Accuracy and clinical use of biliary point-of-care ultrasound: A retrospective cohort study.

Abstract

To assess the utilisation and diagnostic accuracy of biliary point-of-care ultrasounds (POCUS) performed and archived in a single New Zealand ED. A retrospective cohort study of clinically indicated biliary POCUS, archived from 1 October 2019 to 20 May 2020. POCUS interpretation was compared to expert review and radiology imaging when available. The clinician interpretations, diagnostic imaging results and surgical referrals were extracted from electronic medical records. Archived POCUS images were retrospectively reviewed by an expert to determine diagnostic quality, sonographic findings and imaging errors. A total of 133 ultrasounds were included; 118 (90%) archived images were of diagnostic quality. The presence or absence of gallstones was documented in 124 (93%), and other sonographic findings were less consistently documented. There was almost perfect agreement with expert review for gallstones (κ=0.82, 95% confidence interval 0.72-0.93) and substantial agreement for wall thickening (κ=0.63, 95% confidence interval 0.42-0.83). For patients discharged following negative POCUS without radiology imaging or surgical review no patients returned and received a diagnosis of cholecystitis over 6months of follow-up. Imaging errors included failure to visualise the gallbladder neck, misinterpretation of artefacts and technical errors in measurement of the gallbladder wall. Biliary POCUS was primarily utilised for and was accurate for the detection of gallstones. Further quality assurance and training could be directed at earlier credentialing to address technical errors and improved documentation of POCUS findings.