Interobserver Agreement of Ankle Fracture Classification Among Emergency Physicians

To evaluate the interobserver agreement of emergency physicians in classifying ankle fractures using the Danis-Weber system and to assess whether an educational intervention improves diagnostic consistency.

Thirteen emergency physicians from Gosford and Wyong Hospitals in NSW, Australia, independently classified 100 ankle radiographs on two separate occasions, spaced 4 weeks apart. Between the two assessments, participants were emailed an educational PowerPoint presentation on the Danis-Weber classification system. Interobserver agreement was measured using Cohen's Kappa statistic. Accuracy and changes in performance between the two quizzes were also assessed, with statistical significance determined using a paired t-test.

The first test demonstrated moderate interobserver agreement (Kappa = 0.51, p < 0.00001), whereas the second test showed substantial agreement (Kappa = 0.61, p < 0.00001). However, the improvement in agreement between the two rounds was not statistically significant (p = 0.63). Physician accuracy scores were 83% in the first test and 88% in the second, which also did not represent a statistically significant difference (p = 0.63).

Emergency physicians at both Hospitals demonstrated moderate to substantial interobserver agreement in applying the Danis-Weber classification system to ankle fractures, with performance levels comparable to those of orthopaedic and trauma surgeons internationally. However, the educational intervention in the form of an emailed presentation did not significantly improve either agreement or accuracy. Alternative educational strategies may be required to enhance diagnostic consistency in clinical practice.