Radiographic measurement of rotational deformity in pelvic fractures: A novel method with validity and reliability testing

Abstract

Objective: Measuring pelvic ring displacement on plain radiographs has focused on gap and translational displacements with little discussion of rotational displacement, likely because of measurement difficulties. Our hypothesis is that rotational displacement can be accurately measured on computer-reconstructed radiographs (CRRs) with good validity and reliability. Methods: A Sawbones model was used to simulate common pelvic ring fracture patterns. CT scans were obtained and converted to CRRs simulating standard views. A technique of measuring axial rotation on the inlet view and sagittal rotation on the inlet/outlet views was developed. The novel rotational assessment techniques were then used by 8 senior orthopaedic residents and 3 fellowship-trained orthopaedic trauma surgeons on the CRR of each model. These measurements were then assessed for interobserver reliability and validity. Validity was tested by comparing the values obtained by the observers to the actual displacement as measured on CT scans. Results: Newly described techniques for axial and sagittal rotational measurements showed excellent reliability with average confidence intervals of 1.8° and 3.4°, respectively. There was good validity of the technique, with the majority of the "true" measurements falling within the 95% confidence interval of the observer measurements. Conclusions: A new radiographic technique for measuring axial and sagittal rotational deformities shows excellent interobserver reliability and good validity. This technique offers a low-cost and low-radiation examination to assess this deformity, which is readily available in most clinical settings. This measurement system can be used in future clinical studies to assess the clinical implications of residual rotational displacement.