An independent inter- and intraobserver agreement assessment of the AOSpine sacral fracture classification system.

<bold>Background Context: </bold>The AOSpine sacral classification scheme was recently described. It demonstrated substantial interobserver and excellent intraobserver agreement in the study describing it; however, an independent assessment has not been performed.<bold>Purpose: </bold>To perform an independent inter- and intraobserver agreement evaluation of the AOSpine sacral fracture classification system.<bold>Study Design: </bold>Agreement study.<bold>Methods: </bold>Complete computerized tomography (CT) scans, including axial images, with coronal and sagittal reconstructions of 80 patients with sacral fractures were selected and classified using the morphologic grading of the AOSpine sacral classification system by six evaluators (from three different countries). Neurological modifiers and case-specific modifiers were not assessed. After a four-week interval, the 80 cases were presented to the same raters in a random sequence for repeat assessment. We used the Kappa coefficient (κ) to establish the inter- and intraobserver agreement.<bold>Results: </bold>The interobserver agreement was substantial when considering the fracture severity types (A, B, or C), with κ=0.68 (0.63-0.72), but moderate when considering the subtypes: κ=0.52 (0.49-0.54). The intraobserver agreement was substantial considering the fracture types, with κ=0.69 (0.63-0.75), and considering subtypes, κ=0.61 (0.56-0.67).<bold>Conclusion: </bold>The sacral classification system allows adequate interobserver agreement at the type level, but only moderate at the subtypes level. Future prospective studies should evaluate whether this classification system allows surgeons to decide the best treatment and to establish prognosis in patients with sacral fractures. [ABSTRACT FROM AUTHOR]