Screw placement after pinning of slipped capital femoral epiphysis: a postoperative CT scan study.

BACKGROUND

Unrecognized pin penetration in the treatment of slipped capital femoral epiphysis has serious long-term sequelae. The purpose of this study was to use postoperative computerized tomographic (CT) scans to determine the true position of the screw tip when compared with standard radiograph views.

METHODS

Twenty-four patients with 33 slipped capital femoral epiphyses were included in the study. Intraoperative or postoperative radiographs [anteroposterior (AP) and frog lateral] were compared with postoperative CT scans (coronal and axial) to determine (1) distance of the screw tip from the subchondral bone of the femoral head, (2) the number of screw threads across the physis, and (3) the 3-dimensional placement of the screw tip in the femoral head.

RESULTS

The positions of 38 screws in 33 hips were assessed. Screw position within specific quadrants of the femoral head was more anterior and superior than appreciated on radiographs. AP radiographs overestimated the distance between the screw tip and the subchondral bone, the average distance being 5.5 mm on AP radiographs and 3.4 mm on coronal CT (P<0.0001). Bland-Altman analysis confirmed 95% limits of agreement of -5.6 to 1.5 mm, indicating that screws could be up to 5.6 mm closer to the subchondral bone than estimated by the AP radiograph. Closer agreement was found between the frog lateral radiograph and the axial CT views, with the distance from the subchondral bone averaging 4.7 mm on frog lateral radiographs and 4.1 mm on axial CT (P<0.01). Bland-Altman analysis showed 95% limits of agreement between the 2 measures of -3.5 to 2.3 mm, suggesting that some screws were up to 3.5 mm closer to the subchondral bone. Three more anteriorly placed screws seemed to penetrate subchondral bone on CT, findings not shown on standard radiographs.

CONCLUSIONS

Frog lateral radiographs of the hip provide a more accurate estimation of screw placement than AP radiographs. Screws closer than 4 mm to the subchondral bone on frog lateral radiographs or 6 mm on AP radiographs may penetrate subchondral bone.