A polysomnography study examining the association between sleep and postoperative delirium in older hospitalized cardiac surgical patients

Abstract

Hospitalized older patients who undergo elective cardiac surgery with cardiopulmonary bypass are prone to postoperative delirium. Self-reported shorter sleep and longer sleep have been associated with impaired cognition. Few data exist to guide us on whether shorter or longer sleep is associated with postoperative delirium in this hospitalized cohort. This was a prospective, single-site, observational study of hospitalized patients (>60 years) scheduled to undergo elective major cardiac surgery with cardiopulmonary bypass (n = 16). We collected and analysed overnight polysomnography data using the Somté PSG device and assessed for delirium twice a day until postoperative day 3 using the long version of the confusion assessment method and a structured chart review. We also assessed subjective sleep quality using the Pittsburg Sleep Quality Index. The delirium median preoperative hospital stay of 9 [Q1, Q3: 7, 11] days was similar to the non-delirium preoperative hospital stay of 7 [4, 9] days (p = .154). The incidence of delirium was 45.5% (10/22) in the entire study cohort and 50% (8/16) in the final cohort with clean polysomnography data. The preoperative delirium median total sleep time of 323.8 [Q1, Q3: 280.3, 382.1] min was longer than the non-delirium median total sleep time of 254.3 [210.9, 278.1] min (p = .046). This was accounted for by a longer delirium median non-rapid eye movement (REM) stage 2 sleep duration of 282.3 [229.8, 328.8] min compared to the non-delirium median non-REM stage 2 sleep duration of 202.5 [174.4, 208.9] min (p = .012). Markov chain modelling confirmed these findings. There were no differences in measures of sleep quality assessed by the Pittsburg Sleep Quality Index. Polysomnography measures of sleep obtained the night preceding surgery in hospitalized older patients scheduled for elective major cardiac surgery with cardiopulmonary bypass are suggestive of an association between longer sleep duration and postoperative delirium.

Trial registration: ClinicalTrials.gov NCT03498560.

Keywords: cardiac surgery; delirium; hospitalization; older; polysomnography; sleep.

Conflict of interest statement

CONFLICT OF INTEREST

OA has received speaker's honoraria from Masimo Corporation and is listed as an inventor on pending patents on EEG monitoring and oral dexmedetomidine that are assigned to Massachusetts General Hospital. All other authors declare that no competing interests exist.

© 2021 European Sleep Research Society.

Figures

FIGURE 1

Recruitment and patient flow diagram of the study population

FIGURE 2

Schematic of the study protocol. aMOCA, abbreviated Montreal Cognitive Assessment; Long-CAM, long confusion assessment method; PHQ-9, Patient Health Questionnaire-9; PSQI, Pittsburgh Sleep Quality Index

FIGURE 3

Five-state Markov chain model. Transition probabilities computed for 30-min intervals and averaged over three sleep periods, that is, early (7:00–0:00 PM) (a, b), intermediate (0:00–4:00 AM) (c, d) and late (4:00–8:00 AM) sleep (e, f). The top panel (green) corresponds to the non-delirious cohort and the bottom panel (red) corresponds to the delirious cohort. Arrows indicate the directions of sleep stage transitions. non-REM1, non-rapid eye movement (NREM) sleep stage 1; non-REM2, non-rapid eye movement sleep stage 2; non-REM3, non-rapid eye movement sleep stage 3; REM, rapid eye movement sleep stage