Risk factors for avascular necrosis of the femoral head after developmental hip dislocation reduction surgery and construction of Nomogram prediction model

Abstract Background To analyze the risk factors for the development of avascular necrosis (AVN) of the femoral head after reduction surgery in children with developmental hip dysplasia (DDH), and to establish a prediction nomogram. Methods The clinical data of 134 children with DDH (169 hips) treated with closure reduction or open reduction from December 2016 to December 2019 were retrospectively analyzed. Independent risk factors for AVN after DDH reduction being combined with cast external immobilization were determined by univariate analysis and multivariate logistic regression and used to generate nomograms predicting the occurrence of AVN. Results A total of 169 hip joints in 134 children met the inclusion criteria, with a mean age at surgery of 10.7 ± 4.56 months (range: 4–22 months) and a mean follow-up duration of 38.32 ± 27.00 months (range: 12–94 months). AVN developed in 42 hip joints (24.9%); univariate analysis showed that the International Hip Dysplasia Institute (IHDI) grade, preoperative development of the femoral head ossification nucleus, cartilage acetabular index, femoral head to acetabular Y-shaped cartilage distance, residual acetabular dysplasia, acetabular abduction angle exceeding 60°, and the final follow-up acetabular index (AI) were associated with the development of AVN ( P < 0.05). Multivariate logistic regression analysis showed that the preoperative IHDI grade, development of the femoral head ossification nucleus, acetabular abduction angle exceeding 60°, and the final follow-up AI were independent risk factors for AVN development ( P < 0.05). Internal validation of the Nomogram prediction model showed a consistency index of 0.833. Conclusion Preoperative IHDI grade, preoperative development of the femoral head ossification nucleus, final AI, and acetabular abduction angle exceeding 60° are risk factors for AVN development. This study successfully constructed a Nomogram prediction model for AVN after casting surgery for DDH that can predict the occurrence of AVN after casting surgery for DDH.