引用本文
  • 郑嘉祺,王焕,高悠水,等.股骨颈骨折术后并发症风险预测模型的建立和验证[J].同济大学学报(医学版),2020,41(6):739-746.    [点击复制]
  • ZHENG Jia-qi,WANG Huan,GAO You-shui,et al.Establishment and initial validation of the prediction model for postoperative complications of femoral neck fracture[J].同济大学学报(医学版),2020,41(6):739-746.   [点击复制]
【打印本页】 【在线阅读全文】【下载PDF全文】 查看/发表评论下载PDF阅读器关闭

←前一篇|后一篇→

过刊浏览    高级检索

本文已被:浏览 353次   下载 429 本文二维码信息
码上扫一扫!
股骨颈骨折术后并发症风险预测模型的建立和验证
郑嘉祺,王焕,高悠水,艾自胜
0
(同济大学医学院医学统计教研室,上海200092;上海交通大学附属第六人民医院骨科,上海200233)
摘要:
目的建立股骨颈骨折患者内固定术后并发股骨头坏死的风险预测模型并进行内部验证。方法对上海市3家医院2013年3月—2017年1月行内固定手术的新发股骨颈骨折患者378例进行回顾性随访研究。按3∶1比例将全部数据划分为训练集和验证集;训练集数据用于Logistic回归模型及列线图的构建,采用约登指数、灵敏度、特异度等指标评价模型区分度,用Hosmer-Lemeshow检验评价校准度;最后用验证集数据进行模型验证。结果在378例患者中,83例在3年内发生股骨头坏死(22.0%)。发生率为22.0%。二分类Logistic回归分析显示受伤至手术时间、骨折嵌插、VAS疼痛评分、移位情况、复位情况、术后错位距离为坏死预测因子。预测模型约登指数为0.847,C统计量0.874(0.976~0.997),校准度准度较好(χ2=3.62,R2=0.972,P=0.890)。当风险阈值为0%~95%之间时,模型可使患者收益。验证集约登指数为0.941,C统计量为0.999(0.997~1.000),其DCA曲线与用于开发模型的训练集曲线大致吻合。结论该预测模型具有良好的性能,区分度、校准度、可重复性、净收益率较好,能较准确地预测股骨颈骨折术后股骨头坏死的发生。
关键词:  临床预测模型  股骨头坏死  股骨颈骨折  列线图  模型验证
DOI:10.16118/j.1008-0392.2020.06.010
投稿时间:2020-03-04
基金项目:国家自然科学基金(81872718);上海市卫生与健康委员会项目(201840041)
Establishment and initial validation of the prediction model for postoperative complications of femoral neck fracture
ZHENG Jia-qi,WANG Huan,GAO You-shui,AI Zi-sheng
(Dept. of Medical Statistics, Tongji University School of Medicine, Shanghai 200092, China;Dept.of Orthopedic Surgery, Shanghai Jiao Tong University Affiliated Shanghai Sixth People’s Hospital, Shanghai 200233, China)
Abstract:
ObjectiveTo develop a risk prediction model of osteonecrosis of the femoral head(ONFH) in patients with femoral neck fracture after internal fixation and to carry out internal validation. MethodsA retrospective follow-up study was carried out on 378 patients with femoral neck fracture who underwent internal fixation from March 2013 to January 2017 in three hospitals in Shanghai. The patient data were divided into training set and validation set with a ratio of 3∶1. The training set data were used for the construction of Logistic regression model and nomogram. The sensitivity, specificity and Youden index of the model for prediction of ONFH risk were evaluated by Hosmer-Lemeshow test; and the validation of the model was evaluated with validation set. ResultsAmong 378 patients 83 developed ONFH within three years(22.0%). Binary Logistic regression analysis showed that the time from injury to operation, insertion before operation, VAS pain score, displacement, reduction and dislocation distance after operation were predictors. The Youden index was 0.847 and the C-index was 0.874(0.976-0.997), with good calibration(χ2=3.62, R2=0.972, P=0.890). When the risk threshold is between 0% and 95%, the model was beneficial for patients. The intensive index of validation set was 0.941, and the C-index was 0.999(0.997-1.000). The DCA curve was consistent with the training set curve used to develop the model. ConclusionThe model can accurately predict the occurrence of necrosis of femoral head after operation of femoral neck fracture with satisfactory discrimination, reliability, repeatability and net benefiting.
Key words:  clinical prediction model  osteonecrosis of femoral head  femoral neck fracture  nomogram  model validation

您是第4967755位访问者
版权所有《同济大学学报(医学版)》编辑部
主管单位:教育部 主办单位:同济大学
地  址: 上海四平路1239号 邮编:200092 电话:021-65980705 E-mail: yxxb@tongji.edu.cn
本系统由北京勤云科技发展有限公司设计