• ABSTRACT
    • This study aimed to test the usefulness and validity of two versions commonly quoted in the literature of the Milch classification of paediatric supracondylar mass fractures. Anteroposterior and lateral elbow radiographs of 10 consecutive acute paediatric lateral condylar mass fractures requiring open surgical reconstruction were presented to six observers in order to assess inter- and intraobserver reliability of both versions. Accuracy of classification was compared with intraoperative findings. Observers agreed with the operative findings in only 50% of cases. There was poor interobserver (kappa<0.2) and moderate intraobserver agreement (kappa<0.6) for both versions. Thus Milch's classification was not found to be a useful tool.