Background Holter monitors can provide temporal QT analysis, allowing for calculation of QT dispersion and corrected QT dispersion over time. We describe the results of the QT dispersion and corrected QT dispersion over a 24‐hour period in pediatric patients with long QT syndrome (LQTS), compared with a normal cohort, and develop 2 new machine learning models for diagnosis and arrhythmia risk prediction in LQTS. Methods A single‐center retrospective case–control study was conducted comparing patients with congenital LQTS and normal controls who underwent a clinically indicated Holter between January 2022 and May 2024. After identifying patients with cardiac events, 2 neural network models were constructed: (1) LQTS diagnostic prediction model, and (2) LQTS arrhythmia risk prediction model (resuscitated sudden cardiac arrest, ventricular tachycardia/fibrillation). Holter corrected QT interval (QTc) values (maximum, corrected QT dispersion and % time QTc >460 ms) were input variables. Models were trained on 90% of the patients and tested on the remaining 10%. Results A total of 73 patients with LQTS (age 12±6 years, 36% male) and 146 normal patients (age 13±4 years, 32% male) were identified with 193 and 146 Holter records respectively. Five patients had cardiac events. The LQTS diagnosis prediction model achieved an area under the curve of 0.92 with generalizability of area under the curve 0.91. Maximal QTc was the most influential input. The LQTS arrhythmia risk prediction model demonstrated an area under the curve of 0.91 with generalizability area under the curve of 0.87 with QTc dispersion being the most significant feature. Conclusions Holter monitoring QTc and corrected QT dispersion data have value in aiding in the diagnosis of LQTS and greater value in identifying at‐risk patients with LQTS.
Allam et al. (Wed,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: