lavage is often used as an alternative to bronchoscopic bronchoalveolar lavage in the diagnosis of ventilator-associated pneumonia. We have previously reported an improved safety profile for nonbronchoscopic lavage in patients with lung injury, suggesting that this may be a better technique in this patient group. The objective of this study was to determine whether nonbronchoscopic lavage could be used as an alternative to bronchoscopic lavage for the assessment of alveolar permeability and inflammation in patients at risk for acute respiratory distress syndrome (ARDS) or with ARDS.
Design: Prospective randomized crossover trial.
Patients: Intubated patients with ARDS or at risk of ARDS.
Interventions: Bronchoscopic and nonbronchoscopic lavage in the same patient, 15 mins apart.
Measurements and main results: Twenty-one patients with ARDS and 20 patients at risk of ARDS were recruited and underwent nonbronchoscopic and bronchoscopic lavage in randomized order. Despite similar volumes of lavage fluid, nonbronchoscopic lavage had fewer cells and an increased ratio of bronchial epithelial cells to macrophages. Although average concentrations of myeloperoxidase and total protein, the protein permeability index, and the epithelial-lining fluid volume were similar with the two techniques and demonstrated moderate linear associations, Bland and Altman analysis revealed poor comparability, with substantial side-to-side variability and wide 95% limits of agreement. Furthermore, unlike bronchoscopic lavage, nonbronchoscopic lavage was unable to differentiate between patients with ARDS and those at risk of ARDS.
Conclusions: Nonbronchoscopic lavage is not comparable to bronchoscopic lavage and as such cannot be used as an alternative to bronchoscopic lavage for assessing alveolar inflammation in patients with ARDS.