Relationship between angle of femoral varus and length of shortening after proximal femoral osteotomy in children

Abstract

Objective To explore the relationship between varus angle and length of proximal femoral shortening in patients with developmental dysplasia of hip (DDH) using a locking compression pediatric hip plate (LCP-PHP) during pediatric femoral varus osteotomy and to verify the feasibility of our calculation formula for proximal femoral shortening. Methods A total of 55 DDH patients (55 hips) undergoing LCP-PHP proximal femoral varus osteotomy from July 2013 to June 2017 were enrolled. All preoperative and postoperative internal rotation hip radiographic films were collected. There were 12 boys and 43 girls with an average age of 5.6(2-10) years. The involved side was left (n=30) and right (n=25). According to the geometrical relationship of operative procedure, the trigonometric function formula of proximal femoral shortening length was derived as c (sinθ1-sinθ2), c was the center of femoral head to the most medial point of proximal femur, θ1 the angle between preoperative c and osteotomic line, θ2 the angle between c and osteotomic line after varus osteotomy. The radiographs were imported into the Digimizer software for directly measuring the value of proximal femoral shortening length (ΔS) for each child. And Digimizer software was employed for measuring the parameters required for the calculation formula and the value of proximal femoral shortening length (S') was obtained. Paired t-test was used for comparing the results of two methods, and consistency/correlation analysis performed for verifying the feasibility of deducing the calculation formula. Results Fifty-five DDH children had direct measurements and formula calculations. The femoral shortening lengths after varus osteotomy were calculated as follows: varus angle 20° (9.97±2.49) and (9.96±2.37) mm respectively. No significant difference existed between direct measuring and calculation methods of c and proximal femoral length (P>0.05). The Bland-Altman analysis of these radiographs revealed that 95% limits of agreement between femoral shortening length by algorithm and direct measuring were between -0.80mm and + 0.74 mm. The mean bias was -0.03 mm. The results indicated a high agreement between two methods. Pearson’s correlation analysis revealed that the correlation coefficient of two methods was 0.99 (P<0.001) and the results showed a significant correlation between two methods. Conclusions The above trigonometric formula can accurately calculate the corresponding femoral shortening length caused by different varus angles during proximal femoral varus osteotomy in DDH children. Wider clinical application is recommended. Key words: Osteotomy; Femur; Trigonometric function