发育性髋关节脱位石膏固定术后髋臼残余发育不良的Nomogram预测模型构建OA北大核心

Objective To explore the risk factors for residual acetabular dysplasia(RAD)following plaster fixation in children with developmental dysplasia of the hip(DDH)and to construct a nomogram predic-tion model for RAD after plaster fixation.Methods A retrospective analysis was performed on clinical data of DDH children under 24 months old who were treated with plaster fixation at Kunming Children's Hospital from December 2016 to December 2019,totaling 169 hips.Based on the final follow-up X-ray results,hips were di-vided into RAD group(n=44)and non-RAD group(n=125).Comparisons were made between the two groups regarding general data(gender,age at surgery,surgical method,previous harness treatment,side of dislo-cation),preoperative imaging data[preoperative acetabular index(AI),International Hip Dysplasia Institute(IHDI)classification,development of the femoral head ossification nucleus],and postoperative 48-hour MRI data[acetabular labrum inversion,femoral head to tricartilage distance(FTD),cartilage acetabular index(CAI),cartilage coverage].Multivariate logistic regression analysis was used to assess factors influencing RAD after plaster fixation,and statistically significant variables were used to determine the optimal cutoff value,area under the receiver operating characteristic(ROC)curve(AUC),sensitivity,and specificity for predicting RAD.A nomogram prediction model was then constructed.Results A total of 169 hips met the inclusion crite-ria,with 44 hips(26.1％)developing RAD.According to IHDI classification,there were 51 hips in type Ⅱ,75 hips in type Ⅲ,and 43 hips in type Ⅳ.Among these,102 hips underwent closed reduction and 67 hips under-went open reduction.Of the 43 hips with a history of harness treatment,67 had labrum inversion.Preoperatively,114 hips showed ossification nuclei.Univariate analysis showed statistically significant differences in surgical age[13.00(8.00,17.00)months vs.9.00(7.00,15.00)months],IHDI classification,labrum inversion(26/44 vs.41/125),preoperative AI[(39.59±0.65)° vs.(36.11±0.58)°].FTD[2.5(1.4,3.2)mm vs.1.5(0,1.9)mm],and CAI[17.00(13.00±20.50)°vs.12.00(8.25±15.00)°]between RAD and non-RAD groups(P＜0.05).Multivariate logistic regression analysis revealed that CAI＞16.5°(OR=1.112,95％CI:1.029-1.201)and preoperative AI＞39.5°(OR=1.105,95％CI:1.023-1.194)were independent risk factors for RAD(P＜0.05).The sensitivity of CAI and preoperative AI for predicting RAD was 0.523 and 0.568,respectively;the spe-cificity was 0.824 and 0.696,and the AUC was 0.726 and 0.655,respectively.A nomogram prediction model was constructed using the independent risk factors for RAD,with an internal validation consistency index(CI)of 0.745.Conclusions Preoperative AI＞39.5° and CAI＞16.5° are independent risk factors for RAD in DDH children following plaster fixation.The combination of CAI and preoperative AI provides significant predictive val-ue for RAD.This study successfully constructed a nomogram prediction model for RAD after plaster fixation in DDH patients,which can predict the occurrence of RAD following DDH plaster surgery.