Duckworth scaphoid predictors

Predictors of fracture following suspected injury to the scaphoid

Duckworth AD, Buijze GA, Moran M, et al. The Journal of bone and joint surgery. British volume 2012 · PMID 22733954 · doi:10.1302/0301-620x.94b7.28704

Question

Which demographic and clinical examination findings predict a true scaphoid fracture in patients with a suspected scaphoid injury?

Design

Prospective cohort study at a UK trauma centre (Edinburgh) assessing standardised clinical signs at serial review.

Patients

Consecutive patients with a clinically suspected or radiologically confirmed scaphoid fracture, reviewed within 72 hours and again at about two and six weeks.

Intervention

Demographics plus seven standardised clinical tests (including anatomical snuffbox pain on ulnar deviation and scaphoid tubercle tenderness)

Comparison

Final diagnosis of fracture versus no fracture

Primary outcome

Independent predictors of a confirmed scaphoid fracture

Key result

Male sex, sports injury, snuffbox pain on ulnar deviation within 72 hours and scaphoid tubercle tenderness at two weeks independently predicted fracture. Absence of snuffbox pain on ulnar deviation early on was associated with no fracture, and when all four factors were present the fracture risk was very high.

Limitations

Single-centre data with a reference standard based on radiographs, and the prediction rule needed external validation.

Impact on practice

Shows that a structured clinical prediction rule can shift the probability of fracture substantially, helping target supplementary imaging to patients in whom the diagnosis remains uncertain.