Lung Ultrasound Prediction Model for Acute Respiratory Distress Syndrome: A Multicenter Prospective Observational...

Area of Science:

• Critical Care Medicine
• Pulmonology
• Medical Imaging

Background:

• Acute respiratory distress syndrome (ARDS) diagnosis lacks adequately sized, externally validated studies.
• Lung ultrasound (LUS) shows promise for ARDS diagnosis but requires further validation.

Purpose of the Study:

• Develop and validate a data-driven LUS score for ARDS diagnosis.
• Compare the performance of the LUS-ARDS score against chest radiography (CXR).

Main Methods:

• A multicenter prospective observational study of invasively ventilated ICU patients.
• Development of the LUS-ARDS score using logistic regression on 12-region LUS exam results.
• External validation in a separate cohort, with ARDS classification by an expert panel as the reference.

Main Results:

• The LUS-ARDS score demonstrated high diagnostic accuracy (ROC AUC 0.90 in derivation, 0.80 in validation).
• Performance of the LUS-ARDS score was comparable to experienced CXR readers.
• The LUS-ARDS score includes aeration scores and pleural line abnormalities.

Conclusions:

• The LUS-ARDS score provides accurate ARDS diagnosis following external validation.
• This score may serve as a valuable adjunct for ARDS diagnosis, offering objective accuracy.
• Further validation is recommended for widespread clinical adoption.