Sleep Architecture Patterns in Critically Ill Patients and Survivors of Critical Illness - A Retrospective Study.

RATIONALE: Sleep abnormalities are very frequent in critically ill patients during and following ICU stay. Their mechanisms are poorly understood. The odds ratio product (ORP) is a continuous metric (0.0-2.5) of sleep depth measured in 3-sec intervals and derived from the relationship of powers of different EEG frequencies to each other. When expressed as percent of epochs within 10 ORP deciles covering the entire ORP range, it provides information about mechanism(s) of abnormal sleep.

OBJECTIVES: To determine ORP architecture types in critically ill patients and survivors of critical illness who had previously undergone sleep studies.

METHODS: Nocturnal polysomnograms of 47 un-sedated critically-ill patients and 23 survivors of critical illness at hospital discharge were analyzed. Twelve critically-ill patients were monitored also during the day and 15 survivors underwent another polysomnogram 6 months after hospital discharge. In all polysomnograms each 30-sec epoch was characterized by the mean ORP of the ten 3-sec epochs. The number of 30-sec epochs with mean ORP within each of 10 ORP deciles covering the entire ORP range (0.0-2.5) was calculated and expressed as % of total recording time. Thereafter, each polysomnogram was characterized by a 2-digit ORP type with the first digit (range 1-3) reflecting increasing levels of deep sleep (ORP<0.5, deciles 1 and 2) and the second digit (range 1-3) reflecting increasing levels of full wakefulness (ORP>2.25, decile 10). Results of patients were compared to 831 age- and gender-matched community dwellers free of sleep disorders.

RESULTS: In critically ill patients, types 1,1 and 1,2 (little deep sleep and little or average full wakefulness) dominated (46% of patients). In the community, these types are uncommon (<15%) and seen primarily in disorders that preclude progression to deep sleep (e.g., very severe OSA). Next in frequency (22%) was type 1,3, consistent with hyperarousal. Day ORP sleep architecture was similar to night results. Survivors had similar patterns with little improvement after 6 months.

CONCLUSIONS: Sleep abnormalities in critically ill patients and survivors of critical illness result primarily from stimuli that preclude progression to deep sleep or from presence of a hyperarousal state.