An alphanumeric classification of osteoporotic pelvic ring injuries

Abstract

Introduction: Classification and management of osteoporotic pelvic ring injuries (OPRI) continue to pose a considerable challenge to orthopaedic traumatologists. The currently used fragility fractures of the pelvis (FFP) classification of OPRI has recently been shown to have significant weaknesses. The aim of this study therefore was to propose a new, simple, yet comprehensive alphanumeric classification (ANC) of OPRI and to assess its intra- and interobserver reliability. Furthermore, its potential advantages over the FFP classification are discussed. Materials and methods: One hundred consecutive CT scans from patients with OPRI were evaluated by three orthopaedic traumatologists with varying levels of experience and one musculoskeletal radiologist. Intra- and interobserver reliability of the proposed classification system was assessed using weighted kappa (κ) statistics and percentage agreement. In addition, the Fleiss’ kappa statistic was computed to assess interobserver agreement among all four raters. Results: Overall intraobserver reliability of the proposed ANC was substantial [κ ranging from 0.71 to 0.80; percentage agreement: 70% (range, 67–76%)]. Overall interobserver reliability between pairs of raters was substantial as well [κ ranging from 0.61 to 0.68; percentage agreement: 58% (range, 53–61%)]. For ANC types, groups and subgroups, intra- and interobserver reliability were substantial to almost perfect. Interobserver agreement among all four raters was moderate to substantial, with Fleiss’ kappa values of 0.48, 0.69, 0.71 and 0.52 for ANC overall, types, groups and subgroups, respectively. Conclusion: The proposed ANC of OPRI demonstrated overall reliability comparable to that of the FFP classification. The ANC, however, is simple, more comprehensive, and consistently relates to injury severity.