As described in the National Plan to Address Alzheimer Disease,¹⁷ agencies within the Department of Health and Human Services (HHS) have developed and implemented several initiatives aimed at reducing the inappropriate and off-label use of antipsychotics and other potentially harmful medications by patients in nursing homes. In 2012, the Centers for Medicare & Medicaid Services (CMS) launched the National Partnership to Improve Dementia Care in Nursing Homes, which included an initial goal to reduce the national prevalence of off-label antipsychotic use in long-stay nursing home residents, including patients with dementia, by at least 15% by December 31, 2012.^3,18 At the end of 2011, antipsychotics were prescribed to 23.8% of US long-stay nursing home residents.³ Although the goal was not met by the end of 2012, by the end of 2013, the national prevalence rate of long-stay residents in nursing homes who were taking antipsychotics had decreased to 20.2%, which was a 15.1% reduction.³ In late 2014, the CMS announced that new goals were established to reduce off-label antipsychotic use by 25% by the end of 2015 and 30% by the end of 2016 from the initial baseline measure in 2011.^18,19 In addition to these initiatives, the CMS took action aimed at reducing antipsychotic use among older adults in nursing homes, as further described in the US Government Accountability Office's report to congressional requestors.³ 

Other HHS agencies have also launched initiatives to reduce antipsychotic use among older adults with dementia. Most notably, both the Agency for Healthcare Research and Quality and the National Institutes of Health have awarded grants for research related to antipsychotic use in older adults with dementia residing in nursing homes.³ Specifically, the Agency for Healthcare Research and Quality has funded individual grants through its Center for Evidence and Practice Improvement and Comparative Effectiveness programs. Within the National Institutes of Health, the National Institute on Aging and the National Institute of Mental Health have also funded related research, including a number of studies examining the safety of antipsychotic use in older adults. 

Although the National Plan to Address Alzheimer's Disease was established to improve care for all individuals living with dementia regardless of care setting, the HHS has acknowledged that initiatives to reduce antipsychotic use have not focused on care settings outside of nursing homes.³ However, targeting this segment of the population is equally important, given that more than 1.2 million Medicare Part D enrollees residing in the general community have dementia, about 1 in 7 Medicare Part D enrollees with dementia residing in the general community are given an antipsychotic, and Medicare Part D pays for antipsychotics prescribed to these individuals.³ Moreover, the risks associated with antipsychotic use in older adults with dementia are not specific to individuals living in nursing homes.³ Because the HHS does not specifically target its initiatives at reducing antipsychotic use in settings other than nursing homes, older adults residing in the general community may be at greater risk of being affected by antipsychotic use. As the older adult population in the United States increases, it is expected that primary care physicians will be confronted with and challenged by the complexities of treating these patients.²⁰ Physicians must be knowledgeable about and consider the benefits and risks associated with antipsychotic use in older patients with dementia when making decisions on how to best manage the symptoms. 

Overall, few studies have researched the off-label use of typical antipsychotics for the management of BPSD. Two studies were characterized by small sample sizes and possible lack of power, and they focused on the efficacy of haloperidol and thioridazine, thus limiting the generalizability of their findings.^1,21 Data from these studies^1,21 demonstrated a modest benefit of typical antipsychotics over placebo in the management of BPSD, namely, symptoms of aggression and agitation. 

After their FDA approval and introduction into clinical practice more than 2 decades ago, atypical antipsychotics rapidly became the preferred antipsychotic medication class for managing BPSD because of their reported advantages over typical antipsychotics, particularly with respect to extrapyramidal symptoms and tardive dyskinesia.¹ Since then, research has focused on the efficacy of atypical antipsychotics in managing BPSD in patients with dementia.^2,22 However, similar to typical antipsychotics, the evidence shows only modest benefits of about 15% to 20% improvement with the use of atypical antipsychotics over placebo, and the more than 20 studies that have been published have primarily included only 3 atypical antipsychotics: aripiprazole, olanzapine, and risperidone.^1,10,22 

Older adults are particularly vulnerable to drug-drug interactions and drug-disease interactions associated with antipsychotics because of their frequent use of multiple medications (ie, polypharmacy) and the occurrence of multiple comorbidities (ie, polymorbidity), respectively. Also, older adults are more sensitive to the adverse effects of antipsychotics because of age-related physiologic changes, including increased permeability of the blood-brain barrier and reduced central cholinergic activity. 

The widespread off-label use of antipsychotics for the management of BPSD coupled with its associated health risks, especially an increased risk of mortality among patients with dementia, caught the attention of regulators. In 2005, the FDA cited a higher risk of death in older adults with dementia when they issued a public health advisory and asked manufacturers of those drugs to add a boxed warning to their product labeling.⁶ The evidence leading to the boxed warning was derived primarily from a meta-analysis¹¹ that assessed the increased mortality rate in patients with dementia who used antipsychotics. The study evaluated 15 placebo-controlled trials and found that the mortality rate in patients with dementia taking antipsychotics was approximately 1.6 times higher than that in the placebo groups.^6,11 Although the trials did not include all atypical antipsychotics, the FDA concluded that the mortality risk was related to the common pharmacologic effects of all atypical antipsychotics. The mortality risk was considered a class effect because the increase in mortality was seen with all 3 chemical classes of atypical antipsychotics; sensitivity analyses did not show evidence of differential risks for individual drugs.¹¹ 

In 2008, the FDA issued a boxed warning⁷ for typical antipsychotics after 2 observational epidemiologic studies suggested that the mortality rate in patients taking typical antipsychotics for BPSD was comparable to that in patients taking atypical antipsychotics.^12,13 The greatest increase in risk of death occurred within 30 days of initiating typical antipsychotic use and was attributable to cardiovascular-related events.^12,13 The extent to which the findings of increased mortality in observational studies may be attributed to an antipsychotic as opposed to other factors, however, is not clear. 

The use of antipsychotics in older adults carries other health risks aside from increased risk of death. Antipsychotics differ in their drug-drug interaction, drug-disease interaction, and adverse effect profiles.² Regarding the latter item, each antipsychotic adverse effect profile is unique, based on different binding affinities to various neurotransmitter receptors (eg, dopamine, histamine).² A summary of the potentially serious adverse effects of antipsychotics in older adults is shown in the Figure. In general, typical antipsychotics are more likely to cause extrapyramidal symptoms, and atypical antipsychotics are more likely to cause metabolic syndrome.^1,2,23 Still, notable differences exist even within classes. For instance, among atypical antipsychotics, metabolic syndrome is most commonly associated with the use of clozapine, olanzapine, and quetiapine and less commonly associated with aripiprazole and risperidone.^2,23 Similarly, anticholinergic toxicity is most commonly associated with the use of typical antipsychotics, with the exception of haloperidol and the atypical antipsychotic clozapine because of their much higher affinity for the muscarinic acetylcholine receptors compared with other antipsychotics.² 

From 2009 through 2011, 89,094 emergency department (ED) visits were attributed to adverse drug events (ADEs) due to psychiatric medications each year. Antipsychotics were implicated in 24.2% of cases, antidepressants in 28.5%, and sedatives and hypnotics in 34.5%.²⁴ Per 10,000 outpatient visits during which a psychiatric medication was prescribed, antipsychotics were the cause of the second highest ADE rate (11.7 of 10,000 outpatient visits; 95% CI, 10.1-13.2) among all psychiatric medications. Although the types of ADE visits varied across the psychiatric medication classes, typical antipsychotics (26.1 of 10,000 outpatient visits), and specifically haloperidol (43.3 of 10,000 outpatient visits), had especially high rates,²⁴ which was likely attributed to management-emergent extrapyramidal symptoms.²⁵ 

When caring for a patient with dementia, the benefits of both typical and atypical antipsychotics for older adults with BPSD must be carefully weighed against the potential risks. Seventy percent to 80% of patients with dementia will not derive a clinically meaningful benefit from taking an antipsychotic medication.^2,9 The potential side effects and ADEs associated with antipsychotic use in patients with dementia likely outweigh the possible benefits of these medications for most patients.^1,2,8,10 Because the benefits of antipsychotics are only modest in patients with dementia, it is estimated that for every 9 to 25 patients who benefit from antipsychotics, 1 will die.^2,11 Collectively, the data suggest that despite the modest benefits, the serious risks confirm that antipsychotics should not be routinely prescribed in clinical practice to manage BPSD in older adults; rather, these medications should be limited to use in patients whose symptoms are severe enough to put themselves and others at risk.^1,8 For these patients, given the severity of symptoms and the fact that there is no FDA-approved drug for the management of BPSD, the benefits associated with antipsychotic use (about 15% to 20% improvement in symptoms) may outweigh the risks associated with antipsychotic use.^1,10,22 

Evidence-based clinical practice guidelines consistently recommend the judicious use of antipsychotics for the management of BPSD only when other, nonpharmacologic approaches and interventions have failed to ameliorate symptoms, or when patients pose a threat to themselves or to others.^3,19,26,27 Nonpharmacologic interventions are generally grouped into 3 types: (1) behavior management, (2) cognitive or emotion-oriented therapy, and (3) sensory stimulation. Behavioral management interventions usually include a combination of training techniques, such as communication and habit training (reinforced learning of habits related to activities of daily living) accompanied by positive reinforcement.^2,28 Reminiscence therapy is an example of a cognitive/emotion-oriented intervention that involves the recollection of past activities, events, and experiences through old materials (eg, newspaper clippings, photographs) to stimulate memories.^3,28 Sensory stimulation can involve calming techniques, such as aromatherapy, massage therapy, or music therapy. Another example of a sensory stimulation intervention is Snoezelen therapy, which typically involves introducing the patient to a room full of objects designed to stimulate multiple senses.^3,28 

Evidence that demonstrates the success of specific nonpharmacologic interventions is growing, especially for managing symptoms in older adults with BPSD.^29-35 In addition to general nonpharmacologic approaches and interventions, osteopathic physicians are particularly well suited to incorporate the distinctive philosophy and practice of osteopathic medicine in caring for patients with dementia.²⁰ In doing so, it is imperative for osteopathic physicians to develop a collaborative partnership with patients and the patients’ family or caregiver and provide nonpharmacologic interventions that are both suitable and specifically tailored to their needs.^20,36 Ways in which physicians can establish this partnership with patients and their families or caregivers and provide a holistic approach to managing BPSD include involving the patient in health care decisions to the fullest extent possible, informing family members of changes in the patient's health, providing education about the disease and anticipated care needs as the disease progresses, referring patients and family members to community resources (eg, adult activities, nursing homes) and support groups, preparing for and accommodating expected cognitive and functional impairment, implementing safety measures (eg, assessing risk of falling, referring to occupational therapists), assessing for environmental triggers of behavioral symptoms, arranging for in-home support services and respite care, and initiating early discussion about advance care planning (eg, living wills, end-of-life care).^20,36 

Substantial evidence for pharmacologic alternatives to antipsychotics in patients with BPSD is lacking.¹ Nonetheless, before resorting to antipsychotics, some consideration should be given to pharmacologic alternatives.² Table 1 summarizes alternative strategies for the management of BPSD. There are some data to support the use of certain antidepressants (eg, citalopram, trazodone) for the management of BPSD, as evidenced by modest reductions in BPSD and better tolerability when compared with antipsychotics.³⁸ Unfortunately, most of the studies that comprise the evidence for the efficacy and safety of antidepressants in the management of BPSD in older adults were relatively small and of uncertain risk of bias due to methodologic issues.³⁸ Further studies are necessary to determine the role of antidepressants in managing BPSD in older adults. 

Antipsychotics have a useful purpose in clinical practice and, therefore, will never be eliminated. While antipsychotics have many legitimate uses, including management of psychotic and mood disorders (eg, schizophrenia, bipolar disorders), they are often used inappropriately in older adults to manage BPSD.¹⁹ But, for some patients with BPSD, the benefits of using antipsychotics outweigh the risks. Thus, emphasis on reducing inappropriate antipsychotic use should not curtail access to patients who need these medications. Antipsychotic use in patients with BPSD should be considered only if the patient's symptoms are distressing and dangerous to the patient or others, and only after other, less risky nonpharmacologic and pharmacologic interventions have been considered.² 

Some patients with dementia will experience BPSD that do not adequately respond to first-line, nonpharmacologic and pharmacologic interventions. In instances in which physicians have exhausted all other options and after they have had an informed discussion of the benefits and risks of antipsychotic use with the patients and their caregivers, there are several steps that should be implemented. Before initiating an antipsychotic, physicians should document in the medical record the specific diagnosis given to the patient and the symptoms that are being targeted with the antipsychotic, as well as the care plan for initiating, monitoring, and eventually reducing the dose, preferably with a specified (short) timeline. When selecting an antipsychotic, physicians should have an understanding of which ones have some evidence of benefit in managing BPSD symptoms, such as aripiprazole, olanzapine, quetiapine, and risperidone.^2,22 The smallest possible dose should be initially prescribed. The initial and maximum doses of antipsychotics with some evidence of benefit in managing BPSD are provided in Table 2. While a patient is using an antipsychotic, the effectiveness in mitigating symptoms and safety (eg, experience of tolerability with and adverse effects of the medication) should be regularly monitored and documented in the patient's medical record. When the targeted symptoms no longer endanger the patient and others, the physician should implement a strategy to reduce the prescribed dose. 

A goal of medication management for older adults, irrespective of dementia, is to reduce the use of unnecessary medications.⁴ Once prescribed, most older adults with dementia continue to take antipsychotics indefinitely, without attempts to reduce their use.^4,42 Different strategies to reduce unnecessary antipsychotic use, including abrupt discontinuation, gradual dose reduction, and mixed strategies of dose and duration, have been implemented with variable degrees of success.⁴ Yet, there is a paucity of information published in the medical literature to offer specific guidance on how to reduce antipsychotic use in older adults with BPSD. 

Adverse drug withdrawal events are especially prevalent in older adults and include the recurrence of the symptoms that the medication was prescribed to manage, physiologic or psychologic withdrawal reactions, or new symptoms.^4,43-45 Regarding adverse drug withdrawal events associated with antipsychotics, the primary concern is the recurrence of behavioral and psychologic symptoms.^4,19,43 Physiologic withdrawal reactions from antipsychotics, such as insomnia and restlessness, are also concerning, as they can be similar to the underlying condition that prompted the prescription.^4,43 The available data regarding reducing antipsychotic use have demonstrated that abrupt discontinuation or rapid reduction of an antipsychotic is associated with greater risk of adverse drug withdrawal events than gradual reduction over several weeks.^4,46 Therefore, a gradual dose reduction may not only alleviate adverse drug withdrawal events, but may lead to more favorable outcomes.⁴ 

In 2015, a pharmacokinetic-based gradual dose reduction strategy was proposed, constructed on the predicted half-lives (according to age-related physiologic changes) of the individual antipsychotics available in the United States.⁴ This proposed strategy⁴ recommends that physicians reduce antipsychotic dose by 50% in a stepwise progression. For atypical antipsychotics, the interval between dose reductions should be between 2 to 4 weeks with the exception of aripiprazole, which has an extremely long predicted half-life and requires about 2 months between dose reductions. For typical antipsychotics, the time interval between dose reductions should be 2 to 3 weeks, with the exception of pimozide, which has a comparatively long predicted half-life and thus requires about 4 weeks between dose reductions. Overall, this strategy would require at least 2 dose reductions before discontinuation of the antipsychotic, and the estimated time to complete discontinuation of most antipsychotics ranges from 4 to 8 weeks.⁴ Although further research is necessary to determine the effect of this strategy on patients, this strategy is predicted to result in more successful clinical outcomes than a more rapid reduction would.⁴ 

Behavioral and psychologic symptoms of dementia are most commonly managed with antipsychotics.¹ However, these medications are not approved by the FDA for this indication. National initiatives to reduce off-label prescribing of antipsychotics have predominantly focused on older adults residing in nursing homes; yet, research including community-dwelling older adults with dementia suggests that the risks of using antipsychotics for the management of BPSD may outweigh the benefits for most patients. Thus, initiatives to reduce off-label use of antipsychotics to manage BPSD in patients with dementia living in the general community should be introduced as well. Osteopathic physicians are particularly well-suited to help reduce the off-label prescription of antipsychotics by incorporating the distinct philosophy and practice of osteopathic medicine into caring for patients with dementia, especially through nonpharmacologic approaches and interventions and collaborative partnerships with patients and caregivers. Research and clinical innovation in this area should be a national public health priority.