How to Prevent Delirium by Improving Sleep for Patients With Serious Illness

Sleep is essential for rest and repair, but for many patients with serious illness, getting quality sleep is a struggle. Poor sleep can compromise functional status and contribute to or exacerbate the risk of delirium—a concern that's both commonly reported by patients and supported by research.

As clinicians, we must address sleep to help prevent delirium, as it directly impacts mental clarity, healing, and quality of life. In this blog, I share behavioral interventions I've used to help patients improve sleep and discuss when pharmacological treatments may be appropriate.

I typically begin with non-pharmacological interventions, which are often effective in improving sleep quality and reducing the incidence and duration of delirium. A combination of strategies may give the best chance of targeting sleep impairment and delirium.

A key strategy is encouraging good sleep hygiene. I often tell patients that a good night’s sleep starts first thing in the morning—by waking at a consistent time and getting out of bed to natural light and fresh air when possible. A regular bedtime, a cool, dark room, and aiming for at least seven hours of sleep are also well-established principles.

Of course, in the hospital, these basics can be hard to accomplish. Patients may not have access to fresh air, or a window for natural light, and have no control over lighting, noise, or frequent interruptions. These barriers make it even more important to focus on the changes we can make.

As anyone who’s ever spent the night in a hospital knows, it’s hard to sleep when staff come in throughout the night. Whenever possible, cluster non-urgent care activities together at the beginning or end of the night to minimize interruptions and allow for longer stretches of sleep.

At home, patients may rely on nighttime routines to help prepare and cue their brain for sleep (e.g., washing their face and brushing teeth). Encourage them to think about these routines and implement them. This simple act can help to prime a patient’s brain to be ready for and expect sleep.

Light is perhaps the most important driver of the circadian rhythm, giving the brain cues about when it’s time to sleep and wake. At home, patients can use window shades and white noise (or quiet, if they prefer). In the hospital, this is harder to control, but you can ask the nurse to limit light in the room or use dimmers when possible. You can also offer eye masks and ear plugs to help reduce visual and auditory noise. These simple interventions can restore a sense of agency and make a meaningful difference.

Daytime wakefulness is key to better sleep at night. In the hospital, this could mean opening blinds, turning on overhead lights, and getting out of bed during the day. And at home, encourage patients and caregivers to get fresh air in the morning.

It’s important to assess a patient’s willingness to make behavioral changes for improved sleep upon hospitalization. It’s helpful to bring the nurse into the room for this conversation, since they will largely be responsible for managing the environment around the patient while hospitalized. Then, together, you can move forward with strategies that are most appropriate for them.

Achieving these goals requires education and support from nursing staff, caregivers, and loved ones. Take time to explain the impact of sleep hygiene and involve them in creating a supportive environment for rest.

Even with strong prevention efforts, there are times when pharmacological treatment is necessary to manage sleep disruption in a patient with or at risk for delirium. Here’s what I would do when introducing pharmacological treatments.

For a patient who is actively delirious or at high risk for delirium, I first examine the medications they’re already taking. We might need to readjust the timing—for example, stimulants like steroids or caffeine should be avoided late in the day, while sedating drugs given in the morning or mid-day may interfere with daytime alertness and nighttime sleep.

Ask about alcohol, cannabis, and non-prescribed substances that can disrupt sleep and may cause a sleep-disrupting, delirium-inducing, withdrawal syndrome. Deprescribing non-essential medications is also key—a patient’s tolerance can change with disease progression. For example, a patient who has taken alprazolam for insomnia every night for the last thirty years may no longer tolerate it with advanced metastatic lung cancer or an end-stage major neurocognitive disorder.

Note: For any patient with serious illness, and especially older adults, delirium can result from polypharmacy. Patients' medications should be assessed for anticholinergic burden, and for continued efficacy of the medication over time.

Melatonin is an option to try, though it hasn’t always been used effectively to help hospitalized patients. Rather than as a sedative, think of it as a sleep-promoting supplement that primes the brain for sleep. I usually prescribe a smaller dose—1, 2, or 3 mg, rather than standard 5, 10, or even 20 mg doses—and suggest taking it earlier in the evening, about one to three hours before the desired sleep time. Be sure to counsel patients about the potential for vivid dreams, as some find this possible side effect alarming.

For patients who appear sleepy and hypoactively delirious, we generally do not want to add any centrally acting medications. An exception may be appropriate if the patient is quietly experiencing distressing paranoia or another significantly terrifying psychotic experience. In such cases, consider a low-dose trial of a less-sedating antipsychotic (e.g., risperidone or haloperidol).

For patients who are agitated, actively showing signs of psychosis, and not sleeping, I might consider using a sedating antipsychotic. Although olanzapine is often used in palliative care settings, it is unfortunately anticholinergic, so for an agitated patient, I suggest drugs like quetiapine instead.

Depakote (divalproex sodium) can help quiet the mind of a patient with delirium; I may turn to it when someone is primarily impulsive, restless, irritable, or showing signs of affective instability. Its multiple mechanisms may help address some of the underlying drivers of delirium in our patients with serious illness. Many patients also find it to be sedating, so weighting the doses toward bedtime can assist with normalization of the sleep-wake cycle. Unfortunately, Depakote does come with several drug-drug interactions (e.g., carbapenem antibiotics and doxorubicin are two), and important potential adverse reactions (e.g., cytopenia, hyperammonemia, hepatitis, and teratogenesis) that need to be considered.

Regardless of the medication, monitor patients closely for both positive and adverse effects. Common adverse effects to look out for, regarding antipsychotics, are hypotension, especially orthostatic hypotension, QTc prolongation, akathisia and other extrapyramidal symptoms, and excess sedation. And importantly, remember to taper and discontinue any medications that were started during a period of delirium once the patient recovers.

We can help reduce the incidence of delirium from sleep disruptions by promoting a sleep-friendly environment, encouraging good sleep hygiene, and, when necessary, adjusting pre-existing medications and using others judiciously. It must be a team effort. Educating nursing staff, patients, and caregivers—and getting their buy-in—is critical for success.

Edited by Melissa Baron. Clinical review by Andrew Esch, MD, MBA.