Intraobserver and Interobserver Reliability of Radiographic Analysis of Proximal Humerus Fractures in Adolescents.

BACKGROUND: Multiple studies have shown low intrarater and interrater agreement of radiographic classification systems for proximal humerus fractures (PHFs) in adults. There is no standardized method of measuring angulation of pediatric PHFs, nor is there consensus as to the amount of angulation and displacement that require operative fixation of adolescent PHFs. We propose a new standardized method to measure fracture angulation that is similar to the method used to measure the epiphyseal-shaft angle for slipped capital femoral epiphysis. The primary purpose of this study was to evaluate the intraobserver and interobserver reliability of our proposed method compared with a nonstandardized method. The secondary purpose was to evaluate the intrarater and interrater agreement of the Neer and Horowitz (NH), and Salter-Harris (SH) classification systems.

METHODS: Seven raters evaluated 26 deidentified anteroposterior shoulder radiographs of patients 10 to 16 years of age with PHFs. Raters classified each fracture using the NH and SH systems, and used their own method to measure fracture angulation. This process was repeated 2 weeks later. During the second round, raters also measured fracture angulation using our proposed standardized method. Two weeks after the second round, raters reevaluated the radiographs using the standardized method. Intraclass correlation coefficients were calculated.

RESULTS: Excellent intraobserver and interobserver agreement was achieved for the standardized method of measuring fracture angulation. All of the raters had an intrarater reliability classified as excellent (>0.80) using the standardized method. Good intrarater and excellent interrater agreement was achieved when raters used their own fracture angulation measurement method but wide confidence intervals suggested that the results were less precise. Fair to moderate intrarater and interrater reliability was seen for the NH and SH classifications.

CONCLUSIONS: Our standardized method for measuring angulation in adolescent PHFs demonstrated excellent intrarater and interrater reliability. We propose that this technique may be a more precise method of measuring fracture angulation and this method should be used in future studies that evaluate indications for operative management of adolescent PHFs.

LEVEL OF EVIDENCE: Level III-diagnostic.