Assessing the Diagnostic Accuracy of Plain Radiography for Upper Limb Fractures: CT as the Gold Standard

Abstract

Background: Plain radiography is the first-line imaging modality for suspected upper limb fractures; however, its diagnostic performance varies according to fracture location and morphology. Objective: To assess the diagnostic accuracy of plain radiography compared with computed tomography (CT) for detecting upper limb fractures. Methods: This cross-sectional study included 138 patients with suspected upper limb fractures at Aznostics Diagnostic Centre, Lahore, Pakistan. All participants underwent plain radiography and CT. Two radiologists independently assessed the images, with discrepancies resolved by consensus. CT was used as the reference standard. Sensitivity, specificity, and diagnostic classifications were evaluated. Results: Reported sensitivity ranged from 39.3% for scapular body fractures to 100.0% for occult humeral shoulder and radial head fractures. Coracoid process fractures demonstrated 56.3% sensitivity and 100.0% specificity, whereas scapular body fractures showed 39.3% sensitivity and 67.6% specificity. For comminuted ulnar shaft fractures, sensitivity was approximately 89.8%, with 100.0% specificity based on the reported contingency table. Statistically significant p-values were reported for coronoid process (p=0.016) and scapular body fractures (p=0.007). Conclusion: Fractures are more likely to be detected by CT scans. They are able to demonstrate the things that can be missed by X-rays. Plain radiography demonstrated variable diagnostic performance across fracture types. CT may improve fracture detection when radiographic findings are inconclusive.