Errors in Point-of-Care Ultrasound Quality Assurance Data: A Descriptive Study

While the accuracy of specific point-of-care ultrasound (POCUS) examinations has been studied, limited data exist on the characteristics and clinical impact of errors in the emergency department (ED). This study aims to assess the frequency and characteristics of errors identified during the quality assurance (QA) process at a large academic medical center.

A retrospective review was conducted of all POCUS examinations performed from May 2022 to December 2023 across three teaching EDs in a single health system. Errors were categorized as acquisition or interpretive. Critical errors (CEs) were defined by expert consensus as errors with the potential to delay recognition or treatment of a time-sensitive diagnosis associated with serious morbidity or mortality. Data included scan type, indication, operator training, interpretation, and error type; body mass index data were unavailable. Errors were identified via faculty QA review and adjudicated by the consensus of five faculty fellowship-trained in ultrasound (US). Descriptive statistics were used, with Fisher's exact test for comparisons.

Of 14,763 scans, 102 (0.69%) were classified as errors: cardiac (32.4%) and EFAST (16.7%) scans were most common. Among these, 66 scans were clinically indicated, 32 were educational, and 4 had unknown indications. Residents accounted for most errors (81: PGY1 = 37, PGY2 = 39, PGY3 = 5), followed by pediatric emergency medicine fellows (8), students (7), faculty (4), and US fellows (2). Interpretive errors constituted 78% of cases, while 28% were CEs (0.20% of all scans). No significant differences in overall errors were found among all user groups (P = .11). However, PGY1 and PGY2 residents had significantly more errors than PGY3 residents (P = .02), particularly for clinically indicated scans (P = .04). Educational scans showed no significant differences (P = .35).

POCUS errors were rare, with most being interpretive mistakes by junior residents. Nearly one-third of errors were classified as having potential for serious adverse effects, although this classification reflected potential clinical risk rather than documented patient harm. Multi-site, longitudinal studies are needed to identify factors contributing to high-risk errors.