Transcript here - Health Services and Systems Research

Mind the Gap: Supporting
Successful Care Transitions and
Recovery after a Stroke
Janet Prvu Bettger, ScD, FAHA – [email protected]
Associate Professor of Nursing and Senior Fellow in Aging
Faculty Affiliate, Duke Global Health and Clinical Research Institutes
June 24, 2014
Presentation Outline
• Burden of stroke
• Systems perspective of stroke care
• Evidence gaps
• Care models for improved recovery from stroke
All Rights Reserved, Duke Medicine 2007
Projected Deaths by Cause for
High-, Middle- and Low-Income Countries
Other NCDs
Cancers
CVD
All Rights Reserved, Duke Medicine 2007
Burden of Stroke: DALY
Leading cause of serious, longterm disability in the US
All Rights Reserved, Duke Medicine 2007
Johnston et al. Lancet Neurol 2009;8:345-54
Burden of Stroke in the United States (US)
• Incidence: 795,000 new or recurrent stroke each year
• Every 40 seconds someone in the US has a stroke
• Every 4 minutes, someone dies of a stroke
• 3 of 4 stroke survivors are dependent at some level for selfcare
• Over 60% of stroke patients have cognitive impairment
• About 15%-30% are permanently disabled
• Stroke survivors requiring constant care 3 months following
their stroke have a 7-fold increased 1-year mortality risk
All Rights Reserved, Duke Medicine 2007
AHA Heart Disease and Stroke Statistics 2014 Update/ CDC National Vital Statistics Reports 2010
Challenges Specific to Stroke Care
• Average length of acute hospital stay = 4 days
• Episode of care for stroke = 82-109 days
• Almost 80% of stroke patients experience more than
two transitions of care after hospital discharge
• 1 in 3 are rehospitalized within 3 months
• 1 in 3 are institutionalized in a nursing home within 6
months
All Rights Reserved, Duke Medicine 2007
Once someone has a stroke…
WHERE ARE OUR
INTERVENTION POINTS?
All Rights Reserved, Duke Medicine 2007
Stroke System of Care
and Transitions in Care
Community
Response
All Rights Reserved, Duke Medicine 2007
EMS
Acute Care
Rehab &
Recovery
A Critical Intervention Point
General Population
The Transition to
Post-Acute Care
Living in Community
Population
At-Risk
Acute
Care
Post-Acute
Care &
Rehab
Living in LTC
All Rights Reserved, Duke Medicine 2007
Olson DM, Prvu Bettger J, Alexander KP, et al. AHRQ 2011
Medicare Beneficiaries: Comparing patients’ 1st
post-acute setting for all dx to stroke
70
All Medicare
60
Stroke
50
40
%
30
20
10
0
D/C to PAC
IRF
SNF
Inpatient
All Rights Reserved, Duke Medicine 2007
LTAC
HH Outpatient
Community
Evidence-based Information is Lacking to Guide
Delivery of Stroke Care After Hospital Discharge
• What services should a stroke patient receive after
being hospitalized for an acute stroke?
– Compare post-acute and transitional care
treatment options that matter to patients and their
caregivers
– Focus on outcomes of interest to patients and their
caregivers
• What strategies should be in place to improve the
transition from inpatient care and improve longerterm outcomes?
All Rights Reserved, Duke Medicine 2007
All Rights Reserved, Duke Medicine 2007
Services After Acute Care
All Rights Reserved, Duke Medicine 2007
Care Following the Acute Hospitalization
153,775 Acute Ischemic Stroke Patients in GWTG
w/ Medicare FFS Parts A+B Alive at Hospital
Discharge (2006-2008)
51.1% Discharged
to Short-term
Inpatient Postacute Care
24.1% Inpatient
Rehabilitation
Facility or Unit
(N=37,064)
All Rights Reserved, Duke Medicine 2007
27.0% Skilled
Nursing Facility
(N=41,457)
48.9% Discharged
from the Hospital
to the Community
14.9% Home
Health
(N=22,875)
7.1% Outpatient
Rehabilitation
(N=10,982)
26.9% No
Post-acute Care
(N=41,397)
…to generate key evidence that can be used to guide
a critical decision faced by stroke survivors, their
caregivers and health care providers every day,
almost 1 million times a year…
what services to choose following
an acute stroke hospitalization
?
A TREMENDOUS OPPORTUNITY
All Rights Reserved, Duke Medicine 2007
Specific Aims
(what we promised we would do … at a high level)
1. Identify the factors associated with stroke
survivors’ use of rehabilitation and health care
services following hospital discharge (who gets
what services and why based on our data)
2. Compare high intensity rehabilitation (provided in
inpatient rehabilitation facilities; IRF) and low
intensity rehabilitation (provided in skilled nursing
facilities; SNF) on several outcomes
3. Compare outpatient (OP) rehabilitation and home
health (HH), and how either are better than no
rehabilitation.
4. Compare PCP and neurologist follow-up on
outcomes
All Rights Reserved, Duke Medicine 2007
Who are we studying and how?
This is a study using existing data of adults who had a stroke in
2006-2008.
The person had to have
been treated in a
hospital participating
in the Get With The
Guidelines-Stroke
program.
Medicare
FFS
GWTGStroke
AVAIL
Cohort
The person had to be a
Medicare fee-for-service
beneficiary for health care.
Some were in a prospective
cohort study, AVAIL.
All Rights Reserved, Duke Medicine 2007
PCORI Population
Outcomes from Claims Data
Outcomes
Definition
Timing
Hospital readmission
All-cause readmission
30 and 90 days
Time to hospital readmission
Time in days from index hospitalization to
readmission
Up to 12 months
Long-term care placement
Nursing home = residence (nursing facility
assessment CPT code AND place of service
code with no associated SNF claim)
12 months
Survival (analyzed as death)
Mortality (alive/not; will use for death & disability)
12 months
Survival
Time in days from index hospitalization to date of
death
Up to 12 months
Home-time
Number of days from hospital discharge to 12
months without inpatient services, a
rehospitalization or LTC admit
Up to 12 months
Health care utilization
(proxy for cost)
All billable services
(inpatient days, ED visits, observation stays &
provider visits)
30 and 90 days
All Rights Reserved, Duke Medicine 2007
Patient Reported Outcomes
Outcomes
Definition
Primary: living independently
Alive, community-dwelling, modified Rankin (mRS) 0-2
3 &12 mo.
Functionally independent
mRS 0-1
3 &12 mo.
Change in function
+/- 1 or > change in mRS from 3 to 12 months
12 months
Depression
PHQ-8 >10
3 &12 mo.
Persistent depression
PHQ-8 >10 at 3 and 12 months
12 months
Return to work
Employed pre-stroke and returned to work
3 &12 mo.
Medication adherence
Actions correspond with hospital provider
recommendations (warfarin, antihypertensive,
antiplatelet, lipid-lowering, and diabetic medications)
3 &12 mo.
Smoking cessation
Actions correspond with hospital recommendations
3 &12 mo.
Quality of life
Normalized EQ-5D
3 &12 mo.
Change in quality of life
Change in EQ-5D score from 3 to 12 months
12 months
Death and disability
Composite outcome
12 months
All Rights Reserved, Duke Medicine 2007
Timing
Soon we’ll have
clearer evidence of
what services for
which patients…
But how do we support
them along the journey
back home?
All Rights Reserved, Duke Medicine 2007
A look at care across the continuum…
STROKE TRANSITIONS IN CARE
All Rights Reserved, Duke Medicine 2007
Most Common Trajectories
or Patterns of Care
There were 3,016 unique care patterns
in the 120 days after an acute ischemic stroke
All Rights Reserved, Duke Medicine 2007
Most Common Trajectories
or Patterns of Care
Discharged to Short-term Inpatient
Post-acute Care
24.1% Inpatient Rehabilitation Facility or Unit (N=37,064)
5.3% IRF only
19.9% IRF +
HH
8.0% IRF +
HH + OP
16.0% IRF +
OP
6.2% IRF +
HH + Readmit
+ SNF or
+ HH
After Readmit
All Rights Reserved, Duke Medicine 2007
27.0% Skilled Nursing Facility (N=41,457)
8.4% IRF +
SNF
30.5% SNF
only
19.0% SNF +
HH
5.8% IRF +
SNF + HH
21.0% SNF +
Readmit
5.2% SNF +
HH + Readmit
+ SNF or
+ HH/OP
After Readmit
+ SNF
After Readmit
11.5% SNF +
OP
Most Common Trajectories
or Patterns of Care
Discharged from the Hospital to the Community
7.1% Outpatient
Rehabilitation (OP)
(N=10,982)
14.9% Home Health
(N=22,875)
63.2% Home Health only
20.8% HH + Readmit
+ HH or
+ SNF
After Readmit
All Rights Reserved, Duke Medicine 2007
10.1% HH + OP
77.0% OP
Rehabilitation Only
?
Involvement of
Primary and
Specialty Care?
26.9% No Post-acute
Care (N=41,397)
52.6% No Post-acute
Care (no readmission
and alive at 120 days)
27.4% Readmitted after
hospital d/c without postacute care
20.0% Died after hospital
d/c without post-acute
care
A Critical Intervention Point
General Population
The Transition to
Post-Acute Care
Living in Community
Population
At-Risk
Acute
Care
Post-Acute
Care &
Rehab
Living in LTC
All Rights Reserved, Duke Medicine 2007
Olson DM, Prvu Bettger J, Alexander KP, et al. AHRQ 2011
Critical Intervention Points
for Stroke Survivors
Transitions
Living in Community
Acute
Care
Post-Acute
Care &
Rehab
Living in LTC
All Rights Reserved, Duke Medicine 2007
Olson DM, Prvu Bettger J, Alexander KP, et al. AHRQ 2011
Gaps Identified By Observation, Provider
and Patient Reports, and Research
Transitional
Care
Interventions
Stroke Patients’
Needs
Hospital Discharge
Planning
Rehabilitation
Expertise
All Rights Reserved, Duke Medicine 2007
G
A
P
Stroke Patients’ and
Caregiver’s Needs at
Home
Community-based
Care
Rehabilitation
Expertise
Transitions are a National Priority
• HHS Triple Aim: Better Care, Better
Health, Lower Cost
•
HHS Priorities = National Quality
Strategy: Efficiency, population/public
health, clinical effectiveness and
processes, care coordination, patient and
family engagement, patient safety
• CMS: The right care for every person
every time
•
All Rights Reserved, Duke Medicine 2007
Partnership for Patients: Reduce HAC
by 40% and readmissions by 20%
Stroke Readmission:
Opportunity for Improvement
• National
readmission rate:
13.8%
• Hospital riskstandardized
readmission rate
(RSRR) range:
9.1%-20.6%
30
All Rights Reserved, Duke Medicine 2007
Shifting and Narrowing the Curve – How?
All Rights Reserved, Duke Medicine 2007
What is effective for stroke survivors?
TRANSITIONAL CARE
INTERVENTIONS
All Rights Reserved, Duke Medicine 2007
“the set of actions designed to ensure the
coordination and continuity of health care as
patients transfer between different locations or
different levels of care within the same location”
TRANSITIONAL CARE
All Rights Reserved, Duke Medicine 2007
Coleman et al., J Am Geriatr Soc 2003;51(4):556-7.
Transitional care is…
•
•
•
•
Supportive of patients during handoffs
A time-limited service
Focus on continuity
Commonly led by a nurse (more than 50% of interventions
summarized in systematic reviews were nurse-led)
• An emerging key factor in care coordination
All Rights Reserved, Duke Medicine 2007
Why is this important?
Poorly executed or discontinuous health
care transitions increase the risk of
medical and medication errors,
poor patient outcomes,
caregiver stress,
and unnecessary services
All Rights Reserved, Duke Medicine 2007
GRACE
BOOST
All Rights Reserved, Duke Medicine 2007
Care
TCM
Transitions
There are Many “TOC” Models
Which Intervention?
Core Interventions in Evidence-Based Transitional Care Models
Interventions
RED
CTI
Evaluation/risk assessment
Medications reconciled & plan confirmed
X
X
Patient education (with teach-back) on:
Diagnosis (daily)
X
X
Completed tests & appropriate follow-up
X
“Red flags” and response to problems
X
X
Patient education on medication management
X
X
Discharge plan
X
X
Written discharge plan or personal record
X
X
Discharge plan reconciled with national clinical guidelines
X
Appointments made for clinician follow-up, services , tests
X
Appointments to be scheduled by patients
Discharge summary (transition record) sent to postdischarge providers
Documented receipt of information by next provider
Telephone follow-up with patient to ID / resolve problems
BOOST
X
X
TCM
X
Bridge
X
GRACE*
X
INTERACT*
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
N/A
X
X
X
X
In 2-3
Days
X
In 3
Days
X
X
In 2, 30
days
X
N/A
Home visit
X
X
X
N/A
Transitional care point person(s)
X
X
X
Facilitated engagement of patients, families, providers
X
X
across episode
*GRACE is a community-based model; INTERACT is a nursing home-based models
All Rights Reserved, Duke Medicine 2007
Evidence of Effectiveness?
All Rights Reserved, Duke Medicine 2007
KNOWLEDGE GAPS
All Rights Reserved, Duke Medicine 2007
Knowledge Gaps
1. Do these work for stroke patients?
2. Which strategies? (each intervention is multicomponent)
3. Do we replicate these interventions? Adapt locally?
Integrate strategies from different interventions?
4. Which transition or handoff?
5. For what period of time?
6. For which patients?
All Rights Reserved, Duke Medicine 2007
Transitional Care for Stroke: Roots in Policy
National Health
Reform: Reduce
hospital
readmissions
Wave 1: Heart Failure, Pneumonia,
Myocardial Infarction
Proposed Wave 2: Stroke, Chronic
Obstructive Pulmonary Disease
2012 Guidelines International Network
2012 International Stroke Conference
2013 International Association of Gerontology and
Geriatrics
Disseminated
internationally
All Rights Reserved, Duke Medicine 2007
CDC commissioned
a systematic review
Do we have the evidence we need?
Do these work for stroke patients?
Very few of the nationally promoted care transitions
models included stroke patients.
Of those that did, none presented findings for stroke
patients.
All Rights Reserved, Duke Medicine 2007
Evidence of Effectiveness?
All Rights Reserved, Duke Medicine 2007
Team based approach including caregivers to
return stroke patients home earlier but with
continued rehabilitation of similar intensity and
duration to inpatient care
EARLY SUPPORTED
DISCHARGE
All Rights Reserved, Duke Medicine 2007
ESD Components
•
•
•
•
Patient identified in acute care (or inpatient setting)
Discharged earlier
Home visit within 24 hours of hospital discharge
Goal-driven and patient-specific rehabilitation
delivered in the home
• Services provided 4 x day (ESD phase), 6-7 days
week for up to 4 weeks and then reducing to weekly
visits by the point of exit (at most 4-6 weeks)
• Different levels of engagement with stroke specialist
(neurology)
All Rights Reserved, Duke Medicine 2007
Models of ESD
1. Stand-alone acute outreach ESD only
– Prevalent in denser populated urban cities and where there are
large city hospitals
2. ESD with community stroke/neurology team service
– In-hospital component hands off to a usually well established
community-based rehab team partnering with neurology
3. Integrated ESD within community stroke team service
– All the components of models 1 and 2, plus support workers for
rehab every day & multiple visits a day for up to six weeks
4. Integrated ESD within community neurology service
– Often extends beyond stroke but then requires advanced skill set;
prevalent in less urban areas
5. RECOVER trial
– Nurse facilitated and organized, caregiver delivered
All Rights Reserved, Duke Medicine 2007
Evidence for ESD
• Multiple randomized controlled trials
• Meta-analysis confirmed patients who received these services returned
home earlier (shorter inpatient length of stay) and were more likely to
remain at home in the long term (longer “home time”) and to regain
independence in daily activities (reduced death and dependency).
– The best results with well organized discharge teams and patients
with less severe strokes.
• International Consensus Guidelines and considered best practices in
UK and Canada
– Canada ESD: $132.9 million direct cost savings.
• In the U.S.?
– Failed and not feasible given payment model for services
All Rights Reserved, Duke Medicine 2007
Proposed US Model (Govt. focus = ↓ costs)
Reduce costs by
34%, saving $2.4
million over a 3
year period for
300 stroke
patients
Reduce
rehospitalizations
Improve patient
functional status and
reduce secondary
complications of stroke
Implement and optimize uptake of Early
Supported Discharge as the new
health care delivery model for postacute comprehensive stroke
management
Reduce utilization of
post-acute services (in
skilled nursing facilities,
inpatient rehabilitation,
and multipe episodes of
home health care)
Improve selfmanagement of stroke,
co-morbid chronic
conditions and CV risk
factors
Integrate primary care with Early
Supported Discharge to improve
access and transition care to
community-based providers
Reduce long-term care
nursing home placement
Improve patient and
caregiver satisfaction
with post-acute stroke
care
Transition to community-based wellness
and exercise programs, and case
management as needed
“Task” shifting at
3 levels: rehab,
primary and
community care
All Rights Reserved, Duke Medicine 2007
Not Quite a Global Perspective
General Population
Living in Community
Population
At-Risk
Acute
Care
Post-Acute
Care &
Rehab
Living in LTC
All Rights Reserved, Duke Medicine 2007
Olson DM, Prvu Bettger J, Alexander KP, et al. AHRQ 2011
Adapting ESD Globally
• ATTEND trial (Family-Led Rehabilitation after stroke
in India)
• RECOVER trail (A randomized controlled trial on
rehabilitation through caregiver-delivered
• nurse-organized service programs for disabled stroke
patients in rural China)
All Rights Reserved, Duke Medicine 2007
Evidence of Effectiveness?
What is appropriate for rural China?
All Rights Reserved, Duke Medicine 2007
The
Trial
A Randomized Controlled Trial on Rehabilitation
through Caregiver-delivered Nurse-organized
Service Programs for Disabled Stroke Patients in
Rural China
52
All Rights Reserved, Duke Medicine 2007
RECOVER Collaborating Institutions
53
All Rights Reserved, Duke Medicine 2007
RCT Study Description
Intervention and Control
In-hospital
• “Teach-back” on stroke recovery, risk
identification, and management
• Task oriented training
• Joint goal setting
• Evidence-based discharge planning
♦ Intervention Group:
Intervention
Training
Nurse + Patient +
Family Caregiver
Physicians /
Rehabilitation Therapists
Nurses from
County Hospitals
Patients & familynominated caregivers
After hospital discharge
• Call or visit at 2, 4, 6 and 8 weeks
• Blinded researcher measures outcomes
by phone at 3 mo. & in person at 6 mo.
♦ Control Group: conventional care
All Rights Reserved, Duke Medicine 2007
54
Hypotheses
Primary Hypotheses
Primary Outcome
•To improve physical function
•Barthel Index
Secondary Hypotheses
Secondary Outcomes
•To improve physical functioning
•Functional Ambulation Classification
•To improve quality of life
•EQ-5D
•To reduce disability
•modified Rankin Scale
•To reduce depression
•Patients Health Questionnaire-9
Exploratory Hypotheses
Exploratory Outcomes
•To relieve burden of caregivers
•Caregiver Burden Index
•To reduce hospitalization
•Re-admission and hospitalization
•To reduce hospital length of stay &
costs
•Hospital length of stay and medical
costs
55
All Rights Reserved, Duke Medicine 2007
Study Design-Patient Recruitment
Inclusion criteria
• Adults (≥18 years);
• Recent (<1 month) first-ever acute ischemic/hemorrhagic/undifferentiated stroke
patients;
• Expected to survive to discharge from hospital with a reasonable expectation of 6
month survival (i.e. not palliative, no evidence of widespread cancer etc.);
• Residual disability (requiring physical assistance for core activities of daily living
defined as a Barthel Index score of 80 or lower).
Exclusion criteria
• Unable to identify a suitable family-nominated caregiver for training and
subsequent delivery of care;
• Unable to provide informed consent from both the patient (or by proxy) and the
caregiver.
56
All Rights Reserved, Duke Medicine 2007
Pilot and Main Study Samples and Sites
Sites:
• Zhangwu County,
Liaoning Province,
• Qingtongxia County,
Ningxia Province
Qingtongxia
Zhangwu
Pilot Study
• Number of patients: 80
20 (I) + 20 (C) x 2 sites
Main Study
• Number of patients: 200
100 (I) + 100 (C) (1/2 each site)
57
All Rights Reserved, Duke Medicine 2007
Complementary Models
Transitional
Care
Early
Supported
Discharge
Telehealth
RECOVER
2.0
All Rights Reserved, Duke Medicine 2007
Meeting the Needs of Stroke Survivors
and Caregivers Globally
Many possibilities for efficacy and/or implementation
effectiveness trials
• Urban or rural
• Single or multi-component strategy
• Inpatient- or community-based or both
• GPs or nurses or community workers or trained lay people
• Mobile phones and ipads or centrally located computers
• Intervention(s) to focus on functional impairment,
secondary prevention, or prevention of complications
• Patients with or without cognitive impairment
• Different levels of inpatient care
• Different levels of caregiver engagement
All Rights Reserved, Duke Medicine 2007
Why are we committed to this global
research agenda?
•
•
•
•
Improve the lives of stroke survivors
Reduce the burden on informal (family) caregivers
Improve adherence to evidence-based care
Improve the quality of care
• Improve physician-nurse partnership in the care of
patients with stroke, disability, and chronic illnesses
• Build rehabilitation nursing capacity as leaders for
caring for people with disabilities and chronic illness
60
All Rights Reserved, Duke Medicine 2007
Improving Stroke
Outcome: It is
going to take a
village
All Rights Reserved, Duke Medicine 2007
THANK YOU!
Janet Prvu Bettger, ScD, FAHA
[email protected]
All Rights Reserved, Duke Medicine 2007