Comparative Safety of Medications for Severe Agitation: Lessons Learned From Management of Behavioral and Psychological Symptoms of Dementia

The recent publication by Casey et al. highlights an important and challenging clinical area, that is, the management of severe undifferentiated agitation among older adults in ED settings [1]. As outlined in the paper, agitation is common among older adults in ED settings and the potential causes of agitation can be broad and difficult to disentangle, particularly with individuals who may be uncooperative or have difficulty communicating due to underlying neurodegenerative conditions such as dementia, or due to other causes of cognitive impairment such as delirium or acute intoxication with substances. Their study has also identified the paucity of evidence on this topic with only one relatively small randomized controlled trial and eight observational studies [1]. From the limited evidence that is available, their review indicated that adverse events were higher among individuals who received midazolam when compared to haloperidol and that quetiapine may have been associated with lower rates of adverse events when compared to other medications. This topic is important as pharmacological management of agitation among older adults in emergency ED settings is common. Approximately 4% of all older adults presenting to ED between 2014 and 2017 received at least one antipsychotic or sedative medication [2] for management of agitation, with use of these medications being most common among older adults from nursing homes, those with dementia and women [2]. While the current review identified that there may be differences in the safety of various medication classes for the management of undifferentiated agitation, one must balance this evidence with considerations around how effective these medications are in real-world settings. One of the studies included in the review by Casey et al. examined older adults who were hospitalized following ED visits to one of 21 hospitals in the United States, also observed that there were potential safety differences among different medication classes [3]. While overall adverse events were similar for people who received antipsychotics or benzodiazepines, antipsychotics were associated with a reduced likelihood of intubation [3]. Importantly, neither antipsychotics nor benzodiazepines were particularly effective in reducing agitation as defined in their composite measure of medication effectiveness which included repeat use of medications, use of restraints or 1:1 observation for behavioral management. Using this composite, only ~5%–10% of individuals had complete resolution of their agitation which was reported as a treatment failure rate of 94.3% vs. 87.6% for antipsychotics and benzodiazepines, respectively [3]. This same study also indicated that use of antipsychotics or benzodiazepines in ED settings may be associated with longer ED stays and hospitalizations [2]. There is also the risk that psychotropic medications that are intended to be used in the short-term during hospitalizations can get continued long-term without an appropriate indication [4] potentially leading to adverse events far beyond the acute ED or hospitalization time period when treatment was initiated. Why do we need to be cautious about recommending some categories of medications that may be relatively safe compared to other medications, even if the safer alternative has limited effectiveness? There may be lessons to be learned about this topic from where more extensive evidence is available, that being the pharmacological management of behavioral and psychological symptoms of dementia (BPSD). Similar sentiments about medications being relatively safer for the management of agitation in dementia may have contributed to the increased use of atypical antipsychotics for the management of BPSD as these medications were thought to be, and likely are, safer than first generation antipsychotics [5] or other medications such as benzodiazepines. One small study of emergency treatment of agitation in dementia which was not included in the review by Casey et al. found that olanzapine may in fact be more effective and better tolerated than lorazepam for the acute management of agitation in Alzheimer's disease at least in the short-term [6]. Atypical antipsychotics used over longer time periods have shown modest efficacy for specific BPSD such as agitation over 8–12 weeks [7, 8]. The efficacy of antipsychotics needs to be balanced against their known adverse effects including a risk of cerebrovascular adverse events and an increased risk of all cause mortality [7]. However, we have seen that concerns about the relative safety of medications such as those raised around the use of atypical antipsychotics for BPSD can lead clinicians to select other medications which are largely ineffective but presumed to be safer. As an example, the adverse events and warning related to the use of atypical antipsychotics for BPSD have potentially contributed to the increased use of antidepressants among people with dementia [9], which while potentially safer than antipsychotics [10], are not free from side-effects and have been found to have limited efficacy for the treatment of agitation (with the exception of citalopram for Alzheimer's disease) [11]. What are some potential alternatives to medications for the management of agitation in ED settings? A recent scoping review of management strategies for agitation in ED settings identified that the majority of studies have focused on medications for the management of agitation, rather than non-pharmacological or environmental strategies [12]. One suggested approach to approaching the management of agitation in emergency room settings is the ADEPT (assess, diagnose, evaluate, prevent, and treat), emphasizing the importance of understanding the underlying causes of agitation, including delirium, followed by treatments aimed at the underlying causes and strategies to prevent agitation from worsening. While a process such as ADEPT may take time, it emphasizes a person-centered approach to managing agitation that ends with a treatment plan for the cause of agitation, which may or may not involve medications specifically for agitation [13]. Again, lessons from BPSD management indicate that psychosocial approaches for BPSD such as staff education and provision of person-centered care and meaningful activities may be more effective than pharmacological measures in many situations [14]. Other suggested approaches to the ED management of agitation of older adults have also emphasized the importance of assessing and treating pain [15] along with verbal de-escalation strategies and only using medicines if behavioral and other first line approaches fail [16]. Suggested approaches to assessing and managing agitation among older adults in the ED also mirror recommendations provided in recent clinical practice guidelines [17] and clinical approaches for management of BPSD [18, 19]. These recommendations for management of agitation in the context of BPSD emphasize the importance of a thorough assessment of potential contributors to behaviors, implementation of non-pharmacological de-escalation approaches and cautious short-term use of medications for behavioral emergencies to facilitate patient safety and completion of investigations that may be needed to identify causes of agitation [17-19]. We hope that the article by Casey et al. will similarly help guide ED clinicians in selecting the safest medications that may be currently available for management of agitation in older adults in the limited circumstances when emergency treatment of agitation requires pharmacotherapy. We should not, however, lose sight that clinicians should generally avoid the use of medications that have limited efficacy, regardless of their relative safety, and that medications should always be considered an adjunct to, but not a replacement for, person-centered psychosocial approaches to care for vulnerable older adults presenting to ED with acute agitation. D.S. and S.K. contributed to the conceptual development, writing, and critical revision of the editorial. D.S. received honorarium from the Canadian Coalition for Seniors' Mental Health for participating in the development of guidelines for dementia. This publication is linked to a related article by Casey et al. To view this article, visit https://doi.org/10.1111/jgs.19485.