Abstract Introduction Ventricular tachycardia (VT) and ventricular fibrillation (VF) are life- threatening arrhythmias, with non-sustained ventricular tachycardia (NSVT) serving as a potential precursor. Identifying individuals at risk of NSVT is crucial for effective management and prevention of sudden cardiac death in high-risk patients. In daily clinical practice, premature ventricular complex (PVC) and sometimes NSVT are frequently detected on 24-hour ambulatory rhythm monitoring. Purpose In this study, we aimed to predict the development of NSVT by analyzing PVC morphology and intervals in 24- hour ambulatory rhythm monitorings. Methods A retrospective analysis of 24-hour ambulatory rhythm monitoring recordings from 2018 to 2023 was conducted. Inclusion criteria required clear measurement of PVC intervals, at least 100 PVC in the recording, and two preceding normal QRS beats. NSVT diagnosis was based on four consecutive PVC lasting <30 s and a rate >100/min. Various PVC-related parameters were analyzed, and new variables were derived. Statistical analysis included Mann-Whitney U tests and ROC analysis. Results The study included 202 patients, revealing significant differences in several VES temporal variables between groups with only PVC (Group I) and PVC with NSVT (Group II). Notably, Age-Adjusted Sum of Intervals Index (AASI Index-(100 x [(Tp-to- QRSvpa)+(coupling time)+(Qtc1)]/age), long coupling times, extended Tp-to-QRSvpaintervals, and prolonged Tpe-vpa intervals were associated with NSVT development. Age played a pivotal role in predicting NSVT when integrated into a specific variable AASI index. The optimal cut-off value was 4.21, with 84% sensitivity and specificity. Conclusion The study suggests that ventricular arrhythmias may be predicted using temporal parameters of PVC, especially when age is included as a modifier. The AASI index calculated on a PVC and preceeding two normal beats on 24-hour ambulatory rhythm monitoring seems to be promising parameter. These findings hold potential for future development of artificial intelligence-based algorithms for use in any PVC to identify at-risk patient groups, provided larger datasets are utilized to refine predictive models.Variables included in the analysis for cDefinitions of variables and ROC curve
Read full abstract