Treating Agitation in Patients with Dementia
The Carlat Hospital Psychiatry Report, Volume 1, Number 7&8, October 2021
Topics: Agitation | Alzheimers | Antipsychotics | Benzodiazepines | BPSD (behavioral and psychological symptoms of dementia) | Dementia | Lewy Body Dementia | Parkinson’s Disease | Pharmacotherapy | Sundowning
Eran Metzger, MD
Medical Director of Psychiatry, Hebrew SeniorLife, Boston, MA. Assistant Professor of Psychiatry, Harvard Medical School, Boston, MA.
Dr. Metzger has disclosed no relevant financial or other interests in any commercial companies pertaining to this educational activity.
CHPR: Can you tell us about yourself and your background?
Dr. Metzger: I’m the medical director of psychiatry at Hebrew SeniorLife and an assistant professor of psychiatry at Harvard Medical School. My work also includes training residents and geriatric medicine fellows who rotate through our facility.
CHPR: You participated in the psychopharmacology algorithm project for the management of behavioral and psychological symptoms of dementia (BPSD) and co-authored a publication based on that work (Chen A et al, Psychiatry Res 2021;295:113641). Can you give us an overview of BPSD?
Dr. Metzger: We know that 80%–90% of patients who suffer from dementia are likely to experience behavioral and psychological symptoms of dementia, or BPSD. Those symptoms vary; we subdivide them into motor symptoms and verbal symptoms, but they can be disruptive regardless of the category.
CHPR: What’s in your differential diagnosis for these symptoms?
Dr. Metzger: The primary rule-out is an acute medical change. Even in a person who has had these behaviors before, if we see an acute behavioral change, we first look for a change in the patient’s medical condition.
CHPR: And what are your steps for management?
Dr. Metzger: A medical examination is the first part of management. This population, even in supervised settings, is prone to medical changes such as infection and dehydration. Sometimes, treating an acute medical condition alone can resolve the behavioral and psychological symptoms.
CHPR: Right. For example, elderly women with UTIs can get quite agitated, but their behavior often improves once you treat the UTI.
Dr. Metzger: Exactly. UTIs, pneumonia, dehydration, or even exacerbation of a pain syndrome can cause agitation, and just managing the pain can resolve the agitation. When I started working in geriatrics over 20 years ago, one of my mentors stressed the importance of getting a bowel history because of constipation’s effects on behavior. I was skeptical that someone’s constipation could result in clinical changes in behavior, but now I’m a true believer. We often see dramatic improvements in behavior once we remedy a patient’s constipation. In our daily report on each patient, the nursing staff reports whether the patient had a bowel movement that day.
CHPR: We want to look at all possible options to help these patients and not just assume we need to medicate. Are there any other nonmedication approaches we should consider?
Dr. Metzger: Yes. Our preference is always to try nonpharmacologic approaches; in fact, for people in skilled nursing facilities, Medicare requires that we document our attempts at nonpharmacologic approaches before we start medicating our patients. The two approaches with the best evidence are 1) caregiver and staff education and 2) music therapy. Web-based training programs are available, often free of charge, that help staff learn how to identify what need a patient is trying to express and how to respond in a nonconfrontational way even when the patient is agitated.
CHPR: Can you give us the websites for any of these web-based training programs?
Dr. Metzger: Sure. Here are two of them: Alzheimer’s Association person-centered dementia care training (www.alz.org/professionals/professional-providers/dementia-care-training-certification) and the Oasis senior care program (www.susanwehrymd.com/oasis).
CHPR: One example that I saw in your paper was trying to avoid using the word “no” and instead saying, for example, “How about we do this instead?”
Dr. Metzger: Yes. When we see acutely unsafe behavior, we want it to stop right away, so our natural reflex is to use words like “no” and “stop.” But certain patients, particularly men, can escalate in response to that type of language.
CHPR: How about situations where it seems like you need to use a medication to avoid patients hurting themselves or someone else? What’s your medication algorithm?
Dr. Metzger: For emergent agitation, we like to give medication orally if possible, but sometimes that’s not feasible or safe. Some people have recommended using orally disintegrating olanzapine for a patient who isn’t accepting pills, but I know of at least one case where a nurse tried to place an orally disintegrating tablet under a patient’s tongue and had her finger bitten badly. So I don’t recommend trying to involuntarily administer orally disintegrating tablets to an agitated patient. For emergent agitation in a patient who isn’t taking oral medications, our medication of choice would be intramuscular olanzapine. Intramuscular haloperidol hasn’t been studied as well in the older age group, but it’s been used in medical settings safely and is our second option. We do know that safety of these medications is dose dependent, and that generally the higher doses are where you start to get into trouble. In most of our older patients, that risk can be avoided if you give the medication a little time to work.
CHPR: What dose range do you use?
Dr. Metzger: For olanzapine, I would start with 2.5–5 mg intramuscularly depending on the body habitus and would give that a half hour or an hour to have effect. I would repeat that up to three times over the course of the day as necessary. For haloperidol, again depending on the patient’s body habitus, I would use 0.25–1 mg intramuscularly.
CHPR: We sometimes see gabapentin, divalproex, prazosin, and SSRIs used to treat agitation in dementia in geri-psych units. Do you use any of these medications?
Dr. Metzger: Of the medications you mentioned, SSRIs have the strongest evidence to support their use in treating BPSD. The most comprehensive study was the CitAD trial using citalopram up to 30 mg daily (Porsteinsson AP et al, JAMA 2014;311(7):682–691). However, I tend to use either escitalopram or sertraline to minimize the risk of QTc prolongation, based on a 2013 FDA advisory (www.tinyurl.com/4zrya7h4). I use SSRIs for nonemergent agitation because the response latency is similar to when they are used to treat depression. Unfortunately, the evidence base for the other agents is comparatively weak. Well-designed controlled studies of divalproex had negative findings, and the use of gabapentin is supported only by case reports. Prazosin has two small positive trials to support its use, and I hope it will be studied more in the future.
CHPR: What about benzodiazepines? Do you worry about using them in elderly patients?
Dr. Metzger: We worry a lot about benzodiazepines because of their effects on coordination, gait, and cognition. One scenario where I’ve found them to be helpful is in nonambulatory patients who consistently become agitated during caregiving (eg, bathing) despite nonpharmacologic strategies. I’ve found that lorazepam 0.25 mg about 30 minutes beforehand can make the activity safer for both the patient and the caregiver.
CHPR: You refer to nonemergent agitation situations as urgent in your paper. Tell us what you mean by that.
Dr. Metzger: These are situations where a behavior does not immediately place someone at risk, but the behavior really can’t be supported for much longer. The patient might be combative with others, or they might frequently do things that jeopardize their safety, or perhaps they are at greater risk of falling due to psychomotor agitation. So, we don’t have to provide calming within the hour, but we also don’t have several weeks to wait; we’re talking more along the lines of several days at most.
CHPR: What medications do you recommend in these cases?
Dr. Metzger: We again would want to start with oral medications, and the antipsychotics in order of preference would be aripiprazole, followed by risperidone. We favor aripiprazole since it’s less likely to cause extrapyramidal symptoms. I want to emphasize that there are no FDA-approved medications for behavioral and psychological symptoms of dementia. I mean no medications from any class—antipsychotics or otherwise. The antipsychotic pimavanserin is FDA approved for psychosis in Parkinson’s disease only. Canada has approved risperidone for short-term use in psychosis and aggression, while both Britain and Australia have approved risperidone for behavioral and psychological symptoms of dementia.
CHPR: It’s too bad we don’t have more options, especially since antipsychotic meds have a black box warning on top of not being FDA approved.
Dr. Metzger: Exactly. The black box warning was based on a retrospective review of a number of studies that showed an increase in the incidence of cerebrovascular adverse events, but also a statistically significant increase in mortality among patients with dementia who were prescribed these medications. The risk is highest early in treatment, and the hazard ratio across many studies is around 1.55. That sounds like a lot, but looking at the absolute numbers, the difference in 30-day mortality rate in nursing home patients goes up 1.2 actual percentage points. Or if you’re looking at a 180-day study, the absolute risk rises from around 2.9% to about 4.4% (Gill SS et al, Ann Intern Med 2007;146(11):775–786). When we present this information to families, our experience is that they will accept that amount of increased risk in exchange for the possibility of improved quality of life and improved safety, because agitation can place patients at serious risk of physical harm.
CHPR: I’m assuming confounding by indication has been taken into account, right? People who are the most agitated might also be the most medically ill.
Dr. Metzger: That’s an excellent point—studies do try to control for that confounding, but I am not convinced that we’ve been able to design ones that can do that. Unfortunately, the patients who have the most severe symptoms are the ones most likely to get treatment, and there’s a possibility that those severe symptoms are markers for worsened illness and increased mortality associated with that illness.
CHPR: What are your thoughts about minimizing the anticholinergic load that patients are exposed to, considering that anticholinergic medications can exacerbate confusion and agitation?
Dr. Metzger: When we encounter a patient with new or worsened agitation, we review the medication list to make sure there’s not a correlation between the mental status changes and the addition of a new medication, such as one with anticholinergic side effects. An example that comes to mind for male patients is oxybutynin (Ditropan), which is prescribed by our urology colleagues to address urinary frequency. We also see some older patients coming in on tricyclic antidepressants for neuropathic pain or insomnia, so we want to carefully remove those medications, particularly if they’re linked with a deterioration in mental status.
CHPR: What about patients with Parkinson’s or Lewy body dementia? What specific treatment considerations should we keep in mind?
Dr. Metzger: The main concern is exacerbating extrapyramidal symptoms. Although the studies of quetiapine’s efficacy have been very disappointing, it is one of the most popular antipsychotics for behavioral and psychological symptoms of dementia. Clinicians might feel falsely reassured that quetiapine is safe since it has fewer extrapyramidal side effects, but in addition to having questionable efficacy, it has several worrisome adverse effects—including sedation, which places elderly patients at risk for falls as well as aspiration during mealtimes. Also, quetiapine produces more orthostatic hypotension than most other second-generation antipsychotics, and that further increases the risk of falls.
CHPR: What medication would you use instead?
Dr. Metzger: This is a situation where I might first give pimavanserin a try. Because it is still under patent protection as Nuplazid, depending on a patient’s Medicare D plan, cost may be prohibitive. In my limited experience, it is relatively easy to titrate, seems to be well tolerated, and can be effective in curbing BPSD. An alternative would be clozapine, though that can present some of the same adverse effects of sedation and postural hypotension as quetiapine.
CHPR: And what about cholinesterase inhibitors?
Dr. Metzger: That was a controversial topic between the co-authors of our paper. Cholinesterase inhibitors and memantine have only modest effects on behavior but are relatively safe to administer, and some patients do show behavioral improvements that are lost if we stop the medication. So, you may see some benefits among patients with Alzheimer’s, but it’s difficult to predict who specifically will benefit. Also, these medications are no longer as prohibitively expensive as when they first came out.
CHPR: What do you watch for when you have patients on these agents?
Dr. Metzger: Mainly anorexia and weight loss. These are the most common reasons we discontinue cholinesterase inhibitors.
CHPR: And how do you manage sundowning? We sometimes see elderly patients who seem fine during the day but become agitated in the evening.
Dr. Metzger: We’ve used the term sundowning for decades and are no closer to understanding what this syndrome represents. One theory is that it’s related to dysregulation of circadian rhythms. Another is that it represents a reaction to what’s going on in the inpatient milieu, like a change in shift where there’s more commotion and patients are reacting to that. We might try a medication like trazodone because that gets the patient ready for bedtime—something like 25 mg late in the afternoon is enough to calm the patient down and get them ready to go to sleep. But my first choice is melatonin at bedtime.
CHPR: Melatonin levels decline with age, so it makes sense to try it as a first choice for elderly patients.
Dr. Metzger: We also know from a Cochrane Collaboration review that bedtime melatonin can have positive effects over the course of the following day, and that would support a circadian explanation for sundowning (Jansen SL et al, Cochrane Database of Systematic Reviews 2006;(1):CD003802).
CHPR: Are you familiar with dexmedetomidine? The FDA granted breakthrough therapy designation to an oral form of this drug for dementia-related agitation.
Dr. Metzger: Yes, dexmedetomidine has been used for years as an IV sedative and has recently been studied as a sublingual film for agitation associated with dementia. A recent small unpublished Phase 1 study of this alpha-2a adrenergic receptor agonist, conducted among assisted-living patients with dementia, showed improvements in scores of three instruments used to measure agitation with no significant adverse events. A larger study is currently underway, and we’ll have to see.
CHPR: Thank you for your time, Dr. Metzger.