Multivariate predictive model of the therapeutic effects of metoprolol in paediatric vasovagal syncope: a...

Metoprolol therapy for paediatric vasovagal syncope (VVS) has yielded inconsistent results, necessitating predictive markers. We aimed to develop and validate models to identify paediatric VVS patients likely to benefit from metoprolol.

478 metoprolol-treated paediatric patients with VVS were enrolled from three syncope units and divided into retrospective training (March 2017-March 2023, n = 323) and prospective validation cohorts (April 2023-March 2024, n = 155). Fourteen patients (2.9%) were excluded for lacking follow-up data. Patients were classified as responders or non-responders based on symptom improvement after 1-3 months of metoprolol therapy. Univariate analysis and logistic regression were used to select the candidate predictors. A nomogram and a scoring model were established to predict treatment efficacy. The model values were analysed using a receiver operating characteristic (ROC) curve. Consistency was evaluated using the Hosmer-Lemeshow (H-L) test, calibration curve, and concordance index (C-index). The clinical utility of model was assessed through the decision curve analysis (DCA). Internal validation was performed using the bootstrap approach. The predictive model derived from the training cohort was validated in the validation cohort to assess its accuracy and feasibility.

Increased heart rate during positive response in head-up tilt test (ΔHR), corrected QT interval dispersion (QTcd), and standard deviation of all normal-to-normal intervals (SDNN) were selected as independent predictors to develop a predictive model. A nomogram model was built (AUC: 0.900, 95% CI: 0.867-0.932); the H-L test and calibration curves showed a strong alignment between predicted and actual results. The scoring model was established in the training cohort (AUC: 0.941, 95% CI: 0.897-0.985), yielding a sensitivity of 82.8% and a specificity of 96.5%, with a cut-off value of 2.5 points. In the external validation cohort, the scoring model achieved a sensitivity, specificity, and accuracy of 93.6%, 80.9%, and 87.7%, respectively.

The nomogram and scoring model were constructed to predict the efficacy of metoprolol for children with VVS, which will greatly assist paediatricians in the individual management of VVS in children and adolescents.

This research was funded by National High-Level Hospital Clinical Research Funding (Clinical Research Project of Peking University First Hospital, grant number 2022CR59).