Clinical accuracy of imaging techniques for talar neck malunion

To compare the ability of plain radiographs, computed tomography (CT), and radiostereometric analysis (RSA) to detect changes in talus fracture fragment position and alignment using an in vitro model.

Eight cadaveric tali were osteotomized at the talar neck. RSA beads were inserted into each talar fragment. The talus was anatomically reduced and stabilized with a pair of 3.5-mm cortical screws. Plain radiographs and RSA films were obtained. The fragments were then displaced and rotated to create a varus and supination deformity, and screw fixation was repeated in nonanatomic alignment. Displacement and rotation were directly measured. Plain radiographs and RSA were repeated, and CT scans were obtained. The RSA measurements were interpreted in a blinded fashion by an experienced researcher. Two independent blinded orthopedic trauma surgeons measured the displacement and rotation using plain films and CT. The results from each radiographic measurement were compared to the measured displacement and rotation using ANOVA.

Plain radiographs, RSA, and CT all underestimated the measured talar neck displacement and rotation. Radiographs underestimated displacement by 5.0 +/- 2.9 mm, RSA by 5.9 +/- 2.0 mm, and CT scans by 2.4 +/- 4.8 mm (P < 0.05). Rotation was also underestimated by all 3 techniques, but the differences among techniques were not statistically significant.

The most accurate imaging technique to measure displacement in talar neck malunion is CT scan. RSA was less useful as an imaging technique in this study.