Reported rhythm and conduction abnormalities in children and adolescents with SARS-CoV-2 infection: a systematic review and meta-analysis

Background Rhythm and conduction abnormalities have been reported in children with COVID-19 or MIS-C, but their frequency and the extent to which they are attributable to SARS-CoV-2 infection remain uncertain. This systematic review and meta-analysis evaluated reported rhythm and conduction abnormalities and comparative arrhythmia risk in children and adolescents with SARS-CoV-2 infection. Methods We searched PubMed, Scopus, and Web of Science from inception to December 10, 2025, with a final update on September 3, 2026. Observational studies reporting rhythm or conduction abnormalities, or data allowing estimation of relative risk (RR), in children and adolescents with COVID-19 were included. Logit-transformed random-effects models with restricted maximum likelihood estimation and truncated Knapp-Hartung standard errors were used to pool reported proportions. An outcome-definition sensitivity analysis excluded isolated sinus tachycardia and sinus bradycardia where a unique eligible participant-level count could be determined. Results Twenty-three observational studies involving 526,597 children and adolescents were included. Among 18 clinically monitored or clinically ascertained cohorts (n = 5074), the random-effects mean proportion of reported rhythm or conduction abnormalities was 13.5% (95% CI, 7.9%–22.2%; 95% prediction interval, 1.3%–64.4%; I² = 93.45%). In the outcome-definition sensitivity analysis excluding isolated sinus tachycardia and bradycardia, the pooled proportion was 7.8% (95% CI, 4.0%–14.8%) among 13 clinical studies. The crude arrhythmia RR recalculated from Zhang et al.’s overall event counts was 1.26 (95% CI, 1.11–1.43), whereas the RR derived from Gordon et al.’s propensity-score-matched cohorts was 1.24 (95% CI, 1.17–1.33). Conclusions Reported rhythm and conduction abnormalities varied widely across selected pediatric COVID-19 and MIS-C cohorts. The descriptive mean of 13.5% in 18 clinical cohorts had a wide 95% prediction interval (1.3%–64.4%), indicating that this mean provides limited guidance for any individual clinical setting, while population-level database studies reported much lower absolute proportions. The contrast across data sources and the restricted outcome-definition analysis highlight the importance of clinical setting, ascertainment, and outcome reporting when interpreting these estimates. Because the clinical cohorts lacked COVID-negative comparators, their pooled proportion cannot establish SARS-CoV-2-attributable risk or the frequency of arrhythmias requiring intervention.