Barry D. Weiss, MD

The year 2020 was marked by historic protests across the United States and the globe sparked by the deaths of George Floyd, Ahmaud Arbery, Breonna Taylor, and so many other Black people. The protests heightened awareness of racism as a public health crisis and triggered an antiracism movement. Racism is a pervasive and systemic issue that has profound adverse effects on health.^1,2 Racism is associated with poorer mental and physical health outcomes and negative patient experiences in the health care system.^3,4 As evidenced by the current coronavirus pandemic, race is a sociopolitical construct that continues to disadvantage Black, Latinx, Indigenous, and other People of Color.^5,6,7,8 The association between racism and adverse health outcomes has been discussed for decades in the medical literature, including the family medicine literature. Today there is a renewed call to action for family medicine, a specialty that emerged as a counterculture to reform mainstream medicine,⁹ to both confront systemic racism and eliminate health disparities. This effort will require collaboration, commitment, education, and transformative conversations around racism, health inequity, and advocacy so that we can better serve our patients and our communities.

The editors of several North American family medicine publications have come together to address this call to action and share resources on racism across our readerships. We acknowledge those members of the family medicine scholar community who have been fighting for equity consistent with the Black Lives Matter movement by writing about racism, health inequities, and personal experiences of practicing as Black family physicians. While we recognize that much more work is needed, we want to amplify these voices. We have compiled a bibliography of scholarship generated by the family medicine community on the topic of racism in medicine.

The collection can be accessed here.

While this list is likely not complete, it does include over 250 published manuscripts and demonstrates expertise as well as a commitment to addressing these complex issues. For example, in 2016, Dr. J. Nwando Olayiwola, chair of the Department of Family Medicine at Ohio State University, wrote an essay on her experiences taking care of patients as a Black family physician.¹⁰ In January of 2019, Family Medicine published an entire issue devoted to racism in education and training.¹¹ Dr. Eduardo Medina, a family physician and public health scholar, co-authored a call to action in 2016 for health professionals to dismantle structural racism and support Black lives to achieve health equity. His recent 2020 article builds on that theme and describes the disproportionate deaths of Black people due to racial injustice and the COVID-19 pandemic as converging public health emergencies.^12,13 In the wake of these emergencies a fundamental transformation is warranted, and family physicians can play a key role.

We, the editors of family medicine journals, commit to actively examine the effects of racism on society and health and to take action to eliminate structural racism in our editorial processes. As an intellectual home for our profession, we have a unique responsibility and opportunity to educate and continue the conversation about institutional racism, health inequities, and antiracism in medicine. We will take immediate steps to enact tangible advances on these fronts. We will encourage and mentor authors from groups underrepresented in medicine. We will ensure that content includes an emphasis on cultural humility, diversity and inclusion, implicit bias, and the impact of racism on medicine and health. We will recruit editors and editorial board members from groups underrepresented in medicine. We will encourage collaboration and accountability within our specialty to confront systemic racism through content and processes in all of our individual publications. We recognize that these are small steps in an ongoing process of active antiracism, but we believe these steps are crucial. As editors in family medicine, we are committed to progress toward equity and justice.

Simultaneously published in American Family Physician, Annals of Family Medicine, Canadian Family Physician, Family Medicine, FP Essentials, FPIN/Evidence Based Practice, FPM, Journal of the American Board of Family Medicine, The Journal of Family Practice, and PRiMER.

Acknowledgement –

The authors thank Renee Crichlow, MD, Byron Jasper, MD, MPH, and Victoria Murrain, DO, for their insightful comments on this editorial.

The year 2020 was marked by historic protests across the United States and the globe sparked by the deaths of George Floyd, Ahmaud Arbery, Breonna Taylor, and so many other Black people. The protests heightened awareness of racism as a public health crisis and triggered an antiracism movement. Racism is a pervasive and systemic issue that has profound adverse effects on health.^1,2 Racism is associated with poorer mental and physical health outcomes and negative patient experiences in the health care system.^3,4 As evidenced by the current coronavirus pandemic, race is a sociopolitical construct that continues to disadvantage Black, Latinx, Indigenous, and other People of Color.^5,6,7,8 The association between racism and adverse health outcomes has been discussed for decades in the medical literature, including the family medicine literature. Today there is a renewed call to action for family medicine, a specialty that emerged as a counterculture to reform mainstream medicine,⁹ to both confront systemic racism and eliminate health disparities. This effort will require collaboration, commitment, education, and transformative conversations around racism, health inequity, and advocacy so that we can better serve our patients and our communities.

The editors of several North American family medicine publications have come together to address this call to action and share resources on racism across our readerships. We acknowledge those members of the family medicine scholar community who have been fighting for equity consistent with the Black Lives Matter movement by writing about racism, health inequities, and personal experiences of practicing as Black family physicians. While we recognize that much more work is needed, we want to amplify these voices. We have compiled a bibliography of scholarship generated by the family medicine community on the topic of racism in medicine.

The collection can be accessed here.

While this list is likely not complete, it does include over 250 published manuscripts and demonstrates expertise as well as a commitment to addressing these complex issues. For example, in 2016, Dr. J. Nwando Olayiwola, chair of the Department of Family Medicine at Ohio State University, wrote an essay on her experiences taking care of patients as a Black family physician.¹⁰ In January of 2019, Family Medicine published an entire issue devoted to racism in education and training.¹¹ Dr. Eduardo Medina, a family physician and public health scholar, co-authored a call to action in 2016 for health professionals to dismantle structural racism and support Black lives to achieve health equity. His recent 2020 article builds on that theme and describes the disproportionate deaths of Black people due to racial injustice and the COVID-19 pandemic as converging public health emergencies.^12,13 In the wake of these emergencies a fundamental transformation is warranted, and family physicians can play a key role.

We, the editors of family medicine journals, commit to actively examine the effects of racism on society and health and to take action to eliminate structural racism in our editorial processes. As an intellectual home for our profession, we have a unique responsibility and opportunity to educate and continue the conversation about institutional racism, health inequities, and antiracism in medicine. We will take immediate steps to enact tangible advances on these fronts. We will encourage and mentor authors from groups underrepresented in medicine. We will ensure that content includes an emphasis on cultural humility, diversity and inclusion, implicit bias, and the impact of racism on medicine and health. We will recruit editors and editorial board members from groups underrepresented in medicine. We will encourage collaboration and accountability within our specialty to confront systemic racism through content and processes in all of our individual publications. We recognize that these are small steps in an ongoing process of active antiracism, but we believe these steps are crucial. As editors in family medicine, we are committed to progress toward equity and justice.

Simultaneously published in American Family Physician, Annals of Family Medicine, Canadian Family Physician, Family Medicine, FP Essentials, FPIN/Evidence Based Practice, FPM, Journal of the American Board of Family Medicine, The Journal of Family Practice, and PRiMER.

Acknowledgement –

The authors thank Renee Crichlow, MD, Byron Jasper, MD, MPH, and Victoria Murrain, DO, for their insightful comments on this editorial.

The year 2020 was marked by historic protests across the United States and the globe sparked by the deaths of George Floyd, Ahmaud Arbery, Breonna Taylor, and so many other Black people. The protests heightened awareness of racism as a public health crisis and triggered an antiracism movement. Racism is a pervasive and systemic issue that has profound adverse effects on health.^1,2 Racism is associated with poorer mental and physical health outcomes and negative patient experiences in the health care system.^3,4 As evidenced by the current coronavirus pandemic, race is a sociopolitical construct that continues to disadvantage Black, Latinx, Indigenous, and other People of Color.^5,6,7,8 The association between racism and adverse health outcomes has been discussed for decades in the medical literature, including the family medicine literature. Today there is a renewed call to action for family medicine, a specialty that emerged as a counterculture to reform mainstream medicine,⁹ to both confront systemic racism and eliminate health disparities. This effort will require collaboration, commitment, education, and transformative conversations around racism, health inequity, and advocacy so that we can better serve our patients and our communities.

The editors of several North American family medicine publications have come together to address this call to action and share resources on racism across our readerships. We acknowledge those members of the family medicine scholar community who have been fighting for equity consistent with the Black Lives Matter movement by writing about racism, health inequities, and personal experiences of practicing as Black family physicians. While we recognize that much more work is needed, we want to amplify these voices. We have compiled a bibliography of scholarship generated by the family medicine community on the topic of racism in medicine.

The collection can be accessed here.

While this list is likely not complete, it does include over 250 published manuscripts and demonstrates expertise as well as a commitment to addressing these complex issues. For example, in 2016, Dr. J. Nwando Olayiwola, chair of the Department of Family Medicine at Ohio State University, wrote an essay on her experiences taking care of patients as a Black family physician.¹⁰ In January of 2019, Family Medicine published an entire issue devoted to racism in education and training.¹¹ Dr. Eduardo Medina, a family physician and public health scholar, co-authored a call to action in 2016 for health professionals to dismantle structural racism and support Black lives to achieve health equity. His recent 2020 article builds on that theme and describes the disproportionate deaths of Black people due to racial injustice and the COVID-19 pandemic as converging public health emergencies.^12,13 In the wake of these emergencies a fundamental transformation is warranted, and family physicians can play a key role.

We, the editors of family medicine journals, commit to actively examine the effects of racism on society and health and to take action to eliminate structural racism in our editorial processes. As an intellectual home for our profession, we have a unique responsibility and opportunity to educate and continue the conversation about institutional racism, health inequities, and antiracism in medicine. We will take immediate steps to enact tangible advances on these fronts. We will encourage and mentor authors from groups underrepresented in medicine. We will ensure that content includes an emphasis on cultural humility, diversity and inclusion, implicit bias, and the impact of racism on medicine and health. We will recruit editors and editorial board members from groups underrepresented in medicine. We will encourage collaboration and accountability within our specialty to confront systemic racism through content and processes in all of our individual publications. We recognize that these are small steps in an ongoing process of active antiracism, but we believe these steps are crucial. As editors in family medicine, we are committed to progress toward equity and justice.

Simultaneously published in American Family Physician, Annals of Family Medicine, Canadian Family Physician, Family Medicine, FP Essentials, FPIN/Evidence Based Practice, FPM, Journal of the American Board of Family Medicine, The Journal of Family Practice, and PRiMER.

Acknowledgement –

The authors thank Renee Crichlow, MD, Byron Jasper, MD, MPH, and Victoria Murrain, DO, for their insightful comments on this editorial.

Alzheimer’s disease (AD), the most common form of dementia, affects more than 5 million Americans.¹ Estimates suggest that by 2050, the prevalence could triple, reaching 13 to 16 million.¹ To effectively care for patients with AD and their families, family physicians need to be familiar with the latest evidence on all facets of care, from initial detection to patient management and end-of-life care.

This evidence-based review will help you toward that end by answering common questions regarding Alzheimer’s care, including whether routine screening is advisable, what tests should be ordered, which interventions (including nonpharmacologic options) are worth considering, and how best to counsel patients and families about end-of-life care.

Routine screening? Still subject to debate

In considering routine dementia screening in primary care, the key question is whether screening improves outcomes. Advocates note that individuals with dementia may appear unimpaired during office visits and may not report symptoms due to lack of insight; they point out, too, that waiting for an event that makes cognitive impairment obvious, such as a driving mishap, is risky.² Those who advocate routine screening also note that only about half of those who have dementia are ever diagnosed.³

Others, including the US Preventive Services Task Force (USPSTF), disagree. In its 2014 evidence review, the USPSTF indicated that there is “insufficient evidence to assess the balance of benefits and harms of screening for cognitive impairment in older adults.”⁴

Mixed messages

The dearth of evidence is also reflected in the conflicting recommendations of the Affordable Care Act (ACA) and the Centers for Medicare and Medicaid Services (CMS). The ACA requires physicians to assess the cognitive function of Medicare patients during their annual wellness visits. CMS, however, instructs providers to screen for dementia only if observation or concerns raised by the patient or family suggest the possibility of impairment, and does not recommend any particular test.⁵

Cost-effectiveness analyses raise questions about the value of routine screening, as well. Evidence suggests that if a primary care physician screens 300 older patients, 39 will have a positive screen. But only about half of those 39 will agree to a diagnostic evaluation, and no more than 9 will ultimately be diagnosed with dementia. The estimated cost of identifying 9 cases is nearly $40,000—all in the absence of a treatment to cure or stop the progression of the disorder.⁶

The bottom line: Evidence does not support routine dementia screening of older adults. When cognitive impairment is suspected, however, physicians should conduct a diagnostic evaluation—and consider educating patients and families about the Alzheimer’s Association (AA)’s 10 warning signs of AD.⁷ (See “Is it Alzheimer’s? 10 warning signs”⁷ below.) A longer version, available at http://www.alz.org/national/documents/checklist_10signs.pdf, outlines the cognitive changes that are characteristic of healthy aging and compares them to changes suggestive of early dementia.⁷

How to proceed when you suspect AD

Step 1: Screening instrument. The first step in the diagnostic evaluation of a patient with suspected AD is to determine if, in fact, cognitive impairment is present. This can be done by screening with in-office screening instruments, such as the Mini-Cog (available at alz.org/documents_custom/minicog.pdf) or Mini-Mental State Examination (MMSE; health.gov.bc.ca/pharmacare/adti/clinician/pdf/ADTI%20SMMSE-GDS%20Reference%20Card.pdf), among others.⁸

Step 2: Clinical evaluation. If observation and test results suggest cognitive impairment, the next step is to determine whether clinical findings are consistent with the diagnostic criteria for AD (TABLE 1)⁹ developed by workgroups from the National Institute on Aging (NIA)/AA in 2011. A work-up is necessary to identify conditions that can mimic dementia (eg, depression) and behaviors that suggest another type of dementia, such as frontotemporal or Lewy body dementia.¹⁰ Lab testing should be included to rule out potentially reversible causes of cognitive dysfunction (eg, hypothyroidism, vitamin D deficiency).

Step 3: Neuropsychological evaluation. The NIA/AA recommends neuropsychological testing when the brief cognitive tests, history, and clinical work-up are not sufficient for a definitive diagnosis of dementia.⁹When brief cognitive tests, history, and clinical work-up are inconclusive, refer patients for neuropsychological testing.This generally involves a referral to a neuropsychologist, who conducts a battery of standardized tests to evaluate attention, memory, language, visual-spatial abilities, and executive functions, among others. Neuropsychological testing can confirm the presence of cognitive impairment and aid in the differential diagnosis by comparing the patient’s performance in these domains with characteristic features of different dementia syndromes.

Step 4: Brain imaging with either computed tomography or magnetic resonance imaging can be included in the work-up for patients with suspected AD to rule out abnormalities—eg, metastatic cancer, hydrocephalus, or occult chronic subdural hematoma—that could be causing cognitive impairment.^9,10 Clinical features that generally warrant brain imaging include onset of cognitive impairment before age 60; unexplained focal neurologic signs or symptoms; abrupt onset or rapid decline; and/or predisposing conditions, such as cancer or anticoagulant treatment.¹⁰

The role of biomarkers and advanced brain imaging

Biomarkers that might provide confirmation of AD in patients who exhibit early symptoms of dementia have been studied extensively.¹¹ The NIA/AA identified 2 categories of AD biomarkers:

• tests for β-amyloid deposition in the brain, including spinal fluid assays for β-amyloid (Aβ₄₂) and positron emission tomography (PET) scans after intravenous injection of florbetapir or flutemetamol, which bind to amyloid in the brain; and
• tests for neuronal degeneration, which would include spinal fluid assays for tau protein and PET scans after injection of fluorodeoxyglucose (FDG), which shows decreased uptake in patients with AD.⁹

Research reveals the promise of these biomarkers as diagnostic tools, particularly in patients with an atypical presentation of dementia or mild cognitive impairment (MCI) that may be associated with early AD.¹² (More on MCI in a moment.) However, the NIA/AA concluded that additional research is needed to validate these tests for routine diagnostic purposes. Medicare covers PET scans with FDG only for the differential diagnosis of AD vs frontotemporal dementia.¹³

Mild cognitive impairment: How likely that it will progress?

Along with diagnostic criteria for AD, the NIA/AA developed criteria for a symptomatic predementia phase of AD—often referred to as MCI.¹⁴ According to the workgroup, MCI is diagnosed when:

Progression: Less likely than you might think

Patients with MCI are at risk for progression to overt dementia, with an overall annual conversion rate from MCI to dementia estimated at 10% to 15%.^15,16 This estimate must be interpreted with caution, however, because most studies were conducted prior to the 2011 guidelines, when different diagnostic criteria were used. Observers have noted, too, that the numbers largely reflect data collected in specialty clinics and that community-based studies reveal substantially lower conversion rates (3%-6% per year).¹⁶ In addition, evidence suggests that many patients with MCI demonstrate long-term stability or even reversal of deficits.¹⁷

While there is some consideration of the use of biomarkers and amyloid imaging tests to help determine which patients with MCI will progress to AD, practice guidelines do not currently recommend such testing and it is not covered by Medicare.

When evidence indicates an AD diagnosis

When faced with the need to communicate an AD diagnosis, follow the general recommendations for delivering any bad news or discouraging prognosis:

Prioritize and limit the information you provide, determining not only what the patient and family want to hear, but also how much they are able to comprehend.

Confirm that the patient and family understand the information you’ve provided.

Offer emotional support and recommend additional resources¹⁸ (TABLE 2).

Given the progressive cognitive decline that characterizes AD, it is important to address the primary caregiver’s understanding of, and ability to cope with, the disease. It is also important to explore beliefs and attitudes regarding AD. Keep in mind that different cultural groups tend to differ in their beliefs about the nature, cause, and appropriate management of AD, as well as the role of spirituality, help-seeking, and stigma.^19,20

When communicating an Alzheimer's disease diagnosis, prioritize information and offer emotional support. The progressive and ultimately fatal nature of AD also makes planning for the future a priority. Ideally, patients should be engaged in discussions regarding end-of-life care as early as possible, while they are still able to make informed decisions and express their preferences. Discussing end-of-life care can be overwhelming for newly diagnosed patients and their families, however, so it is important that you address issues—medical, financial, and legal planning, for example—that families should be considering.

TABLE 2

Drugs address cognitive and behavioral function

No current treatments
 can cure 
or significantly alter the progression of AD, but 2 classes of medications are used to improve cognitive function. No currently available treatments can cure or significantly alter the progression of AD, but 2 classes of medications are used in an attempt to improve cognitive function. One is cholinesterase inhibitors (ChEIs), which potentiate acetylcholine synaptic transmission. The other is N-methyl-D-aspartate (NMDA) glutamate receptor blockers. Other classes of drugs are sometimes used to treat behavioral symptoms of dementia, such as agitation, aggression, mood disorders, and psychosis (eg, delusions and hallucinations).

Cognitive function. Results from studies of pharmacologic management of MCI vary widely, but recent reviews have found no convincing evidence that either ChEIs or NMDA receptor blockers have an effect on progression from MCI to dementia.^21,22 Neither class is FDA-approved for treating MCI.

In patients with dementia, the effects of ChEIs and NMDA receptor blockers on cognition are statistically significant but modest, and often of questionable clinical relevance.²³ Nonetheless, among ChEIs, donepezil is approved by the US Food and Drug Administration (FDA) for mild, moderate, and severe dementia and galantamine and rivastigmine are approved for mild and moderate dementia. There is no evidence that any one ChEI is more effective than any other,²⁴ and the choice of drugs is often guided by cost, adverse effects, and health plan formularies. Memantine, the only FDA-approved NMDA receptor blocker, is approved for moderate to severe dementia and can be used alone or in combination with a ChEI.

In patients with dementia, the effects of ChEIs and NMDA receptor blockers on cognition are statistically significant but modest, and are often of questionable clinical relevance. If these drugs are used in an attempt to improve cognition in AD, guidelines recommend the following approach for initial therapy: Prescribe a ChEI for the mild stage, a ChEI plus memantine for the moderate stage, and memantine (with or without a ChEI) for the severe stage.²⁵ The recommendations also include monitoring every 6 months.

There is no consensus about when to discontinue medication. Various published recommendations call for continuing treatment until the patient has “lost all cognitive and functional abilities;”²² until the patient’s MMSE score falls below 10 and there is no indication that the drug is having a “worthwhile effect;”²¹ or until he or she has reached stage 7 on the Reisberg Functional Assessment Staging scale, indicating nonambulatory status with speech limited to one to 5 words a day.¹⁰

Behavioral function. A variety of drugs are used to treat behavioral symptoms in AD. While not FDA-approved for this use, the most widely prescribed agents are second-generation antipsychotics (aripiprazole, olanzapine, quetiapine, and risperidone). The main effect of these drugs is often nothing more than sedation, and one large multi­site clinical trial concluded that the adverse effects offset the benefits for patients with AD.²⁶ Indeed, the FDA has issued an advisory on the use of second-generation antipsychotics in AD patients, stating that they are associated with an increased risk of death.²⁷ The recently updated Beers Criteria strongly recommend avoiding these drugs for treating behavioral disturbances in AD unless nonpharmacologic options have failed and the patient is a threat to self or others.²⁸

The FDA has issued an advisory on the use of second-generation antipsychotics in Alzheimer’s patients, stating that they are associated with an increased risk of death. Because of the black-box warning that antipsychotics increase the risk of death, some physicians have advocated obtaining informed consent prior to prescribing such medications.²⁹ At the very least, when family or guardians are involved, a conversation about risks vs benefits should take place and be documented in the medical record.

Other drug classes are also sometimes used in an attempt to improve behavioral function, including anti-seizure medications (valproic acid, carbamazepine), antidepressants (trazodone and selective serotonin reuptake inhibitors), and anxiolytics (benzodiazepines and buspirone). Other than their sedating effects, there is no strong evidence that these drugs are effective for treating dementia-related behavioral disorders. If used, caution is required due to potential adverse effects.

Nonpharmacologic management is “promising”

A recent systematic review of nonpharmacologic interventions for MCI evaluated exercise, training in compensatory strategies, and engagement in cognitively stimulating activities and found “promising but inconclusive” results. The researchers found that studies show mostly positive effects on cognition but have significant methodologic limitations.³⁰ Importantly, there is no evidence of delayed or reduced conversion to dementia.

For patients who already have mild-to-moderate dementia, cognitive stimulation seems to help in the short term.³¹ There is also some evidence that exercise and occupational therapy may slow functional decline,³² but the effects are small to modest and their actual clinical significance (eg, the ability to delay institutionalization) is unclear. There is promising but preliminary evidence that cognitive rehabilitation (helping patients devise strategies to complete daily activities) may improve functioning in everyday life.³³

While behavioral symptoms are often due to the dementia itself, it is important to identify and treat medical and environmental causes that may be contributing, such as infection, pain, and loud or unsafe environments.Interventions such as massage therapy, aromatherapy, exercise, and music therapy may be effective in the short term for agitated behavior. As noted before, nonpharmacologic treatments are generally preferred for behavioral problems and should be considered prior to drug therapy. Approaches that identify and modify both the antecedents and consequences of problem behaviors and increase pleasant events have empirical support for the management of behavioral symptoms.³⁴ Interventions including massage therapy, aromatherapy, exercise, and music therapy may also be effective in the short term for agitated behavior.³⁵

Caregivers should be encouraged to receive training in these strategies through organizations like AA. Caregiver education and support can reduce caregivers’ distress and increase their self-efficacy and coping skills.³⁶

End-of-life care must be addressed

Perhaps the most important aspect of end-of-life care in AD is assuring that families (or health care proxies) understand that AD is a fatal illness, with most patients dying within 4 to 8 years of diagnosis.¹ Evidence indicates that patients whose proxies have a clear recognition of this are less likely to experience “burdensome” interventions such as parenteral therapy, emergency department visits, hospital admissions, and tube feedings in their last 3 months of life.³⁷

Overall, decisions regarding discontinuing medical treatments in advanced AD should be made by balancing the likelihood of benefit with the potential for adverse effects.³⁸The progressive and ultimately fatal nature of Alzheimer's disease makes planning for the future a priority. For example, the American Geriatrics Society recently recommended against feeding tubes because they often result in discomfort due to agitation, use of restraints, and worsening pressure ulcers.³⁹

Unfortunately, only a minority of families receives straightforward information on the course and prognosis of AD, including the fact that patients eventually stop eating and that the natural cause of death is often an acute infection. Studies also show that patients with dementia are at risk for inadequate treatment of pain.⁴⁰ Assuring adequate pain control is an essential component of end-of-life care.

Hospice. End-of-life care can often be improved with hospice care. This service is underused by patients with dementia, even though hospice care is available at no cost through Medicare. Hospice eligibility criteria for patients with AD are shown in
TABLE 3.^41,42

Finally, a word about prevention

Numerous risk factors have been associated with an increased risk of AD (TABLE 4)^2,3. Some, like age and genetics, are nonmodifiable, while others—particularly cardiovascular risk factors—can be modified.¹ There are also factors associated with decreased risk, most notably, physical exercise and participating in cognitively stimulating activities.³ Identification of these factors has led to the hope that addressing them can prevent AD.

But association does not equal causation. In 2010, a report from the National Institutes of Health concluded that, although there are modifiable factors associated with AD, there is insufficient evidence that addressing any of them will actually prevent AD.⁴³ In fact, there is good evidence that some of these factors (eg, statin therapy) are not effective in reducing the incidence of dementia, and that others (eg, vitamin E and estrogen therapy) are potentially harmful.⁴⁴

The absence of empirically supported preventive interventions does not mean, however, that we should disregard these risks and protective factors. Encouraging social engagement, for example, may improve both emotional health and quality of life. Addressing cardiovascular risk factors can reduce the rate of coronary and cerebrovascular disease, potentially including vascular dementia, even if it does not reduce the rate of AD.

Studies are evaluating the use of monoclonal antibodies with anti-amyloid properties for preventing AD in individuals who have APOE ε4 genotypes or high amyloid loads on neuroimaging.⁴⁵ It will be several years before results are available, however, and the outcome of these studies is uncertain as the use of anti-amyloid agents for treating established dementia has not been effective.^46,47

CORRESPONDENCE
Marisa Menchola, PhD, Department of Psychiatry, University of Arizona College of Medicine, 1501 N. Campbell Ave., 7OPC. Tucson, AZ 85724; marisam@psychiatry.arizona.edu

Alzheimer’s disease (AD), the most common form of dementia, affects more than 5 million Americans.¹ Estimates suggest that by 2050, the prevalence could triple, reaching 13 to 16 million.¹ To effectively care for patients with AD and their families, family physicians need to be familiar with the latest evidence on all facets of care, from initial detection to patient management and end-of-life care.

This evidence-based review will help you toward that end by answering common questions regarding Alzheimer’s care, including whether routine screening is advisable, what tests should be ordered, which interventions (including nonpharmacologic options) are worth considering, and how best to counsel patients and families about end-of-life care.

Routine screening? Still subject to debate

In considering routine dementia screening in primary care, the key question is whether screening improves outcomes. Advocates note that individuals with dementia may appear unimpaired during office visits and may not report symptoms due to lack of insight; they point out, too, that waiting for an event that makes cognitive impairment obvious, such as a driving mishap, is risky.² Those who advocate routine screening also note that only about half of those who have dementia are ever diagnosed.³

Others, including the US Preventive Services Task Force (USPSTF), disagree. In its 2014 evidence review, the USPSTF indicated that there is “insufficient evidence to assess the balance of benefits and harms of screening for cognitive impairment in older adults.”⁴

Mixed messages

The dearth of evidence is also reflected in the conflicting recommendations of the Affordable Care Act (ACA) and the Centers for Medicare and Medicaid Services (CMS). The ACA requires physicians to assess the cognitive function of Medicare patients during their annual wellness visits. CMS, however, instructs providers to screen for dementia only if observation or concerns raised by the patient or family suggest the possibility of impairment, and does not recommend any particular test.⁵

Cost-effectiveness analyses raise questions about the value of routine screening, as well. Evidence suggests that if a primary care physician screens 300 older patients, 39 will have a positive screen. But only about half of those 39 will agree to a diagnostic evaluation, and no more than 9 will ultimately be diagnosed with dementia. The estimated cost of identifying 9 cases is nearly $40,000—all in the absence of a treatment to cure or stop the progression of the disorder.⁶

The bottom line: Evidence does not support routine dementia screening of older adults. When cognitive impairment is suspected, however, physicians should conduct a diagnostic evaluation—and consider educating patients and families about the Alzheimer’s Association (AA)’s 10 warning signs of AD.⁷ (See “Is it Alzheimer’s? 10 warning signs”⁷ below.) A longer version, available at http://www.alz.org/national/documents/checklist_10signs.pdf, outlines the cognitive changes that are characteristic of healthy aging and compares them to changes suggestive of early dementia.⁷

How to proceed when you suspect AD

Step 1: Screening instrument. The first step in the diagnostic evaluation of a patient with suspected AD is to determine if, in fact, cognitive impairment is present. This can be done by screening with in-office screening instruments, such as the Mini-Cog (available at alz.org/documents_custom/minicog.pdf) or Mini-Mental State Examination (MMSE; health.gov.bc.ca/pharmacare/adti/clinician/pdf/ADTI%20SMMSE-GDS%20Reference%20Card.pdf), among others.⁸

Step 2: Clinical evaluation. If observation and test results suggest cognitive impairment, the next step is to determine whether clinical findings are consistent with the diagnostic criteria for AD (TABLE 1)⁹ developed by workgroups from the National Institute on Aging (NIA)/AA in 2011. A work-up is necessary to identify conditions that can mimic dementia (eg, depression) and behaviors that suggest another type of dementia, such as frontotemporal or Lewy body dementia.¹⁰ Lab testing should be included to rule out potentially reversible causes of cognitive dysfunction (eg, hypothyroidism, vitamin D deficiency).

Step 3: Neuropsychological evaluation. The NIA/AA recommends neuropsychological testing when the brief cognitive tests, history, and clinical work-up are not sufficient for a definitive diagnosis of dementia.⁹When brief cognitive tests, history, and clinical work-up are inconclusive, refer patients for neuropsychological testing.This generally involves a referral to a neuropsychologist, who conducts a battery of standardized tests to evaluate attention, memory, language, visual-spatial abilities, and executive functions, among others. Neuropsychological testing can confirm the presence of cognitive impairment and aid in the differential diagnosis by comparing the patient’s performance in these domains with characteristic features of different dementia syndromes.

Step 4: Brain imaging with either computed tomography or magnetic resonance imaging can be included in the work-up for patients with suspected AD to rule out abnormalities—eg, metastatic cancer, hydrocephalus, or occult chronic subdural hematoma—that could be causing cognitive impairment.^9,10 Clinical features that generally warrant brain imaging include onset of cognitive impairment before age 60; unexplained focal neurologic signs or symptoms; abrupt onset or rapid decline; and/or predisposing conditions, such as cancer or anticoagulant treatment.¹⁰

The role of biomarkers and advanced brain imaging

Biomarkers that might provide confirmation of AD in patients who exhibit early symptoms of dementia have been studied extensively.¹¹ The NIA/AA identified 2 categories of AD biomarkers:

• tests for β-amyloid deposition in the brain, including spinal fluid assays for β-amyloid (Aβ₄₂) and positron emission tomography (PET) scans after intravenous injection of florbetapir or flutemetamol, which bind to amyloid in the brain; and
• tests for neuronal degeneration, which would include spinal fluid assays for tau protein and PET scans after injection of fluorodeoxyglucose (FDG), which shows decreased uptake in patients with AD.⁹

Research reveals the promise of these biomarkers as diagnostic tools, particularly in patients with an atypical presentation of dementia or mild cognitive impairment (MCI) that may be associated with early AD.¹² (More on MCI in a moment.) However, the NIA/AA concluded that additional research is needed to validate these tests for routine diagnostic purposes. Medicare covers PET scans with FDG only for the differential diagnosis of AD vs frontotemporal dementia.¹³

Mild cognitive impairment: How likely that it will progress?

Along with diagnostic criteria for AD, the NIA/AA developed criteria for a symptomatic predementia phase of AD—often referred to as MCI.¹⁴ According to the workgroup, MCI is diagnosed when:

Progression: Less likely than you might think

Patients with MCI are at risk for progression to overt dementia, with an overall annual conversion rate from MCI to dementia estimated at 10% to 15%.^15,16 This estimate must be interpreted with caution, however, because most studies were conducted prior to the 2011 guidelines, when different diagnostic criteria were used. Observers have noted, too, that the numbers largely reflect data collected in specialty clinics and that community-based studies reveal substantially lower conversion rates (3%-6% per year).¹⁶ In addition, evidence suggests that many patients with MCI demonstrate long-term stability or even reversal of deficits.¹⁷

While there is some consideration of the use of biomarkers and amyloid imaging tests to help determine which patients with MCI will progress to AD, practice guidelines do not currently recommend such testing and it is not covered by Medicare.

When evidence indicates an AD diagnosis

When faced with the need to communicate an AD diagnosis, follow the general recommendations for delivering any bad news or discouraging prognosis:

Prioritize and limit the information you provide, determining not only what the patient and family want to hear, but also how much they are able to comprehend.

Confirm that the patient and family understand the information you’ve provided.

Offer emotional support and recommend additional resources¹⁸ (TABLE 2).

Given the progressive cognitive decline that characterizes AD, it is important to address the primary caregiver’s understanding of, and ability to cope with, the disease. It is also important to explore beliefs and attitudes regarding AD. Keep in mind that different cultural groups tend to differ in their beliefs about the nature, cause, and appropriate management of AD, as well as the role of spirituality, help-seeking, and stigma.^19,20

When communicating an Alzheimer's disease diagnosis, prioritize information and offer emotional support. The progressive and ultimately fatal nature of AD also makes planning for the future a priority. Ideally, patients should be engaged in discussions regarding end-of-life care as early as possible, while they are still able to make informed decisions and express their preferences. Discussing end-of-life care can be overwhelming for newly diagnosed patients and their families, however, so it is important that you address issues—medical, financial, and legal planning, for example—that families should be considering.

TABLE 2

Drugs address cognitive and behavioral function

No current treatments
 can cure 
or significantly alter the progression of AD, but 2 classes of medications are used to improve cognitive function. No currently available treatments can cure or significantly alter the progression of AD, but 2 classes of medications are used in an attempt to improve cognitive function. One is cholinesterase inhibitors (ChEIs), which potentiate acetylcholine synaptic transmission. The other is N-methyl-D-aspartate (NMDA) glutamate receptor blockers. Other classes of drugs are sometimes used to treat behavioral symptoms of dementia, such as agitation, aggression, mood disorders, and psychosis (eg, delusions and hallucinations).

Cognitive function. Results from studies of pharmacologic management of MCI vary widely, but recent reviews have found no convincing evidence that either ChEIs or NMDA receptor blockers have an effect on progression from MCI to dementia.^21,22 Neither class is FDA-approved for treating MCI.

In patients with dementia, the effects of ChEIs and NMDA receptor blockers on cognition are statistically significant but modest, and often of questionable clinical relevance.²³ Nonetheless, among ChEIs, donepezil is approved by the US Food and Drug Administration (FDA) for mild, moderate, and severe dementia and galantamine and rivastigmine are approved for mild and moderate dementia. There is no evidence that any one ChEI is more effective than any other,²⁴ and the choice of drugs is often guided by cost, adverse effects, and health plan formularies. Memantine, the only FDA-approved NMDA receptor blocker, is approved for moderate to severe dementia and can be used alone or in combination with a ChEI.

In patients with dementia, the effects of ChEIs and NMDA receptor blockers on cognition are statistically significant but modest, and are often of questionable clinical relevance. If these drugs are used in an attempt to improve cognition in AD, guidelines recommend the following approach for initial therapy: Prescribe a ChEI for the mild stage, a ChEI plus memantine for the moderate stage, and memantine (with or without a ChEI) for the severe stage.²⁵ The recommendations also include monitoring every 6 months.

There is no consensus about when to discontinue medication. Various published recommendations call for continuing treatment until the patient has “lost all cognitive and functional abilities;”²² until the patient’s MMSE score falls below 10 and there is no indication that the drug is having a “worthwhile effect;”²¹ or until he or she has reached stage 7 on the Reisberg Functional Assessment Staging scale, indicating nonambulatory status with speech limited to one to 5 words a day.¹⁰

Behavioral function. A variety of drugs are used to treat behavioral symptoms in AD. While not FDA-approved for this use, the most widely prescribed agents are second-generation antipsychotics (aripiprazole, olanzapine, quetiapine, and risperidone). The main effect of these drugs is often nothing more than sedation, and one large multi­site clinical trial concluded that the adverse effects offset the benefits for patients with AD.²⁶ Indeed, the FDA has issued an advisory on the use of second-generation antipsychotics in AD patients, stating that they are associated with an increased risk of death.²⁷ The recently updated Beers Criteria strongly recommend avoiding these drugs for treating behavioral disturbances in AD unless nonpharmacologic options have failed and the patient is a threat to self or others.²⁸

The FDA has issued an advisory on the use of second-generation antipsychotics in Alzheimer’s patients, stating that they are associated with an increased risk of death. Because of the black-box warning that antipsychotics increase the risk of death, some physicians have advocated obtaining informed consent prior to prescribing such medications.²⁹ At the very least, when family or guardians are involved, a conversation about risks vs benefits should take place and be documented in the medical record.

Other drug classes are also sometimes used in an attempt to improve behavioral function, including anti-seizure medications (valproic acid, carbamazepine), antidepressants (trazodone and selective serotonin reuptake inhibitors), and anxiolytics (benzodiazepines and buspirone). Other than their sedating effects, there is no strong evidence that these drugs are effective for treating dementia-related behavioral disorders. If used, caution is required due to potential adverse effects.

Nonpharmacologic management is “promising”

A recent systematic review of nonpharmacologic interventions for MCI evaluated exercise, training in compensatory strategies, and engagement in cognitively stimulating activities and found “promising but inconclusive” results. The researchers found that studies show mostly positive effects on cognition but have significant methodologic limitations.³⁰ Importantly, there is no evidence of delayed or reduced conversion to dementia.

For patients who already have mild-to-moderate dementia, cognitive stimulation seems to help in the short term.³¹ There is also some evidence that exercise and occupational therapy may slow functional decline,³² but the effects are small to modest and their actual clinical significance (eg, the ability to delay institutionalization) is unclear. There is promising but preliminary evidence that cognitive rehabilitation (helping patients devise strategies to complete daily activities) may improve functioning in everyday life.³³

While behavioral symptoms are often due to the dementia itself, it is important to identify and treat medical and environmental causes that may be contributing, such as infection, pain, and loud or unsafe environments.Interventions such as massage therapy, aromatherapy, exercise, and music therapy may be effective in the short term for agitated behavior. As noted before, nonpharmacologic treatments are generally preferred for behavioral problems and should be considered prior to drug therapy. Approaches that identify and modify both the antecedents and consequences of problem behaviors and increase pleasant events have empirical support for the management of behavioral symptoms.³⁴ Interventions including massage therapy, aromatherapy, exercise, and music therapy may also be effective in the short term for agitated behavior.³⁵

Caregivers should be encouraged to receive training in these strategies through organizations like AA. Caregiver education and support can reduce caregivers’ distress and increase their self-efficacy and coping skills.³⁶

End-of-life care must be addressed

Perhaps the most important aspect of end-of-life care in AD is assuring that families (or health care proxies) understand that AD is a fatal illness, with most patients dying within 4 to 8 years of diagnosis.¹ Evidence indicates that patients whose proxies have a clear recognition of this are less likely to experience “burdensome” interventions such as parenteral therapy, emergency department visits, hospital admissions, and tube feedings in their last 3 months of life.³⁷

Overall, decisions regarding discontinuing medical treatments in advanced AD should be made by balancing the likelihood of benefit with the potential for adverse effects.³⁸The progressive and ultimately fatal nature of Alzheimer's disease makes planning for the future a priority. For example, the American Geriatrics Society recently recommended against feeding tubes because they often result in discomfort due to agitation, use of restraints, and worsening pressure ulcers.³⁹

Unfortunately, only a minority of families receives straightforward information on the course and prognosis of AD, including the fact that patients eventually stop eating and that the natural cause of death is often an acute infection. Studies also show that patients with dementia are at risk for inadequate treatment of pain.⁴⁰ Assuring adequate pain control is an essential component of end-of-life care.

Hospice. End-of-life care can often be improved with hospice care. This service is underused by patients with dementia, even though hospice care is available at no cost through Medicare. Hospice eligibility criteria for patients with AD are shown in
TABLE 3.^41,42

Finally, a word about prevention

Numerous risk factors have been associated with an increased risk of AD (TABLE 4)^2,3. Some, like age and genetics, are nonmodifiable, while others—particularly cardiovascular risk factors—can be modified.¹ There are also factors associated with decreased risk, most notably, physical exercise and participating in cognitively stimulating activities.³ Identification of these factors has led to the hope that addressing them can prevent AD.

But association does not equal causation. In 2010, a report from the National Institutes of Health concluded that, although there are modifiable factors associated with AD, there is insufficient evidence that addressing any of them will actually prevent AD.⁴³ In fact, there is good evidence that some of these factors (eg, statin therapy) are not effective in reducing the incidence of dementia, and that others (eg, vitamin E and estrogen therapy) are potentially harmful.⁴⁴

The absence of empirically supported preventive interventions does not mean, however, that we should disregard these risks and protective factors. Encouraging social engagement, for example, may improve both emotional health and quality of life. Addressing cardiovascular risk factors can reduce the rate of coronary and cerebrovascular disease, potentially including vascular dementia, even if it does not reduce the rate of AD.

Studies are evaluating the use of monoclonal antibodies with anti-amyloid properties for preventing AD in individuals who have APOE ε4 genotypes or high amyloid loads on neuroimaging.⁴⁵ It will be several years before results are available, however, and the outcome of these studies is uncertain as the use of anti-amyloid agents for treating established dementia has not been effective.^46,47

CORRESPONDENCE
Marisa Menchola, PhD, Department of Psychiatry, University of Arizona College of Medicine, 1501 N. Campbell Ave., 7OPC. Tucson, AZ 85724; marisam@psychiatry.arizona.edu

Alzheimer’s disease (AD), the most common form of dementia, affects more than 5 million Americans.¹ Estimates suggest that by 2050, the prevalence could triple, reaching 13 to 16 million.¹ To effectively care for patients with AD and their families, family physicians need to be familiar with the latest evidence on all facets of care, from initial detection to patient management and end-of-life care.

This evidence-based review will help you toward that end by answering common questions regarding Alzheimer’s care, including whether routine screening is advisable, what tests should be ordered, which interventions (including nonpharmacologic options) are worth considering, and how best to counsel patients and families about end-of-life care.

Routine screening? Still subject to debate

In considering routine dementia screening in primary care, the key question is whether screening improves outcomes. Advocates note that individuals with dementia may appear unimpaired during office visits and may not report symptoms due to lack of insight; they point out, too, that waiting for an event that makes cognitive impairment obvious, such as a driving mishap, is risky.² Those who advocate routine screening also note that only about half of those who have dementia are ever diagnosed.³

Others, including the US Preventive Services Task Force (USPSTF), disagree. In its 2014 evidence review, the USPSTF indicated that there is “insufficient evidence to assess the balance of benefits and harms of screening for cognitive impairment in older adults.”⁴

Mixed messages

The dearth of evidence is also reflected in the conflicting recommendations of the Affordable Care Act (ACA) and the Centers for Medicare and Medicaid Services (CMS). The ACA requires physicians to assess the cognitive function of Medicare patients during their annual wellness visits. CMS, however, instructs providers to screen for dementia only if observation or concerns raised by the patient or family suggest the possibility of impairment, and does not recommend any particular test.⁵

Cost-effectiveness analyses raise questions about the value of routine screening, as well. Evidence suggests that if a primary care physician screens 300 older patients, 39 will have a positive screen. But only about half of those 39 will agree to a diagnostic evaluation, and no more than 9 will ultimately be diagnosed with dementia. The estimated cost of identifying 9 cases is nearly $40,000—all in the absence of a treatment to cure or stop the progression of the disorder.⁶

The bottom line: Evidence does not support routine dementia screening of older adults. When cognitive impairment is suspected, however, physicians should conduct a diagnostic evaluation—and consider educating patients and families about the Alzheimer’s Association (AA)’s 10 warning signs of AD.⁷ (See “Is it Alzheimer’s? 10 warning signs”⁷ below.) A longer version, available at http://www.alz.org/national/documents/checklist_10signs.pdf, outlines the cognitive changes that are characteristic of healthy aging and compares them to changes suggestive of early dementia.⁷

How to proceed when you suspect AD

Step 1: Screening instrument. The first step in the diagnostic evaluation of a patient with suspected AD is to determine if, in fact, cognitive impairment is present. This can be done by screening with in-office screening instruments, such as the Mini-Cog (available at alz.org/documents_custom/minicog.pdf) or Mini-Mental State Examination (MMSE; health.gov.bc.ca/pharmacare/adti/clinician/pdf/ADTI%20SMMSE-GDS%20Reference%20Card.pdf), among others.⁸

Step 2: Clinical evaluation. If observation and test results suggest cognitive impairment, the next step is to determine whether clinical findings are consistent with the diagnostic criteria for AD (TABLE 1)⁹ developed by workgroups from the National Institute on Aging (NIA)/AA in 2011. A work-up is necessary to identify conditions that can mimic dementia (eg, depression) and behaviors that suggest another type of dementia, such as frontotemporal or Lewy body dementia.¹⁰ Lab testing should be included to rule out potentially reversible causes of cognitive dysfunction (eg, hypothyroidism, vitamin D deficiency).

Step 3: Neuropsychological evaluation. The NIA/AA recommends neuropsychological testing when the brief cognitive tests, history, and clinical work-up are not sufficient for a definitive diagnosis of dementia.⁹When brief cognitive tests, history, and clinical work-up are inconclusive, refer patients for neuropsychological testing.This generally involves a referral to a neuropsychologist, who conducts a battery of standardized tests to evaluate attention, memory, language, visual-spatial abilities, and executive functions, among others. Neuropsychological testing can confirm the presence of cognitive impairment and aid in the differential diagnosis by comparing the patient’s performance in these domains with characteristic features of different dementia syndromes.

Step 4: Brain imaging with either computed tomography or magnetic resonance imaging can be included in the work-up for patients with suspected AD to rule out abnormalities—eg, metastatic cancer, hydrocephalus, or occult chronic subdural hematoma—that could be causing cognitive impairment.^9,10 Clinical features that generally warrant brain imaging include onset of cognitive impairment before age 60; unexplained focal neurologic signs or symptoms; abrupt onset or rapid decline; and/or predisposing conditions, such as cancer or anticoagulant treatment.¹⁰

The role of biomarkers and advanced brain imaging

Biomarkers that might provide confirmation of AD in patients who exhibit early symptoms of dementia have been studied extensively.¹¹ The NIA/AA identified 2 categories of AD biomarkers:

• tests for β-amyloid deposition in the brain, including spinal fluid assays for β-amyloid (Aβ₄₂) and positron emission tomography (PET) scans after intravenous injection of florbetapir or flutemetamol, which bind to amyloid in the brain; and
• tests for neuronal degeneration, which would include spinal fluid assays for tau protein and PET scans after injection of fluorodeoxyglucose (FDG), which shows decreased uptake in patients with AD.⁹

Research reveals the promise of these biomarkers as diagnostic tools, particularly in patients with an atypical presentation of dementia or mild cognitive impairment (MCI) that may be associated with early AD.¹² (More on MCI in a moment.) However, the NIA/AA concluded that additional research is needed to validate these tests for routine diagnostic purposes. Medicare covers PET scans with FDG only for the differential diagnosis of AD vs frontotemporal dementia.¹³

Mild cognitive impairment: How likely that it will progress?

Along with diagnostic criteria for AD, the NIA/AA developed criteria for a symptomatic predementia phase of AD—often referred to as MCI.¹⁴ According to the workgroup, MCI is diagnosed when:

Progression: Less likely than you might think

Patients with MCI are at risk for progression to overt dementia, with an overall annual conversion rate from MCI to dementia estimated at 10% to 15%.^15,16 This estimate must be interpreted with caution, however, because most studies were conducted prior to the 2011 guidelines, when different diagnostic criteria were used. Observers have noted, too, that the numbers largely reflect data collected in specialty clinics and that community-based studies reveal substantially lower conversion rates (3%-6% per year).¹⁶ In addition, evidence suggests that many patients with MCI demonstrate long-term stability or even reversal of deficits.¹⁷

While there is some consideration of the use of biomarkers and amyloid imaging tests to help determine which patients with MCI will progress to AD, practice guidelines do not currently recommend such testing and it is not covered by Medicare.

When evidence indicates an AD diagnosis

When faced with the need to communicate an AD diagnosis, follow the general recommendations for delivering any bad news or discouraging prognosis:

Prioritize and limit the information you provide, determining not only what the patient and family want to hear, but also how much they are able to comprehend.

Confirm that the patient and family understand the information you’ve provided.

Offer emotional support and recommend additional resources¹⁸ (TABLE 2).

Given the progressive cognitive decline that characterizes AD, it is important to address the primary caregiver’s understanding of, and ability to cope with, the disease. It is also important to explore beliefs and attitudes regarding AD. Keep in mind that different cultural groups tend to differ in their beliefs about the nature, cause, and appropriate management of AD, as well as the role of spirituality, help-seeking, and stigma.^19,20

When communicating an Alzheimer's disease diagnosis, prioritize information and offer emotional support. The progressive and ultimately fatal nature of AD also makes planning for the future a priority. Ideally, patients should be engaged in discussions regarding end-of-life care as early as possible, while they are still able to make informed decisions and express their preferences. Discussing end-of-life care can be overwhelming for newly diagnosed patients and their families, however, so it is important that you address issues—medical, financial, and legal planning, for example—that families should be considering.

TABLE 2

Drugs address cognitive and behavioral function

No current treatments
 can cure 
or significantly alter the progression of AD, but 2 classes of medications are used to improve cognitive function. No currently available treatments can cure or significantly alter the progression of AD, but 2 classes of medications are used in an attempt to improve cognitive function. One is cholinesterase inhibitors (ChEIs), which potentiate acetylcholine synaptic transmission. The other is N-methyl-D-aspartate (NMDA) glutamate receptor blockers. Other classes of drugs are sometimes used to treat behavioral symptoms of dementia, such as agitation, aggression, mood disorders, and psychosis (eg, delusions and hallucinations).

Cognitive function. Results from studies of pharmacologic management of MCI vary widely, but recent reviews have found no convincing evidence that either ChEIs or NMDA receptor blockers have an effect on progression from MCI to dementia.^21,22 Neither class is FDA-approved for treating MCI.

In patients with dementia, the effects of ChEIs and NMDA receptor blockers on cognition are statistically significant but modest, and often of questionable clinical relevance.²³ Nonetheless, among ChEIs, donepezil is approved by the US Food and Drug Administration (FDA) for mild, moderate, and severe dementia and galantamine and rivastigmine are approved for mild and moderate dementia. There is no evidence that any one ChEI is more effective than any other,²⁴ and the choice of drugs is often guided by cost, adverse effects, and health plan formularies. Memantine, the only FDA-approved NMDA receptor blocker, is approved for moderate to severe dementia and can be used alone or in combination with a ChEI.

In patients with dementia, the effects of ChEIs and NMDA receptor blockers on cognition are statistically significant but modest, and are often of questionable clinical relevance. If these drugs are used in an attempt to improve cognition in AD, guidelines recommend the following approach for initial therapy: Prescribe a ChEI for the mild stage, a ChEI plus memantine for the moderate stage, and memantine (with or without a ChEI) for the severe stage.²⁵ The recommendations also include monitoring every 6 months.

There is no consensus about when to discontinue medication. Various published recommendations call for continuing treatment until the patient has “lost all cognitive and functional abilities;”²² until the patient’s MMSE score falls below 10 and there is no indication that the drug is having a “worthwhile effect;”²¹ or until he or she has reached stage 7 on the Reisberg Functional Assessment Staging scale, indicating nonambulatory status with speech limited to one to 5 words a day.¹⁰

Behavioral function. A variety of drugs are used to treat behavioral symptoms in AD. While not FDA-approved for this use, the most widely prescribed agents are second-generation antipsychotics (aripiprazole, olanzapine, quetiapine, and risperidone). The main effect of these drugs is often nothing more than sedation, and one large multi­site clinical trial concluded that the adverse effects offset the benefits for patients with AD.²⁶ Indeed, the FDA has issued an advisory on the use of second-generation antipsychotics in AD patients, stating that they are associated with an increased risk of death.²⁷ The recently updated Beers Criteria strongly recommend avoiding these drugs for treating behavioral disturbances in AD unless nonpharmacologic options have failed and the patient is a threat to self or others.²⁸

The FDA has issued an advisory on the use of second-generation antipsychotics in Alzheimer’s patients, stating that they are associated with an increased risk of death. Because of the black-box warning that antipsychotics increase the risk of death, some physicians have advocated obtaining informed consent prior to prescribing such medications.²⁹ At the very least, when family or guardians are involved, a conversation about risks vs benefits should take place and be documented in the medical record.

Other drug classes are also sometimes used in an attempt to improve behavioral function, including anti-seizure medications (valproic acid, carbamazepine), antidepressants (trazodone and selective serotonin reuptake inhibitors), and anxiolytics (benzodiazepines and buspirone). Other than their sedating effects, there is no strong evidence that these drugs are effective for treating dementia-related behavioral disorders. If used, caution is required due to potential adverse effects.

Nonpharmacologic management is “promising”

A recent systematic review of nonpharmacologic interventions for MCI evaluated exercise, training in compensatory strategies, and engagement in cognitively stimulating activities and found “promising but inconclusive” results. The researchers found that studies show mostly positive effects on cognition but have significant methodologic limitations.³⁰ Importantly, there is no evidence of delayed or reduced conversion to dementia.

For patients who already have mild-to-moderate dementia, cognitive stimulation seems to help in the short term.³¹ There is also some evidence that exercise and occupational therapy may slow functional decline,³² but the effects are small to modest and their actual clinical significance (eg, the ability to delay institutionalization) is unclear. There is promising but preliminary evidence that cognitive rehabilitation (helping patients devise strategies to complete daily activities) may improve functioning in everyday life.³³

While behavioral symptoms are often due to the dementia itself, it is important to identify and treat medical and environmental causes that may be contributing, such as infection, pain, and loud or unsafe environments.Interventions such as massage therapy, aromatherapy, exercise, and music therapy may be effective in the short term for agitated behavior. As noted before, nonpharmacologic treatments are generally preferred for behavioral problems and should be considered prior to drug therapy. Approaches that identify and modify both the antecedents and consequences of problem behaviors and increase pleasant events have empirical support for the management of behavioral symptoms.³⁴ Interventions including massage therapy, aromatherapy, exercise, and music therapy may also be effective in the short term for agitated behavior.³⁵

Caregivers should be encouraged to receive training in these strategies through organizations like AA. Caregiver education and support can reduce caregivers’ distress and increase their self-efficacy and coping skills.³⁶

End-of-life care must be addressed

Perhaps the most important aspect of end-of-life care in AD is assuring that families (or health care proxies) understand that AD is a fatal illness, with most patients dying within 4 to 8 years of diagnosis.¹ Evidence indicates that patients whose proxies have a clear recognition of this are less likely to experience “burdensome” interventions such as parenteral therapy, emergency department visits, hospital admissions, and tube feedings in their last 3 months of life.³⁷

Overall, decisions regarding discontinuing medical treatments in advanced AD should be made by balancing the likelihood of benefit with the potential for adverse effects.³⁸The progressive and ultimately fatal nature of Alzheimer's disease makes planning for the future a priority. For example, the American Geriatrics Society recently recommended against feeding tubes because they often result in discomfort due to agitation, use of restraints, and worsening pressure ulcers.³⁹

Unfortunately, only a minority of families receives straightforward information on the course and prognosis of AD, including the fact that patients eventually stop eating and that the natural cause of death is often an acute infection. Studies also show that patients with dementia are at risk for inadequate treatment of pain.⁴⁰ Assuring adequate pain control is an essential component of end-of-life care.

Hospice. End-of-life care can often be improved with hospice care. This service is underused by patients with dementia, even though hospice care is available at no cost through Medicare. Hospice eligibility criteria for patients with AD are shown in
TABLE 3.^41,42

Finally, a word about prevention

Numerous risk factors have been associated with an increased risk of AD (TABLE 4)^2,3. Some, like age and genetics, are nonmodifiable, while others—particularly cardiovascular risk factors—can be modified.¹ There are also factors associated with decreased risk, most notably, physical exercise and participating in cognitively stimulating activities.³ Identification of these factors has led to the hope that addressing them can prevent AD.

But association does not equal causation. In 2010, a report from the National Institutes of Health concluded that, although there are modifiable factors associated with AD, there is insufficient evidence that addressing any of them will actually prevent AD.⁴³ In fact, there is good evidence that some of these factors (eg, statin therapy) are not effective in reducing the incidence of dementia, and that others (eg, vitamin E and estrogen therapy) are potentially harmful.⁴⁴

The absence of empirically supported preventive interventions does not mean, however, that we should disregard these risks and protective factors. Encouraging social engagement, for example, may improve both emotional health and quality of life. Addressing cardiovascular risk factors can reduce the rate of coronary and cerebrovascular disease, potentially including vascular dementia, even if it does not reduce the rate of AD.

Studies are evaluating the use of monoclonal antibodies with anti-amyloid properties for preventing AD in individuals who have APOE ε4 genotypes or high amyloid loads on neuroimaging.⁴⁵ It will be several years before results are available, however, and the outcome of these studies is uncertain as the use of anti-amyloid agents for treating established dementia has not been effective.^46,47

CORRESPONDENCE
Marisa Menchola, PhD, Department of Psychiatry, University of Arizona College of Medicine, 1501 N. Campbell Ave., 7OPC. Tucson, AZ 85724; marisam@psychiatry.arizona.edu

Older adults are taking more medications than ever before. Nearly 9 out of 10 US residents who are 60 years of age or older take at least one prescription drug, more than a third take 5 to 9 medications, and 12% take 10 or more.¹

The increase is largely driven by newer medications to effectively treat a variety of medical conditions, and by practice guidelines that often recommend multidrug regimens.²

As a result, the term “polypharmacy,” which once referred to a specific number of medications, is now used more broadly to mean “a large number” of drugs.

From a safety standpoint, the number of medications a patient takes matters. The risk of adverse drug effects and dangerous drug-drug interactions increases significantly when an individual takes ≥5 medications.³

More than 4.5 million adverse drug effects occur each year in the United States, and nearly three quarters of them are initially evaluated in outpatient settings.⁴ Research suggests that about 80% of the time, these adverse effects are not recognized as such by the patient’s physician. So instead of discontinuing the offending medication, physicians treat the drug-related symptoms by adding yet another medication—a phenomenon known as “the prescribing cascade.”⁵

This review can help you safeguard older patients taking multiple medications by recognizing and responding to drug-related problems, identifying drugs that can be safely eliminated (or, in some cases, drugs that should be added), and checking regularly to ensure that the medication regimen is appropriate and up to date.

CASE Mrs. R, a 79-year-old woman who recently moved to town, is brought to your office by her daughter and son-in-law. The patient has a hard time reporting her medical history, but her daughter tells you her mother has chronic obstructive pulmonary disease (COPD), heart failure, type 2 diabetes, and mild urinary incontinence, and was recently diagnosed with early dementia.

Mrs. R’s daughter has brought in a bagful of medications, but she’s not sure which ones her mother takes regularly. The medications are an albuterol inhaler, alprazolam, digoxin, diphenhydramine, donepezil, furosemide, glargine insulin, guaifenesin, levothyroxine, metformin, extended-release metoprolol, naproxen, omeprazole, simvastatin, tolterodine, and zolpidem—a total of 16 different drugs.

If Mrs. R were your patient, how would you manage her multidrug regimen?

Start with a medication review

The first step in evaluating a patient’s medication regimen is to find out whether the drugs in the patient’s possession and/or in the medical record are the ones he or she is actually taking. Ask older patients who haven’t brought in their medications, or the caregiver of a confused patient, to bring them to the next visit.

The next step: Determine whether the medication regimen is right for the patient.

Polypharmacy may be indicated
Despite the risks associated with polypharmacy, do not assume that it is inappropriate. For some conditions, multiple medications are routinely recommended. Patients with heart failure, for example, have been shown to have better outcomes when they take 3 to 5 medications, including beta-blockers, angiotensin-converting enzyme (ACE) inhibitors, and diuretics.²

Some treatment guidelines also call for multiple medications. Achieving the more stringent blood pressure goals recommended in the Seventh Report of the Joint National Committee on Prevention, for instance, often requires 2 or more antihypertensive agents.⁶ In many cases, however, patients end up taking more drugs than necessary.

Is the patient taking the right drugs?
Medication reconciliation (determining whether the treatment regimen is appropriate for the patient’s diagnoses) is the way to find out.

The most widely recommended approach to medication reconciliation is to create a table and do a systematic review.⁷ List all the patient’s medical conditions in the first column and all current medications in the second column. Use the third column to note whether each medication is one the patient should be on, based not only on his or her medical conditions and other drugs being taken but also on current renal and hepatic function and body size, and contraindications.

A medication may be inappropriate if it duplicates, cancels out the action of, or otherwise interacts with another drug the patient is taking; is contraindicated in older patients; or is ineffective for the condition for which it was prescribed. In one key study of nearly 200 patients 65 years and older who took 5 or more medications, more than half had been prescribed at least one drug that was ineffective for the patient’s condition or that duplicated the action of another medication.⁸

In addition to finding drugs that the patient should not be taking, medication reconciliation may also reveal that the patient is not receiving optimal therapy and that one or more drugs should be added to his or her treatment regimen.

Check meds after transitions. A move from home to hospital, from emergency department to home, or any other transition relating to patient care should prompt a medication reconciliation. Medications are often added or inadvertently discontinued at such times,^9,10 and instructions relating to medication are often misunderstood.¹¹ In one study of 384 frail elderly patients being discharged from a hospital, for example, 44% were found to have been given at least one unnecessary prescription—most commonly for a medication that was neither indicated nor effective for any of the patient’s medical problems.¹² It was also common for patients to be given drugs that duplicated the action of others they were already taking.

Even in the absence of such transitions, medication reconciliation should occur at regular intervals. Many physicians do a medication reconciliation at every visit to ensure that the medical record is accurate and the patient’s medication regimen is optimal.

Managing polypharmacy: These resources can help

Numerous tools are available to help you evaluate and monitor patients’ medication regimens, including some that were developed specifically for older patients.

START (Screening Tool to Alert doctors to Right Treatment) identifies drugs and drug classes that are underused with older patients.¹³ START criteria (TABLE 1)^13-17 focus on medications that should be used yet are often omitted in older patients who have the appropriate indications.

TABLE 1
START criteria: Drug therapy that should be given to older patients^13-17

In using START or any other drug-related tool, it is important to keep in mind that therapy should be individualized. Not all the medications in the START criteria are appropriate for every patient, and a medication that is indicated for a given medical condition may or may not provide real benefit for a particular patient. That would depend on the individual’s overall health and life expectancy, the goals of treatment, and how long it would take for the patient to realize any benefit from the drug in question.¹⁸ A vigorous 79-year-old might benefit from statin therapy for prevention of cardiovascular events, for instance, while a patient like Mrs. R, who is also 79 but has dementia and multiple other medical problems, would be unlikely to live long enough to realize such a benefit.

”Age” assessment tool. One criterion in deciding whether medication(s) are appropriate for an older patient is his or her “physiologic age”—calculated on the basis of the individual’s chronological age and self-reported health status (TABLE 2).¹⁹

TABLE 2
Calculating your patient’s “real” age¹⁹

Flagging drugs that may be inappropriate
Several tools have been developed to aid clinicians in identifying medications that are potentially inappropriate for older adults, although here, too, decisions about their use must be individualized. Two of the most widely used tools are the Beers criteria and STOPP (Screening Tool of Older Persons’ potentially inappropriate Prescriptions).

Beers criteria were developed by Mark Beers et al in 1991²⁰ and have been updated at regular intervals, most recently by the American Geriatrics Society in 2012.²¹ The drugs and drug classes included in the Beers criteria should not be prescribed for older patients in most cases, either because the risk of using them outweighs the benefit or because safer alternatives are available. Key components are listed in TABLE 3.²¹

TABLE 3
Beers criteria:* Drug classes that may be inappropriate for older adults²¹

One limitation of the Beers criteria has been its all-or-nothing approach, with many of the medications on the list deemed inappropriate for all older adults regardless of their circumstances. The 2012 update does a better job of individualizing recommendations: Medications are now categorized as those that should be avoided in older patients regardless of their diseases or conditions, those that should be avoided only in patients with certain diseases or conditions, and those that may be used for this patient population but require caution.²¹

STOPP is similar to the Beers criteria, but uses a different approach: Most medications on this list are considered in the context of specific medical problems.²² While the Beers criteria classify digoxin >0.125 mg/d as generally inappropriate for older adults, for example, STOPP criteria state that long-term dosing at that level is inappropriate only for those with impaired renal function.²² A list of medications identified by STOPP as contributing to hospitalization due to adverse drug effects is available at http://ageing.oxfordjournals.org/content/37/6/673.

Both tools address this drug category. Cumulative anticholinergic burden is a concept applied to the use of anticholinergic medications, which are included in both the Beers and STOPP criteria. Although isolated short-term exposure to a drug with anticholinergic properties may be tolerated by a healthy and cognitively intact older patient, repetitive exposure to such drugs, even if separated in time, has negative effects. One study evaluated more than 500 community-dwelling older adults and found that the more exposure an individual had to anticholinergic medications over the course of a year, the greater the impairment in short-term memory and activities of daily living.²³ Another study, this one involving more than 13,000 community-dwelling and institutionalized patients, showed that the longer an older patient takes an anticholinergic medication, the more likely there is to be a measurable decline in performance on the Mini-Mental State Examination.²⁴

Programs that flag potential interactions
Drug-drug interactions are a key concern of polypharmacy, and electronic medical records and prescribing systems that flag potential drug-drug interactions when a new medication is ordered are designed to help physicians avoid them. Unfortunately, clinicians only react to 3% to 9% of such notifications, overriding them because computerized systems often fail to distinguish between important and unimportant interactions.^25-27 Thus, clinicians often must decide whether to react to or override warnings, an often difficult decision with patient safety and medicolegal implications. The best advice we can offer is to carefully evaluate drug interaction warnings using common sense, and seek consultation with a clinical pharmacist when uncertainty exists. This approach should prevent prescribing medications that have potentially harmful interactions with drugs the patient is already taking.

For physicians who do not have access to an electronic prescribing system that provides such notification, several online resources are available, some by subscription (eg, Lexicomp, www.lexi.com; Micromedex, www.micromedex.com/index.html; and Pepid, www.pepid.com) and others with free access (eg, AARP, healthtools.aarp.org/drug-interactions; Drugs.com (www.drugs.com/drug_interactions.php; and HealthLine, www.healthline.com/druginteractions).

CASE After doing a medication reconciliation for Mrs. R, you find that she is taking tolterodine, an anticholinergic medication for urge urinary incontinence, and donepezil, a procholinergic medication for dementia. This type of drug-drug interaction, in which the action of one drug effectively cancels out the effect of another, should not be ignored.

Overall, you identify 8 of her medications that could be discontinued: The list includes guaifenesin (a nonessential medication of questionable efficacy); naproxen (inappropriate per Beers criteria; inappropriate in patients with heart failure, according to STOPP); alprazolam, zolpidem, and diphenhydramine (duplicate medications that are all on the Beers criteria as inappropriate for chronic use and ill-advised in patients with cognitive impairment); and omeprazole and levothyroxine (for which nothing in the patient’s history suggests a need), as well as tolterodine. Depending on dose, digoxin is yet another candidate for discontinuation.

Discontinuing medications: Proceed carefully

Physicians are often reluctant to discontinue chronic medications in older patients—even in those with advanced disease who are not likely to benefit from treatment. Focus groups have identified a number of reasons for their hesitation, including:

• the assumption that patients have no problem taking large numbers of drugs
• the fear that patients may misinterpret a plan to discontinue medications as evidence that the physician is giving up on them
• the belief that physicians must comply with practice guidelines that recommend multiple drug treatments
• concern that proposing discontinuation of medications often leads to a discussion of life expectancy and end-of-life care.²⁸

Physicians may also fear that discontinuation of certain drugs will increase the risk of adverse outcomes. More than 30 studies have evaluated discontinuation of chronic medications in older adults, however, and found that drugs as diverse as antihypertensives, antipsychotics, benzodiazepines, and selective serotonin reuptake inhibitors (SSRIs) can often be discontinued without adverse outcomes. In many cases, improvement in patient function results.²⁹ Medications that present the most difficulty are those that patients often become physically or psychologically dependent on, such as benzodiazepines, guaifenesin, proton-pump inhibitors, nonsteroidal anti-inflammatory drugs, and SSRIs. Some (eg, benzodiazepines, SSRIs) require a gradual reduction; for others, no taper is required
(TABLE 4).^30-37

TABLE 4
Recommendations for discontinuing hard-to-stop drugs

CASE You trim down Mrs. R’s regimen by discontinuing each of the 8 drugs, one at a time, and carefully monitor the patient during the withdrawal period. Because she had been taking alprazolam daily, the dose is tapered slowly to avoid withdrawal. Omeprazole also requires a gradual taper to avoid rebound hyperacidity.³

After confirming that Mrs. R has heart failure and COPD, you identify 2 medications that should be added to her drug regimen—an ACE inhibitor for heart failure and an inhaled anticholinergic for COPD.

Going from 16 medications to 10 saves money, decreases the likelihood of adverse events and drug-drug interactions, and helps with adherence. Mrs. R’s new drug regimen is expected to lead to improvements in memory and overall quality of life, as well.

CORRESPONDENCE 
Barry D. Weiss, MD, Department of Family and Community Medicine, University of Arizona College of Medicine, Tucson, AZ 85724; bdweiss@email.arizona.edu

Older adults are taking more medications than ever before. Nearly 9 out of 10 US residents who are 60 years of age or older take at least one prescription drug, more than a third take 5 to 9 medications, and 12% take 10 or more.¹

The increase is largely driven by newer medications to effectively treat a variety of medical conditions, and by practice guidelines that often recommend multidrug regimens.²

As a result, the term “polypharmacy,” which once referred to a specific number of medications, is now used more broadly to mean “a large number” of drugs.

From a safety standpoint, the number of medications a patient takes matters. The risk of adverse drug effects and dangerous drug-drug interactions increases significantly when an individual takes ≥5 medications.³

More than 4.5 million adverse drug effects occur each year in the United States, and nearly three quarters of them are initially evaluated in outpatient settings.⁴ Research suggests that about 80% of the time, these adverse effects are not recognized as such by the patient’s physician. So instead of discontinuing the offending medication, physicians treat the drug-related symptoms by adding yet another medication—a phenomenon known as “the prescribing cascade.”⁵

This review can help you safeguard older patients taking multiple medications by recognizing and responding to drug-related problems, identifying drugs that can be safely eliminated (or, in some cases, drugs that should be added), and checking regularly to ensure that the medication regimen is appropriate and up to date.

CASE Mrs. R, a 79-year-old woman who recently moved to town, is brought to your office by her daughter and son-in-law. The patient has a hard time reporting her medical history, but her daughter tells you her mother has chronic obstructive pulmonary disease (COPD), heart failure, type 2 diabetes, and mild urinary incontinence, and was recently diagnosed with early dementia.

Mrs. R’s daughter has brought in a bagful of medications, but she’s not sure which ones her mother takes regularly. The medications are an albuterol inhaler, alprazolam, digoxin, diphenhydramine, donepezil, furosemide, glargine insulin, guaifenesin, levothyroxine, metformin, extended-release metoprolol, naproxen, omeprazole, simvastatin, tolterodine, and zolpidem—a total of 16 different drugs.

If Mrs. R were your patient, how would you manage her multidrug regimen?

Start with a medication review

The first step in evaluating a patient’s medication regimen is to find out whether the drugs in the patient’s possession and/or in the medical record are the ones he or she is actually taking. Ask older patients who haven’t brought in their medications, or the caregiver of a confused patient, to bring them to the next visit.

The next step: Determine whether the medication regimen is right for the patient.

Polypharmacy may be indicated
Despite the risks associated with polypharmacy, do not assume that it is inappropriate. For some conditions, multiple medications are routinely recommended. Patients with heart failure, for example, have been shown to have better outcomes when they take 3 to 5 medications, including beta-blockers, angiotensin-converting enzyme (ACE) inhibitors, and diuretics.²

Some treatment guidelines also call for multiple medications. Achieving the more stringent blood pressure goals recommended in the Seventh Report of the Joint National Committee on Prevention, for instance, often requires 2 or more antihypertensive agents.⁶ In many cases, however, patients end up taking more drugs than necessary.

Is the patient taking the right drugs?
Medication reconciliation (determining whether the treatment regimen is appropriate for the patient’s diagnoses) is the way to find out.

The most widely recommended approach to medication reconciliation is to create a table and do a systematic review.⁷ List all the patient’s medical conditions in the first column and all current medications in the second column. Use the third column to note whether each medication is one the patient should be on, based not only on his or her medical conditions and other drugs being taken but also on current renal and hepatic function and body size, and contraindications.

A medication may be inappropriate if it duplicates, cancels out the action of, or otherwise interacts with another drug the patient is taking; is contraindicated in older patients; or is ineffective for the condition for which it was prescribed. In one key study of nearly 200 patients 65 years and older who took 5 or more medications, more than half had been prescribed at least one drug that was ineffective for the patient’s condition or that duplicated the action of another medication.⁸

In addition to finding drugs that the patient should not be taking, medication reconciliation may also reveal that the patient is not receiving optimal therapy and that one or more drugs should be added to his or her treatment regimen.

Check meds after transitions. A move from home to hospital, from emergency department to home, or any other transition relating to patient care should prompt a medication reconciliation. Medications are often added or inadvertently discontinued at such times,^9,10 and instructions relating to medication are often misunderstood.¹¹ In one study of 384 frail elderly patients being discharged from a hospital, for example, 44% were found to have been given at least one unnecessary prescription—most commonly for a medication that was neither indicated nor effective for any of the patient’s medical problems.¹² It was also common for patients to be given drugs that duplicated the action of others they were already taking.

Even in the absence of such transitions, medication reconciliation should occur at regular intervals. Many physicians do a medication reconciliation at every visit to ensure that the medical record is accurate and the patient’s medication regimen is optimal.

Managing polypharmacy: These resources can help

Numerous tools are available to help you evaluate and monitor patients’ medication regimens, including some that were developed specifically for older patients.

START (Screening Tool to Alert doctors to Right Treatment) identifies drugs and drug classes that are underused with older patients.¹³ START criteria (TABLE 1)^13-17 focus on medications that should be used yet are often omitted in older patients who have the appropriate indications.

TABLE 1
START criteria: Drug therapy that should be given to older patients^13-17

In using START or any other drug-related tool, it is important to keep in mind that therapy should be individualized. Not all the medications in the START criteria are appropriate for every patient, and a medication that is indicated for a given medical condition may or may not provide real benefit for a particular patient. That would depend on the individual’s overall health and life expectancy, the goals of treatment, and how long it would take for the patient to realize any benefit from the drug in question.¹⁸ A vigorous 79-year-old might benefit from statin therapy for prevention of cardiovascular events, for instance, while a patient like Mrs. R, who is also 79 but has dementia and multiple other medical problems, would be unlikely to live long enough to realize such a benefit.

”Age” assessment tool. One criterion in deciding whether medication(s) are appropriate for an older patient is his or her “physiologic age”—calculated on the basis of the individual’s chronological age and self-reported health status (TABLE 2).¹⁹

TABLE 2
Calculating your patient’s “real” age¹⁹

Flagging drugs that may be inappropriate
Several tools have been developed to aid clinicians in identifying medications that are potentially inappropriate for older adults, although here, too, decisions about their use must be individualized. Two of the most widely used tools are the Beers criteria and STOPP (Screening Tool of Older Persons’ potentially inappropriate Prescriptions).

Beers criteria were developed by Mark Beers et al in 1991²⁰ and have been updated at regular intervals, most recently by the American Geriatrics Society in 2012.²¹ The drugs and drug classes included in the Beers criteria should not be prescribed for older patients in most cases, either because the risk of using them outweighs the benefit or because safer alternatives are available. Key components are listed in TABLE 3.²¹

TABLE 3
Beers criteria:* Drug classes that may be inappropriate for older adults²¹

One limitation of the Beers criteria has been its all-or-nothing approach, with many of the medications on the list deemed inappropriate for all older adults regardless of their circumstances. The 2012 update does a better job of individualizing recommendations: Medications are now categorized as those that should be avoided in older patients regardless of their diseases or conditions, those that should be avoided only in patients with certain diseases or conditions, and those that may be used for this patient population but require caution.²¹

STOPP is similar to the Beers criteria, but uses a different approach: Most medications on this list are considered in the context of specific medical problems.²² While the Beers criteria classify digoxin >0.125 mg/d as generally inappropriate for older adults, for example, STOPP criteria state that long-term dosing at that level is inappropriate only for those with impaired renal function.²² A list of medications identified by STOPP as contributing to hospitalization due to adverse drug effects is available at http://ageing.oxfordjournals.org/content/37/6/673.

Both tools address this drug category. Cumulative anticholinergic burden is a concept applied to the use of anticholinergic medications, which are included in both the Beers and STOPP criteria. Although isolated short-term exposure to a drug with anticholinergic properties may be tolerated by a healthy and cognitively intact older patient, repetitive exposure to such drugs, even if separated in time, has negative effects. One study evaluated more than 500 community-dwelling older adults and found that the more exposure an individual had to anticholinergic medications over the course of a year, the greater the impairment in short-term memory and activities of daily living.²³ Another study, this one involving more than 13,000 community-dwelling and institutionalized patients, showed that the longer an older patient takes an anticholinergic medication, the more likely there is to be a measurable decline in performance on the Mini-Mental State Examination.²⁴

Programs that flag potential interactions
Drug-drug interactions are a key concern of polypharmacy, and electronic medical records and prescribing systems that flag potential drug-drug interactions when a new medication is ordered are designed to help physicians avoid them. Unfortunately, clinicians only react to 3% to 9% of such notifications, overriding them because computerized systems often fail to distinguish between important and unimportant interactions.^25-27 Thus, clinicians often must decide whether to react to or override warnings, an often difficult decision with patient safety and medicolegal implications. The best advice we can offer is to carefully evaluate drug interaction warnings using common sense, and seek consultation with a clinical pharmacist when uncertainty exists. This approach should prevent prescribing medications that have potentially harmful interactions with drugs the patient is already taking.

For physicians who do not have access to an electronic prescribing system that provides such notification, several online resources are available, some by subscription (eg, Lexicomp, www.lexi.com; Micromedex, www.micromedex.com/index.html; and Pepid, www.pepid.com) and others with free access (eg, AARP, healthtools.aarp.org/drug-interactions; Drugs.com (www.drugs.com/drug_interactions.php; and HealthLine, www.healthline.com/druginteractions).

CASE After doing a medication reconciliation for Mrs. R, you find that she is taking tolterodine, an anticholinergic medication for urge urinary incontinence, and donepezil, a procholinergic medication for dementia. This type of drug-drug interaction, in which the action of one drug effectively cancels out the effect of another, should not be ignored.

Overall, you identify 8 of her medications that could be discontinued: The list includes guaifenesin (a nonessential medication of questionable efficacy); naproxen (inappropriate per Beers criteria; inappropriate in patients with heart failure, according to STOPP); alprazolam, zolpidem, and diphenhydramine (duplicate medications that are all on the Beers criteria as inappropriate for chronic use and ill-advised in patients with cognitive impairment); and omeprazole and levothyroxine (for which nothing in the patient’s history suggests a need), as well as tolterodine. Depending on dose, digoxin is yet another candidate for discontinuation.

Discontinuing medications: Proceed carefully

Physicians are often reluctant to discontinue chronic medications in older patients—even in those with advanced disease who are not likely to benefit from treatment. Focus groups have identified a number of reasons for their hesitation, including:

• the assumption that patients have no problem taking large numbers of drugs
• the fear that patients may misinterpret a plan to discontinue medications as evidence that the physician is giving up on them
• the belief that physicians must comply with practice guidelines that recommend multiple drug treatments
• concern that proposing discontinuation of medications often leads to a discussion of life expectancy and end-of-life care.²⁸

Physicians may also fear that discontinuation of certain drugs will increase the risk of adverse outcomes. More than 30 studies have evaluated discontinuation of chronic medications in older adults, however, and found that drugs as diverse as antihypertensives, antipsychotics, benzodiazepines, and selective serotonin reuptake inhibitors (SSRIs) can often be discontinued without adverse outcomes. In many cases, improvement in patient function results.²⁹ Medications that present the most difficulty are those that patients often become physically or psychologically dependent on, such as benzodiazepines, guaifenesin, proton-pump inhibitors, nonsteroidal anti-inflammatory drugs, and SSRIs. Some (eg, benzodiazepines, SSRIs) require a gradual reduction; for others, no taper is required
(TABLE 4).^30-37

TABLE 4
Recommendations for discontinuing hard-to-stop drugs

CASE You trim down Mrs. R’s regimen by discontinuing each of the 8 drugs, one at a time, and carefully monitor the patient during the withdrawal period. Because she had been taking alprazolam daily, the dose is tapered slowly to avoid withdrawal. Omeprazole also requires a gradual taper to avoid rebound hyperacidity.³

After confirming that Mrs. R has heart failure and COPD, you identify 2 medications that should be added to her drug regimen—an ACE inhibitor for heart failure and an inhaled anticholinergic for COPD.

Going from 16 medications to 10 saves money, decreases the likelihood of adverse events and drug-drug interactions, and helps with adherence. Mrs. R’s new drug regimen is expected to lead to improvements in memory and overall quality of life, as well.

CORRESPONDENCE 
Barry D. Weiss, MD, Department of Family and Community Medicine, University of Arizona College of Medicine, Tucson, AZ 85724; bdweiss@email.arizona.edu

Older adults are taking more medications than ever before. Nearly 9 out of 10 US residents who are 60 years of age or older take at least one prescription drug, more than a third take 5 to 9 medications, and 12% take 10 or more.¹

The increase is largely driven by newer medications to effectively treat a variety of medical conditions, and by practice guidelines that often recommend multidrug regimens.²

As a result, the term “polypharmacy,” which once referred to a specific number of medications, is now used more broadly to mean “a large number” of drugs.

From a safety standpoint, the number of medications a patient takes matters. The risk of adverse drug effects and dangerous drug-drug interactions increases significantly when an individual takes ≥5 medications.³

More than 4.5 million adverse drug effects occur each year in the United States, and nearly three quarters of them are initially evaluated in outpatient settings.⁴ Research suggests that about 80% of the time, these adverse effects are not recognized as such by the patient’s physician. So instead of discontinuing the offending medication, physicians treat the drug-related symptoms by adding yet another medication—a phenomenon known as “the prescribing cascade.”⁵

This review can help you safeguard older patients taking multiple medications by recognizing and responding to drug-related problems, identifying drugs that can be safely eliminated (or, in some cases, drugs that should be added), and checking regularly to ensure that the medication regimen is appropriate and up to date.

CASE Mrs. R, a 79-year-old woman who recently moved to town, is brought to your office by her daughter and son-in-law. The patient has a hard time reporting her medical history, but her daughter tells you her mother has chronic obstructive pulmonary disease (COPD), heart failure, type 2 diabetes, and mild urinary incontinence, and was recently diagnosed with early dementia.

Mrs. R’s daughter has brought in a bagful of medications, but she’s not sure which ones her mother takes regularly. The medications are an albuterol inhaler, alprazolam, digoxin, diphenhydramine, donepezil, furosemide, glargine insulin, guaifenesin, levothyroxine, metformin, extended-release metoprolol, naproxen, omeprazole, simvastatin, tolterodine, and zolpidem—a total of 16 different drugs.

If Mrs. R were your patient, how would you manage her multidrug regimen?

Start with a medication review

The first step in evaluating a patient’s medication regimen is to find out whether the drugs in the patient’s possession and/or in the medical record are the ones he or she is actually taking. Ask older patients who haven’t brought in their medications, or the caregiver of a confused patient, to bring them to the next visit.

The next step: Determine whether the medication regimen is right for the patient.

Polypharmacy may be indicated
Despite the risks associated with polypharmacy, do not assume that it is inappropriate. For some conditions, multiple medications are routinely recommended. Patients with heart failure, for example, have been shown to have better outcomes when they take 3 to 5 medications, including beta-blockers, angiotensin-converting enzyme (ACE) inhibitors, and diuretics.²

Some treatment guidelines also call for multiple medications. Achieving the more stringent blood pressure goals recommended in the Seventh Report of the Joint National Committee on Prevention, for instance, often requires 2 or more antihypertensive agents.⁶ In many cases, however, patients end up taking more drugs than necessary.

Is the patient taking the right drugs?
Medication reconciliation (determining whether the treatment regimen is appropriate for the patient’s diagnoses) is the way to find out.

The most widely recommended approach to medication reconciliation is to create a table and do a systematic review.⁷ List all the patient’s medical conditions in the first column and all current medications in the second column. Use the third column to note whether each medication is one the patient should be on, based not only on his or her medical conditions and other drugs being taken but also on current renal and hepatic function and body size, and contraindications.

A medication may be inappropriate if it duplicates, cancels out the action of, or otherwise interacts with another drug the patient is taking; is contraindicated in older patients; or is ineffective for the condition for which it was prescribed. In one key study of nearly 200 patients 65 years and older who took 5 or more medications, more than half had been prescribed at least one drug that was ineffective for the patient’s condition or that duplicated the action of another medication.⁸

In addition to finding drugs that the patient should not be taking, medication reconciliation may also reveal that the patient is not receiving optimal therapy and that one or more drugs should be added to his or her treatment regimen.

Check meds after transitions. A move from home to hospital, from emergency department to home, or any other transition relating to patient care should prompt a medication reconciliation. Medications are often added or inadvertently discontinued at such times,^9,10 and instructions relating to medication are often misunderstood.¹¹ In one study of 384 frail elderly patients being discharged from a hospital, for example, 44% were found to have been given at least one unnecessary prescription—most commonly for a medication that was neither indicated nor effective for any of the patient’s medical problems.¹² It was also common for patients to be given drugs that duplicated the action of others they were already taking.

Even in the absence of such transitions, medication reconciliation should occur at regular intervals. Many physicians do a medication reconciliation at every visit to ensure that the medical record is accurate and the patient’s medication regimen is optimal.

Managing polypharmacy: These resources can help

Numerous tools are available to help you evaluate and monitor patients’ medication regimens, including some that were developed specifically for older patients.

START (Screening Tool to Alert doctors to Right Treatment) identifies drugs and drug classes that are underused with older patients.¹³ START criteria (TABLE 1)^13-17 focus on medications that should be used yet are often omitted in older patients who have the appropriate indications.

TABLE 1
START criteria: Drug therapy that should be given to older patients^13-17

In using START or any other drug-related tool, it is important to keep in mind that therapy should be individualized. Not all the medications in the START criteria are appropriate for every patient, and a medication that is indicated for a given medical condition may or may not provide real benefit for a particular patient. That would depend on the individual’s overall health and life expectancy, the goals of treatment, and how long it would take for the patient to realize any benefit from the drug in question.¹⁸ A vigorous 79-year-old might benefit from statin therapy for prevention of cardiovascular events, for instance, while a patient like Mrs. R, who is also 79 but has dementia and multiple other medical problems, would be unlikely to live long enough to realize such a benefit.

”Age” assessment tool. One criterion in deciding whether medication(s) are appropriate for an older patient is his or her “physiologic age”—calculated on the basis of the individual’s chronological age and self-reported health status (TABLE 2).¹⁹

TABLE 2
Calculating your patient’s “real” age¹⁹

Flagging drugs that may be inappropriate
Several tools have been developed to aid clinicians in identifying medications that are potentially inappropriate for older adults, although here, too, decisions about their use must be individualized. Two of the most widely used tools are the Beers criteria and STOPP (Screening Tool of Older Persons’ potentially inappropriate Prescriptions).

Beers criteria were developed by Mark Beers et al in 1991²⁰ and have been updated at regular intervals, most recently by the American Geriatrics Society in 2012.²¹ The drugs and drug classes included in the Beers criteria should not be prescribed for older patients in most cases, either because the risk of using them outweighs the benefit or because safer alternatives are available. Key components are listed in TABLE 3.²¹

TABLE 3
Beers criteria:* Drug classes that may be inappropriate for older adults²¹

One limitation of the Beers criteria has been its all-or-nothing approach, with many of the medications on the list deemed inappropriate for all older adults regardless of their circumstances. The 2012 update does a better job of individualizing recommendations: Medications are now categorized as those that should be avoided in older patients regardless of their diseases or conditions, those that should be avoided only in patients with certain diseases or conditions, and those that may be used for this patient population but require caution.²¹

STOPP is similar to the Beers criteria, but uses a different approach: Most medications on this list are considered in the context of specific medical problems.²² While the Beers criteria classify digoxin >0.125 mg/d as generally inappropriate for older adults, for example, STOPP criteria state that long-term dosing at that level is inappropriate only for those with impaired renal function.²² A list of medications identified by STOPP as contributing to hospitalization due to adverse drug effects is available at http://ageing.oxfordjournals.org/content/37/6/673.

Both tools address this drug category. Cumulative anticholinergic burden is a concept applied to the use of anticholinergic medications, which are included in both the Beers and STOPP criteria. Although isolated short-term exposure to a drug with anticholinergic properties may be tolerated by a healthy and cognitively intact older patient, repetitive exposure to such drugs, even if separated in time, has negative effects. One study evaluated more than 500 community-dwelling older adults and found that the more exposure an individual had to anticholinergic medications over the course of a year, the greater the impairment in short-term memory and activities of daily living.²³ Another study, this one involving more than 13,000 community-dwelling and institutionalized patients, showed that the longer an older patient takes an anticholinergic medication, the more likely there is to be a measurable decline in performance on the Mini-Mental State Examination.²⁴

Programs that flag potential interactions
Drug-drug interactions are a key concern of polypharmacy, and electronic medical records and prescribing systems that flag potential drug-drug interactions when a new medication is ordered are designed to help physicians avoid them. Unfortunately, clinicians only react to 3% to 9% of such notifications, overriding them because computerized systems often fail to distinguish between important and unimportant interactions.^25-27 Thus, clinicians often must decide whether to react to or override warnings, an often difficult decision with patient safety and medicolegal implications. The best advice we can offer is to carefully evaluate drug interaction warnings using common sense, and seek consultation with a clinical pharmacist when uncertainty exists. This approach should prevent prescribing medications that have potentially harmful interactions with drugs the patient is already taking.

For physicians who do not have access to an electronic prescribing system that provides such notification, several online resources are available, some by subscription (eg, Lexicomp, www.lexi.com; Micromedex, www.micromedex.com/index.html; and Pepid, www.pepid.com) and others with free access (eg, AARP, healthtools.aarp.org/drug-interactions; Drugs.com (www.drugs.com/drug_interactions.php; and HealthLine, www.healthline.com/druginteractions).

CASE After doing a medication reconciliation for Mrs. R, you find that she is taking tolterodine, an anticholinergic medication for urge urinary incontinence, and donepezil, a procholinergic medication for dementia. This type of drug-drug interaction, in which the action of one drug effectively cancels out the effect of another, should not be ignored.

Overall, you identify 8 of her medications that could be discontinued: The list includes guaifenesin (a nonessential medication of questionable efficacy); naproxen (inappropriate per Beers criteria; inappropriate in patients with heart failure, according to STOPP); alprazolam, zolpidem, and diphenhydramine (duplicate medications that are all on the Beers criteria as inappropriate for chronic use and ill-advised in patients with cognitive impairment); and omeprazole and levothyroxine (for which nothing in the patient’s history suggests a need), as well as tolterodine. Depending on dose, digoxin is yet another candidate for discontinuation.

Discontinuing medications: Proceed carefully

Physicians are often reluctant to discontinue chronic medications in older patients—even in those with advanced disease who are not likely to benefit from treatment. Focus groups have identified a number of reasons for their hesitation, including:

• the assumption that patients have no problem taking large numbers of drugs
• the fear that patients may misinterpret a plan to discontinue medications as evidence that the physician is giving up on them
• the belief that physicians must comply with practice guidelines that recommend multiple drug treatments
• concern that proposing discontinuation of medications often leads to a discussion of life expectancy and end-of-life care.²⁸

Physicians may also fear that discontinuation of certain drugs will increase the risk of adverse outcomes. More than 30 studies have evaluated discontinuation of chronic medications in older adults, however, and found that drugs as diverse as antihypertensives, antipsychotics, benzodiazepines, and selective serotonin reuptake inhibitors (SSRIs) can often be discontinued without adverse outcomes. In many cases, improvement in patient function results.²⁹ Medications that present the most difficulty are those that patients often become physically or psychologically dependent on, such as benzodiazepines, guaifenesin, proton-pump inhibitors, nonsteroidal anti-inflammatory drugs, and SSRIs. Some (eg, benzodiazepines, SSRIs) require a gradual reduction; for others, no taper is required
(TABLE 4).^30-37

TABLE 4
Recommendations for discontinuing hard-to-stop drugs

CASE You trim down Mrs. R’s regimen by discontinuing each of the 8 drugs, one at a time, and carefully monitor the patient during the withdrawal period. Because she had been taking alprazolam daily, the dose is tapered slowly to avoid withdrawal. Omeprazole also requires a gradual taper to avoid rebound hyperacidity.³

After confirming that Mrs. R has heart failure and COPD, you identify 2 medications that should be added to her drug regimen—an ACE inhibitor for heart failure and an inhaled anticholinergic for COPD.

Going from 16 medications to 10 saves money, decreases the likelihood of adverse events and drug-drug interactions, and helps with adherence. Mrs. R’s new drug regimen is expected to lead to improvements in memory and overall quality of life, as well.

CORRESPONDENCE 
Barry D. Weiss, MD, Department of Family and Community Medicine, University of Arizona College of Medicine, Tucson, AZ 85724; bdweiss@email.arizona.edu