Diagnostic accuracy of plain radiographs and cineradiography in diagnosing traumatic scapholunate dissociation

Abstract

Objective: Missed tears of the scapholunate ligament (SLL) and scapholunate dissociation (SLD) after wrist injuries pose a high risk of post-traumatic osteoarthritis of the carpus. Plain X-rays and dynamic radiographic studies are frequently used for initial diagnostic work-up. Given the limited evidence of their accuracy, we retrospectively compared the radiographic findings of patients with suspected traumatic SLD with wrist arthroscopy as the accepted reference standard.

Design and methods: During a 6-year period, plain radiographs and DSA cineradiography scans were obtained from 198 patients who had sustained a hyperextension injury to their hand. Of those, 102 (72 men, 30 women, mean age of 42+/-12 years) subsequently underwent diagnostic arthroscopy. Digital images were reevaluated by experienced radiologists unaware of the arthroscopic findings for the presence or absence of SLD. We calculated the sensitivity (SN) and specificity (SP) with 95% confidence intervals (CI), and computed areas under the receiver operating characteristic curves (AUC/ROC).

Results: Arthroscopy revealed 42 SLL ruptures, 13, 10, and 19 of which were classified as grade I (partial), II (complete with dynamic instability), or III (complete with static instability) respectively. Plain radiographs correctly identified 24 injuries (SN 57.1%, 95% CI 41.0-72.3%), and produced 1 false-positive result (SP 98.3%, 95% CI 91.1-100.0%). The scapholunate distance and the SL angle contributed independently to the diagnostic variance, with an AUC/ROC of 85.7% (95% CI 76.8-92.2%). Cineradiography had a sensitivity of 36 out of 42 (85.7%, 95% CI 71.5-94.6%), and a specificity of 57 out of 60 (95.0%, 95% CI 86.1-99.0%).

Conclusion: Pathological results on plain radiographs and cineradiography reliably indicate the presence of SLD after wrist trauma. Although non-conclusive, a negative dynamic study markedly reduces the pre-test probability of disease. Both methods remain key elements of primary diagnostic strategies for suspected traumatic SLD, and may facilitate the selection of additional tests.