[Predictive value of ventricular repolarization heterogeneity parameters for malignant arrhythmias in children with...

Objective: To explore the construction and clinical value of a predictive model for malignant ventricular arrhythmias (MVA) in children with viral myocarditis (VMC) based on parameters of ventricular repolarization heterogeneity. Methods: We retrospectively analyzed 99 children with VMC admitted to Henan Provincial People's Hospital from January 2019 to February 2023 as the training set, and 38 children with VMC admitted from March 2023 to May 2024 as the external validation set in a cross-sectional study. Patients were divided into MVA and non-MVA groups based on the occurrence of MVA. We measured QT interval (QT), corrected QT interval (QTc), T-wave peak-end interval (TpTe), and heart rate-corrected TpTe (TpTec) in all children using conventional electrocardiography and Holter monitoring. Multi-variate logistic regression analysis was used to identify factors influencing the occurrence of MVA in children with VMC and the efficacy of models predicting MVA with cardiac troponin I (cTnI) and different electrocardiographic parameters was evaluated using the receiver operating characteristic (ROC) curve. The predictive power of the models was internally and externally validated through C-index, Hosmer-Lemeshow goodness-of-fit test, ROC curves, and decision curve analysis. Results: There were 99 children in the training set and 38 in the external validation set, with no significant differences in age [(6.9±3.7) vs (6.9±3.4) years] and gender (boys: 46.5% vs 50.0%) between the two groups (all P>0.05). In the training set, there were 47 cases (47.5%) in the MVA group, and there were 20 cases (52.6%) in the MVA group in the external validation set. In both the training and external validation sets, there were statistically significant differences in cTnI, QTc, TpTec, and TpTe/QT between the MVA and non-MVA groups (P<0.01). TpTec [OR (95%CI): 1.123 (1.067-1.183)] and TpTe/QT [OR (95%CI): 1.026 (1.008-1.045)] were both associated with the occurrence of MVA in children with VMC. The Hosmer-Lemeshow goodness-of-fit test showed that there was no statistically significant difference between the predicted and observed values (P=0.294). The area under the ROC curve (AUC) (95%CI) for cTnI to predict MVA in VMC children was 0.651 (0.540-0.763), with a sensitivity of 0.617 and a specificity of 0.865 when the Youden's index was maximized (2 475.000). For TpTec, the AUC (95%CI) was 0.917 (0.859-0.975), with a sensitivity of 0.723 and a specificity of 0.981 when the Youden's index was maximized (110.500). For TpTe/QT, the AUC (95%CI) was 0.825 (0.746-0.905), with a sensitivity of 0.745 and a specificity of 0.750 when the Youden's index was maximized (0.265). The AUC (95%CI) for the combination of TpTec and TpTe/QT was 0.939 (0.892-0.987), with a sensitivity of 0.894 and a specificity of 0.865 when the Youden's index was maximized (0.472), which was higher than that of TpTe/QT (P<0.01). In the validation set, the accuracy and fit of the TpTec combined with TpTe/QT model were also good (P=0.800), and the AUC (95%CI) for predicting MVA was 0.992 (0.972-1.000), with a sensitivity of 0.936 and a specificity of 0.904 when the Youden's index was maximized (0.557). Conclusions: TpTec≥110.50 ms and TpTe/QT≥0.265 indicate a potential risk of MVA in children with VMC, requiring close clinical attention. The predictive value of the combined TpTec and TpTe/QT for MVA in children with VMC is higher than that of TpTe/QT alone.