Assessment of articular fragment displacement in acetabular fractures: a comparison of computerized tomography and plain radiographs

OBJECTIVES: Quantitatively evaluate plain radiographs, relative to computed tomography (CT) scans, for assessment of articular fragment displacement (step and gap) in displaced acetabular fractures and in canine osteotomized acetabular specimens. DESIGN: Retrospective evaluation of a consecutive series of CT scans and plain radiographs of patients with displaced acetabular fractures and canine acetabulae osteotomized to represent acetabular fractures with displacement. SETTING: Level I trauma center. PATIENTS/PARTICIPANTS: Computed tomography scans and plain radiographic images of 62 consecutive patients with displaced acetabular fractures were reviewed; 20 patients met the inclusion criteria regarding location, fracture pattern, availability of plain radiographs and CT scans, and the position of the femoral head at the time of radiographic assessment. The hemipelvi of five adult mongrels underwent osteotomy of the acetabulum and similar radiographic evaluation. INTERVENTION: Three independent reviewers measured step and gap deformity on plain radiographs and CT scans utilizing a standardized measurement technique. MAIN OUTCOME MEASUREMENTS: Sensitivity and specificity of plain radiographs for detecting step and gap displacement (2 mm and 4 mm) relative to CT scans were determined. Intraclass correlation coefficient and intraobserver reliability was also calculated. For the canine specimens, sensitivities of each imaging method were determined relative to actual fragment displacement measurements. RESULTS: In the clinical images, when compared to CT, plain radiographs showed poor sensitivity at detecting step deformity (sensitivity = 25%). When analyzed by fracture type, plain radiographs were particularly poor at detecting step deformities in fractures involving a single column of the acetabulum (sensitivity = 0%). Excellent intraobserver and intraclass reliability existed among the three reviewers. Computed tomography scans were considerably more accurate in measuring step and gap displacement relative to actual measurements than plain radiographs in the canine specimens. CONCLUSIONS: 1) Plain radiographs showed poor sensitivity for the detection of step and gap deformities in patients with acetabular fractures and in osteotomized canine specimens, relative to CT scans; 2) differences between CT and plain radiographs in both specimen types were greatest with the most clinically relevant deformity-i.e., step; and 3) CT scans are essential and should continue to be used in conjunction with plain radiographs in the preoperative evaluation of displaced acetabular fractures and perhaps should be considered in the postoperative assessment of fracture reduction.