Ultrasound for the early diagnosis of clinically suspected scaphoid fracture

Abstract

Purpose To test the ability of ultrasound to diagnose clinically suspected occult scaphoid fractures. Methods Eighteen wrists in 18 patients with an average age of 35 years (range, 10–77 years) were seen in the emergency room, each with a single traumatic wrist injury, snuffbox tenderness, swelling, and a negative wrist x-ray result. They were evaluated in this prospective, blind, controlled study by physical examination, x-ray, and high-resolution ultrasound. One hand surgeon performed the examination, and ultrasounds were read by a musculoskeletal radiologist. Patients were immobilized in a thumb spica splint and then seen in the office 1 to 14 days after the emergency room visit, at which time a repeat physical examination, wrist x-ray, and the single investigative ultrasound were done using the opposite wrist as a control. All patients were immobilized and evaluated until symptoms resolved or x-ray showed scaphoid fracture site resorption or callus, in which case they were kept immobilized until healed. Results Ultrasound identified correctly 7 of 9 cases that were eventually positive for scaphoid fracture on plain x-ray. Ultrasound was read correctly as negative in 8 of 9 x-ray—negative cases; this was statistically significant. The 1 false-positive case had radioscaphoid arthrosis and radial wrist swelling. Sensitivity was 78% and specificity was 89%. The positive predictive value was 88% and negative predictive value was 80%. Conclusions We recommend that high-frequency ultrasound be used to investigate occult suspected scaphoid fractures because of its ability to allow early diagnosis and to eliminate the need for a more invasive or expensive diagnostic test in most cases.