Skip to content

ICU sleep protocol shows an uncertain signal for less delirium

A randomized trial in older surgical patients suggested possible benefit, without a measurable improvement in reported sleep quality.

By 100HP editorialAbstract-based explanation checked

Based on the published abstract. The full paper may contain additional methods, results and limitations.

The 30-second takeaway

Older surgical ICU patients assigned to a sleep-support protocol had fewer delirium cases than those receiving usual care. The estimate remained uncertain, with a credible interval compatible with benefit or harm. Sleep-quality scores and other clinical outcomes did not differ significantly, making this an early signal that needs confirmation.

Keep in mind

The authors call these findings hypothesis-generating and seek larger multicenter trials. Sleep was assessed subjectively, and use of earplugs and eye masks was only moderate.

THE NUMBERS, WITH CONTEXT

What researchers found

Bayesian OR 0.55; 95% credible interval 0.22–1.31

Delirium during ICU care

Odds ratio for the sleep protocol versus usual care; delirium occurred in 10.1% versus 17.0% over 7 days.

J Intensive Care, 2025 · Original source ↓

What the protocol included

The protocol combined changes to noise, light, doors and alarms with earplugs and eye masks. Because these measures were delivered together, the abstract cannot identify whether a particular component contributed to the delirium pattern.

How to read the uncertainty

The Bayesian analysis favored a reduction in delirium, but uncertainty remained wide. The absence of a significant sleep-score difference also means the trial did not establish improved perceived sleep as the explanation for that pattern.

CHECK THE ORIGINAL

The original publication

Effects of a multicomponent sleep enhancement protocol on delirium incidence in elderly critically ill surgical patients: a randomized controlled trial.

Kitisin N, Lao-Amornphunkul S, Hemtanon N et al.
J Intensive Care · 2025

PubMed ID
41185060
Record checked

AI-assisted research and writing. This explains one selected publication; it is not a complete review of everything known. Our approach.