Valproate is NOT recommended for routine treatment of ICU delirium based on current evidence
Valproate is not FDA-approved for delirium treatment and lacks high-quality evidence supporting its use in ICU delirium, though emerging retrospective data suggest it may be safe as an adjunctive agent when standard approaches fail. 1
Guideline-Based Approach
The 2018 Society of Critical Care Medicine (SCCM) guidelines do not recommend routine pharmacologic treatment for ICU delirium 2, 3. The guidelines specifically suggest:
- Not routinely using haloperidol, atypical antipsychotics, or statins to treat delirium (conditional recommendation, low quality evidence) 2, 3
- Prioritizing multicomponent nonpharmacologic interventions (reorientation, sleep optimization, early mobilization, sensory aids) 3
- Reserving antipsychotics only for patients with distressing hallucinations/delusions or dangerous agitation, then discontinuing immediately after symptom resolution 3
Valproate is not mentioned in these guidelines, reflecting the absence of high-quality evidence at the time of publication.
FDA Labeling Considerations
Valproate is FDA-approved only for seizures and bipolar disorder—not delirium 1. Key safety concerns in critically ill patients include:
- Thrombocytopenia risk increases significantly at trough levels >110 µg/mL (females) or >135 µg/mL (males) 1
- Somnolence is particularly problematic in elderly patients 1
- Requires monitoring for fluid/nutritional intake and dehydration 1
- Multiple drug interactions with common ICU medications (phenobarbital, carbamazepine, phenytoin) 1
Emerging Research Evidence
The most recent 2024 study (highest quality available) showed valproate reduced concomitant psychoactive medication use by 33% by day 3 in 87 ICU patients with hyperactive delirium 4. However, this was a retrospective cohort study—not a randomized controlled trial. Key findings:
- Loading dose (10 mg/kg) decreased risk of needing additional psychoactive medications (OR 2.8, p=0.047) 4
- Adverse events occurred in only 10.3% of patients 4
- Benefits primarily seen in cardiogenic shock and respiratory failure patients 4
Contradictory evidence: A 2024 comparative study found no significant difference in delirium-free days between valproate and other antipsychotics 5. This suggests valproate may not be superior to existing options.
Safety Profile from Retrospective Data
Across multiple retrospective studies (2017-2024):
- Hyperammonemia: 12-19% incidence 6, 7
- Thrombocytopenia: 9-13% incidence 6, 7
- No deaths directly attributed to valproate 6
- Typical dosing: 750-1500 mg/day (median 23 mg/kg/day) 7, 8
Clinical Algorithm When Considering Valproate
If standard approaches fail:
- First-line: Implement multicomponent nonpharmacologic bundle (ABCDEF) 3
- Second-line: Consider dexmedetomidine if agitation precludes ventilator weaning 3
- Third-line: Short-term antipsychotic (haloperidol/atypical) only for distressing symptoms or dangerous agitation 3
- Off-label consideration: Valproate as adjunctive therapy only when:
Critical Pitfalls
- No RCT evidence exists for valproate in delirium—all data are retrospective or case series 6
- Median initiation on ICU day 7 suggests it's being used as rescue therapy, not first-line 7
- Thrombocytopenia risk is particularly concerning in critically ill patients with existing coagulopathy
- Drug interactions with common ICU medications require careful monitoring 1
- One comparative study showed no advantage over standard antipsychotics 5
Bottom Line
Valproate should not be used routinely for ICU delirium. Prioritize guideline-recommended nonpharmacologic interventions first. If pharmacologic treatment is necessary, follow the SCCM algorithm (dexmedetomidine for agitation precluding extubation, short-term antipsychotics only for distressing/dangerous symptoms). Valproate may be considered as an off-label adjunctive agent in refractory cases with appropriate monitoring, but recognize this is based solely on retrospective data with no mortality or quality-of-life benefit demonstrated.