Ryan DeSanti, DO
Saint Christopher's Hospital for Children
Philadelphia, Pennsylvania
Disclosure information not submitted.
Kara Gill, MD
Associate professor of Radiology
University of Wisconsin
Madison, Wisconsin, United States
Disclosure information not submitted.
Jonathan Swanson, MD
Professor of Radiology
University of Wisconsin
Madison, Wisconsin, United States
Disclosure information not submitted.
Michael Lasarev, M.S.
Biostatistician
University of Wisconsin School of Medicine and Public Health, United States
Disclosure information not submitted.
Eileen Cowan, MD
Assistant professor of Pediatrics
University of Wisconsin
Madison, Wisconsin, United States
Disclosure information not submitted.
Pierre Kory, MD
Associate professor of Medicine
Mount Sinai Beth Israel, United States
Disclosure information not submitted.
Jessica Schmidt, MD
Assistant professor of Emergency Medicine
University of Wisconsin
Madison, Wisconsin, United States
Disclosure information not submitted.
Awni Al-Subu, MD
Associate Professor of Pediatrics
University of Wisconsin Hospital and Clinics
Madison, Wisconsin
Disclosure information not submitted.
Title: Comparison of radiograph and ultrasound in the assessment of children with acute respiratory failure
Background: Lung ultrasound (LUS) is increasingly used to evaluate children in acute respiratory failure (ARF). While chest x-ray (CXR) is the most commonly used imaging technique in pediatric ARF, its accuracy and reliability has never been rigorously evaluated. Yet it is the gold standard against which new imaging techniques such as LUS are increasingly being evaluated in pediatrics. This study aimed to quantify imaging agreement between CXR and LUS in the evaluation of children with ARF. Our hypothesis was that LUS would demonstrate substantial agreement with CXR.
Methods: A planned secondary analysis of an observational study of children admitted to a 21-bed PICU undergoing LUS for evaluation of ARF. Children older than 37 weeks gestational age and ≤18 years of age admitted to the PICU with ARF were evaluated from December 2018 to February 2020. CXR and LUS completed within 6 hours of each other were evaluated for the presence common imaging abnormalities associated with ARF across 4 lung regions: interstitial lung disease, consolidation, pleural effusion, and pneumothorax. Kappa statistics adjusted for maximum attainable agreement (k) were used to quantify agreement between imaging techniques.
Results: 88 children had LUS completed for the primary study, of whom 32 had concomitant lung imaging completed within 6 hours of each other. The cohort included 2 patients (6%) with status asthmaticus, 17 (53%) with bronchiolitis/viral pneumonitis, 12 (37%) with pneumonia, and 1 (3%) with multiple concurrent diagnoses as determined by an independent review of the medical record following hospital discharge. There was slight agreement between LUS and CXR derived diagnoses with 58% agreement (k/kmax=0.36, 95% CI -0.01–0.80). Evaluation of specific imaging patterns included: normal pattern, 57% agreement (k=0.32); interstitial pattern, 47% agreement (k=0.003); consolidation, 65% agreement (k=0.29); pleural effusion, 94% agreement (k=0.00); and pneumothorax, 99% agreement (k=0.00).
Conclusions: There is slight agreement between CXR and LUS derived diagnoses in children with ARF. LUS should not replace use of CXR in children admitted with ARF. Further study is necessary to determine the role of LUS in the PICU.