Objective:To establish and verify a novel fulminant myocarditis(FM)diagnostic model based on LASSO regression.Methods:One hundred and seven cases of FM definitely diagnosed in Tongji Hospital affiliated to Huazhong University of Science and Technology Tongji Medical College from January 2021 to March 2023 were retrospectively analyzed and divided into a modeling group and a validation group.There were sixty-nine patients with FM and thirty-eight patients with non-fulminant myocarditis(NFM).The LASSO regression algorithm was used to identify parameters associated with the highest probability for FM.Multivariate Logistic regression analyses based on Akaike information criterion(AIC)was constructed to determine important variables and receiver operating characteristic curve was drawn to evaluate the model effectiveness.To establish a nomogram model which then was verified internally.Results:Five predictive factors were screened by the LASSO regression and AIC and a nomogram prediction model were established;the five predictors included systolic blood pressure(OR=0.92,95％CI 16.57-18.49),cardiac troponin I(OR=1.55,95％CI 0.95-2.53),N-terminal pro-brain natriuretic peptide(OR=18.92,95％CI 1.82-196.49),the rate of change in global longitudinal strain of the left ventricle within 24 h(OR=8.71,95％CI 1.26-60.05)no less than 30％,the rate of change in left ventricular ejection fraction within 24 h no less than 30％(OR=17.51,95％CI 0.05-17.12).The area under the curve of the modeling group and the validation group were respectively 0.960(95％CI 0.919-1.000)and 0.907(95％CI 0.793-1.000).Conclusion:This study established and validated an early prediction model consisting of five clinical indexes to identify potential fulminant myocarditis patients in early stage,which is of some predictive ability.
NETL
NSTL
万方数据
