Gou Pan, Gao Kai, Wang Xiaoting, Liu Xing
Orthopedic Center of Children, Chongqing Medical University, Chongqing, China.
Front Pediatr. 2022 Oct 3;10:1018641. doi: 10.3389/fped.2022.1018641. eCollection 2022.
The aim of our study was to develop a nomogram predicting the early failure rate of Pavlik harness in infants under 6 months of age with developmental dysplasia of the hip (DDH).
We retrospectively analyzed the clinical data of 227 patients (372 hips) with DDH who were treated with Pavlik harness at our institution from August 2019 to January 2022. Fifty-eight patients (102 hips) failed the Pavlik harness treatment, and 169 patients (270 hips) were successfully treated. Then, the independent risk factors for treatment failure were determined univariate and multivariate logistic regression and used to generate the nomogram predicting the failure rate of the Pavlik harness.
It was found that age at initial treatment (OR 1.031, 95% CI 1.022-1.040, < 0.001), angle α (OR 0.723, 95% CI 0.671-0.779, < 0.001), and concomitant deformity (OR 0.129, 95% CI 0.036-0.459, = 0.002) were independent risk factors for treatment failure. The nomogram showed good discrimination [the area under the curve (AUC): 0.862], good calibration, and a net benefit in the range of probabilities between 5 and 90% according to the decision curve analysis.
This study successfully established the nomogram prediction model based on three independent risk factors. Due to the high level of predicting accuracy, this nomogram could be a useful resource for pediatric orthopedic surgeons to identify patients at major risk of Pavlik harness failure who might need more reliable treatments.
我们研究的目的是开发一种列线图,用于预测6个月以下发育性髋关节发育不良(DDH)婴儿使用 Pavlik 吊带治疗的早期失败率。
我们回顾性分析了2019年8月至2022年1月在我院接受 Pavlik 吊带治疗的227例(372髋)DDH 患者的临床资料。58例(102髋)患者 Pavlik 吊带治疗失败,169例(270髋)患者治疗成功。然后,通过单因素和多因素逻辑回归确定治疗失败的独立危险因素,并用于生成预测 Pavlik 吊带失败率的列线图。
发现初次治疗时的年龄(OR 1.031,95%CI 1.022 - 1.040,P < 0.001)、α角(OR 0.723,95%CI 0.671 - 0.779,P < 0.001)和合并畸形(OR 0.129,95%CI 0.036 - 0.459,P = 0.002)是治疗失败的独立危险因素。列线图显示出良好的区分度[曲线下面积(AUC):0.862]、良好的校准度,并且根据决策曲线分析在概率范围为5%至90%时有净效益。
本研究基于三个独立危险因素成功建立了列线图预测模型。由于预测准确性高,该列线图可为小儿骨科医生识别可能需要更可靠治疗的、有Pavlik吊带治疗失败重大风险的患者提供有用的参考。