Radiographic Assessment of Ankle Fracture Displacement: A Validation Study

Abstract

Ankle fracture displacement is an important outcome in clinical research examining the effectiveness of surgical and rehabilitation interventions. However, the assessment of displacement remains subjective without well-described or validated measurement methods. The aim of this study was to assess inter- and intrarater reliability of ankle fracture displacement radiographic measures and select measurement thresholds that differentiate displaced and acceptably reduced fractures. Eight fellowship-trained orthopaedic surgeons evaluated a set of 26 postoperative ankle fracture radiographs on 2 occasions. Surgeons followed standardized instructions for making 5 measurements: coronal displacement (3) talar tilt (1), and sagittal displacement (1). Inter- and intraobserver interclass correlations were determined by random effects regression models. Logistic regression was used to determine the optimal sensitivity and specificity for the measurements with the highest correlation. Three of the 5 measures had excellent interobserver reliability (correlation coefficient > 0.75): (1) coronal plane distance between the lateral border of tibia and lateral border of talus, (2) coronal plane talar tilt, and (3) sagittal plane displacement. The threshold that best discriminated displaced from well-aligned fractures was 2 mm for coronal plane distance (sensitivity 82.1%, specificity 85.4%), 3 degrees for talar tilt (sensitivity 80.4%, specificity 82.2%), and 5 mm for sagittal plane distance (sensitivity 83.9%, specificity 84.9%). This study identified 3 reliable measures of ankle fracture displacement and determined optimal thresholds for discriminating between displaced and acceptably reduced fractures. These measurement criteria can be used for the design and conduct of clinical research studying the impact of surgical treatment and rehabilitation interventions.