Establishment And Evaluation Of Early Warning Scoring System For Children With Fulminant Myocarditis

Objective:Acute myocarditis(AM)in children is a rare but serious disease.Fulminant myocarditis(FM)is a special subtype of AM,characterized by rapid progression of acute congestive heart failure,fatal arrhythmia,cardiogenic shock and even sudden death.There is a very high mortality rate in the acute phase,and children who survive tend to have a better prognosis.Early identification,emergency intervention and advanced mechanical circulation support can improve the survival rate and prognosis of FM children.Based on this,we designed this study in order to summarize the risk factors affecting AM from the clinical characteristics of AM and FM children,and then construct AM risk prediction model and FM early warning scoring system,which provide a useful assessment tool for clinical practice and contribute to the early identification of FM.Methods:Demographic,first symptom,laboratory examination,auxiliary examination of instruments,prognosis and other data of children hospitalized and diagnosed with AM in the children’s hospital of suzhou university from January 2017 to December 2018 were collected.The subjects were divided into AM group and FM group.Univariate analysis showed significant difference between the two groups(P<0.01).Lasso regression was used to analyze the clinical characteristics of the two groups.The simplest model at lambda=lambda.1se was used to screen the high-related risk factors of FM.Logistic regression was used to screen the independent risk factors.Finally,regression equation prediction model was established.C index,calibration plot and calibration curve were used to evaluate the discrimination and calibration of the model.Partial regression coefficient was rounded after adjustment,the FM early warning scoring system was constructed and the prediction probability was calculated.Finally,the effectiveness of clinical decision curve was analyzed(DCA analysis).Results:A total of 103 children with AM were enrolled in this study,including 56 males and 47 females,with a male-to-female ratio of 1.2:1.The age of onset showed a bimodal distribution.Infants under 3 years old and children under 8 years old had the highest incidence.A total of 37 risk factors were found by univariate analysis between the two groups.Lasso regression variable screening analysis was used to screen out 8 high risk factors,including hepatomegaly,shock,anhelation,mild atrioventricular block,high atrioventricular block,BNP,IgM and D dimer.After logistic regression,the four independent risk factors were hepatomegaly,shock,anhelation and high atrioventricular block.The regression equation was In(p/(1-p)=0.02588+0.51147*(hepatomegaly)+0.72441*(shock)+0.30488*(anhelation)+0.27977*(high atrioventricular block).The area under characteristic curve(AUC)of model subjects was 0.964,with 93.6%specificity and 96.6%sensitivity.Calbration plot and calibration curve analysis show that the model has a high degree of discrimination and calibration.An early-warning scoring system consisting of 4 indicators with a total score of 7 was constructed.The DCA curve analysis shows that the scoring system is effective and can be used in clinical practice.Conclusion:The incidence of acute myocarditis in children showed a bimodal distribution.The highest incidence was found in infants under 1 year old and 8 years old.hepatomegaly,shock,anhelation and high atrioventricular block are independent risk factors for AM.The predictive model has high discrimination and calibration,concise,convenient and accurate clinical use,and has high practical value.Effectiveness analysis of clinical decision curve of early warning scoring system shows that patients with this scoring system have a high net benefit rate,from which patients can benefit significantly.