A modification of Klein's Line to improve sensitivity of the anterior-posterior radiograph in slipped capital femoral epiphysis.

Abstract

Radiographs can diagnose slipped capital femoral epiphysis (SCFE) on the anterior-posterior (AP) pelvis view and the frog-leg lateral view of the hips. On the AP radiograph, the lack of intersection between a line drawn parallel to the superior edge of the femoral neck (Klein's Line) and the epiphysis confirms a slip. Despite broad knowledge of the Klein's Line principle, application is difficult and inaccurate, especially in mild cases. On the frog-leg lateral radiograph, Southwick head/shaft angle and Wilson percent epiphyseal displacement commonly quantify the slip. Here, we set out to evaluate the intraobserver and interobserver reliability and the efficacy of these methods. Five separate observers on 2 separate occasions evaluated 30 AP and 30 frog-leg lateral radiographs of patients with unilateral SCFE for head/shaft angle, percent epiphyseal displacement, and width of epiphysis lateral to Klein's Line. We calculated the minimum difference required for 2 measurements to be considered different with 95% confidence ("minimum agreement difference"). For head/shaft angle, the intraobserver minimum agreement difference was +/-7.27 degrees and the interobserver +/-8.80 degrees; for percent epiphyseal displacement, the intraobserver was +/-7.18% and the interobserver was +/-7.27%; and for width of epiphysis lateral to Klein's Line, the intraobserver was +/-1.98 mm and the interobserver +/-2.16 mm. For each of these measures, the slipped hips significantly differed from the control hips (P<0.001 in all cases). Our analysis of the classical definition of Klein's Line--the lack of intersection between Klein's Line and the epiphysis as a sign of a slip--identified only 40.3% of slips. By modifying Klein's Line such that we measure the width of epiphysis lateral to Klein's Line, we improved sensitivity to 79% if a difference of 2 mm between hips indicated a slip. On the basis of our findings, head/shaft angle and percent epiphyseal displacement provide reliable and efficacious means of radiographic SCFE diagnosis. The classic definition of Klein's line fails to identify 60% of slips. We propose a modification of Klein's Line to improve sensitivity.