J Electrocardiol. 2025 Aug 20;92:154101. doi: 10.1016/j.jelectrocard.2025.154101. Online ahead of print.
ABSTRACT
INTRODUCTION: Age-related shifts in pediatric ECG morphology are well described, yet the classic maturation milestones have never been tested head-to-head in a case-control design that yields diagnostic-accuracy metrics for congenital heart disease (CHD).
METHODS: In this retrospective diagnostic-accuracy study we reviewed 1637 12‑lead ECGs recorded in 2023 at a Brazilian tertiary cardiac center and its tele-ECG network. CHD was confirmed in 349 children; 1288 children with structurally normal hearts served as controls. An ECG was classified negative when R > S in V6 and an inferior axis were present and ≤ 35 months: no rsR’ > 120 ms, pure-R, qR or pure-q in V1; 36-95 months: above criteria plus either R < S in V1 or R ≥ S but <6 mm; ≥ 96 months: above criteria plus S > R in V1. Any other pattern was deemed positive. Sensitivity, specificity and likelihood ratios (LR+, LR-) were calculated for three prespecified age strata.
RESULTS: Infants (1-35 months): sensitivity 67.5 % (95 % CI 60.4-74.5), specificity 88.8 % (86.0-91.7), LR+ 6.05, LR- 0.37. Transition pattern (36-95 months): sensitivity 66.7 % (57.4-76.0), specificity 87.0 % (83.7-90.2), LR+ 5.11, LR- 0.38. Adult-like pattern (≥ 96 months): sensitivity 71.6 % (61.8-81.4), specificity 90.5 % (87.6-93.4), LR+ 7.52, LR- 0.31.
CONCLUSIONS: A rapid age-tiered analysis delivers moderate rule-in (LR+ ≈ 5-8) and rule-out power (LR- ≈ 0.3-0.4). The tool is not a stand-alone diagnostic test. Full waveform interpretation remains essential once a tracing is flagged.
CLINICAL TRIAL REGISTRATION: None.
PMID:40850022 | DOI:10.1016/j.jelectrocard.2025.154101
