CHEST Guidelines-Integrated-Use-of-Bedside-Lung-Ultrasound-and-Echo

Integrated-Use-of-Bedside-Lung-Ultrasound-and-Echo

Pdf Summary

This research explores the integration of bedside lung ultrasound (LUS) with echocardiography in diagnosing acute respiratory failure (ARF) in ICU patients. Traditionally, LUS is valued for its specificity in diagnosing respiratory disorders quickly and accurately at bedside. However, its individual diagnostic sensitivity is variable and coupling it with echocardiography could potentially enhance diagnostic accuracy.<br /><br />The study prospectively enrolled 136 patients aged 18 and over with severe ARF at a university hospital over 12 months. The study compared the diagnostic efficiency of LUS alone against a combined cardiopulmonary ultrasound approach, referred to as thoracic ultrasonography (TUS), supported by machine learning models. TUS demonstrated better diagnostic accuracy than LUS, particularly in distinguishing between cardiogenic pulmonary edema and other conditions like pneumonia and pulmonary embolism.<br /><br />Diagnostic efficacy was assessed using machine learning models like partial least squares regression, and outcomes were measured using metrics like receiver operating characteristic (ROC) curves. Results indicated TUS significantly enhanced diagnostic capability over LUS in diagnosing cardiogenic edema and pneumonia (P < .001 in both learning and test samples).<br /><br />The study concludes that TUS markedly outperforms LUS in identifying ARF causes, placing emphasis on echocardiography's role in ARF diagnostics and management when integrated with pulmonary assessments. The use of machine learning reinforces this distinction by facilitating more nuanced differential diagnoses, potentially improving treatment protocols.<br /><br />The researchers suggest that implementing artificial intelligence in developing integrative diagnostic models could offer more precise characterizations of ARF at the bedside, thus enabling more informed clinical decisions. Overall, this integration aims to optimize diagnostic accuracy, leading to improved patient prognosis through early and accurate ICU interventions.

Keywords

bedside lung ultrasound

echocardiography

acute respiratory failure

ICU patients

thoracic ultrasonography

machine learning

diagnostic accuracy

cardiogenic pulmonary edema

artificial intelligence

integrative diagnostic models