Design: Systematic review and meta-analysis of randomised, controlled and pilot studies
Population: 8 studies, 1,190 hospitalised adults
Primary outcome: Delirium incidence
Effect: RR 0.64 (0.38–1.07), I² 52%, very low certainty

A systematic review in BMJ Open (28 September) found eight studies, 1,190 hospitalised adults, testing bright light therapy — mostly morning light of at least 5,000 lux for one to two hours over three or four days — to prevent delirium.

The pooled risk ratio for delirium was 0.64 (95% CI 0.38 to 1.07), an interval that includes no effect. Heterogeneity was moderate, several studies carried a high risk of bias, adherence and adverse events were poorly reported, and GRADE certainty was very low. Only three studies measured circadian rhythm, with mixed results.

The direction is interesting and the intervention is cheap, but the evidence is not strong enough to buy light boxes. Daylight exposure and a normal day-night routine are already part of multicomponent delirium prevention, which has a stronger evidence base.

Do not invest in bright light therapy for delirium prevention on current evidence.
Keep curtains open by day and lights low at night as part of routine ward care.
Multicomponent prevention remains the approach with the best evidence.

Why it matters

A plausible, cheap intervention can look attractive before the evidence supports it.

The statistics, in plain English

A risk ratio of 0.64 would mean about a third fewer cases of delirium, but the confidence interval runs up to 1.07, so no effect — or a slight increase — cannot be excluded. Very low GRADE certainty means the true effect could be substantially different.