摘要
目的探讨恙虫病不同临床转归的临床特征,及重症恙虫病发生的预测因素。方法回顾性分析恙虫病383例,对比不同临床转归的差异并对各因素进行多因素Lasso回归分析。利用各个危险因素构建ROC曲线和诺模图,并对模型进行评估。结果重症恙虫病组谷草转氨酶(AST)、白蛋白(ALB)、总胆红素(TBIL)、直接胆红素(DBIL)、肌酐(CRE)、白细胞计数(WBC)、中性粒细胞比值(NEUT)、单核细胞比值(MONO)和非重症恙虫病组比较差异均有统计学意义(P均<0.05)。Lasso回归分析显示,从发病到多西环素开始使用的时间间隔、谷丙转氨酶(ALT)、ALB、球蛋白(GLO)、TBIL、CRE、血小板(PLT)、淋巴细胞比值(LYMPH)、MONO等为重症恙虫病组发生的预测因素。在这些因素构建的ROC预测曲线中,CRE和PLT的AUC分布为0.800和0.724。这些因素综合构建的ROC曲线AUC值最高,为0.931。并且,由它们构建的诺模图可量化预测发生恙虫病加重情况的可能性。结论ALT、ALB、TBIL、DBIL、CRE、WBC、NEUT、MONO有预警恙虫病病情加重发生的价值,且本研究所构建的诺模图可量化评估恙虫病的病情。
Objective To explore the clinical characteristics of different clinical outcomes of tsutsugamushi disease and its predictive factors.Methods 383 patients with tsutsugamushi disease were analyzed retrospectively.The difference of different clinical outcomes between the two groups was analyzed,and Lasso regression analysis was carried out on relevant indexes.ROC curves were constructed using various risk factors,and nomograph was constructed using these factors.Finally,the model was evaluated and verified.Results The laboratory indexes of tsutsugamushi disease patients in the two groups,such as aspartate transaminase(AST),albumin(ALB),total bilirubin(TBIL),direct bilirubin(DBIL),creatinine(CRE),white blood cell count(WBC)and neutrophil ratio(NEUT),had obvious differences,and the differences were statistically significant(P<0.05).Lasso regression analysis showed that DAY(Time interval from onset of tsutsugamushi disease to onset of doxycycline use),alanine transaminase(ALT),ALB,globulin(GLO),TBIL,CRE,PLT,LYMPH and MONO were predictors of tsutsugamushi disease.In the ROC prediction curve constructed by these factors,the AUC distributions of CRE,PLT were 0.8 and 0.724,respectively.The AUC value of ROC curve constructed by these factors was the highest,which was 0.931.Moreover,the nomogram constructed by them could quantitatively predict the possibility of tsutsugamushi disease exacerbation.Conclusion ALT,ALB,TBIL,DBIL,CRE,WBC,NEUT,MONO had the value of early warning of tsutsugamushi disease exacerbation.And the nomogram constructed could quantitatively evaluate the condition of tsutsugamushi disease.
作者
吴李贤
李湛东
李冠琼
舒琦
WU Li xian;LI Zhan dong;LI Guan qiong;SU Qi(Department of Infectious,the First People′s Hospital of Foshan,Foshan,Guangdong 528000,China)
出处
《热带医学杂志》
CAS
2019年第6期776-780,共5页
Journal of Tropical Medicine
关键词
恙虫病
诺模图
死亡
Tsutsugamushi disease
Nomogram
Death