Reliability of lung ultrasound interpretation by physiotherapists in premature infants

This study is a retrospective analysis of Lung Ultrasound (LUS) scans collected from preterm infants as part of a non-randomised, exploratory proof of concept study (Mater HREC Project 56566) investigating the effect of nebulised surfactant (AeroFact) on lung aeration and regional ventilation distribution (Electrical impedance tomography, thoracic breath sensor and LUS measurements of neonates with respiratory distress syndrome during surfactant administration: Mater HREC Project 56566). LUS scans being analysed were originally collected commencing in March 2021 and recruitment is ongoing. The study setting is Mater Mother’s Hospital Neonatal Critical Care Unit (NCCU) and Physiotherapy Department, South Brisbane, Queensland Australia. Anticipated completion on data collection is December 2023.. Scans previously collected by a neonatologist in the Aerofact study will have been de-identified with the infant’s study identification at the time of collection. Two LUS scans will have been collected, prior to and within 4 hours following, administration of nebulised surfactant. The LUS will be independently assessed by three novice physiotherapist assessors who have been trained in analysing LUS and are blinded to the history of the infants from whom the scans were taken. The assessors will all be currently working in the neonatal nursery and scans will have been previously deidentified and saved securely on the ultrasound machine. Interpretation will be undertaken by all assessors using a pre-determined standardised LUS scoring tool. based on previous studies (Brat et al, 2015). Each LUS scan will consist of 6 separate regional lung images that will be scored individually and independently on a dedicated score sheet by each assessor. A score between 0-3 will be given for each regional lung image depending on the specific signs identified in the following order: 0 = normal lung aeration (presence of A lines only), 1 = interstitial lung pathology (multiple, well-spaced B lines), 2 = alveolar patterning (crowded or coalescent B lines, with or without consolidation adjacent to the pleura) or 3 = extensive consolidation (Brat et al, 2015). Additionally, any other pathology identified (pleural effusion or pneumothorax) will also be documented. Total scores will also be calculated for each LUS scan by adding each of the separate regional lung scores together, with total scores ranging from 0-18.