5/21/19 Geriatrics Literature Updates Disclosures Kenneth - - PDF document

5/21/19 1

Geriatrics Literature Updates

Kenneth Covinsky, MD @Geri_Doc Eric Widera, MD @EWidera

University of California San Francisco San Francisco VA Medical Center

Disclosures

• Eric Widera
• Associate Editor, Social Media Editor, for the Journal of the American

Geriatrics Society (JAGS)

• Ken Covinsky
• Editorial Board for the Journal of the American Geriatrics Society (JAGS);

Associate Editor, JAMA Internal Medicine

• Alex Smith
• Deputy Editor, Journal of the American Geriatrics Society (JAGS)

5/21/19 2

Methods for AGS Updates Talk

• Search of leading journals
• January 2018-December 2018
• JAGS, NEJM, JAMA, JAMA-IM, Annals, Health Affairs, Lancet, BMJ, Academic

Medicine, JGIM, J Geron-Med Sci, JPM, JPSM, Annals of Family Medicine

• Search of social media:
• Twitter (i.e. @AGSJournal), Blogs, PC-FACS, podcasts, Health In Aging

Research Summaries (healthinaging.org)

• Selection Criteria
• Impact and Interest

5/21/19 3

BMJ 2018. http://dx.doi.org/10.1136/bmj.k3503

Methods

• Retrospective cohort study
• 14,915 Veterans aged 65 years or over (median age 76) with HTN

admitted to hospital with non-cardiac conditions:

• Pneumonia, urinary tract infection, or venous thromboembolism.
• Exclusions:
• Secondary discharge diagnoses that might necessitate an acute change in

antihypertensive treatment (eg. A fib, ACS)

• Primary Outcome:
• A new or higher dose antihypertensive agent at discharge compared with

drugs used before admission.

Percentage with elevated inpatient BP

Normoten sive inpatient BP Elevated Inpatient BP 23%

5/21/19 4

Percentage with elevated inpatient BP

Elevated Pre- hospital BP Well Controlled Pre-hospital BP 47% Normoten sive inpatient BP Elevated Inpatient BP 23%

Measuring Matters!!!!

• How SPRINT did it
• Average of 3 office BP readings taken with proper cuff size
• Participants seated with their back supported
• 5 minutes of rest before measurement
• No conversation during the rest period or BP determinations.
• Conventional Auscultatory SBP
• Up to 20mmHg higher than this technique

Myers MG, et tal. Hypertension. 2010;55:195–200

Percentage of intensified antihypertensive treatment post-hospitalization

Intensified HTN treatmet 14% No change In te n sifie d HT N tre atm e t

No Intensification

Percentage of intensified antihypertensive treatment post-hospitalization

Intensified HTN treatmet 14% No change In te n sifie d HT N tre atm e t

62% started on one new antihypertensive 14% started on multiple new antihypertensives

5/21/19 5

Percentage of intensified antihypertensive treatment post-hospitalization

Intensified HTN treatmet 14% No change In te n sifie d HT N tre atm e t

No Intensification

Well Controlled Pre- hospital BP 52% E le vate d Outp a ti e n t BP We ll contro ll e d Outp t BP

Care driven by numbers and not the clinical context

Take Home Points

• 1 in 7 older adults admitted to hospital for common non-cardiac

conditions were discharged with intensified antihypertensive treatment.

• More than half of intensifications occurred in patients with previously

well controlled outpatient blood pressure.

• Decisions to intensify antihypertensive in the hospital seem to be

driven by numbers and not the clinical context

5/21/19 6

Clinicians would be wise to adopt Sin City’s famous tagline, “What happens in Vegas, stays in Vegas;” often the safest approach to inpatient chronic disease management should be to let what happens in hospital stay in hospital. Nathan Stall & Chaim Bell

JAGS 67:11–16, 2019

Warning: Hospitals Are Bad For Your Health

• Hospital-Acquired Disability
• One-Third of persons over age 70 leave hospital with new ADL disability
• Less than half will recover after discharge
• May be related to processes of hospitalization
• Immobilization
• Delirium induction
• Malnutrition
• Evidence that changing hospital processes reduced risk of hospital

acquired disability

5/21/19 7

JAMA IM. 2019 doi:10.1001/jamainternmed.2018.4869

Multi Component Hospital Exercise Intervention

• Setting: ACE Unit: Pamplona Spain
• Enrollment Criteria (n=370)
• Medical admissions over age 75
• Barthel Index at least 60
• Able to ambulate (assistance OK)
• Able to communicate and collaborate with research team
• LOS at least six days
• Enrolled subjects
• Mean age = 87
• 57% women
• Almost all with multimorbidity
• CHF, Infection most common reasons for admission

The Intervention

• Two daily 20 minute sessions
• Am session with fitness specialist
• Progressive resistance: 8-10 reps X 2, 30-60% capacity, 4 muscle groups
• Balance and gait training: Semi tandem foot standing, line walking,

walking around obstacles, walking on unstable surfaces

• Unsupervised PM session
• Lower extremity exercises with anklets, corridor walking
• Costs
• Fitness specialist
• $5000 worth of exercise equipment

Effect of Intervention

Control Intervention Difference Barthel Index Change

• 5.0

+1.9 +6.9 (+4.4, +9.5) SPPB change +0.2 +2.4 +2.2 (+1.7, +2.6) MMSE Change +0.3 +2.1 +1.8 (+1.3, 2.3)

• No difference in falls
• No difference in 3 month mortality
• No difference in 3 month readmission
• No difference in 3 month institutionalization

5/21/19 8

Summary

• An intensive hospital based exercise intervention combining

strength training, balance, and walking had substantial positive effect on post discharge function in very old medicine inpatients

• Caveats
• Careful selection process (excludes most physically and cognitively

challenged)

• Requires extra staffing
• Much longer length of stay than typical US hospital patient
• Probably requires intervention continuation in SNF/home
• Very important proof of concept that hospital acquired

disability can be prevented in some patients

5/21/19 9

Pimavanserin (Nuplazid)

Cost: $3,000/month

Lancet Neurology 2018; 17: 213–22

Background

• Pimavanserin serotonin 5HT2A inverse agonist and antagonist with no

dopamine D2 affinity

• Available in the US since 2016 for treatment of psychosis in patients

with Parkinson Disease related psychosis

• Largely based on one trial with marginal benefit (a gain of 3 points on a 45-

point scale)

• Does it work for Alzheimer’s disease psychosis?

Methods

• Study Design: Phase 2, randomized, double-blind, placebo-controlled
• Nursing home residents age >=50 with possible or probable AD
• Psychotic symptoms including hallucinations or delusions severe enough to

warrant treatment.

• Randomly assigned to pimavanserin (n=90) or placebo (n=91) for 12

weeks

• Primary outcome:
• Mean change in the NPI–NH psychosis score (hallucinations + delusions) at 6

weeks

• NPI-NH ranges from 0 to 24 with lower scores correspond to less severity

Lancet Neurology 2018; 17: 213–22

5/21/19 10

Adjusted mean change in the psychosis score

*Range 0 to

• 24. Lower

scores correspond to less severity

Adjusted mean change in the psychosis score

*Range 0 to

• 24. Lower

scores correspond to less severity

Primary Outcome Measures in ClinicalTrials.org

2014

• NPI-NH, CMAI-Short form, ADCS-Clinical Global Impression of Change at 12 weeks

2014

• NPI-NH at 12 weeks

2016

• NPI-NH at 12 weeks

2016

• NPI-NH at 12 weeks

2017

• NPI-NH Psychosis domains at 6 weeks (dosed total of 12 weeks)

2017

• “Change from Baseline to Day 43 in the NPI-NH psychosis score”

Subgroup analysis for psychosis score based

• n severity

Baseline NPI-NH < 12 NPI-NH >=12

5/21/19 11

Subgroup analysis for psychosis score based

• n severity

Baseline NPI-NH < 12 NPI-NH >=12

Adjusted mean change in psychosis score for patients with baseline score ≥12 Adverse Events

Pimavanserin (n=90) Placebo (n=91) Any serious adverse event 17% 11% Agitation 21% 14% Aggression 10% 4% Peripheral edema 8% 2% Anxiety 6% 2% Behavioral and psychiatric symptoms of dementia 6% 2%

Pimavanserin (Nuplazid)

Cost: $3,000/month

5/21/19 12

Cummings et al. JAMA. 2015;314(12):1242-1254

Ken’s Conclusions from 2016

• Conclusion A:
• Dextromethorphan efficacious for reducing agitation with tolerable side

effects

• Conclusion B:
• Dextromethorphan efficacious in increasing falls with some additional mild

benefits of reduced agitation

• Should you consider Dextromethorphan for off-label use for patients

with dementia agitation?

• Evidence of effectiveness is very modest, and risk for harm substantial

JAMA Intern Med. 2019;179(2):224-230.

5/21/19 13

JAMA Intern Med. 2019;179(2):224-230. doi:10.1001/jamainternmed.2018.6112 JAMA Intern Med. 2019;179(2):224-230. doi:10.1001/jamainternmed.2018.6112

Maust et al. JAMA Intern Med. 2018

The National Partnership to Improve Dementia Care in Nursing Homes

• CMS program to reduce antipsychotic use in persons with dementia
• Limited effectiveness reducing behavioral symptoms
• Bad side effects
• Partnership between CMS, state agencies, surveyors, advocacy groups
• Public reporting of antipsychotic use
• No reporting of other mood stabilizing meds
• Anticonvulsants such as carbamazepine, valproic acid
• Less evidence of efficacy, side effects at least as bad
• No measures of nonpharmacologic treatment

5/21/19 14

Evaluation of the National Partnership

• Identify nationwide cohort of long stay nursing home residents
• Mean age 80, 70% women, 83% white, 10% black, 5% Latino,
• Use Part D data to compare psychotropic drug use before and after

national partnership

• Overall use, antipsychotics, mood stabilizers
• Hypotheses
• No good deed goes unpunished
• Law of unintended consequences
• You get what you measure (and don’t get what you don’t measure)

Adopted from M aust DT et al. JAM A Intern M ed. 2018;178(5):640–647. 10 20 30 40 50 60 70 80 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4

Patients, % Time, y (Quarters)

Percentage of Long-Stay Nursing Home Residents With Dementia Prescribed an Antipsychotic

• r Other Psychotropic Medication

Antipsychotics Mood Stabilizer

2009 2010 2011 2012 2013 2014

Prepartnership period Postpartnership period

Psychotropics Overall (non-BZD)

Bottom Line

• The partnership to improve dementia care was NOT successful
• Reduction in antipsychotic use
• Primarily continuation of trends already occurring
• Counterbalanced by increase in alternative drugs that are just as bad, maybe worse
• We got what we measured
• When dealing with hard problem, nursing homes shifted to treatment not

measured

• Hope was that they would shift nonpharmacological behavioral treatment
• Doubtful they did

Broad Implications For Blunt Brute Force CMS Quality Initiatives

• Improving care for frail older persons very important, but it is

complex

• Consider the larger social and cultural context of care
• Focus as much on making the right things happen as avoiding wrong things
• Measure what matters
• Message to CMS
• Thank you for caring about vulnerable persons with dementia
• But please recognize impact of blunt QI initiatives focused on

shaming/punishment will not get us far

• Your data are not very good, yet that is the basis for most of what you do

5/21/19 15

• Maharani. JAGS 66:1130–1136, 2018.

Background

• 3 out of 5 adults 70 years and older have hearing loss
• Hearing loss is also associated with cognitive decline
• ? due to the AD itself vs sensory deprivation, social

isolation, or cognitive overload

• Can hearing aids alter the negative cognitive
• utcomes of hearing loss?
• Frank et al. The Journals of Gerontology: Series A.

2011

• Davies et al. JAGS. 2017

5/21/19 16

Methods

• Health and Retirement Study Data
• 2,040 adults aged 50 and older
• No dementia or hearing aids a baseline
• Self-Reported use of hearing aids during the study
• Measured cognitive performance repeatedly every 2

years over 18 years (1996–2014).

• Episodic memory scores: sum of immediate and delayed

recall of 10 words (range 0-20)

* Slope adjusted for demographic & socioeconomic characteristics, lifestyle behavior, and health status. * Slope adjusted for demographic & socioeconomic characteristics, lifestyle behavior, and health status. * Slope adjusted for demographic & socioeconomic characteristics, lifestyle behavior, and health status.

5/21/19 17

The Big Limitation

Episodic memory tests were presented orally

Tw Two Big Things Coming

Over-the-Counter Hearing Aid Act Aging and Cognitive Health Evaluation in Elders (ACHIEVE) Trial

JAMA Intern Med. 2018;178(10):1301-1310. doi:10.1001/jamainternmed.2018.3915

Tai Chi Intervention For Fall Reduction

• Tai Chi
• Chinese martial art
• Exercises based on slow movement with breathing and meditative components
• Tai Chi for Fall Reduction
• Stabilizing and destabilizing postural actions
• Integration of musculoskeletal, sensory, cognitive systems
• Weight bearing and weight shifting, trunk/pelvic rotation, ankle sways, eye-head-

hand movements

• Implementation
• Taught by trained instructors in 15 persons classes; 2x/wk for 24 weeks

($1000/person)

5/21/19 18

Multimodal Exercise and Stretching

• Multimodal exercise
• Walking: Long strides, heel to toe, side stepping
• Strength training: ankle dorsiflexion, knee extensors, hip abductors
• Balance training: tandem foot standing, line walking, single leg

standing, weight transfers

• Flexibility: Stretching exercises
• Stretching group
• Seated breathing, stretching, and relaxation exercises
• For both groups, 15 person classes, 2x/wk for 24 weeks

Subjects

• Community-dwelling older persons (age > 70) across Oregon
• Fall risk (History of fall or abnormal timed get up and go)
• Able to walk one block
• MMSE > 20
• Randomized subjects (n=470)
• Mean age =78
• 91% white, 6% African-American
• 56% college educated
• 72% fell in past 6 months
• 53% > 3 chronic conditions

27 16 11

5 10 15 20 25 30 Stretching Mutlimodal E xercise Tai Chi Falls/100 Person Months

Results

P <.001 P <.04 P <.001

Summary

• Multimodal Exercise Intervention and Tai Chi are both very effective

at reducing falls

• Tai Chi somewhat more effective than Multimodal exercise
• We have two good choices for reducing falls
• Patients at high risk of falls should be offered a choice of Tai Chi or

Multimodal Exercise

5/21/19 19

Lomas-Vega et al. JAGS 2017

Reid et al. NEJM. 2018

Should We Treat Osteopenia

• Treatment generally indicated for most women with
• steoporosis
• BMD T score < -2.5 or osteoporotic fracture
• But, most fractures occur in women with osteopenia
• T score between -1.0 and -2.5
• Potential large reservoir of benefit
• But lots of treatment for an asymptomatic condition

Study Design

• Double blind RCT in 2000 women over 6 years
• Zoledronate 5 mg IV every 18 months
• Placebo saline infusion
• Women > 65 years (mean=71 years)
• At least one hip with T score of -1.0 to -2.5
• Not currently on osteoporosis rx
• Primary Outcome: Time to first fragility fracture
• Reports of symptomatic fractures and spine radiographs
• Secondary Outcome: symptomatic fractures, mortality

5/21/19 20

Effects Of Zoledronate Over 6 years

• Fragility Fracture (Zoledronate vs Placebo)
• 12% vs 19% (HR=0.63, 95% CI =0.50-0.79)
• NNT to prevent one woman with fracture =15
• Effect on symptomatic fractures
• 27% risk reduction (P<.001), NNT =20
• Zoledronate side effects
• Acute phase reaction 6%
• Iritis 0.6%
• Mortality (Zoledronate vs Placebo)
• 4.5% vs 7.0%, HR 0.65, 95% CI 0.40-1.05

Less is More Except When More is More

• Treating Osteopenia in Women with Zoledronate over Age 65

without Osteoporosis for six years leads to

• 37% reduced incidence of fracture
• 27% reduced risk of symptomatic fracture
• Women with osteopenia should be given option of

treatment

• Possible exception: Limited life expectancy

Message

• “You will need to come to clinic every 18 months for infusion”
• “For every 20 women treated, 19 will not see a benefit, and
• ne will avoid a fracture. The fracture is likely to be very

bothersome, possibly disabling”

• “You have a one in 17 chance of a flu like response that will

probably be very bothersome for several days from which you will fully recover”

Statin Use in Older Adults

Johansen ME, Green LA . JAMA Intern Med.2015;175(10):1715–1716.

5/21/19 21

5/21/19 22

Methods of the Metanalysis

• Individual participant data of 28 randomized clinical trials
• Both statin vs control & more intensive versus less intensive statin
• Both primary and secondary prevention trails
• 186,854 total participants
• 15% age 71-75
• 8% age >75
• Median follow-up duration was 5 years
• Results were weighted by the absolute LDL cholesterol difference the

trials at 1 year

Benefits of statin therapy or a more intensive statin regimen

• 21% proportional reduction in major vascular events per 1

mmol/L reduction in all age groups

• Diminished slightly with increasing age but this trend was not

statistically significant (p trend=0.06)

• Secondary vs Primary Prevention for Major Vascular events
• Secondary prevention: similar reductions across age group (p

trend=0.2)

• Primary prevention: significant trend towards smaller risk

reductions with increasing age (p trend=0.05)

Figure 4

Effects on major vascular events: Primary Prevention

5/21/19 23

Summary of subgroup analyses looking at

• lder age and primary prevention

Initial Study Inclusion (Primary vs Secondary Prevention) Age for Subgroup analysis Benefit of Primary Prevention PROSPER Primary & Secondary 70-82 None ALLHAT-LLT Primary > 65 or >75 None JUPITER Primary >70 Decrease in CV outcomes

Initiating a moderate-intensity statin may be reasonable Older Patients and Statins > 75 years of Age

Limited life expectancy -> History of CVD

• >

Primary prevention

• >

STAREE

Atorvastatin (40 mg) vs placebo in healthy people older than 70 year

5/21/19 24

Aspirin in Reducing Events in the Elderly (ASPREE) trial…

Methods

• Randomized placebo-controlled trial
• 19,000 community-dwelling adults in Australia and US without CV disease
• ≥ 70 years of age (≥ 65 years for black and Hispanic participants in US)
• Expected survival of at least 5 years
• No dementia, physical disability, or high risk for bleeding
• Randomized to:
• Enteric-coated aspirin, 100 mg/d
• Placebo
• Primary Outcome (Geriatrics relevant!!)
• composite of death, dementia, or persistent ADL disability
• Follow-up period: Median 4.7 years

Who were these patients?

• Median age 74 years
• 56% women
• 9% non-white
• 11% with diabetes

Results

Event rates/1000 person-y At 4.7 years Outcomes Aspirin Placebo Hazard ratio (95% CI) Disability-free survival

(death, dementia, disability)

22 21 1.01 (0.92 to 1.11) All-cause mortality 13 11 1.14 (1.01 to 1.29) Cancer Mortality 3 2 1.31 (1.10 to 1.56) Cardiovascular disease

(including CV death, MI, stroke, HF admit)

11 11 0.95 (0.83 to 1.08) Major hemorrhage 9 6 1.38 (1.18 to 1.62)

5/21/19 25

Major Large Primary Prevention Studies

• ASPREE: In healthy, community-dwelling elderly people aged 70 and
• lder, aspirin does not prevent cardiovascular disease and does

increase one’s risk of major hemorrhage

• ARRIVE: among younger individuals with moderate risk of coronary

heart disease, the use of aspirin was not beneficial.

• ASCEND: among diabetic patients, aspirin reduced the incidence of

major adverse cardiovascular events but was counterbalanced by an increase in major bleeding.

Bottom Line

• Definite
• Aspirin should not be started for primary prevention in persons over age 70
• Unclear
• Should aspirin be continued for primary prevention in persons over age 70

who have been using for a long time?

• Should aspirin be used for secondary prevention in persons over age 70?

Phillips et al. Academic Medicine. 2019

$0 $2 0,000 $4 0,000 $6 0,000 $8 0,000 $1 00,00 0 $1 20,00 0 $1 40,00 0 $1 60,00 0 $1 80,00 0 $2 00,00 0

1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Median Education Debt

Graduating Debt for Medical Students by Year

$200K in 2018 $50K in 1992

5/21/19 26

$0 $2 0,000 $4 0,000 $6 0,000 $8 0,000 $1 00,00 0 $1 20,00 0 $1 40,00 0 $1 60,00 0 $1 80,00 0 $2 00,00 0

1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Median Education Debt

Graduating Debt for Medical Students by Year

$200K in 2018 $50K in 1992

Does educational debt affect physician’s career choice?

Methods

• Cross-sectional secondary analysis of 6,229 family medicine residents

taking the 2014 and 2015 ABFM examination

• registration questionnaire data and ABFM administrative data
• Excluded residents who were undecided
• 40% of residents had not yet chosen their practice path
• 22% had not decided on fellowship plans

Percentage with High Educational Debt

<150K in Debt 40%

>150K in Debt 60%