Accuracy of Point-of-Care Ultrasonography for Diagnosis of Elbow Fractures in Children
Rabiner JE, Khine H, et al
Ann Emerg Med
This prospective observational study by two urban pediatric EDs used a convenience sample of 130 pediatric patients presenting with elbow injuries. Patients were included if they were under age 21 and had an elbow injury requiring radiography for evaluation. Prior to x-ray and ultrasonography, clinician pretest probability for fracture was assessed using a six-point scale from very unlikely (<1% probability) to very likely (>99% probability). Twenty-six pediatric emergency physicians and fellows with focused ultrasound training consisting of one hour of lecture and workshop performed and interpreted the ultrasounds. These were compared with radiology attendings' interpretations of the x-ray and those of an expert sonographer blinded to the clinical exam. Follow-up consisted of a review of the patient chart or a structured telephone interview at least one week after the initial visit.
The procedure was performed by placing a 5- to 10-MHz linear transducer on the patient's distal humerus with the patient's elbow in 90-degree flexion. Long and short axis views were obtained. An elevated posterior fat pad was defined as “rise of the fat pad” above the extension of the anterior humeral line on long axis view and above the imaginary line between the epicondyles on short axis view. Lipohemarthrosis was defined on ultrasonography as heterogeneous echodensity in the effusion underlying the fat pad.
Thirty-eight (29%) of 130 patients had evidence of fracture on initial x-ray, with five additional fractures diagnosed on subsequent follow-up. Fracture patterns were representative of the literature. An elevated posterior fat pad on ultrasonography was detected in 36 of 38 patients with evidence of fracture on initial x-ray. Four of five patients, however, who had a fracture not seen on the initial x-ray had an elevated posterior fat pad on ultrasonography. Sensitivity and specificity of an elevated posterior fat pad on ultrasonography was 93% (95% CI, 81%-98%) and 76% (95% CI, 66%-84%), respectively. The three patients with fracture on x-ray did not have an elevated posterior fat pad on either x-ray or ultrasonography. Presence of lipohemarthrosis increased the sensitivity to 98% but decreased specificity to 70%.
Overall, the results showed good agreement between novice and expert sonographer interpretations, with a Kappa coefficient of 0.78, with an increase from 0.74 to 0.94 between the first and last quartile of ultrasounds. Discordant interpretations tended to involve subtle lipohemarthrosis and slight rises of the posterior fat pad on transverse views. Of seven negative novice interpretations on ultrasonography that a blinded expert sonographer said were positive, none had a fracture. Of disagreements where novice sonographer was positive but expert sonographer was negative, only one in seven were true positives. This suggests that clinical correlation may be a factor in helping to rule out fracture while it may have biased novice sonographers to interpret their scans as falsely positive.
The authors propose that point-of-care ultrasonography by emergency physicians may reduce x-rays in pediatric elbow injuries by 50 percent. Two points, in addition to the high negative predictive value of ultrasonography in this study, suggest this procedure may be feasible in many EDs. A large group of novice clinician sonographers performed the ultrasounds, which can be generalized to many academic and nonacademic EDs, and the median time of 90 seconds for performing ultrasonography would not encumber physician flow, and in most EDs would be more time efficient than obtaining an x-ray.
Limitations include the convenience sampling, observational nature of the study and the small but adequately powered sample size. Cortical disruption also was not assessed, which may have detected the missed fractures. This would have added significantly to the level of difficulty and procedure time of the ultrasonography, however.
The American Medical Association has supported the use of point-of-care ultrasonography by emergency physicians for more than 10 years, with emphasis on leaving the scope of practice and training to be defined solely by emergency physicians. This technique has the potential to supplant x-rays for evaluating pediatric elbow injuries in daily practice if more studies support the use of ultrasonography in pediatric elbow injuries.
Thanks to James Tsung, MD, for the images of elevated posterior fat pad as well as his and Dr. Luis Lovato's comments about the study and this review. Thanks also to Jessie Wall, MD, and Minette Brown for helping to obtain images for this article.
Click and Connect! Access the links in EMN by reading this issue on our website or in our iPad app, both available on www.EM-News.com.© 2013 Lippincott Williams & Wilkins, Inc.