Non-directed bronchoalveolar lavage: improving the quality and timing of pneumonia diagnosis in mechanically ventilated patients in the intensive care unit.

Abstract

Early respiratory sampling in mechanically ventilated patients with suspected community-acquired pneumonia (CAP) is essential for guiding antimicrobial therapy, but endotracheal aspirate (ETA) is frequently contaminated with upper airway flora and bronchoscopic bronchoalveolar lavage (BAL) is invasive. Non-directed BAL (ND-BAL) can be performed at the bedside without bronchoscopy and has comparable diagnostic accuracy to bronchoscopic BAL. To implement ND-BAL as a bedside procedure for mechanically ventilated intensive care unit (ICU) patients admitted within 48 hours of hospital admission with suspected CAP, increasing the proportion receiving a respiratory sample within 6 hours of ICU admission from 38% to ≥70% over 12 months and to compare oral flora contamination between ETA and ND-BAL. A multidisciplinary team introduced a standardised ND-BAL protocol in a tertiary ICU. Baseline audit (January-December 2023) was compared with postimplementation data (March 2024-February 2025). The proportion of eligible patients with a respiratory sample collected within 6 hours increased from 14/40 (35%) to 34/44 (77%). ND-BAL samples demonstrated lower oral flora contamination than ETA (50% vs 90.9%). Legionella culture testing increased from 15% to 56.8%, while empiric methicillin-resistant Staphylococcus aureus and Legionella antimicrobial coverage remained similar. No ND-BAL-related adverse events were identified. ND-BAL was feasibly integrated as a joint nursing-medical bedside procedure and improved the timeliness and quality of respiratory sampling in mechanically ventilated patients with suspected CAP. This provides a platform for future rapid molecular diagnostics and antimicrobial stewardship.