The accuracy of simple, feasible alternatives to polysomnography for assessing sleep in intensive care: An observational study

Abstract

Background

Sleep disturbance is common in intensive care patients. Understanding the accuracy of simple, feasible sleep measurement techniques is essential to informing their possible role in usual clinical care.

Objective

The aim of the study was to investigate whether sleep monitoring techniques such as actigraphy (ACTG), behavioural assessments, and patient surveys are comparable with polysomnography (PSG) in accurately reporting sleep quantity and quality among conscious, intensive care patients.

Methods

An observational study was conducted in 20 patients admitted to the intensive care unit (ICU) for a minimum duration of 24 h, who underwent concurrent sleep monitoring via PSG, ACTG, nursing-based observations, and self-reported assessment using the Richards–Campbell Sleep Questionnaire.

Results

The reported total sleep time (TST) for the 20 participants measured by PSG was 328.2 min (±106 min) compared with ACTG (362.4 min [±62.1 min]; mean difference = 34.22 min [±129 min]). Bland–Altman analysis indicated that PSG and ACTG demonstrated clinical agreement and did not perform differently across a number of sleep variables including TST, awakening, sleep-onset latency, and sleep efficiency. Nursing observations overestimated sleep duration compared to PSG TST (mean difference = 9.95 ± 136.3 min, p > 0.05), and patient-reported TST was underestimated compared to PSG TST (mean difference = −51.81 ± 144.1 7, p > 0.05).

Conclusions

Amongst conscious patients treated in the ICU, sleep characteristics measured by ACTG were similar to those measured by PSG. ACTG may provide a clinically feasible and acceptable proxy approach to sleep monitoring in conscious ICU patients.