Older woman playing solitaire on an ipad, face is not visible.

Long-term Cognitive Impairment after ICU Delirium: Common but Under-Recognized

Case example

Sandra, a 75 year old single female, was found lying on the floor of her home by her long-time neighbor. She was not oriented to who she was, where she lived, or the year. She was brought to the emergency room and found to have a urinary tract infection, sepsis (a blood infection), and an exacerbation of pre-existing congestive heart failure.

Sandra was hospitalized on the intensive care unit (ICU) for three days and treated with antibiotics for her infection. While she was in the hospital, she thought she was in her childhood home and only sometimes recognized her neighbor. By the end of the week, she was able to recognize her neighbor and knew the date. She was discharged back home with a home health 12 hours a day 3 times a week.

Three months later, Sandra’s neighbor brought her to her primary care doctor for a follow-up appointment. The neighbor was quite concerned that she was now having some difficulties with keeping track of her new medication regimen, which had become more complicated since her ICU stay. Sandra was able to get to familiar places but could not figure out how to go new places even when she used her GPS. She also seemed slightly withdrawn and would get easily frustrated about “small things.”

Long-term cognitive impairment (LTCI) from ICU delirium

Researchers still do not fully understand the long-term cognitive impairment (LTCI) from ICU delirium. About 50-70% of all ICU survivors suffer from a cognitive, psychological or functional impairment after their ICU hospitalization. About 30-80% of ICU patients develop cognitive deficits, making it a serious but under recognized problem. Risk factors for ICU-acquired long-term cognitive impairment (LTCI) in older adults include neurologic dysfunction, infection or severe sepsis, and acute dialysis. Duration of delirium is a risk factor for LTCI for adults of any age.

Case continued

Older woman playing solitaire on an ipad, face is not visible.Sandra was referred to a neuropsychologist for a lengthier evaluation. The evaluation found that Sandra had mild to moderate cognitive difficulties, most notably in executive functioning and memory. Sandra also was noted to have mild depressive symptoms. She was referred to cognitive rehabilitation, at which she learned how to set up automated processes (e.g. automatic online bill pay) and reminders to help reduce her stress levels. She also participated in cognitive behavioral therapy to address her depressive symptoms.

 Over the next few months, Sandra noticed an improvement in her ability to learn information, as well as her in her mood. She was able to use the GPS to find new places, although sometimes she would make some wrong turns and need to correct herself. Her mood also improved.

 However, she had another hospitalization for congestive heart failure exacerbation. It turned out that Sandra was having difficulty keeping track of her salt and fluid intake due to her persistent executive functioning deficits. At the recommendation of her primary care doctor, she agreed to move into an assisted living facility to help monitor her medical conditions more closely.

How is long-term cognitive impairment from ICU delirium different from Alzheimer’s disease?

LTCI from ICU delirium is quite different from Alzheimer’s disease. Proposed mechanisms for how LTCI arises include a lack of oxygen and inflammation. This theory is consistent with the observation that the onset of LTCI from delirium appears to depend on the duration of exposure to the pathophysiologic processes of delirium. This mechanism is quite different from Alzheimer’s disease, in which the key proteins thought to be involved in Alzheimer’s disease are beta amyloid and tau. Autopsies have shown deposition of beta amyloid plaques and neurofibrillary tangles in the brains of patients with Alzheimer’s disease, but not in patients who have developed in dementia after an episode of delirium.

The progression of LTCI is quite different from Alzheimer’s disease. The deficits from LTCI can improve one year after the initial insult, although some patients continue to have persistent deficits. In Alzheimer’s disease, patients’ cognition continue to progressively decline. LTCI also appears to affect multiple cognitive domains (executive functioning, memory, and attention). Alzheimer’s disease primarily affects memory, although other cognitive domains can be affected. Finally, LTCI may be associated with smaller superior frontal lobes, thalamus, and cerebellar volumes and white matter damage, whereas AD is classically associated with hippocampal atrophy. Interestingly, however, smaller hippocampal volumes are also associated with longer duration of delirium.

What is the workup and management for LTCI?

The Mini-Mental State Examination can often be normal in patients with LTCI because it is not sensitive to the executive functioning deficits. The Montreal Cognitive Assessment can sometimes show deficits, but longer cognitive screens such as Repeatable Battery for Neuropsychological Assessment (RBANS) in combination with executive functioning measures such as Trails A and B may be needed to capture the more subtle deficits seen in LTCI. Clinicians may also consider a detailed neuropsychological evaluation if this service is available, but this battery should perform at least a few months after the hospitalization to capture a more accurate picture of the new baseline. Also, patients with significant physical deficits may not be able to tolerate a lengthy test battery that can last up to three hours. There are no FDA approved treatments for LTCI. Cognitive rehabilitation to address certain type of deficits can be helpful in some individuals, and the IMPROVE trial is looking at the efficacy of a cognitive and physical training program to ameliorate LTCI deficits.

Although post-ICU depression and LTCI are separate phenomena, clinicians should also carefully examine their post-ICU patients with LTCI for depression. Depression is classically associated with worse executive functioning, and can worsen already existing cognitive deficits from LTCI. Antidepressants and depression-focused psychotherapies such as cognitive behavioral therapy should be considered. If patients do not respond to these initial approaches or have a history of being in mental health treatment, a mental health referral should also be made.

This post was contributed by . Dr. Wang is a geriatric psychiatrist and Assistant Professor of Clinical Psychiatry at Indiana University.

Posted in NIDUS Blog and tagged , , , .


  1. First let me say I am enjoying these vignettes. Presently, I work in a geropsych unit as an APRN which is housed in a rural critical access hospital. A common reaction is to refer any behaviors to geropsych; I am always reminding the clinical staff to search for the etiology of a presenting delirium to prevent unnecessary geropsych inpatient admissions. This gets muddy when the patient is status post an ICU hospitalization or even more so, status post anesthesia. Recently, I’ve had two patient’s who were clearly LTCI verses an undiagnosed dementia. As delineated above, there are not any FDA approved treatments for LTCI–matter of fact, I am not certain if it can be coded for; does the diagnosis exists In the DSM V under mild or major neurocognitive disorders?

    In practice, we definitely assess for any concomitant mental health concerns such as depression. Is there any promise that cholinesterase inhibitors might prove semi-effective? Does the LTCI trajectory show improvement with time….or are we seeing stagnation or even worse, development into a major neurocognitive disorder? I have posted this response prior to searching out the clinical trial, IMPROVE—will access!

    Thank you for illuminating a clearer picture on a very common issue,

    Solange Marcel, APRN-BC
    Delirium U13 conference

    • Hi Solange – glad you are so interested in this topic! Please see response from the blog author, Dr. Sophia Wang, below.

      “There are no FDA approved treatments for long-term cognitive impairment from delirium, and it is not defined in the DSM-5 as its own identity in the same fashion as Alzheimer’s disease or vascular dementia. The delirium superimposed on a previously undiagnosed dementia is a thorny issues many clinicians struggle with.
      With regards to the question about anesthesia exposure and cognitive decline, post-operative cognitive decline has been recognized as an issue. There is an international working group looking at how to define post-operative cognitive decline to clarify its criteria and how it differs from delirium and long-term cognitive impairment from delirium.

      It is not likely that AChI will prove effective for the depressive symptoms, but a lot more work is needed to understand the complex relationships between cognition, depression, and delirium and how to treat it. In 2016, Dr. Sharon Inouye published a paper in Alzheimer’s Dementia suggesting a biphasic trajectory after delirium in post-operative patients. This is a very intriguing and exciting finding from both a mechanistic and treatment standpoint. First, are the mechanisms driving the first wave of damage versus the second wave the same or different? Why is the brain recovering from the first wave but then declining again? From a treatment standpoint, will we need to develop two separate phases of treatment to stop the long-term damage of delirium?

      Here’s the citation to the Inouye paper.
      Alzheimers Dement. 2016 Jul;12(7):766-75. doi: 10.1016/j.jalz.2016.03.005. Epub 2016 Apr 18.
      The short-term and long-term relationship between delirium and cognitive trajectory in older surgical patients.
      Inouye SK1, Marcantonio ER1, Kosar CM2, Tommet D3, Schmitt EM2, Travison TG1, Saczynski JS4, Ngo LH5, Alsop DC6, Jones RN7.

      Thanks for your interest! This is such an exciting field, and a lot more research needs to be done to answer these important questions.”

Leave a Reply