Abstract
This single-center prospective observational study aimed to evaluate sleep architecture
in mechanically ventilated pediatric intensive care unit (PICU) patients receiving
protocolized light sedation. We enrolled 18 children, 6 months to 17 years of age,
receiving mechanical ventilation and standard, protocolized sedation for acute respiratory
failure, and monitored them with 24 hours of limited (10 channels) polysomnogram (PSG).
The PSG was scored by a sleep technician and reviewed by a pediatric sleep medicine
physician. Sixteen children had adequate PSG data for sleep stage scoring. All received
continuous opioid infusions, 15 (94%) received dexmedetomidine, and 7 (44%) received
intermittent benzodiazepines. Total sleep time was above the age-matched normal reference
range (median 867 vs. 641 minutes, p = 0.002), attributable to increased stage N1 and N2 sleep. Diurnal variation was
absent, with a median of 47% of sleep occurring during night-time hours. Rapid eye
movement (REM) sleep was observed as absent in most patients (n = 12, 75%). Sleep was substantially disrupted, with more awakenings per hour than
normal for age (median 2.2 vs. 1.1, p = 0.008), resulting in a median average sleep period duration (sleep before awakening)
of only 25 minutes (interquartile range [IQR]: 14–36) versus normal 72 minutes (IQR:
65–86, p = 0.001). Higher ketamine and propofol doses were associated with increased sleep
disruption. Children receiving targeted, opioid-, and dexmedetomidine-based sedation
to facilitate mechanical ventilation for acute respiratory failure have substantial
sleep disruption and abnormal sleep architecture, achieving little to no REM sleep.
Dexmedetomidine-based sedation does not ensure quality sleep in this population.
Keywords
intensive care units - pediatric - sleep stages - polysomnography - sleep - rapid
eye movement - dexmedetomidine - hypnotics and sedatives
