Accuracy of point-of-care ultrasonography for diagnosis of elbow fractures in children.

STUDY OBJECTIVE

We determine the test performance characteristics for point-of-care ultrasonography performed by pediatric emergency physicians compared with radiographic diagnosis of elbow fractures and compare interobserver agreement between enrolling physicians and an experienced pediatric emergency medicine sonologist.

METHODS

This was a prospective study of children aged up to 21 years and presenting to the emergency department (ED) with elbow injuries requiring radiographs. Before obtaining radiographs, pediatric emergency physicians performed focused elbow ultrasonography. An ultrasonographic result positive for fracture at the elbow was defined as the pediatric emergency physician's determination of an elevated posterior fat pad or lipohemarthrosis of the posterior fat pad. All patients received an elbow radiograph in the ED and clinical follow-up. The criterion standard for fracture was fracture on initial or follow-up radiographs.

RESULTS

One hundred thirty patients with a mean age of 7.5 years were enrolled by 26 sonologists. Forty-three (33%) patients had a radiograph result positive for fracture. A positive elbow ultrasonographic result had a sensitivity of 98% (95% confidence interval [CI] 88% to 100%), specificity of 70% (95% CI 60% to 79%), positive likelihood ratio of 3.3 (95% CI 2.4 to 4.5), and negative likelihood ratio of 0.03 (95% CI 0.01 to 0.23) for fracture. The interobserver agreement (κ) was 0.77. The use of elbow ultrasonography would reduce radiographs in 48% of patients but would miss 1 fracture.

CONCLUSION

Point-of-care ultrasonography is highly sensitive for elbow fractures, and a negative ultrasonographic result may reduce the need for radiographs in children with elbow injuries. Elbow ultrasonography may be useful in settings in which radiography is not readily accessible or is time consuming to obtain.