Evaluation of the reliability of three classification systems for the distal radius fractures along with CT imaging

Abstract

Aim Along the years, different classification systems and treatment options have been proposed for the distal radius fractures. The most commonly utilised are the AO (Arbeitsgemeinschaft fur Osteosynthesefragen/Association for the Study of Internal Fixation), the Fernandez and the Universal System. There has be no consensus in literature when these systems where compared. The aim of this study was to evaluate the reliability of these systems by determining inter-observer (agreement between assessors) and intra-observer (i.e. agreement with the initial assessment) accuracy based on plain radiographs and subsequent use of CT scans. Mehtods A prospective randomized study was performed using 26 patients who presented with a displaced distal radial fracture. Five orthopeadic surgeons and two hand consultants were asked to classify the patients’ plain radiographs and CT scans. Reproducibility was assessed by the use of the proportion of agreement and kappa coefficient between pairs of observers. Results All classification systems generate fair to moderate interobserver reproducibility (Kappa = 0.19-0.43) when using not only Xray but also CT imaging. The highest values were obtained with the Fernandez classification using CT scan (Kappa 0.43- moderate agreement, 55% inter-observer and 11.5% agreement with the “golden rule”). On the contrary, the lowest values were obtained with the Universal classification after evaluation of the x-rays (Kappa 0.19 - slight agreement, 27% inter-observer and 0% agreement with the “golden rule”). There was no significant improvement of outcomes with the use of CT imaging. Conclusion Despite the fact that Fernandez system demonstrates greater reproducibility among the panel of experts and was less affected after the evaluation of the CT scan, , its value is still moderate and we believe that this could be attributed to its relatively vague description of complex injuries. Overall, no classification system generates satisfactory interobserver reproducibility sufficient for clinical application. Intra-observer reproducibility did not improve [dramatically] after addition of CT scans, [potentially] questioning the use/role of CT imaging as a preoperative assessment tool.