成人斯蒂尔病患者死亡的风险因素分析及个体化预测方法探讨OA
Risk factors for mortality in adult-onset Still's disease and development of an individualized nomogram predic-tion model
目的 分析成人斯蒂尔病(AOSD)患者死亡的风险因素,构建个体化预测Nomogram模型并验证效能.方法 回顾性分析2018年1月至2024年12月收治的213例AOSD患者临床资料,采用7∶3比例将其随机分为建模组(n=149)、验证组(n=64),比较2组临床资料.根据建模组患者治疗结局将其分为死亡组(n=25)、存活组(n=124),比较2组临床资料,采取多因素logistic回归模型分析AOSD患者死亡的风险因素,应用R软件构建AOSD死亡的个体化预测Nomogram模型,绘制受试者工作特征曲线(ROC)、校准曲线、临床决策曲线(DCA)评价该模型的预测效能.结果 AOSD患者病死率为16.43％(35/213);建模组中死亡组年龄、复发、MAS、血清铁蛋白、胸膜炎高于存活组(t=5.077,P＜0.001;x2=7.652,P=0.006;t=15.892,P＜0.001;t=5.050,P＜0.001;t=4.583,P=0.032),年龄(OR=1.236,95％CI:1.014～1.507,P＜0.001)、复发(OR=3.001,95％CI:1.297～6.944,P＜0.001)、MAS(OR=3.758,95％CI:1.730～8.167,P＜0.001)、胸膜炎(OR=2.807,95％CI:1.001～7.869,P＜0.001)为AOSD死亡的风险因素;将上述多因素logistic回归分析获得的影响因素作为预测指标,构建AOSD患者死亡的个体化预测Nomogram模型,其预测建模组、验证组AOSD 患者死亡的曲线下面积分别为 0.936(95％CI:0.884～0.970)、0.847(95％CI:0.779～0.901);内部验证显示C-index值为0.917、0.845,模型预测死亡风险与实际死亡风险的差异均无统计学意义(x2=1.348,P＞0.05;x2=1.432,P＞0.05);该模型预测建模组阈值概率在10％～84％、89％～100％时可获得临床净收益,预测验证组阈值概率在12％～79％、81％～95％、97％～100％时可获得临床净收益.结论 AOSD患者死亡的危险因素包括年龄、复发、MAS、胸膜炎,据此构建的AOSD死亡个体化预测Nomogram模型具有良好的预测效能.
Objective To identify risk factors for mortality in patients with adult-onset Still's disease(AOSD)and to develop and validate an individualized nomogram for risk prediction.Methods A retrospective study included 213 patients with AOSD treated between January 2018 and December 2024.Patients were randomly split into a training cohort(n=149)and a validation cohort(n=64)at a ratio of 7∶3.In the training cohort,patients were categorized into a death group(n=25)and a survival group(n=124).Clinical variables were compared,and multivariable logistic regression was used to identify independent risk factors for mortality.A nomogram was constructed using R software.Model perform-ance was evaluated by receiver operating characteristic(ROC)curves,calibration plots,and decision curve analysis(DCA).Internal validation was performed using the concordance index(C-index).Results The overall mortality rate was 16.43％(35/213).In the training cohort,patients who died were older and had higher rates of relapse,macrophage activation syndrome(MAS),serum ferritin levels,and pleuritis than survivors(all P＜0.05).Multivariable analysis i-dentified age(OR=1.236,95％CI:1.014-1.507),relapse(OR=3.001,95％CI:1.297-6.944),MAS(OR=3.758,95％CI:1.730-8.167),and pleuritis(OR=2.807,95％CI:1.001-7.869)as independent risk factors for mortality(all P＜0.05).A nomogram incorporating these variables demonstrated excellent discrimination,with AUCs of 0.936(95％CI:0.884-0.970)in the training cohort and 0.847(95％CI:0.779-0.901)in the validation cohort.The C-index values were 0.917 and 0.845,respectively.Calibration curves showed good agreement between predicted and observed risks(Hosmer-Lemeshow test,P＞0.05).DCA indicated that the model provided clinical net benefit across a wide range of threshold probabilities.Conclusion Age,relapse,MAS,and pleuritis are independent risk fac-tors for mortality in patients with AOSD.The proposed nomogram demonstrates good predictive performance and may assist in individualized risk stratification and clinical decision-making.
王晓莹;杨晓佩;王喜甲;张磊;刘升云
郑州大学第一附属医院风湿免疫科(河南郑州 450000)郑州大学第一附属医院风湿免疫科(河南郑州 450000)郑州大学第一附属医院风湿免疫科(河南郑州 450000)郑州大学第一附属医院风湿免疫科(河南郑州 450000)郑州大学第一附属医院风湿免疫科(河南郑州 450000)
医药卫生
成人斯蒂尔病死亡风险因素预测效能Nomogram模型
adult-onset Still's diseasemortalityrisk factorspredictive performancenomogram model
《广东医学》 2026 (4)
523-529,7
河南省医学适宜技术推广项目(SYJS2022099,SYJS2022100)
评论