ICU delirium treatment focuses primarily on non-drug strategies, because the medications most people would expect to help, like antipsychotics, have not been shown to shorten delirium or improve survival in rigorous clinical trials. The cornerstone of treatment is a structured care approach called the ABCDEF bundle, which combines daily sedation breaks, breathing trials, consistent delirium screening, careful medication choices, early physical activity, and family involvement. About half of all ICU patients develop delirium at some point during their stay, making it one of the most common and consequential complications of critical illness.
How ICU Delirium Is Detected
Treatment starts with reliable detection, and ICU teams use standardized screening tools rather than relying on bedside impressions alone. The two most common tools are the Confusion Assessment Method for the ICU (CAM-ICU) and the Intensive Care Delirium Screening Checklist (ICDSC). The CAM-ICU catches about 80% of delirium cases and correctly rules it out roughly 96% of the time. The ICDSC has slightly lower accuracy, detecting about 74% of cases with an 82% specificity. Most guidelines recommend screening at least once per nursing shift.
Detection matters because the most common form of ICU delirium is easy to miss. Unlike the agitated, hallucinating version most people picture, the hypoactive and mixed subtypes account for the vast majority of cases in the ICU. Hypoactive delirium, which makes up roughly 25 to 44% of cases, looks like extreme drowsiness, withdrawal, slowed movement, and flat affect. Mixed delirium, the single most common subtype at about 53% of cases, cycles between agitated and quiet phases. The purely hyperactive form, with restlessness, emotional swings, and hallucinations, accounts for only about 23% of ICU delirium.
The ABCDEF Bundle
The primary treatment framework for ICU delirium is the ABCDEF bundle, a coordinated set of practices that address the major drivers of delirium. Each letter represents a specific intervention.
A stands for assessing, preventing, and managing pain. Uncontrolled pain is a potent trigger for delirium, so teams prioritize keeping patients comfortable with the lightest effective approach. B covers both spontaneous awakening trials and spontaneous breathing trials. Every day, the care team stops sedative medications (as long as pain is controlled) to let the patient wake up, then pairs that with a test of whether the patient can breathe without the ventilator. In a major randomized trial, pairing these two steps led to earlier liberation from mechanical ventilation and shorter ICU stays.
C is for choice of sedation. Light sedation is conditionally recommended over deep sedation for mechanically ventilated adults. The type of sedative matters too. Compared to older sedative classes, one newer approach to sedation reduced the risk of delirium by about 33% and shortened time on the ventilator by nearly two hours on average, though it can cause drops in heart rate and blood pressure. Benzodiazepines, by contrast, are independently associated with triggering delirium. D means routine delirium screening using the tools described above.
E stands for early mobility and exercise. Getting patients sitting, standing, or walking as soon as safely possible is a core delirium intervention. While meta-analyses show a trend toward shorter delirium duration with early mobilization, the clearest benefits are in preventing delirium from developing and improving functional recovery. F involves family engagement and empowerment, including keeping familiar faces present, reorienting the patient to time and place, and reducing the disorientation that fuels delirium.
Why Antipsychotics Haven’t Worked
Many people assume that medications like haloperidol or other antipsychotics are standard treatments for ICU delirium. They are not. Two landmark trials published in the New England Journal of Medicine tested this directly. The MIND-USA trial randomized over 560 ICU patients with delirium to receive haloperidol, ziprasidone (another antipsychotic), or placebo. There was no significant difference in days alive without delirium, days in coma, or 90-day mortality between any of the groups.
A second large multicenter trial confirmed these findings: haloperidol did not lead to significantly more days alive and out of the hospital at 90 days compared to placebo. While there was a numerical trend toward lower 90-day mortality in the haloperidol group (36.3% vs. 43.3%), the primary outcome showed no benefit. Based on this evidence, the Society of Critical Care Medicine’s most recent guideline update was unable to issue a recommendation for or against using antipsychotics to treat delirium. In practice, antipsychotics are sometimes still used short-term when a patient’s agitation poses an immediate safety risk, but they are not considered a treatment for the delirium itself.
Medications That Can Make Delirium Worse
A critical part of treatment is reviewing every medication a patient receives and removing anything that could be fueling the delirium. Several common drug classes are known contributors: benzodiazepines (particularly lorazepam and midazolam, which are independent risk factors), anticholinergic medications, corticosteroids, opioids, certain antibiotics, antihistamines, and a nausea drug called metoclopramide. Anticholinergic drugs, which block a key chemical messenger in the brain and are found in many over-the-counter sleep aids and allergy medications, are especially problematic.
The goal is not necessarily to stop all of these medications, since some may be essential for treating the patient’s underlying illness. Instead, teams weigh each drug’s benefit against its potential to worsen or prolong delirium, substituting safer alternatives where possible.
Environmental and Supportive Strategies
The ICU environment itself promotes delirium. Constant noise, round-the-clock lighting, frequent interruptions, and the absence of familiar cues all disrupt sleep and orientation. Treatment guidelines conditionally recommend strategies to restore a more natural day-night cycle: dimming lights at night, clustering care activities to allow uninterrupted rest periods, reducing alarm volumes where safe, and offering earplugs or eye masks. Keeping clocks, calendars, and family photos within sight helps patients stay oriented. Ensuring patients have their glasses or hearing aids, which are often removed on admission, makes a measurable difference in reducing confusion.
Long-Term Cognitive Effects
ICU delirium is not just a temporary problem that resolves at discharge. A landmark study tracking ICU survivors found striking rates of lasting cognitive decline. At three months after leaving the hospital, 40% of patients had thinking and memory scores comparable to someone with a moderate traumatic brain injury. A full 26% scored at levels typically seen in mild Alzheimer’s disease. These deficits persisted: at 12 months, 34% still had scores in the traumatic brain injury range and 24% in the mild Alzheimer’s range. These impairments affected younger and older patients alike.
This long-term picture is part of what’s called post-intensive care syndrome, which can include cognitive problems, physical weakness, and psychological symptoms like depression or PTSD. It underscores why preventing and aggressively treating delirium matters so much. Every additional day of delirium in the ICU is associated with worse cognitive outcomes months later, which is why the non-pharmacological strategies in the ABCDEF bundle, despite sounding simple, represent the most evidence-backed approach available.