What is the safest evidence‑based sleep aid for an adult ICU patient with delirium?

Best Sleep Aid for ICU Patients with Delirium

For an ICU patient with active delirium, prioritize non-pharmacologic sleep interventions (earplugs, eyeshades, noise/light reduction, care clustering) over any medication, as no pharmacologic sleep aid has proven efficacy in this population and some may worsen delirium. 1

Why Pharmacologic Sleep Aids Are Not Recommended in Active Delirium

The 2018 Society of Critical Care Medicine guidelines make no recommendation for any pharmacologic sleep aid specifically for ICU patients, and this applies even more strongly when delirium is already present 1:

• Melatonin receives no recommendation due to insufficient evidence—three small trials (n=60 total) showed no significant improvement in sleep quality or quantity in ICU patients 1

• The lack of FDA regulation in the U.S. creates concerns about product quality and consistency, preventing many hospitals from formulary addition 1

• A 2025 multicenter RCT (n=334) found that low-dose melatonin (0.3 mg) achieved better pharmacokinetics than high-dose (3 mg) but did not reduce delirium incidence (54.4% vs 55.2%, p=NS) in mechanically ventilated ICU patients 2

• Dexmedetomidine may be considered only if the patient requires sedation for agitation that precludes ventilator weaning, but it is not recommended solely for sleep promotion 1

• It increases stage 2 sleep but does not reduce sleep fragmentation or increase restorative deep/REM sleep 1

• Clinical concerns include high cost, hemodynamic side effects (hypotension, bradycardia), and limited generalizability 1

• Propofol is explicitly discouraged—it suppresses REM sleep, causes hemodynamic instability, and can precipitate respiratory depression requiring mechanical ventilation 1

The Evidence-Based Approach: Non-Pharmacologic Sleep Protocol

Implement a multicomponent sleep-promoting protocol, which has demonstrated a 38% reduction in delirium prevalence (RR 0.62,95% CI 0.42–0.91) 1:

• Provide earplugs and eyeshades to all patients—this combination with relaxing music improved self-reported sleep quality in cardiac surgery patients 1, 3
• Reduce environmental noise and light during nighttime hours 3
• Cluster patient care activities to minimize nighttime interruptions and protect sleep cycles 3
• Designate a protected quiet period from 12:00 AM to 5:00 AM as the time most likely to remain uninterrupted 3
• Offer relaxing music if requested by the patient, though evidence for this alone is limited 1, 3

Special Consideration: If Sedation Is Already Required

If your delirious ICU patient is mechanically ventilated and requires sedation for agitation that prevents weaning/extubation, dexmedetomidine may be preferable to benzodiazepines 1:

• A single RCT (n=71) showed dexmedetomidine increased ventilator-free hours within 7 days compared to placebo, though it did not affect ICU or hospital length of stay 1
• For hemodynamically stable patients requiring overnight sedation, dexmedetomidine may improve sleep architecture compared to alternatives 1
• Use continuous low-dose infusion (0.5–0.6 µg/kg/h) without bolus to avoid severe hypotension and bradycardia 4

Critical Pitfalls to Avoid

• Never use benzodiazepines for sleep in delirious ICU patients—they are more deliriogenic than dexmedetomidine and worsen outcomes 4
• Avoid first-generation antihistamines (e.g., hydroxyzine) due to anticholinergic effects and heightened delirium risk in critically ill patients 3, 4
• Do not use antipsychotics for sleep promotion—they should only be used for distressing delirium symptoms (hallucinations, fearfulness) or agitation that poses physical harm, and must be discontinued immediately once symptoms resolve 1

When Melatonin Might Be Considered (Delirium Prevention, Not Treatment)

If your patient does not yet have delirium and you are attempting prevention, melatonin 3 mg at 9:00 PM may be reasonable despite weak evidence 1, 5:

• A 2025 meta-analysis (32 RCTs, n=3895) suggested melatonin may reduce delirium incidence (RR 0.72,95% CI 0.58–0.89) and slightly improve perceived sleep quality, though certainty of evidence was low 5
• A 2019 retrospective cohort (n=232) found significantly lower delirium development in patients receiving melatonin (7.7% vs 24.3%, p=0.001) 6
• However, the most rigorous 2025 RCT contradicted this, showing no benefit 2

The bottom line: Once delirium is present, focus on the non-pharmacologic sleep protocol and address the underlying causes of delirium rather than adding sleep medications.