Pediatric Rapid Ultrasound for Shock and Hypotension Phenotype Differentiation in the Emergency Department: Evaluation of Feasibility and Reliability in a Malawi Cohort.

Objective: To evaluate the feasibility, reliability, and diagnostic implications of performing the pediatric Rapid Ultrasound for Shock and Hypotension (p-RUSH) in children with undifferentiated shock upon hospital presentation in a low-resource setting (LRS).

Methods: Prospective observational study from February 2019 to December 2019. Methods: Pediatric emergency department (ED) of a large academic referral hospital in Blantyre, Malawi. Methods: Children (2 mo to 16 yr old) with shock upon presentation to the pediatric ED. Methods: None.

Results: Thirty children with shock were enrolled, of whom 14 died. The p-RUSH was performed upon admission to the ED, before administration of an IV fluid bolus. The p-RUSH was performed within a median time of 11.5 minutes, and 92.3% of the image frames in 4-second video clips were interpretable. Images were scored by two independent reviewers and the qualitative and quantitative assessments were compared and showed positive correlations as follows: 1) qualitative assessments of cardiac performance vs. left ventricle ejection fraction and fractional shortening measurements (r = 0.684 and r = 0.616, respectively, both p < 0.05) and 2) qualitative assessment of inferior vena cava (IVC) collapsibility vs. IVC collapsibility index (r = 0.470; p < 0.05). The interobserver agreement between cardiac and IVC qualitative assessments yielded a kappa statistic of up to 0.850 (cardiac views) and 0.275 (IVC collapsibility). Both reviewers applied a novel algorithmic flow diagram to diagnose the shock phenotype. In 23 of 30 children, the flowchart could be completed, which suggested either hypovolemic or distributive shock as the most common phenotype.

Conclusions: In a Malawian pediatric ED, the p-RUSH was feasible and qualitative assessments were reliable. This 2019 proof-of-concept test provides a basis for further external validation of the p-RUSH and our algorithm for identifying shock phenotypes, which may lead to individualizing care of children presenting with shock in LRSs.