Objective Characterise lung ultrasound (LUS) findings, diagnostic accuracy and agreement between novice and expert interpretations in young children with respiratory tract infections and wheeze.

Methods Prospective cross-sectional study in a paediatric ED. Patients ≤2 years with a respiratory tract infection and wheeze at triage were recruited unless in severe respiratory distress. Prior to clinical management, a novice sonologist performed the LUS using a six-zone scanning protocol. The treating physician remained blinded to ultrasound findings; final diagnoses were extracted from the medical record. An expert sonologist, blinded to all clinical information, assessed the ultrasound video clips at study completion. Positive LUS was defined as the presence of ≥1 of the following findings: ≥3 B-lines per intercostal space, consolidation and/or pleural abnormalities.

Results Ninety-four patients were enrolled (median age 11.1 months). LUS was positive in 42% (39/94) of patients (multiple B-lines in 80%, consolidation in 64%, pleural abnormalities in 23%). The proportion of positive LUS, along with their diagnostic accuracy (sensitivity (95% CI), specificity (95% CI)), were as follows for children with bronchiolitis, asthma, pneumonia and asthma/pneumonia: 46% (45.8% (34.0% to 58.0%), 72.7% (49.8% to 89.3%)), 0% (0% (0.0% to 23.3%), 51.3% (39.8% to 62.6%)), 100% (100% (39.8% to 100.0%), 61.1% (50.3% to 71.2%)), 50% (50% (6.8% to 93.2%), 58.9% (48.0% to 69.2%)), respectively. There was good agreement between the novice and expert sonographers for a positive LUS (kappa 0.68 (95% CI 0.54 to 0.82)).

Conclusions Among children with respiratory tract infections and wheeze, a positive LUS seems to distinguish between clinical syndromes by ruling in pneumonia and ruling out asthma. If confirmed in future studies, LUS may emerge as a point-of-care tool to guide diagnosis and disposition in young children with wheeze.