Systematic review of diagnostic test accuracy of point-of-care gastric ultrasound by emergency physicians.

Gastric ultrasound (GUS) has been used by anesthesiologists, critical care physicians, and radiologists for various clinical applications. Emergency physicians have begun to adopt this technique to the emergency setting. This study aims to determine the diagnostic test characteristics of GUS for prandial status by emergency physicians. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis-Diagnostic Test Accuracy (PRISMA-DTA) guideline, we conducted a systematic review of diagnostic test characteristics of emergency physician-performed GUS for prandial status. We identified nine studies (n = 659) of emergency physician-performed GUS. Four studies evaluated adults (n = 173) and five studies evaluated pediatric patients (n = 486). Based on QUADAS-2 definitions, studies were highly applicable. However, studies were of low to very low quality and at a high risk of bias. In adults, GUS was 86-100% sensitive and 70-92% specific for prandial status, with 100% feasibility. In pediatric patients, GUS was 82-84% sensitive and 22-74% specific for prandial status, with 85-100% feasibility. In most cases, GUS was performed in less than 5 min. In the studies identified by this systematic review, point-of-care GUS by emergency physicians was highly feasible across adult and pediatric populations. This imaging modality has fair to excellent test characteristics for assessing prandial status in adults. Test characteristics in pediatric patients ranged from very poor to good. Accordingly, GUS is a potentially helpful adjunct for assessing gastric content and volume (i.e., full vs. empty stomach) that would benefit from further prospective investigation. Nevertheless, caution should be adopted prior to incorporating GUS into clinical decision making.