Impacted intraarticular fragments of distal radius fractures: A radiographic characterization and analysis of reliability and diagnostic accuracy

Abstract

Background: Reduction using ligamentotaxis may not be effective enough to treat impacted intraarticular fragments of distal radius fractures. Articular incongruence resulting from the loss of reduction is a risk factor for postoperative osteoarthritis and worse clinical outcome. This study aimed to analyze the radiographic characterization of the impacted intraarticular fragments of distal radius fractures using two/three-dimensional computed tomography (CT). Further, we assessed the reliability and diagnostic accuracy in detecting the fragments using plain radiographs.

Methods: We analyzed 167 three-dimensional CT images of the intraarticular distal radius fractures and selected 12 fractures with impacted intraarticular fragments. We recorded the location, size, and displacement of the fragment using CT images. In addition, six examiners evaluated 25 fractures including those 12 fractures having the fragments using plain radiographs for detecting the fragments and their displacements. Further, we evaluated the reliability and diagnostic accuracy of the plain radiographs in the detection of the fragment.

Results: Fifteen impacted intraarticular fragments were found in 12 wrists. The displacement of the scaphoid facet fragment was significantly larger than that of the lunate facet fragment in CT measurement (7.0 mm and 3.6 mm). Inter and intraobserver reliability of the diagnosis for the fragment in plain radiographs were poor and fair (κ: 0.14 and κ:0.27). Diagnostic accuracy in detecting the fragment in plain radiographs generated mean sensitivity: 0.4, mean specificity: 0.73, and mean accuracy: 0.58. The mean sensitivity in detecting a lunate facet fragment was lower than that of a scaphoid facet fragment in plain radiographs (0.24 and 0.44).

Conclusion: Impacted intraarticular fragments were found in 7% of intraarticular distal radius fractures. We observed low reliability and sensitivity in detecting the fragment using plain radiographs. Preoperative recognition of the fragments using plain radiograph were difficult, even though the magnitude of step-off of the scaphoid facet fragment was large.