Measurement Reliability in Pediatric Lateral Condyle Fractures of the Humerus

Background: Lateral condyle fractures of the distal humerus are the second most common pediatric elbow fracture. A difference of 1 or 2mm in fracture displacement measurement may change the operative or nonoperative management of a minimally displaced fracture. The goal of this study was to determine the interobserver and intraobserver agreement in the measurements of displacement of lateral condyle fractures of the humerus. Methods: A retrospective radiographic review of 60 patients was performed by 3 reviewers. Fracture displacement was measured in millimeters on all radiographic projections and repeated by a single author 1 month later. Intraclass correlation coefficients (ICC) were calculated for intraobserver and interobserver agreement. We then calculated the percentage of measurements that differed by >1 or 2mm. Analysis was performed on the entire cohort of displaced fractures and again separately for a cohort of minimally displaced fractures (defined as <5mm). Results: The combined ICC for intraobserver reliability across all radiographic projections was 0.96 (range, 0.952 to 0.977), but there was disagreement between a single reviewer's own measurements 17.3% of the time. Within the cohort of fractures displaced <5mL, this intrarater disagreement (>2mm) dropped to 4%. Interobserver reliability was also high with an ICC of 0.90 (range, 0.861 to 0.924). Within the minimally displaced cohort (<5mm), the interrater disagreement was 6.6% when using a 2mm cutoff and 22.2% when using a 1mm cutoff. Conclusions: Intraobserver and interobserver reliability are high when measuring lateral condyle fractures and the percent of clinical disagreement is low when measuring minimally displaced fractures for which treatment depends on measuring to the millimeter. The internal oblique projection is the most reliable view for measuring lateral condyle fracture displacement, whereas the external oblique view offers little in terms of reliability. This study may be used as a platform for further research, verifying that reliable measurements of lateral humeral condyle fracture displacement can be obtained within 1 to 2mm from adequate plain films.