Evaluation of the role of comparison radiographs in the diagnosis of traumatic elbow injuries.

This study attempted to determine whether comparison radiographic views of the uninjured elbow result in increased diagnostic accuracy. Physicians (one junior and one senior orthopaedic resident and an orthopaedic surgeon) were provided with a short clinical summary and asked to interpret radiographs of the injured elbow (IE) or both the injured and uninjured elbow (UE) in a randomized fashion from 25 children with elbow injuries. The overall percentage of correct diagnoses (one vs. two elbow radiographs) were as follows: orthopaedic surgeon, 80 versus 96%; for junior resident, 80 versus 84%; and for senior resident, 84 versus 88% (p > 0.05). Kappa scores for interobserver variability and intrarater agreement were in the high range (0.756 to 0.904, kappa = 0.08). Clinically relevant diagnoses were missed by trainees (not the orthopaedic surgeon) whether radiographs of IE or both IE and UE were interpreted. Incorrect radiograph interpretations were due to false positives in 39 of 40 cases. Comparison radiographs of the UE were not useful in improving diagnostic accuracy in elbow trauma. However, although they may be necessary in some instances, routine radiographic examination of the UE is not warranted.