王玲, 刘妍, 张琴. 急性心肌炎短期预后不良的影响因素及预测模型构建[J]. 心脏杂志, 2020, 32(4): 354-359. DOI: 10.12125/j.chj.202005053
    引用本文: 王玲, 刘妍, 张琴. 急性心肌炎短期预后不良的影响因素及预测模型构建[J]. 心脏杂志, 2020, 32(4): 354-359. DOI: 10.12125/j.chj.202005053
    Ling WANG, Yan LIU, Qin ZHANG. Analysis of risk factors and construction of prediction model for short-term poor prognosis of acute myocarditis patients[J]. Chinese Heart Journal, 2020, 32(4): 354-359. DOI: 10.12125/j.chj.202005053
    Citation: Ling WANG, Yan LIU, Qin ZHANG. Analysis of risk factors and construction of prediction model for short-term poor prognosis of acute myocarditis patients[J]. Chinese Heart Journal, 2020, 32(4): 354-359. DOI: 10.12125/j.chj.202005053

    急性心肌炎短期预后不良的影响因素及预测模型构建

    Analysis of risk factors and construction of prediction model for short-term poor prognosis of acute myocarditis patients

    • 摘要:
        目的  探讨急性心肌炎住院期间预后不良的影响因素,并进一步构建预测模型。
        方法  回顾性分析2009年12月~2019年12月我院诊治的178例急性心肌炎患者的临床资料,依据住院期间转归分为主要不良心脏事件(major adverse cardiac events, MACE)组23例、非MACE组155例。比较2组一般资料、实验室检测指标、超声心动图、心电图指标及治疗情况,多因素logistic回归分析急性心肌炎患者住院期间发生MACE的危险因素,并构建预测模型,采用ROC曲线对预测模型的价值进行验证。
        结果  单因素分析显示,MACE组年龄、女性比例、重症心肌炎比例以及入院时呼吸频率、白细胞、中性粒细胞、血肌酐(serum creatinine, SCr)、空腹血糖、N末端脑钠尿肽、肌钙蛋白I、肌酸激酶同工酶MB(creatine kinase MB, CK-MB)、肌酸激酶、乳酸脱氢酶、丙氨酸氨基转氨酶、C反应蛋白、LVEF<40%、QRS间期>120 ms比率,应用IABP、ECMO、CRRT、呼吸机辅助、β受体阻滞剂、正性肌力/升压药比率均高于非MACE组,入院时收缩压低于非MACE组(P<0.05);多因素logistic回归分析显示,重症心肌炎(OR=2.89, 95%CI: 1.55-6.36, P<0.05)、女性(OR=2.55, 95%CI: 1.73-5.28, P=0.010)、LVEF<40%(OR=2.38, 95%CI:1.17-3.53, P<0.05)、QRS间期>120 ms(OR=3.23, 95%CI: 1.86-9.21, P<0.05)、CK-MB>23.56 u/L(OR=2.78, 95%CI: 1.35-5.76, P<0.05)、SCr>135.85 μmol/L(OR=1.49, 95%CI:1.16-2.83, P<0.05)是急性心肌炎住院期间发生MACE的危险因素;根据多因素logistic回归分析结果构建回归方程,并拟合为新变量Predict,ROC曲线分析显示,Predict预测急性心肌炎住院期间出现MACE的AUC为0.830(95%CI:0.771-0.888, P<0.01),截断点为0.576,敏感度78%,特异度为76%。
        结论  急性心肌炎患者院内发生MACE的危险因素较多,涉及患者性别、疾病类型、心肌损伤程度、心功能降低、心电图改变及并发肾损伤。根据危险因素构建的预测模型对急性心肌炎短期预后有较高的预测价值。

       

      Abstract:
        AIM  To explore the influencing factors for poor prognosis of acute myocarditis patients during hospitalization and to construct a predictive model.
        METHODS  The clinical data of 178 patients with acute myocarditis diagnosed and treated in our hospital from December 2009 to December 2019 were retrospectively analyzed and based on the outcome during hospitalization, they were divided into major adverse cardiac events (MACE) group (n=23) non-MACE group (n=155). General data, laboratory test indexes, echocardiogram, electrocardiogram indexes and treatment status were compared between the two groups. Multivariate logistic regression analysis was used to evaluate the risk factors of MACE during the hospitalization and a prediction model was constructed based on multivariate logistic regression analysis. ROC curve was used to verify the value of the model.
        RESULTS  Univariate analysis showed that age, female ratio, severe myocarditis ratio, respiratory rate, white blood cells, neutrophils, serum creatinine (SCr), fasting blood glucose, N-terminal B-type natriuretic peptide, Troponin I, creatine kinase MB (CK-MB), creatine kinase, lactate dehydrogenase, alanine aminotransferase, C-reactive protein, LVEF <40%, QRS interval> 120 ms, and the ratios of IABP, ECMO, CRRT, ventilator assist, β-blocker and positive inotropic/boosting drugs in MACE group were higher than those in non-MACE group, and systolic blood pressure in MACE group at admission was lower than that in non-MACE group (P<0.05). Multivariate logistic regression analysis showed that severe myocarditis (OR=2.89, 95%CI:1.55-6.36, P<0.05), female (OR=2.55, 95%CI:1.73-5.28, P=0.010), LVEF <40% (OR=2.38, 95%CI:1.17-3.53, P<0.05), QRS interval> 120 ms (OR=3.23, 95%CI:1.86-9.21, P<0.05), CK-MB>23.56 u/L (OR=2.78, 95% CI: 1.35-5.76, P< 0.05) and SCr>135.85 μmol/L (OR=1.49, 95% CI: 1.16-2.83, P<0.05) were the risk factors of MACE in acute myocarditis patients during hospitalization. Based on multivariate logistic regression analysis results, a regression equation was conducted fit it as a new variable Predict. ROC curve analysis showed that Predict predicted the occurrence of MACE in acute myocarditis patients during hospitalization was 0.830 (95%CI:0.771-0.888, P<0.01), sensitivity 78 %, the specificity 76% and the critical value 0.562.
        CONCLUSION  There are many risk factors for MACE in acute myocarditis patients, which are related to the patient’s gender, disease types, degrees of myocardial injury, reduced cardiac functions, electrocardiogram changes and renal injury. The prediction model constructed according to risk factors has a high predictive value for the short-term prognosis of acute myocarditis.

       

    /

    返回文章
    返回