Practice Signal
Reviewed by Ben Cottongim, PharmD
Antipsychotics do not resolve ICU delirium, shorten hospitalization, or improve survival. Five large randomized trials (MIND-USA, AID-ICU, REDUCE, MENDS2, DahLIA), two meta-analyses, and the 2025 PADIS guideline update converge on the same conclusion: haloperidol, ziprasidone, and atypical antipsychotics lack efficacy as delirium treatments in critically ill patients. Pharmacists should challenge reflexive antipsychotic orders for ICU delirium and instead champion the ABCDEF bundle, reserving antipsychotics only for brief management of severe agitation that threatens patient safety.
ICU delirium affects 30-80% of mechanically ventilated patients and is independently associated with prolonged hospitalization, cognitive impairment, and increased mortality. For decades, haloperidol was the reflexive pharmacologic response, anchored more by tradition and expert opinion than by rigorous evidence. Atypical antipsychotics such as quetiapine and ziprasidone followed haloperidol into common use on the assumption that dopamine D2 antagonism addressed an underlying neurochemical disturbance.
That assumption has not survived controlled testing. Between 2016 and 2025, five landmark RCTs enrolling more than 4,400 critically ill patients tested haloperidol, ziprasidone, quetiapine, olanzapine, and dexmedetomidine against placebo or each other for established ICU delirium. None met its primary efficacy endpoint. The MIND-USA trial (2018, N=566) found no reduction in delirium or coma duration with haloperidol or ziprasidone, and its long-term follow-up confirmed that null result persisted at 3 and 12 months across cognitive, functional, and quality-of-life domains. The AID-ICU trial (2022, N=1,000), the largest single RCT, showed haloperidol added zero additional days alive outside the hospital at 90 days. The REDUCE trial (2018, N=1,789) tested prophylactic haloperidol and found no survival benefit whatsoever.
What has emerged from this evidence is a clearer distinction between treating the symptom of agitation and treating the syndrome of delirium. The DahLIA trial (2016) showed that dexmedetomidine increased ventilator-free hours in agitated delirium patients ready for extubation, and the 4D trial (2025) confirmed dexmedetomidine controls agitation faster than placebo in non-intubated patients. Neither trial showed that dexmedetomidine resolves delirium itself. Meanwhile, two 2024 meta-analyses synthesized the cumulative RCT data and reached the same conclusion: antipsychotics provide no meaningful outcome benefit for ICU delirium.
Subscribe to PharmD Signal for the complete evidence synthesis and practice recommendations on every topic, plus every deep analysis we publish.
Become a Subscriber →