Abstract
Background:sEMGdia is currently a qualitative measure of assessing respiratory effort in children during sleep. A novel method for quantifying sEMGdia has been developed. Aim:To determine the interobserver reliability and agreement of analysing sEMGdia using this method. Method: Excerpts from different stages of sleep (light, deep, and REM sleep) during routine children's sleep studies were exported from the Compumedics Profusion 3 system to the Spike2 system for analysis. All sEMGdia signals from each excerpts were analysed by a Paediatric sleep physician(expert) and a medical student (novice). Inter-rater reliability and agreement were assessed using intra-class correlation coefficient (ICC) and Bland-Altman plot. Result: Excerpts from 20 children's sleep studies were analysed. The expert-novice inter-rater ICC was 0.93 (CI 0.56-0.98), 0.93 (CI 0.53-0.98), and 0.92 (CI 0.38-0.98) for excerpts from light, deep, and REM sleep respectively. The difference between the sEMGdi value calculated were graphed against the average sEMGdi
. The mean expert-novice difference for sEMGdi is 1.38uV +/- 1.20 (CI 0.80-1.97). Conclusion:Our method of analysing sEMGdi has an acceptable level of interobserver reliability between assessors with different levels of experience. These data support the potential use of sEMGdi as a non-invasive marker of respiratory effort in children.
- © 2014 ERS