Establishment and validation of a prediction model for myocarditis in Chinese children below 14 years old: a protocol...

Paediatric myocarditis, a rare inflammatory disease, often presents without clear early symptoms. Although cardiac troponin I levels can aid in diagnosing myocarditis, they are not definitive indicators. Troponin I levels frequently fluctuate within and outside the reference range, potentially causing misinterpretations by clinicians. Although a negative troponin I result is valuable for excluding myocarditis, its specificity is low. Moreover, the clinical diagnosis of paediatric myocarditis is exceptionally challenging, and accurate early-stage diagnosis and treatment pose difficulties. Currently, the Dallas criteria, involving cardiac biopsy, serves as the gold standard for myocarditis diagnosis. However, this method has several drawbacks and is unsuitable for children, resulting in its limited use.

In this study, we will employ multiple logistic regression analysis to develop a predictive model for early childhood myocarditis. This model will assess the patient's condition at onset and provide the probability of a myocarditis diagnosis. Model performance will be evaluated for accuracy and calibration, and the results will be presented through receiver operating characteristic (ROC) curves and calibration plots. Clinical decision curve analysis, in conjunction with ROC curve analysis, will be employed to determine the optimal cut-off value and calculate the net clinical benefit value for assessing clinical effectiveness. Finally, internal model validation will be conducted using bootstrapping.

Approval from the Clinical Research Ethics Committee of The Third Affiliated Hospital of Wenzhou Medical University has been obtained. The research findings will be disseminated through presentations at scientific conferences and publication in peer-reviewed journals.