Utility of incomplete right bundle branch block as an isolated ECG finding in children undergoing initial cardiac evaluation.

Abstract

This study evaluates the ability of experienced pediatric electrophysiologists (EPs) to reliably classify incomplete right bundle branch block (IRBBB) and assesses its clinical utility as an isolated ECG finding in a group of healthy outpatient children without prior cardiac evaluation. We performed a retrospective analysis of all electrocardiographic and echocardiographic records at Boston Children's Hospital between January 1, 2005, and December 31, 2014. Echocardiographic diagnoses were identified if registered between the date of the index electrocardiogram and the ensuing year. A selected subset of 473 ECGs was subsequently reanalyzed in a blinded manner by six pediatric EPs to determine the consistency with which the finding of IRBBB could be assigned. Of the 331278 ECGs registered in the BCH database, 32127 (9.7%) met inclusion criteria and were analyzed for the prevalence of isolated right bundle conduction disturbance findings. The mean age was 12.1±4.0years, and the population was 49% male. Of the 32127 ECGs, 72.5% were coded normal, 3.0% were coded IRBBB, and 0.5% were coded complete right bundle branch block (CRBBB). A total of 7.3% of patients coded as normal had an ensuing echocardiogram, compared to 12.5% coded IRBBB. Echo findings were recorded in 0.1% of normal and 0.2% of IRBBB. Patients with ASD-secundum type were no more likely to have isolated IRBBB on previous ECG than the general population (2.5% vs 3.0%). Analysis of inter-reader variability in ECG findings and conduction disturbance identification was high (range of IRBBB prevalence 1-20% among readers). Reinterpretation of ECGs using explicit diagnostic criteria did not demonstrate consistent discrimination of IRBBB and Normal ECGs. IRBBB is not uncommon in a healthy school age population and is observed to have high inter-reader variability. It was associated with increased use of echocardiographic exam but was not associated with increased rate of echocardiographic findings when compared with rates for normal ECGs.