Comparison of fluoroscopic techniques for assessment of femoral rotational alignment

Anatomic rotational reduction of diaphyseal femur fractures is essential in restoring limb mechanics. Errors in reproducing anteroposterior (AP) or lateral knee reference radiographs of the contralateral limb could result in inaccuracies during rotational reduction. The objective of this study was to examine whether fluoroscopic rotational variation can be observed with the same degree of precision with AP and lateral distal femur projections.

AP and lateral radiographs were obtained from intact knees of 7 cadaveric specimens using fluoroscopy. The lateral condylar width and coronal femoral width from the AP images and the posterior condylar offset and sagittal femoral width from the lateral images were measured by 3 reviewers. Interclass correlation coefficients (ICCs) among the 3 reviewers were calculated. The mean data from all reviewers were plotted against angle of rotation, and the slope (M) and regression of the line were then determined.

ICCs were 0.997 (lateral) and 0.994 (AP), demonstrating excellent interobserver agreement. The mean (±SD) M value for lateral images was 0.016 ± 0.001 and for AP images was 0.009 ± 0.001 (P < .0001). The higher lateral M value represents a more appreciable difference in size of the measured segment for the same rotational change.

The observed rotational change was 1.76 times greater on lateral images compared to AP images; thus, the lateral images may be more precise as a reference for rotation. The routine use of lateral knee radiographs to guide intraoperative rotational alignment of the femur may therefore be justified.