Accuracy of point-of-care ultrasound examination of the lung in primary care performed by general practitioners: a cross-sectional study.

Background: Lung ultrasonography (LUS) is a point-of-care imaging modality with growing potential in primary care.

Objectives: While its use is well established in hospital settings, data on its accuracy when performed by general practitioners (GPs) remain limited. This study aimed to assess the diagnostic accuracy of LUS conducted by GPs following structured training.

Methods: We recruited 17 GPs from various regions of the Czech Republic. They completed a two-day educational course focused on LUS. Patients with current dyspnoea (NYHA II-IV) or a history of dyspnoea within the last four weeks were included and underwent LUS to assess the presence of pleural effusion and interstitial syndrome. An independent expert sonographer, blinded to clinical data, evaluated recorded LUS video loops as the reference standard. LUS findings were categorized into A profile (presence of A lines and intact lung sliding, indicating normal aeration), B profile (three or more B lines per intercostal space in at least two intercostal spaces per hemithorax, suggesting interstitial syndrome), pulmonary consolidation and pleural effusion.

Results: A total of 128 patients were enrolled in the study. A total of 768 thoracic segments were examined. A profile was identified in 642 (83.6%) segments, B profile in 108 (14.1%), pulmonary consolidation in 8 (1.0%), and pleural effusion in 12 (1.6%). For the identification of A profile, the sensitivity was 97.51% (95% CI 95.98-98.57), and the specificity was 88.10% (95% CI 81,13-93,18); for B profile, the sensitivity was 87.04% (95% CI 79,21-92,73), and the specificity was 97.73% (95% CI96,28-98,72); for pulmonary consolidation, the sensitivity was 100.0% (95% CI 63,06-100,00), and the specificity was 100.0% (95% CI 99,52-100,0); for pleural effusion, the sensitivity was 83.33% (95% CI 51,59-97,91), and the specificity was 99.87% (95% CI 99,27-100,00).

Conclusion: Our findings provide important preliminary data, demonstrating that GPs can perform LUS accurately after a structured training program. THE TRIAL REGISTRATION IDENTIFIER: is NCT04905719.