Does sleep or sleep deprivation increase epileptiform discharges in pediatric electroencephalograms?

Abstract

Sleep deprivation before obtaining an electroencephalogram (EEG) is believed both to increase the likelihood of sleep during an EEG and to increase the detection of interictal epileptiform discharges. However, depriving a child of sleep poses a burden on both the parent and the child. The objective of this study was to compare the effects of sleep, standard sleep deprivation, partial sleep deprivation, and no sleep deprivation on the odds of an epileptiform abnormality in outpatient pediatric EEGs. Data were collected from all pediatric EEGs performed at a busy, university-based neurologic practice during two 2-month periods. During the first period, all EEGs were performed as ordered, either standard sleep-deprived (SSD) or non-sleep-deprived (NSD). During the second 2 months, SSD EEGs were performed per routine. However, non-SSD families were instructed to keep their children awake 2 hours later the night before the EEG. Those who complied were classified as partially sleep-deprived (PSD). Patient characteristics across protocols were compared with chi(2) and analysis of variance tests as appropriate. The odds of epileptiform and abnormal findings associated with sleep, NSD, PSD, and SSD EEGs were calculated using logistic regression. Of 820 eligible EEGs, sleep occurred in 22% of NSD, 44% of PSD, and 57% of SSD EEGs. The sample size of this study allowed for an 85% power, with alpha of.05, to detect an absolute increased EEG yield of 10%. Neither the presence of sleep (odds ratio [OR]: 0.99; 95% confidence interval [CI]: 0.69-1.42) nor the use of PSD (OR: 0.90; 95% CI: 0.50-1.62) or SSD (OR: 0.96; 95% CI: 0.63-1.47) protocols increased the odds of epileptiform EEGs. Sleep deprivation should not be used routinely to increase the yield of pediatric EEGs.