Bright light for delirium prevention: a hopeful but uncertain signal

Design

Systematic review and random-effects meta-analysis

Population

8 studies, 1,190 hospitalised adults

Primary outcome

Delirium incidence

Effect

RR 0.64 (95% CI 0.38–1.07), I² 52%; very low certainty

This systematic review and meta-analysis, published in BMJ Open on 28 September, found eight studies with 1,190 hospitalised adults testing bright light therapy to prevent delirium. Most used morning light of at least 5,000 lux for one to two hours over three to four days.

The pooled estimate favoured light (RR 0.64) but the confidence interval crossed 1 (0.38 to 1.07), heterogeneity was moderate and several studies were at high risk of bias. Certainty was very low. Adherence and adverse effects were poorly reported.

Multicomponent delirium prevention, including orientation, sleep, mobility, hydration and sensory aids, remains the approach with evidence. Daylight exposure and a normal day-night rhythm on the ward cost little and fit within that, but dedicated light devices are not yet justified.

• Keep multicomponent delirium prevention as the standard on wards.
• Open curtains by day and dim lights at night to support sleep-wake rhythm.
• Do not buy dedicated light devices on this evidence alone.
• Record delirium screening results to judge any local change.

Why it matters

A cheap, safe idea still lacks the evidence to justify dedicated equipment.

Don't overread it

The confidence interval crosses no effect, and certainty was very low.

The statistics, in plain English

A risk ratio of 0.64 would mean about a third less delirium, but the interval from 0.38 to 1.07 includes no benefit at all. The result is compatible with anything from a large benefit to none.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.