Sleep Quality and Delirium Occurrence in Intensive Care Unit Patients: A Prospective Observational Study

Patients in the intensive care unit (ICU) frequently experience sleep disturbances. Although circadian rhythm disruption caused by poor sleep is known to increase the risk of delirium, empirical research on this association remains limited.

To investigate the association between sleep quality and the development of delirium in ICU patients.

This prospective observational study was conducted in six ICUs across three university hospitals in South Korea. Sleep parameters were measured using the Fitbit Sense wearable device, and delirium was assessed with the confusion assessment method for the ICU. We performed multivariable logistic regression to examine the association between sleep quality and delirium occurrence.

A total of 357 non-delirious patients were included in the study. Delirium occurred in 19.3% of patients during their ICU stay. On the second day of the ICU stay, 9.3% screened positive for delirium. Shorter sleep time measured in minutes (OR = 0.997, 95% CI: 0.993-1.000, p = 0.046) and a higher wake after sleep onset ratio (OR = 1.050, 95% CI: 1.010-1.091, p = 0.014) on the first night were statistically significantly associated with delirium on the second day. On the third day, 14.9% of patients screened positive. Delirium on the third day was associated with greater disease severity (OR = 2.879, 95% CI: 1.417-5.847, p = 0.003) and the presence of delirium on the second day (OR = 262.133, 95% CI: 14.247-4823.137, p < 0.001). There was no statistically significant association between sleep parameters from the previous night and delirium on the third day.

Poor sleep quality during the first night in the ICU-especially reduced sleep time and increased wake after sleep onset-may be linked to the development of delirium.

Critical care nurses should closely monitor patients' sleep and consider early interventions to support circadian rhythm preservation as part of delirium prevention strategies.

The study protocol was registered with the Clinical Research Information Service (https://cris.nih.go.kr); KCT0007851.