A Pilot Assessment of 3 Point-of-Care Strategies for Diagnosis of Perioperative Lung Pathology

BACKGROUND: Lung ultrasonography is superior to clinical examination and chest X-ray (CXR) in diagnosis of acute respiratory pathology in the emergency and critical care setting and after cardiothoracic surgery in intensive care. Lung ultrasound may be useful before cardiothoracic surgery and after discharge from intensive care, but the proportion of significant respiratory pathology in this setting is unknown and may be too low to justify its routine use. The aim of this study was to determine the proportion of clinically significant respiratory pathology detectable with CXR, clinical examination, and lung ultrasound in patients on the ward before and after cardiothoracic surgery. METHODS: In this prospective observational study, patients undergoing elective cardiothoracic surgery who received a CXR as part of standard care preoperatively or after discharge from the intensive care unit received a standardized clinical assessment and then a lung ultrasound examination within 24 hours of the CXR by 2 clinicians. The incidence of collapse/atelectasis, consolidation, alveolar-interstitial syndrome, pleural effusion, and pneumothorax were compared between clinical examination, CXR, and lung ultrasound (reference method) based on predefined diagnostic criteria in 3 zones of each lung. RESULTS: In 78 participants included, presence of any pathology was detected in 56% of the cohort by lung ultrasound; 24% preoperatively and 94% postoperatively. With lung ultrasound as a reference, the sensitivity of the 5 different pathologies ranged from 7% to 69% (CXR), 7% to 76% (clinical examination), and 14% to 94% (combined); the specificity of the 5 different pathologies ranged from 91% to 98% (CXR), from 90% to 99% (clinical examination), and from 82% to 97% (combined). For clinical examination and lung ultrasound, intraobserver agreements beyond chance ranged from 0.28 to 0.70 and from 0.84 to 0.97, respectively. The agreements beyond chance of pathologic diagnoses between modalities ranged from 0.11 to 0.64 (CXR and lung ultrasound), from 0.08 to 0.7 (CXR and lung ultrasound), and from 0 to 0.58 (clinical examination and CXR). CONCLUSIONS: Clinically important respiratory pathology is detectable by lung ultrasound in a substantial number of noncritically ill, pre or postoperative cardiothoracic surgery participants with high estimate of interobserver agreement beyond that expected by chance, and we showed clinically significant diagnoses may be missed by the contemporary practice of clinical examination and CXR.