Expert Agreement in the Interpretation of Lung Ultrasound Studies Performed on Mechanically Ventilated Patients

Abstract

The frequency of agreement was lower than might have been expected in this study. Several potential reasons are identified, chief among them the fact that ICU patients often develop multiple pulmonary insults, making agreement on a specific primary diagnosis challenging. This finding suggests that the utility of lung US in identifying the main contributing lung condition in ICU patients may be lower than in dyspneic patients encountered in the emergency department. It also raises the possibility that the clinical context is more important for lung US than other imaging modalities.

Interobserver Agreement of Lung Ultrasound Findings of COVID-19

“…Previous investigations have demonstrated moderate to substantial agreement for B-lines [22][23][24][25] . In contrast, there is only moderate to fair agreement for consolidation 11,12,22 , pleural irregularity 22 , and effusions. 11,22 Others have shown that there is substantial interrater reliability for LUS across differing probes (e.g.…”

“…In contrast, there is only moderate to fair agreement for consolidation 11,12,22 , pleural irregularity 22 , and effusions. 11,22 Others have shown that there is substantial interrater reliability for LUS across differing probes (e.g. linear vs. phased array) 25,26 , which is important given different portable devices were used in this investigation.…”

“…Outside of COVID-19, previous investigations have found moderate to excellent interobserver agreement for many lung ultrasound (LUS) findings, including B-lines, consolidations, and effusions. 2,11,12 In this study, we characterize the interobserver agreement of LUS findings that have been described for COVID-19. These images were collected using portable ultrasound devices, which may be more commonly utilized as a result of the COVID-19 pandemic.…”

“…Outside of COVID-19, previous investigations have found moderate to excellent interobserver agreement for many lung ultrasound (LUS) findings, including B-lines, consolidations, and effusions. 2,11,12…”

Development of a deep learning classifier to accurately distinguish COVID-19 from look-a-like pathology on lung ultrasound

“…5 While an accompanying thick pleural line is helpful in differentiating cardiogenic from non-cardiogenic causes of B lines, 6 reliable methods to differentiate non-cardiogenic causes from one another on LUS have not been established. Additionally, user dependent interpretation of LUS contributes to wide variation in disease classification, 7,8 creating urgency for techniques that improve diagnostic precision and reducing user-dependence.…”