Background: Our center evaluates all distal radial fractures with traction radiographs before splinting. Although investigations of various imaging modalities to evaluate distal radial fractures have been presented in the literature, to our knowledge the use of traction radiographs has not been well described. We hypothesized that the addition of traction radiographs to standard radiographs increases interobserver and intraobserver reliability for injury descriptions, affects the choice of treatment plan, and decreases the perceived need for computed tomography.
Methods: Radiographs for fifty consecutive eligible patients with distal radial fractures that were treated at a level-1 trauma center were used to create two image sets for each patient. Set 1 included injury and splint radiographs, and Set 2 included the images from Set 1 plus traction radiographs. The image sets were stripped of all demographic data and were presented in random order to seven fellowship-trained orthopaedic surgeons. The surgeons independently reviewed each of the 100 image sets and answered ten questions regarding the description and treatment of the injury. Analyses were conducted with kappa statistics to evaluate interobserver reliability. Intraobserver variability was assessed with the McNemar test after adjusting for clustering.
Results: Traction radiographs improved interobserver reliability for four of ten questions. With regard to intraobserver variability, responses to two questions were significantly changed. With the addition of traction radiographs, the observation of intra-articular fragments requiring reduction increased from 38.3% to 53.1% (p < 0.05) and the perceived need to order computed tomography for further evaluation decreased from 21.7% to 5.1% (p < 0.001). No other changes reached significance.
Conclusion: The addition of traction radiographs appeared to affect surgeons' interobserver reliability in the evaluation of distal radial fractures. In addition, traction radiographs changed the rate of detection of intra-articular fragments requiring reduction and the perceived need for computed tomography. These data indicate that traction radiographs may provide some of the same information as computed tomographic scans at a lower cost and argue for additional research comparing computed tomographic scans and traction radiographs of the distal part of the radius.