Radiological measurement of pelvic fractures using a pelvic deformity measurement software program

BACKGROUND: It is difficult for the surgeon to measure pelvic displacement in the closed reduction operation for unstable pelvic fracture. We therefore developed a pelvic deformity measurement software program based on standardized radiographs. The objectives of the present study were to evaluate the inter-observer reliability of the program for measuring specific fracture types on preoperative pelvic films and to assess the validity of the measurement software program by comparing it with a gold standard. METHODS: Twenty-five patients diagnosed with AO/OTA type B or C pelvic fractures with the unilateral pelvis fractured and dislocated were included in this study. Four separate observers repeatedly determined the translational and rotational patterns and outcomes using the software program and hand measurement, and calculated the displacement using computed tomography (CT) coupled with a three-dimensional (3D) CT model. The validity of the measurement software was calculated by assessing the consistency between the software measurements and the gold standard. Additionally, inter-observer reliability was assessed for the software. The software was also applied in preliminary clinical practice for closed reduction procedures. RESULTS: The overall inter-observer reliabilities of the software program, CT coupled with 3D reconstruction, and hand measurements were high, with kappa values of 0.956, 0.958, and 0.853, respectively. The software showed validity similar to that of CT coupled with 3D reconstruction (0.939 vs. 0.969), and better than that of hand measurement (0.939 vs. 0.858). A preliminary clinical application demonstrated that the software is effective for guiding closed reduction of pelvic fractures. CONCLUSIONS: Our newly established pelvic deformity measurement program is a reliable and accurate tool for analyzing pelvic displacement patterns and can be used for guidance of closed reduction and planning of the reduction pathway. LEVEL OF EVIDENCE: III