Electrocardiographic Predictors of Left Ventricular Hypertrophy in Pediatric Hypertension

Prema Ramaswamy, Ekta Patel, Michael Fahey, Joseph Mahgerefteh, Irene D. Lytrivi, Juan C. Kupferman

Research output: Contribution to journalArticlepeer-review

10 Scopus citations

Abstract

Objective: To determine the efficacy of electrocardiography (ECG) in detecting left ventricular hypertrophy (LVH) in pediatric hypertension (HT). Study design: Concomitant echocardiograms and electrocardiograms in 108 children with HT were reviewed. Left ventricular mass (LVM), assessed by echocardiography, was used as a basis for a diagnosis of LVH (echo LVH) using accepted pediatric criteria. Using Wilcoxon's rank-sum test, 14 ECG variables were compared between subjects with and without echo LVH. Spearman correlations were used to examine the linear association between echo LVH and these ECG variables. The sensitivity and specificity of ECG in diagnosing LVH were computed. Results: Of the 108 subjects studied, 35 (32%) met the pediatric criteria for LVH; of these, 8 (7.4%) also met the adult criteria (>51 g/m2.7) for LVH. Mean values for only 5 ECG criteria differed significantly among the groups: RI, SaVR, RaVL, RI+SIII, and SVI+RV6 (P < .05). Significant correlations were found for several ECG criteria and at least 1 measure of LVM, but the magnitudes were modest. Standard ECG criteria predicted LVH with high specificity (>90%) but low sensitivity (<35%). RI >10 mm was identified as demonstrating a modestly improved positive likelihood ratio of ∼3. Conclusions: ECG is not an adequate predictor of LVH for clinical use in HT.

Original languageEnglish (US)
Pages (from-to)106-110
Number of pages5
JournalJournal of Pediatrics
Volume154
Issue number1
DOIs
StatePublished - Jan 2009
Externally publishedYes

ASJC Scopus subject areas

  • Pediatrics, Perinatology, and Child Health

Fingerprint

Dive into the research topics of 'Electrocardiographic Predictors of Left Ventricular Hypertrophy in Pediatric Hypertension'. Together they form a unique fingerprint.

Cite this