Abstract
BACKGROUND: The aim of this study was to compare plain lateral radiographs and 3D surface rendering (SR) CT imaging, in the characterization of posterior malleolar fracture (PMF) morphology using the Mason and Molloy classification. The null hypothesis was that there was no difference in characterization of morphology between plain radiographs and 3D SR CT. METHODS: Morphology of the PMF was categorized initially by the CT scan as classified by Mason and Molloy on 180 trimalleolar ankle fractures. PM fracture fragment size on the lateral radiograph were compared to their respective 3D surface rendering CT reconstructions, by two independent observers. Morphology of the PMF was assessed using all preoperative radiographs as compared to 3D SR CT. RESULTS: On comparison of fracture fragment morphology, all fractures had poor categorization by plain radiographs although rotational pilon fractures (type 2A and 2B fractures) had the worse sensitivity and specificity (below 33% and below 50%, respectively). Radiographs underestimated joint involvement in type 2B fracture patterns because of the underappreciation of the posteromedial fragment. CONCLUSION: This study shows that the use of plain radiographs to categorize morphology of PMFs is poor. The study adds to the ever-growing body of evidence on the inaccuracy of using plain radiographs in PMFs to plan treatment. Additional CT imaging is imperative to allow for appropriate treatment planning in the management of PMF. Smaller fracture fragments are more susceptible to inaccuracies, especially the rotational pilon subtypes (2A and 2B) because of the obliquity of the posteromedial fragment to the plane of the X-ray source. LEVEL OF EVIDENCE: Level III, retrospective comparative study.