发热伴血小板减少综合征28天死亡风险预测模型的建立和验证
摘要
月至2024年8月确诊SFTS患者作为建模队列,前瞻性纳入2024年9月至2025年11月患者作为验证队列。以28天死
亡为结局,采用Cox比例风险回归构建模型,并通过AUROC、AUPRC、Brier评分、校准曲线、DCA和KM曲线评
价模型性能;采用L1 Logistic联合Platt Scaling进行敏感性分析。结果:建模队列303例,验证队列131例。最终模型
纳入年龄、中枢神经系统症状、尿素氮、APTT和D-二聚体。Cox模型在建模组和验证组中的AUROC分别为0.851
和0.852,AUPRC分别为0.677和0.628,Brier评分分别为0.125和0.134;验证组阴性预测值为0.920。敏感性分析显示,
Platt校准Logistic模型在验证组中AUROC为0.892,Brier评分为0.106。结论:基于入院早期常规指标构建的Cox模
型可有效预测SFTS患者28天死亡风险,具有较好的判别能力和临床解释性。
关键词
全文:
PDF参考
[1]Liu Q, He B, Huang S, et al.Severe fever with
thrombocytopenia syndrome, an emerging tick-borne
zoonosis[J].The Lancet Infectious Diseases, 2014, 14(8):763-772.
[2]Xue-Jie Y, Mi-Fang L, Shou-Yin Z, et al.Fever
with thrombocytopenia associated with a novel bunyavirus
in China.[J].The New England journal of medicine, 2011,
364(16):1523-32.
[3]Toru T, Ken M, Tadaki S, et al.The first
identification and retrospective study of Severe Fever with
Thrombocytopenia Syndrome in Japan.[J].The Journal of
infectious diseases, 2014, 209(6):816-27.
[4]Sup W O, Rae J Y, Tae K K, et al.Effect of Early
Plasma Exchange on Survival in Patients with Severe Fever
with Thrombocytopenia Syndrome: A Multicenter Study.[J].
Yonsei medical journal, 2017, 58(4):867-871.
[5]Jia K, HyoJeong H, JiHye H, et al.Risk factors associated
with death due to severe fever with thrombocytopenia syndrome
in hospitalized Korean patients (2018-2022).[J].Osong public
health and research perspectives, 2023, 14(3):151-163.
[6]Ying X, Mingran S, Ning L, et al.Clinical feature
of severe fever with thrombocytopenia syndrome (SFTS)-
associated encephalitis/encephalopathy: a retrospective
study[J].BMC Infectious Diseases, 2021, 21(1):904-904.
Refbacks
- 当前没有refback。
