BMJ Open Respir Res. 2026 Aug 31;13(1):e004249. doi: 10.1136/bmjresp-2026-004249.
ABSTRACT
BACKGROUND: Bronchial washing can be inefficient or unsafe due to aerosolisation, poor yield and loss of sample. The setup and connectivity of sampling containers is often challenging. Furthermore, the process of bronchoscopy in the intensive care unit is inadequately reported.We tested whether a novel closed bronchoscopy sampling system is safer and more efficient than the standard sampling technique. We recorded variation in bronchoscopy practice in mechanically ventilated patients.
METHODS: A multicentre randomised controlled trial saw 188 bronchoscopies performed in mechanically ventilated patients requiring diagnostic sampling. The closed-system Ambu bronchosampler (ABS) was compared to standard sampling containers (SSC) with single-use bronchoscopes (Ambu A4) for primary outcomes of safety (spill, unplanned disconnection) and efficiency (ease of connectivity, setup, loss of sample and % volume retrieved), secondary efficiency outcomes and the tertiary outcomes of the bronchoscopy process.
RESULTS: ABS was safer than SSC; spills 0% versus 23.7% (0/94 vs 22/93), p<0.001; unplanned tube disconnections 5.3% vs 21.3% (5/95 vs 20/93), p<0.001; unplanned container disconnections 2.1% vs 7.5% (2/95 vs 7/93), p=0.08. Efficiency was significantly higher in ABS than SSC (acceptable/(very) easy) setup 98.9% vs 82.8% (94/95 vs 77/93), connectivity 100% vs 54.7% (88/88 vs 47/86), upright 98.9% vs 10.8% (94/95 vs 10/93) and sample loss 11.6% vs 48.4% (11/95 vs 45/93), all p<0.001. Volume retrieved percentage unchanged (33.0% vs 35.4%, p>0.2).Bronchoscopy for infection in 80.9%, therapeutic lavage in 51% and infrequently otherwise. Prior imaging in 93.2%, preprocedural checklists/timeouts in 83.5% and post-procedure chest X-rays in 78.2%.Mandatory ventilation was most common peribronchoscopy. Neuromuscular blockade in 73.8% and recruitment manoeuvres in 34.7%. The bronchoscope was not removed intraprocedurally in 80.2% of cases. Positive microbiology was found in 45.7% of samples. Adverse events were rare (<5%).
CONCLUSIONS: Closed-system bronchoscopic sampling was safer (less spill and unplanned disconnections) and more efficient (setup, connectivity and loss prevention) than standard containers. Closed-system diagnostic sampling devices should be the standard during bronchoscopy.
TRIAL REGISTRATION NUMBER: ClinicalTrials.gov NCT05964075.
PMID:42674806 | DOI:10.1136/bmjresp-2026-004249
