Establishment Of Risk Prediction Model And Risk Score For In-hospital Mortality In The Patients With Myocarditis

Objective: Risk prediction model and risk score for in-hospital mortality in the patients withmyocarditis were established to quickly and objectively evaluate the criticality of the inpatients withmyocarditis. According to the evaluated score, physicians could save the patients on the verge of death bypositive and predictable treatments in a great extent.Method: A retrospective cohort study was done toanalysis the clinical medical records of403patients with myocarditis in Ningbo First Hospital from Jan.1,2003to Jan.1,2014. There were238males (59%) and165females (41%) in the include patients. On thebasis of clinical outcomes, they were divided into two groups: death and survival. Firstly, we established arisk prediction model for death by means of the univariate and multivariate logistic regression analysis inSPSS21version. Secondly, Hosmer-Lemeshow (H-L) test and the area of under the Receiver OperatingCharacteristic (ROC) were assessed to investigate the Goodness of Fit and Forecast Accuracy respectively.Finally, the risk score of death in hospital was made in accordance with the coefficient β in the logisticregression model. Results: Six independent risk factors associated with death in hospital and each scoreaccording to the regression model in the patients with myocarditis were presented as followed:Ccr<60ml/min (OR:19.936，95%CI:5.657-70.255，3points), age≥50(OR:7.425,95%:2.176-25.339,2points), history of ventricular tachycardia (OR:6.886,95%CI:1.864-25.436,2points), NYHA Ⅲ-Ⅳ(OR:4.029,95%CI:1.133-14.324,1point)、 male (OR:3.476,95%CI:0.990-12.198,1point) andtroponin-I≥50μg/L (OR:3.101,95%CI:0.906-10.618,1point). In accord with the score, each patient wouldcalculate his/her points. The following regions revealed different risk:0-3points in low risk,4-6points inmiddle risk and7-10points in high risk. Hosmer-Lemeshow (H-L) and the area of under the ReceiverOperating Characteristic (ROC) test revealed the approving outcomes of Goodness of Fit and ForecastAccuracy respectively. In addition, we found that69.2%patients died in the three days from admission indeath group. The data prompted that it was vital to give the inpatients active treatments in the three days,which would transfer the outcome in the future probably. Conclusion: Ccr<60ml/min, age≥50, history ofventricular tachycardia, NYHA Ⅲ-Ⅳ、male and troponin-I≥50μg/L in admission were the independent riskfactors association with in-hospital mortality in the patients with myocarditis. Moreover, the interval ofthree days after admission was vital to the outcome of patients in the future probably.