Interobserver reliability and intraobserver reproducibility of three radiological classification systems for intra-articular calcaneal fractures.

Abstract

The management of intra-articular calcaneal fractures is difficult. One aspect for successful management is the use of a reliable and reproducible fracture classification system (FCS). The purpose of this study was to evaluate the interobserver reliability and intraobserver reproducibility of Letournel, Sanders, and Zwipp classification systems on CT scan and the Bohler's angle measurement on plain X-ray. Furthermore, we studied if the addition of a CT scan to the plain X-ray influenced the evaluation of fracture extension to the calcaneocuboid joint. The CT scan and plain X-ray images of 51 intra-articular calcaneal fractures were evaluated two times by three observers (two radiologists and one orthopedic surgeon) within a 5-month interval. The interobserver reliability was measured using the Fleiss kappa while the intraobserver reproducibility was measured using the Cohen's kappa. The mean kappa values for the interobserver reliability and intraobserver reproducibility of the Sanders classification were 0.25 and 0.39, respectively, of Zwipp classification were 0.24 and 0.16, respectively, while those of the Letournel classification were 0.50 and 0.42, respectively. For the Böhler's angle, the mean kappa values for the interobserver reliability and intraobserver reproducibility were 0.34 and 0.32, respectively. The addition of CT scan images to plain X-ray found a higher incidence of calcaneocuboid joint involvement. Clinicians should be aware of the limitation regarding the interobserver reliability and intraobserver reproducibility of the Letournel, Sanders and Zwipp classification systems for calcaneal fractures. Future studies should attempt to improve the present classification systems.