P101 CNS-active medications do not increase sedation-like EEG amplitudes during sleep.

Abstract Background Sedation and sleep have distinct electroencephalogram (EEG) properties. The presence of EEG patterns suggestive of sedation during sleep may have negative health effects. It has been shown that sedated intensive care unit (ICU) patients can be differentiated from matched sleep study patients not on Central Nervous System (CNS)-active medications using the amplitude criteria derived from cumulative frequency curves (CFCs). Aim To determine whether CFCs can be used to distinguish EEG patterns resembling sedation in patients with and without CNS-active medications. Methods Sleep study EEG data was analysed from 163 adult patients. Patients were categorised based on CNS medication status. 92 patients had no CNS medications, while 71 had CNS medications. Medications were sorted into the 7 CNS drug classes. Studies with poor signal quality, AHI > 5, were excluded. An automated analysis was applied to the left frontal lobe EEG trace and patients were compared on peak amplitude. Results Females and patients (<30yrs) spent more time at higher amplitudes. Patients with one or more CNS medications spent slightly more time (mean 7.46; SD 0.93 hours) asleep, vs unmedicated patients (mean 7.36; SD 1.01 hours; not statistically significant). Discussion Contrary to our original hypothesis, there was no significant difference in time at higher EEG amplitudes between CNS-medicated and non-medicated patients. It appears that in this population sample, CNS medications do not significantly affect sleep architecture but do slightly (not significantly) increase sleep duration. This study had limitations – we aim to use adequate sample sizes for each class of drug in future.