凝血指标对COVID-19病情发展的预测价值

Prediction of severe corona virus disease 2019 by coagulation indicators

目的探讨凝血酶原时间(PT)、活化部分凝血活酶时间(APTT)、纤维蛋白原(Fib)、凝血酶时间(TT)及D-二聚体(DD)对新型冠状病毒肺炎(COVID-19)病情发展的预测和诊断的价值。方法选取2020年1—2月阜阳市第二人民医院就诊的COVID-19患者120例,分为重症组(17例)和轻症组(103例)。比较2组PT、APTT、Fib、TT和DD 5项指标的差异,用Logistic回归分析建立预测模型,采用受试者工作特征(ROC)曲线进行评价。结果重症组TT和DD均高于轻症组,差异均有统计学意义(P<0.01)。重症组Fib高于轻症组,差异有统计学意义(P<0.05)。DD和TT是COVID-19患者重症化的独立危险因素,建立预测模型Logit(P)=-12.460+1.760DD+0.601TT,预测模型拟合优度检验结果提示模型拟合度良好。ROC曲线预测效能分析结果显示,DD、TT以及联合检测的曲线下面积(AUC)分别为0.854、0.817、0.875,敏感性分别为76.47%、70.59%、70.59%,特异性分别为83.50%、86.41%、98.06%。DD、TT预测COVID-19患者重症化的最佳临界值分别为0.48μg/L、16 s。结论DD>0.48μg/L、TT>16 s以及二者联合能有效预测和诊断COVID-19患者病情重症化的发展。

Objective To investigate the roles of prothrombin time(PT),activated partial thromboplastin time(APTT),fibrinogen(Fib),thrombin time(TT)and D-dimer(DD)in the prediction and diagnosis of severe corona virus disease 2019(COVID-19).Methods A total of 120 patients with COVID-19 were recruited from Fuyang Second People's Hospital from January 2020 to February 2020.There were 17 cases of severe group and 103 cases of mild group.The differences of PT,APTT,Fib,TT and DD were compared between the 2 groups,and a predictive model was established by Logistic regression analysis and evaluated by receiver operating characteristic(ROC)curve analysis.Results Compared with mild group,severe group had higher TT,DD and Fib(P<0.01,P<0.05).DD and TT were independent risk factors for severe COVID-19.The predictive model Logit(P)=-12.460+1.760DD+0.601TT was established,and the goodness of fit test results of the predictive model showed P>0.05.The areas under curves(AUC)of DD,TT and combined determinations were 0.854,0.817 and 0.875,respectively.The sensitivities were 76.47%,70.59%and 70.59%,and the specificities were 83.50%,86.41%and 98.06%,respectively.DD predicted that the cut-off value of severe COVID-19 was 0.48μg/L,and the cut-off value of TT was 16 s.Conclusions DD>0.48μg/L and TT>16 s can effectively predict and diagnose the occurrence of severe COVID-19.The performance of combined determination is better.