Quantifying Clinically Meaningful Point-of-Care Ultrasound Interpretation Discrepancies Using an Emergency Department Quality Assurance Program

Objectives: Emergency medicine professional associations recommend that quality assurance (QA) programs be implemented wherever emergency department (ED) point-of-care ultrasound (POCUS) is in use. The purpose of this study is to identify the rate of clinically meaningful interpretation discrepancies between initial ED POCUS interpretation and a gold standard using a QA program in a Canadian academic ED. Methods: All POCUS examinations completed in our ED are subject to a QA process. The results of all POCUS examinations undergoing this process from July 1, 2014, to June 30, 2015, were retrospectively reviewed. Four blinded abstractors collected data with a standardized tool after a training session. Information regarding patient demographics, POCUS indication, emergency physician initial POCUS interpretation, physician POCUS expertise, the presence of an interpretation discrepancy, and whether the discrepancy was clinically meaningful was abstracted. The proportion of interpretation discrepancies, clinically meaningful discrepancies, discrepancies requiring remedial action, and differences in discrepancy rates between non-expert and expert sonographers were analyzed. Results:A total of 2,869 POCUS studies were included for review, with 2,668 in the final data set after exclusions. In total, only 1.4% of all scans contained an interpretation discrepancy. The rate of clinically meaningful discrepancies was 0.5%, and the rate of scans requiring remedial action was 0.1%. Overall, 85.5% of all scans were performed by four POCUS expert physicians, with the remainder by a non-expert. Scans performed by non-expert sonographers were significantly more prone to discrepancies than those performed by experts. No single scan indication was more prone to discrepancy. Conclusions: The overall ED POCUS interpretation discrepancy rate and clinically meaningful discrepancy rate identified using our QA process were very low. The findings are limited by the small group of expert sonographers completing most scans.