Total talar fracture - inter- and intra-observer reproducibility of two classification systems (Hawkins and AO) for central talar fractures.

INTRODUCTION: We evaluated the inter- and intra-observer reproducibility of two classification systems for central talar fractures (Hawkins, as modified by Canal and Kelly and then by us; AO/AOT). HYPOTHESIS: The analysis and classification of these fractures will be better with CT scans than with X-rays.
MATERIAL AND METHODS: Four observers evaluated 39 X-ray and CT scan files twice in the span of six weeks; each evaluation entailed classifying the fractures and describing their main features. Cohen's Kappa coefficient for inter-rater agreement was calculated and analysed.
RESULTS: The inter- and intra-observer reproducibility with CT scans was better with X-rays for most of the parameters evaluated. The modified Hawkins classification provided better reproducibility than the AO/AOT one. However, this classification system was not perfect, even after modifications and use of CT scans. DISCUSSION: CT scans are an essential tool for the analysis of all talar fractures. We modified the Hawkins classification (as modified by Canal and Kelly) to include a Type 0 (no displacement or less than 2mm), include frontal body fractures that are displaced like neck fractures and take into account comminuted fractures and other trauma in the area. LEVEL OF PROOF: IV - retrospective clinical study.