History and radiographic findings associated with clinically suspected radial head subluxations.
OBJECTIVES: To determine: 1) physician practices regarding the use of radiographs for radial head subluxations (RHS), 2) the prevalence of missed fractures in children with a clinical diagnosis of RHS, 3) the relative risk of a fracture with a nonclassic history for mechanism of injury for RHS, and 4) radiographic findings associated with RHS that are difficult to reduce. METHODS: This study began with a physician survey that addressed the integration of radiographs into the management of RHS. We subsequently conducted a prospective randomized trial with a consecutive sampling of children less than 6 years of age who presented to one of 2 urban pediatric emergency departments and 2 suburban pediatric urgent care centers with a clinical diagnosis of RHS. After informed consent was obtained, reduction was undertaken with a maximum of four attempts (two by hyperpronation and two by supination/flexion), 15 minutes apart. Failure to reduce the RHS resulted in the procurement of a radiograph of the elbow. At the conclusion of the study, all radiographs were evaluated by a radiologist blinded to the diagnosis. Patients receiving radiographs were contacted 2 weeks after discharge for verification of the diagnosis. RESULTS: Eighty-four percent of 224 physicians returned completed surveys. Fifty-six percent reported using radiographs for failed reduction attempts. In the second phase of the study, 136 patients were enrolled prospectively: 127 were reduced successfully and 9 patients failed attempts at reduction. Of the nine patients receiving radiographs: four had fractures (prevalence of 2.9% with 95% confidence interval (CI) = 0.8-7.4), two had no radiographic findings and normal function on follow up, and three had isolated posterior fat pads on radiograph and normal function on follow-up. The relative risk of a fracture in children with a nonclassic history defined as any mechanism other than "pull" was 1.200 (95% CI = 0.441-3.264); the relative risk was 1.886 (95% CI = 0.680-5.231) when defining a nonclassic history as any mechanism other than "pull" or "fall." CONCLUSIONS: 1) Physicians tend to order radiographs for elbow injuries they initially perceive to be radial head subluxations when attempts at reduction fail. 2) In our study, fractures in children who presented with the classic flexed elbow/pronated wrist position were rare. 3) The relative risk of a fracture in children with a nonclassic history for mechanism of injury was not significant. 4) An isolated finding of a posterior fat pad in a child with RHS that is difficult to reduce was not associated with a fracture in our small sample of children with radiographic findings.