Temperature rhythms and ICU sleep: the TRIS study

BACKGROUND: Core body temperature (CBT) patterns associated with sleep have not been described in the critically ill. This study aimed to characterize night-time sleep and its relationship to CBT in ICU patients. METHODS: A prospective study was performed in a 27-bed tertiary adult intensive care unit of 20 mechanically ventilated patients in the weaning stage of their critical illness. The study assessed sleep by polysomnography (PSG) during the evening between 21:00-7:00 hours, nursing interventions using the Therapeutic Intervention Scoring System (TISS), illness severity using SOFA and APACHE II scores and CBT 24-hour pattern. RESULTS: Patients were awake for approximately half the study period (45.04%. IQR 13.81-77-17) with no REM (0%, IQR 0-0.04%) and median arousals of 19.5/hour (IQR 7.1-40.9). The 24-hour CBT had a rhythmic pattern in 13 (65%) patients with a highly variable phase of median peak time at 17:35 hours (IQR 12:40-19:39). No significant associations were found between CBT rhythmicity, sleep stages, sleep EEG frequency density, illness severity scores or TISS on the day of PSG. There was no relationship between time awake and CBT rhythmicity (P=0.48) or CBT peak time (P=0.82). The relationship between circadian rhythms and sleep patterns in the critically ill is complex. CONCLUSIONS: Patients recovering in ICU commonly have CBT loss of rhythmicity or a significant phase shift with loss of normal night-time patterns of sleep architecture. Appropriate care plans to promote sleep and circadian rhythm require further investigation of contributing factors such as environment, clinical care routines, illness type and severity.