Pneumothorax diagnosis with lung sliding quantification by speckle tracking: A prospective multicentric observational study.

Abstract

Lung ultrasound is commonly used for the diagnosis of pneumothorax. However, recognition of pleural sliding is subjective and can be difficult for novice. The primary objective was to compare a novices physician's performance in diagnosing pneumothorax from ultrasound (US) scans either with visual evaluation or with maximum longitudinal pleural strain (MLPS). The secondary objective was to compare the diagnostic relevance of US with visual evaluation or MLPS to diagnose pneumothorax with an intermediately experienced and an expert physician. We conducted a prospective, observational study in two emergency department and two intensive care unit, between February 2019 and June 2020. We included 99 adult patients with suspected pneumothorax, who received a chest computed tomography (CT). Three physicians with different experience of interpreting US scans (a novice physician, an intermediately experienced physician, and an expert) analyzed the US scans of 99 patients with suspected pneumothorax (50 (51%) with confirmed pneumothorax), which were confirmed by CT scan. With a threshold of 5%, the MLPS sensitivity was 94% (95% CI [83%; 98%]), and the specificity was 100% (95% CI [93%; 100%]). The novice physician had an area under the curve (AUC) with visual analysis of 0.75 (95% CI [0.67; 0.83]) vs 0.86 (95% CI [0.79; 0.94]) with MLPS (p = 0.04). The intermediate physician's AUC for diagnosing pneumothorax with visual analysis was 0.93 (95% CI [0.88; 0.99]) vs 1.00 (95% CI [1.00; 1.00]) with MLPS (p < 0.01) and for the expert physician it was 0.98 (95% CI [0.95;1.00]) vs 0.97 (95% CI [0.93; 1.00]), respectively (p = 0.69). In our study, speckle tracking analysis improved the accuracy of US for the novice and the intermediate but not the expert sonographer in the diagnosis of pneumothorax.