Background: Impairment of sleep quality and quantity has been described in critically ill patients. Delirium, an organ dysfunction that affects outcome of the critically ill patients, is characterized by an acute onset of impaired cognitive function, visual hallucinations, delusions, and illusions. These symptoms resemble the hypnagogic hallucinations and wakeful dreams seen in patients with neurological degenerative disorders and suffering of disorders of rapid eye movement (REM) sleep. We assessed the characteristics of sleep disruption in a cohort of surgical critically ill patients examining the hypothesis that severe impairments of rapid eyes movement (REM) sleep are associated to delirium.
Methods: Surgical patients admitted to the intensive care units of the San G. Battista Hospital (University of Turin) were enrolled. Once weaning was initiated, sleep was recorded for one night utilizing standard polysomnography. Clinical status, laboratory data on admission, co-morbidities and duration of mechanical ventilation were recorded. Patients were a priori classified as having a "severe REM reduction" or "REM reduction" if REM was higher or lower than 6% of the total sleep time (TST), respectively. Occurrence of delirium during intensive care unit (ICU) stay was identified by CAM-ICU twice a day. Multivariate forward stepwise logistic regression analysis was performed with sleep ("severe REM reduction" vs. "REM reduction") as the a priori dependent factor.
Results: REM sleep amounted to 44 (16-72) minutes [11 (8-55) % of the TST] in 14 patients ("REM reduction") and to 2.5 (0-36) minutes [1 (0-6) % of the TST] in the remaining 15 patients ("severe REM reduction") (P = 0.0004). SAPS II on admission was higher in " severely REM deprived" then in "REM deprived" patients. Delirium was present in 11 patients (73.3%) of the patients with "severe REM reduction" and lasted for a median of 3 (0-11) days before sleep assessment, while only one patient having "REM reduction" developed delirium that lasted for 1 day. The factors independently associated with a higher risk of developing "severe REM reduction" were delirium and daily dosage of lorazepam.
Conclusion: The present study shows that while all critically ill patients present a profound fragmentation of sleep with a high frequency of arousals and awakenings and a reduction of REM sleep, a percentage of patients present an extremely severe reduction of REM sleep. Delirium and daily dosage of lorazepam are the factors independently associated to extremely severe REM sleep reduction.