A comparison of conventional radiography and computed tomography in the evaluation of spiral fractures of the tibia.

Spiral fractures of the distal third of the tibial shaft can have problems of malunion or non-union that may be related to inability to assess the quality of the initial reduction of the fracture. We created typical two-fragment torsional tibial fractures in cadaver bones. The fracture fragments were mounted in a specially constructed jig that allowed simulation of various malreduction positions including varus and valgus angulation, antecurvatum, recurvatum, malrotation, and shortening. Anteroposterior and lateral radiographs and computed tomographic scans of the specimens were made. Measurements of the fracture gap were made at defined distances along the fractures, directly on the specimens as well as on the radiographs and computed tomographic scans. Excellent correlation was obtained between measurements of the gap on the specimens and on the computed tomographic scans, but plain radiographic measurements in thirty of thirty-four instances underestimated the true width of the gap. When shortening was introduced larger maximum gaps (of as much as eighteen millimeters) as well as greater discrepancies between measurements on the scans and plain radiographs were seen. The sizes of the fracture gaps were also greater for a given degree of shortening when the pitch of the spiral fracture was greater.