Analysis of four-fragment fractures of the proximal humerus: the interest of 2D and 3D imagery and inter- and intra-observer reproducibility.

The treatment of the four-part fractures of the proximal humerus remains a therapeutic challenge. The decision-making is based on preoperative criteria concerning the displacement of the fracture, the stability of the fracture and the risk of avascular necrosis of the humeral head. The aim of this study was to analyse the inter- and intra-observer reproducibility of those criteria previously described by Hertel. Three observers analysed three times 20 radiologic files comprising 2D X-rays, 2D CT scan and 3D reconstructions for the intra-observer study, and an expert committee was used to assess the inter-observer reproducibility. The Kappa coefficient was used to measure agreement. The Kappa coefficient founded poor to moderate agreement for the majority of the criteria after the 2D X-ray analysis. This coefficient was improved with the use of 2D CT scan and 3D reconstructions, in particular for the medial hinge assessment, the humeral head fracture and the metaphyseal extension. The reproducibility of the criteria described by Hertel on 2D X-rays is at least moderate. Reproducibility could be considerably improved by associating 2D scans and 3D reconstruction, in particular for the criteria related to prognosis for the vascularisation of the humeral head.