The evaluation of the sensitivity and specificity of wrist examination findings for predicting fractures.

Background: The aim of this study was to evaluate the sensitivity and specificity of physical examination findings and functional tests in adult acute wrist trauma patients who presented to the emergency department (ED) and to create a reliable and practical clinical decision rule for determining the necessity of radiography in wrist trauma.

Methods: This prospective observational study was conducted in a tertiary ED. Each patient was checked for 18 physical examination findings and functional tests. Patients with suspected fracture were enrolled consecutively. Antero-posterior and lateral wrist views were performed for each patient. All radiographical studies were interpreted by an orthopedic surgeon. The prevalence, sensitivity and specificity, negative and positive predictive values of each finding were calculated. A modeling for predicting fractures was created using computer.

Results: 207 patients were evaluated and 69 patients (33.3%) had fractures. The most common encounterd fracture site was distal radius (29.5%). The most sensitive examination finding was pain in dorsiflexion (95.7%) and the most specific finding was ecchymosis (97.8%). Wrist edema, deformity and pain aggravated by pronation were found to be strong predictors of fracture. The area under the receiver operating characteristic curve at internal validation for a prediction model based on these three predictors was 0.88 (95% CI: 0.83-0,93). The overall sensitivity and specificity of this model were 94% (95% CI: 85-98%) and 51% (95% CI 43-60%) respectively. According to the model created in this study, 34% of acute blunt wrist trauma patients do not require any X-ray imaging.

Conclusions: This triple modeling may be used as an effective decision rule for predicting all wrist fractures in the ED and in the disaster setting.