Tuesday, June 9, 2015

Delirium Reinvented

One of my colleagues posted an article from the The Atlantic on delirium to her Facebook feed a few days ago.  Most of my colleagues in that venue are hospital, consultation-liaison, addiction or geriatric psychiatrists and we diagnose a lot of delirium.  Entitled the Overlooked Danger of Delirium in Hospitals it makes it seem like this is some kind of new and strange diagnostic category.  The article talks about the prevalence, the association with critical illness and advanced age, and the diagnostic overlap of dementia and delirium.  We hear from an Internal Medicine specialist Sharon Inouye, MD about the need to correctly diagnose and prevent delirium.  She mentions that as opposed to a decade ago, physician and nurses are all taught about delirium.  There is mention of the CAM (Confusion Assessment Method) that Inouye developed.  Like all health care articles there are estimates of the massive cost of delirium as well some prevention techniques.  There is also political concern that Medicare will declare delirium a "never" event with penalties for any hospital with cases of delirium.  That would be unfortunate because it makes a mistake that also seems to be made in this article - that delirium is a manifestation of many illnesses, especially the kind of illnesses that patient's are hospitalized for.

The article seemed odd to me because it was written from the perspective that delirium is an iatrogenic preventable event!  Certainly that can be the case. Delirium is a primary feature of hundreds of different disorders and recognizing delirium and those etiologies is potentially life saving.  Delirium can mimic psychiatric conditions due to the presence of hallucinations and delusional thinking.  For example, it is entirely possible to see a patient in the emergency department with apparent paranoid delusions and miss the fact that they happen to be delirious.  Sometimes the only sign is that the patient is inattentive and when vital signs are checked they have an elevated temperature.  This can be a common presentation of viral encephalitis in younger patients or urinary tract infections in the elderly.  It is bad form to miss either of those diagnoses and attribute the symptoms to a psychiatric disorder.  Another common form of delirium that is missed is drug or alcohol intoxication or withdrawal states.  Some intoxicants will render the patient totally unable to care for themselves until they are detoxified.  Other deliriums from alcohol or sedative withdrawal are life threatening and can be associated with seizures and other life-threatening states.  An acute change in a person's mental state resulting in delirium needs to be recognized and assessed as a medical emergency.    

One of the first cases of delirium that I ran into after residency was a case of cerebral edema that I was consulted on because of "hysterical behavior".  After that, I worked in and eventually ran a Geriatric Psychiatry and Memory Disorders Clinic for about 8 years.  The majority of people coming to that clinic had dementia of some sort.  They would see me and a neurologist.  We started out with an internist who was also a geriatric specialist, but that turned out to be overkill in terms of the number of medical specialists seeing each person in an outpatient clinic.  We eventually opted for records from the patient's primary care physician.  One of the most valuable functions of that clinic was our ability to follow people with prolonged deliriums.  Once a delirium has been established by a disease state and that state has resolved the delirium can persist for months.  Some of the outliers in that clinic took up to 6 months to clear.  We found that in many cases, the patients were extensively tested for intellectual ability and functional capacity when they were in the delirious state and told that they had dementia.  It was always instructive for the patient and family to get the testing repeated when we were sure the delirium had resolved and find that they had been restored to baseline.  Many people know their full scale IQ score and were relieved to see that they were back to that level of functioning.

A valuable lesson from working in that clinic and in hospital settings was the use of the electroencephalogram (EEG) as a possible test for delirium.   EEGs are commonly viewed as diagnostic tools to determine if a person is having seizures, but they also contain a lot of information about brain metabolism.  EEGs can be difficult to interpret especially if the patient is on a number of medications that affects cerebral metabolism. There are two broad categories of EEG patterns for delirium: one with a predominance of slow frequencies (designated theta and delta) and one with faster frequencies (designated beta).  We found a number of people with very significant cognitive impairment that was thought to be either a psychiatric disorder or a dementia but with a profound degree of slowing more consistent with a delirium.    

Delirium is an augenblick diagnosis for most psychiatrists.  The patient could appear disinterested, apathetic, agitated, or overtly confused.  It occurs in situations where brain physiology is compromised such as post surgical/anaesthesia states, drug intoxication states, drug reaction states, or possible physical illness delirium should be high on the differential diagnosis.  The Atlantic article makes it seem like knowledge about delirium is something very recent, but psychiatrists have been focused on it for a long time.  In the first two iterations of the DSM, delirium was subsumed under the categories of acute and chronic brain syndromes (DSM-I 1952) and organic brain syndromes (DSM-II 1968).  The current diagnostic code and name has been with us since the DSM III in 1980.  One of the early experts in delirium was Zbigniew J. Lipowski, MD, FRCP(C) - a Professor of Psychiatry from the University of Toronto.  His first text on the condition was Delirium: Acute Brain Failure in Man published in 1980.  That was followed by his classic text,  Delirium: Acute Confusional States published in 1990.  A comparable text from a neurological standpoint was Arieff and Griggs Metabolic Brain Dysfunction in Systemic Disorders published in 1992.

Any psychiatrist trained in the past 30 years should be able to diagnose delirium and come up with a differential diagnosis and monitoring or treatment plan.  A significant number of people can be followed on an outpatient basis as long as they are in a safe environment with the appropriate level of assistance.  The main goal of treatment is to make sure that the primary medical illness that led to the problem has been treated.  There are no known medications that will accelerate the resolution of these symptoms and medical management usually involves getting rid of medications that can lead to cognitive problems.  That can include benzodiazepines, antidepressants and antipsychotics but also more common medications like antihistamines and anticholinergic medications that are used for various purposes.  Like most psychiatric interventions in our health care system, clinics with staff interested in doing this work are few and far between generally because they are rationed resources.

There is a current movement underway to train Family Physicians and Internists (like Dr. Inouye) to recognize and prevent delirium.  In the minority of hospitals where psychiatrists work they are also a clear resource.  A delirium in a previously healthy person should signal a fairly comprehensive evaluation to figure out what happened.

And whenever there is a question of whether a person has a delirium or a psychiatric disorder - call a psychiatrist.  Psychiatrists know a lot about delirium and have for decades.

George Dawson, MD, DFAPA


Sandra G. Boodman.  Overlooked Danger of Delirium in Hospitals.  The Atlantic.  June 7, 2015.

Supplementary 1:  The graphic is a standard EEG.  I tried to post a slowed EEG seen in delirium, but the publisher wanted what I consider to be an exorbitant fee for a non-commercial blog.  If anyone has a slow anonymous EEG laying around, send me a copy and I will post it.


  1. Delirium is exhibit A as to why the collabo-Care model is clinically unsound. Try diagnosing it and formulating a tx plan from a distance. I can only imagine what my old C and L profs would have said had I told them I hadn't actually seen the patient.

  2. Good point.

    Very brief list of DSM symptoms - seems ideal for a checklist.

    Also a great example of the limitations of verbal descriptions and checklists. Until you have seen a number of cases in training and experienced those patterns, the symptoms list is essentially useless.

    I can remember driving in to the hospital and thinking about the diagnosis based on hearing the patient scream over the phone when the nurses called me. When I heard that scream, I was back on the transplant unit diagnosing encephalitis in the immunocompromised.

    There is a reason that a book like the DSM ≠ real live training.

  3. I'm increasingly convinced this whole collabo-care think if flat out malpractice. If you prescribe medicine to a patient you do not examine, you can get hauled in front of the state boards and get sued to high heaven if anything goes wrong. I've heard arguments that there is no increased risk because no doctor patient relationship and you aren't directly prescribing, but really no one knows what will happen until there is a high profile test case.