Diagnostic performance of radiographs and computed tomography for displacement and instability of acute scaphoid waist fractures.

BACKGROUND: Fracture displacement is the most important factor associated with nonunion of a scaphoid waist fracture.We evaluated the performance characteristics of radiographs and computed tomography (CT) in the diagnosis of intraoperative displacement and instability of scaphoid waist fractures using wrist arthroscopy as the reference standard.
METHODS: During a six-year period (2004 to 2010) at two institutions, forty-four adult patients with a scaphoid waist fracture underwent arthroscopy-assisted operative fracture treatment at a mean of nine days (range, two to twenty-two days) after injury. Subjects included all of those with a displaced scaphoid fracture seen on radiographs and a selection of patients with a nondisplaced scaphoid fracture. All patients had preoperative radiographs and CT. Arthroscopy with up to 5 kg of traction was the reference standard for fracture displacement and instability.
RESULTS: The reference standard (arthroscopy) led to a diagnosis of twenty-two displaced fractures (all unstable) and twenty-two nondisplaced fractures (seven unstable). Displacement was diagnosed in eleven patients (25%) with the use of radiographs and in twenty (45%) with CT. The sensitivity, specificity, and accuracy for diagnosing intraoperative displacement were 45%, 95%, and 70%, respectively, with the use of radiographs and 77%, 86%, and 82%, respectively, with CT. The sensitivity, specificity, and accuracy for diagnosing intraoperative instability were 34%, 93%, and 55%, respectively,with the use of radiographs and 62%, 87%, and 70%, respectively, with CT. Assuming a 10% prevalence of fracture displacement and instability among all scaphoid waist fractures, the positive and negative predictive values for displacement were 53% and 94%, respectively, with the use of radiographs and 39% and 97% with CT whereas the positive and negative predictive values for instability were 36% and 93%, respectively, with radiographs and 34% and 95% with CT.
CONCLUSIONS: Radiographs and CT scans cannot be relied on to accurately diagnose intraoperative scaphoid fracture displacement or instability compared with arthroscopic examination. The influence, with regard to the risk of nonunion, of intraoperative instability of a scaphoid fracture that is seen to be nondisplaced on radiographs or CT is currently unknown. LEVEL OF EVIDENCE: Diagnostic Level III.