Sleep and Delirium in the Intensive Care: A Pre-post Cohort Study Following Implementation of a Unit-wide Sleep Bundle

Abstract

Introduction: Intensive Care Unit (ICU) patients commonly report disrupted sleep that has been linked to development of delirium. Implementation of a sleep bundle may reduce sleep disturbances and delirium incidence and duration. Objectives: To compare ICU patients’ self-reported sleep assessments and delirium status before and after implementation of a unit-wide sleep bundle. Methods: A before-and-after study incorporating implementation of a sleep bundle was conducted. Patients reported sleep using the RichardsCampbell Sleep Questionnaire [RCSQ], a validated five-item visual analogue scale where aspects of sleep are scored from 0 (poor) to 100 (excellent). Patients also reported strategies that helped or hindered their sleep. Delirium was assessed using the Confusion Assessment Method-ICU. Differences between groups were compared using median (interquartile range), p < 0.05 was considered significant. Results: 188 participants were recruited [101 pre-intervention (237 observations); 87 post-intervention (210 observations)]. Participants were generally in their late 50s and more than half were male. Participants in the post-intervention period were able to report sleep on a greater proportion of ICU days (pre:36% vs post:54%, p ¼ 0.02), although no change in overall sleep score was reported [46(24e66) vs 49(20e80), p ¼ 0.30]. Although there was no change in the number of patients who had delirium during their ICU stay, there was a reduction in the days of delirium patients experienced [pre-intervention 0(0e2) vs post-intervention 0(0-1), p < 0.01]. Post-intervention, patients with no delirium reported better scores for “falling asleep” [59(20e80) vs 33(20e65); p ¼ 0.02] and for “returning to sleep” [50(23e80) vs 30(15e70); p ¼ 0.02] than those who experienced delirium. They also reported less noise-related items (n ¼ 51) that woke them or kept them awake overnight compared with 114 reports in the preintervention period. Conclusion(s): There was no significant improvement in the patient reported sleep quality, however there was a reduction in days of delirium following the implementation of a unit-wide sleep bundle