Transcript What can states do to improve chronic illness care?

Organization of Health Care and
Delivery System Design
Alan Glaseroff MD
CMO, Humboldt IPA
IHI National Forum 2007
Orlando, Florida
12/10/07
[email protected]
Redesigning Chronic Illness Care: Evidence,
Experiences, and Stakeholders
IHI National Forum December 10, 2007
Health of Populations and Individuals
• Delivery system exists within communities
• Many other stakeholders with interests
– Patients, employers, public health/government, community groups,
educational system, payers
• Chronic disease affects certain populations disproportionately
• Collaboration needed (spectrum of relationships) to improve
outcomes and reduce disparities
• Collective accountability/responsibility the only answer
• “If not us, who? If not now, when?”
Disparities: Life Expectancy at Birth
Overall White
Overall Black
85
without Homicides White
without Homicides Black
without AIDS White
80
without AIDS Black
without Both White
without Both Black
Years
75
70
65
Tony Iton MD, Alameda County Public Health Director
2001
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
1977
1975
1973
1971
1969
1967
1964
1962
1960
60
Life Expectancy by Census Tract
Alameda County
100
95
Life Expectancy (Years)
90
85
80
75
70
65
60
55
50
0%
10%
20%
30%
Pov erty Rate
Tony Iton MD, Alameda County Public Health Director
40%
50%
60%
Causes of Differences in
Health Outcomes By Race
• Genetics*
10-15%
• Access to health care 10-15%
15% + 15% = only 30%
What causes the other 70%???
*genes ≠ race
Tony Iton MD, Alameda County Public Health Director
Medical Model
HEALTH
CARE
ACCESS
Socio-Ecological
-Bay Area Regional Health Inequities Initiative
Expanded Model
Chronic Care Model
Community
Resources and Policies
SelfManagement
Support
Informed,
Activated
Patient
Health System
Health Care Organization
Delivery
System
Design
Productive
Interactions
Decision
Support
Clinical
Information
Systems
Prepared,
Proactive
Practice Team
Functional and Clinical Outcomes
Health Care Organization
• Quality as core strategy
• Visibly support improvement at all levels, starting with
senior leaders.
• Promote effective improvement strategies aimed at
comprehensive system change.
• Encourage open and systematic handling of
problems.
• Provide incentives based on quality of care.
• Develop agreements for care coordination.
Chronic Care Model
Community
Resources and Policies
SelfManagement
Support
Informed,
Activated
Patient
Health System
Health Care Organization
Delivery
System
Design
Productive
Interactions
Decision
Support
Clinical
Information
Systems
Prepared,
Proactive
Practice Team
Functional and Clinical Outcomes
Delivery System Design
• Multiple levels
– Regional/National: macrosystem
– Integrated Medical Care Organization:
mesosystem
– Practice level: microsystem
• Alignment required for breakthrough
improvement in community health
Delivery System Design
• Define population of patients
• Define roles and distribute tasks amongst team
members.
• Use planned interactions to support evidence-based
care.
• Provide clinical care management services.
• Ensure regular follow-up.
• Give care that patients understand and that fits their
culture
Mesosystem: Practice Environment
in Humboldt
• 29 primary care practices in various sizes,
types and stages of transformation (all in
the Humboldt IPA)
– 5 community health centers
– Many 1-3 clinician practices in private practices (one 17
MD Internal Medicine practice)
– No large integrated multispecialty group
– Managed care covering 5% of population
• How to rapidly improve chronic disease
care in the community?
Humboldt Diabetes Project
• CHCF-funded research project started 11/02
• County-wide effort coordinated by IPA (>95% of all clinicians in the
county, including MDs, advanced-practice clinicians, behavioral
health providers) but…
• IPA manages only 10% of lives in Humboldt County
…but systems must apply to most patients in a practice
• Problem:
– Lack of access to most administrative data
• Solution:
– Information must come from clinical setting
Getting Started
• “Burning Platform” to capture hearts and minds (disease focus
vs. abstract “redesign”)
• Grant support for concept
• Clinical champion presenting own data making it safe for
others
• “Inviting the implementers into the planning process”
• Piloting systems
• Kick-off conference (including patient voices)
• Site champion network supervised by ½-time FNP
• Feedback on practice-level and individual performance
Humboldt Diabetes Project
Clinician
Education
Patient
Education
reted
Integ ion
s
D e ci r t
o
Supp
Pat
Invo ients
lv
Self ed In
-C a
re
s
a ye r
Get P spitals
Ho
and Play
to
Registry
and Flow
Sheet
Prompts
and
Reminders
Case
Management
Buil
d an
C h r d Ma i n
o
t
Infra nic Ca ain a
re
stru
ctur
e
N e xt
Chronic
Diseas
e
es
ct Offic
Conne rnet for
nte
to the I
l
Clinica nd
tion a
Informa ication
un
C o mm
DIABETIC
PROJECT
FLOW
DIAGRAM
Data from
office visit
PACES, CHCF
chart audits
Staged Diabetes
Management
Guideline
Feedback to
clinicians
yes
Pt. enrolled in
diabetic study
Data entered
in registry
Monthly audit of
data in registry
Prompts and
reminders
Office visit:
scheduled,
random
Services
delivered?
Registry note
returned?
Data analysis
Patient
visit
sheet
no
A1c > 9
Pharmacy
data
yes
Risk stratification
Lab
data
BASICS
Case management
To improve outcomes in chronic
illness…
• Patients must be prescribed and taking
proven therapies
• Patients must be managing their illness
well
Microsystem: Frustration
• Patients are frustrated by waits and
discontinuities, often don’t receive proven
services and often feel they are not
heard.
• Providers feel they have little control over
their work life, are stressed by demands
for productivity despite older, sicker
clientele and the reduced variability in
their clinical day.
Is There Time for Management of Patients
With Chronic Diseases in Primary Care?
•
METHODS
– Applied guideline recommendations for 10 common chronic diseases
to a panel of 2,500 primary care patients with an age-sex distribution
and chronic disease prevalence similar to those of the general
population, estimated the minimum physician time required to deliver
high-quality care for these conditions.
•
RESULTS
– Top 10 chronic diseases (STABLE) 828 hours per year, or 3.5 hours a
day
– Top 10 chronic diseases (Poor Control) 2,484 hours, or 10.6 hours a
day.
•
CONCLUSION?
Ann Fam Med 2005;3:209-214.
Duke University Dept. of Community and Family Medicine
What we know about primary care visits
• 50-70% are largely informational or
informative (including check-backs for
chronic illness care) yet they are
organized like acute visits
• US average is 16.3 minutes
• Patients are given an average of 20
seconds to tell their story before they are
interrupted
How would I recognize a
productive interaction?
Informed,
Activated
Patient
Productive
Interactions
Prepared
Practice
Team
•Assessment of self-management skills and
confidence as well as clinical status
•Tailoring of clinical management by stepped
protocol
•Collaborative goal-setting and problem-solving
resulting in a shared care plan
•Active, sustained follow-up
Microsystem: Defining roles and
tasks across team to achieve
productive interactions
“It is naïve to bring together a highly diverse
group of people and expect that, by calling them
a team, they will in fact behave as a team. It is
ironic indeed to realize that a football team
spends 40 hours a week practicing teamwork for
the two hours on Sunday afternoon when their
them work really counts. Teams in organizations
seldom spend two hours per year practicing
when their ability to function as team counts 40
hours per week."
Harold Wise, Making Health Teams Work
Team Meetings
• Regular intervals
• All members of care team (groups of < 10)
• Agenda:
– Old business
– New Business
– What isn’t working?
– Opportunities for excellence?
Defining Tasks
Example of task distribution
Microalbuminuria testing
• Receptionist recognizes patient has diabetes, attaches
requisition to chart
• MA collects specimen
• RN reviews slip, recognizes out-of-range tests, orders
confirmatory test, discusses possible need for ACE
inhibitor
• MD discusses and prescribes ACE inhibitor
• RN calls pt. to check on med. adherence and side
effects
Use planned interactions to
support evidence-based care
One-on-one, group, telephone, email,
outreach….the possibilities are endless
What is a Planned Visit?
• A Planned Visit is an encounter with the patient initiated by the
practice to focus on aspects of care that typically are not delivered
during an acute care visit.
• Planned care elements can be inserted into acute visits if needed
(small practices, patients refusing to come in for planned care, etc.)
• All visits contain elements of both (patient agenda/clinician
agenda)
• The more planned care functions done by other members of the
team, the more time for the patient agenda in the exam room
(improves clinician-patient relationship, higher patient satisfaction)
What does a Planned Visit look like?
• The provider team proactively calls in patients
for a longer visit (individual or group) to
systematically review care priorities.
• Visits occur at regular intervals as determined
by provider and patient.
• Team members have clear roles and tasks.
• Delivery of clinical management and patient
self-management support are the key aspects
of care.
How do you do a Planned Visit?
You Plan It!
Example: Diabetes
• Choose a patient sub-population, e.g., all
patients with diabetes not seen in 6 months
with A1c > 7
• Identify patients from registry
• MD reviews list for patients at highest risk (via
evidence-based guidelines): BP>130/80;
LDL>100, etc and prioritizes visits
Patient Outreach
• Have receptionist or provider call patient and
explain the need for planned visit using script
explaining different nature of visit
• Personal appeal by clinician works best
• Ask patient to bring in bag of all medications they
are taking (including OTCs and herbals)
Preparing for the Visit
• “Team Huddle” at start of clinic session
• RN/LPN/MA prints any relevant patient
summaries from registries and attaches
to front of chart
• MD reviews medications/labs prior to visit
REGISTRY
FLOW
OFFICEVISITS
Planned Visit
Chart visit in
normal fashion
Prep chart
no
Registry
Patient?
yes
Download
most recent
Progress Note
Place PN on
front of chart
Medical
Assistant tells
patient: “Take
off your shoes”
Progress Note
employed in visit
Pull chart
Progress Note
returned to
office staff
Patient Info sheet sent
to patient
Walk-in Visit
Updated Progress
Note faxed to office
and placed on Registry
web-site
IPA updates
registry
Progress Note
faxed to IPA
The Visit
• Ask patient open-ended questions
– “How’s your health? Any issues you want to discuss?
• Review patient’s data
• Identify interventions, labs, referrals and selfmanagement needs
• Problem solve adherence/other issues with patient
• Create an patient action plan (if indicated)
• Schedule follow-up
Group Visits: Introduction
• Fun and efficient
• Patients can receive:
Self-management support training
Social support
Specialty service as needed/available
One-on-one with medical provider
Medication counseling
• Multiple models for Group Visit agendas: open-ended vs.
curriculum-based; single disease vs. multiple; newly
diagnosed vs. range of experience; professional vs. peer-led
Patient Survey: Less Frustrated?
In general, would you say your health is: (check one box)
B
F
:
Excellent
Very Good
Good
Fair
Poor
4%
19%
37%
30%
10%
5%
27%
42%
22%
4%
How effective do you believe your health care provider is in managing
your diabetes?
Not
effective at
all
Not very effective
Somewhat
effective
Effective
Very
effective
B
1%
3%
18%
45%
34%
F
<1%
1%
13%
44%
41%
Clinician Survey: Less Overwhelmed?
How effective do you believe you are in caring for your diabetic
patients?
Not
effective
at all
Not very
effective
Somewhat
effective
Effective
Very
effective
Baseline -
3%
32%
57%
8%
F/U
-
27%
56%
17%
-
Compared to a year ago, how effective are you in caring
for your diabetic patients?
Less
effective
F/U
-
Somewhat
less effective
Somewhat
more effective
Same
-
27%
41%
Note: The sum of the categories may not add to 100% due to rounding.
More
effective
33%
What is care management?
Many different things to different people
• Resource coordination
• Utilization management
• Follow-up
• Patient education
• Clinical management
Features of effective care management
• Regularly assess disease control,
adherence, and self-management status
• Either adjust treatment (best practice) or
communicate need to physician
immediately (less effective)
• Provide self-management support
• Provide more intense follow-up
• Assist with navigation through the health
care process
Effect of Group SMS on HbA1c
Group SMS (Basics) vs. Control and Sustainability
7.80
7.60
7.40
7.20
7.00
Pr
e
Ba
sic
No s
v0
De 5
c0
Ja 5
n0
Fe 6
b0
M 6
ar
-0
Ap 6
r-0
M 6
ay
-0
Ju 6
n0
Ju 6
l-0
Au 6
g0
Se 6
p06
6.80
Basics Oct 05
Grads(n=135)
non Basics
(n=3400+)
Ensure regular follow-up by
the primary care team
•The trick is noticing when it isn’t happening
•Can be accomplished in many different ways
Humboldt Diabetes Project Data
Measure
October, 2003
October, 2004
January, 2007
Results
Results
Results
(n=802)
(n=778)
(n=4330)
5.2%
HbA1c control: >9% (poor control)
7.7%
6.9%
HbA1c control: <7% (good control)
52%
55%
Patients with BP <140/90
62%
59%
59%
67%
Patients with BP <130/80
32%
33%
37%
Patients with LDL<130
60%
73%
78%
Patients with LDL <100
32%
44%
49%
Contact us:
•www.improvingchroniccare.org
New Methods for Teaching
the Chronic Care Model
IHI National Forum December 10, 2007
Breakthrough Series Collaborative
Participants
Select
Topic
Planning
Group
Prework
Identify
Change
Concepts
P
A
P
D
A
S
LS 1
P
D
A
S
LS 2
D
S
LS 3
Action Period Supports
(12 months time frame)
E-mail
Visits
Web-site
Phone
Assessments
Senior Leader Reports
Event
Experience with Collaboratives
• More than 1,000 different health care organizations
and various diseases involved to date
• Began with national BTS, now regional, state-based
& facility specific
• HRSA’s Health Disparities Collaboratives600+ community and migrant health centers, now
academic medical centers & small practices
• External evaluations of early efforts by Chin et al.,
RAND
Lessons Learned from the Teams
• Teams spent considerable time searching
for/developing tools
• Some teams felt intimidated by taking on the whole
model – asked for a sequence
• Collaboratives were time & resource intensive
• Many changes were made in ways that were not
sustainable financially
Challenges Remaining
Reaching
beyond
early
adopters
Try less
timeintensive
learning
Target
small
practices
Create
supportive
systems
Trying New Ideas
• Virtual
• Collaboratives On Wheels
• Coaching
• Combos
Integrating Chronic Care and Business
Strategies in the Safety Net
The Intervention
Practice Coach
_________________
STEP-UP Methodology
PLUS
Toolkit
______________
Business & Clinical
Tools
The Toolkit Sequence
Getting
Started
Organize Your Improvement Team
Familiarize Team With Strategies
Use Data To Set Priorities
Assess Data &
Set Priorities
Select Performance Measures
Build Measurement Capacity
Redesign
Care and
Business
Systems
Organize YourYour
Care Team
Organize
Care Team
Clearly Define
PatientPatient
Panels
Clearly
Define
Panels
Create Infrastructure Support Pts
Create Infrastructure To Support Patients
Plan Care
Plan Care
Support Self-Management
Assure Support For Self-Management
Improve &
Sustain
Changes
• Integrated
CCM &
Business
Changes
Reexamine Outcomes & Adjust
Reexamine Outcomes & Make Adjustments
Capture Incentives
• Organized
into four
phases
The Toolkit
Business
CaseCase
The
Toolkit&&theThe
Business
Thank you!
Katie Coleman, MSPH
[email protected]
www.improvingchroniccare.org
Coaching Outline
Tasks
Philosophy
Assessment Day
Focus on motivation,
consultation & education
½ day presentation on CCM
& PDSA
On-going meetings by phone,
email & in-person
Coaching of the leaders & the
teams
Be mindful of the timing of
interventions
Fix processes relevant to the
task at hand
Well-structured & supported
groups benefit most
Redesigning Chronic Illness Care:
Taking Improvement to Scale
Ed Wagner, MD, MPH
MacColl Institute for Healthcare Innovation
Center for Health Studies
Group Health Cooperative
Improving Chronic Illness Care
A national program of the Robert Wood Johnson Foundation
IHI National Forum December 10, 2007
CCM Developments
• Guides several state programs
• Adaptations undertaken by European countries, World
Health Organization, and several Canadian provinces.
• Foundation for NCQA and JCAHO certification for chronic
disease programs.
• Part of new Patient-centered Medical Home Models of
Primary Care proposed by AAFP, ACP, AAP, AOA.
• Several practice assessment tools now available for large
and small practices.
• Assessments now used in some pay for performance
programs (NCQA).
Challenges Remaining
Early
Adopters
Try less
timeintensive
learning
Target
small
practices
Create
supportive
systems
What will it Take to Improve Care for
Chronic Illness for the Population?
Three Options When Selecting a Strategy
1. Assume that competition, financial incentives
and computers will improve care.
2. Rely on direct to patient disease
management.
3. Improve medical care by helping practices
change care systems.
What can we learn from successful larger
health systems?
Organizational factors supportive of high quality chronic
care:
• Strategic values and leadership support long-term
investment in managing chronic diseases
• Well-aligned goals between physicians and corporate
managers
• Investment in information technology systems and other
infrastructure to support chronic care
• Use of performance measures and financial incentives to
shape clinical behavior
• Active programs of Quality Improvement based on
explicit models
BMJ 2004;328:223-225
What’s needed to improve chronic
illness care for the population?
Build a regional healthcare “system”
• Collaboration and Leadership
• Measurement (& incentives)
• Infrastructure
• Active program of
practice change
•Care will not improve unless we
change the systems of care
•The goal is to transform health care
delivery across a region
Leadership
•Someone needs to take and then
assure leadership
•Major stakeholders need to be
involved and committed to
improvement
•Need outcome and patient experience data
as well as process data to assess effort,
performance, and improvement
•Practices will have to be able to provide
valid and complete data on these indicators;
claims will not suffice
•Practices should be able to use these data
in clinical care, not just periodically send them off
•Smaller practices need info. and technical
support to develop such data systems
•Need strategies and
infrastructure to help
ALL practices change
their delivery systems
•Strategies – QI methods,
Provider networks
•Infrastructure—IT, guidelines, care managers
•More activated and informed
consumers may help push
improvement
•Public disclosure of performance
data may spur improvement
•Create incentives for providers
to make the investments needed
to improve chronic care
•Create benefit plans that reward
consumers for making costeffective choices
A Framework for Regional Quality Improvement
Leadership
Is geographic improvement possible?
State efforts
Is geographic improvement possible?
Indiana
• Health Commissioner and Medicaid Director to improve
care for 80,000 chronically ill Medicaid recipients
• State leadership and money creating a Medicaid care
system
• Statewide Collaborative Program PLUS
-call center
-community-based nurse care managers linked to
practices
-statewide Web-based patient registry
-registry updated with claims data
-considering performance incentives
-embedded RCT
• Reported cost-savings to the Governor
Is geographic improvement possible?
North Carolina
• State leadership and money has created a
visionary Medicaid care system
• Measurement system, Guidelines, Physician
networks, Care Managers, Collaboratives
• Financial rewards for participating
• Early results promising
• Plans to extend to include all patients
regardless of insurance coverage
Is geographic improvement possible?
Washington State
•Diabetes
Surveillance
•Regional
Collaboratives
•Laid
groundwork for
PSHA
Is geographic improvement possible?
Pennsylvania
•Governor brought
disparate interests
together
•All the major
players at the table
•Timeline &
?budget to make it
happen
Lessons Learned
Indiana
Make your effort bipartisan & protect it from
political winds.
North
Carolina
Reach out! Provider networks can engage
small practices in quality improvement
Rhode Island Bring all the “p”s to the table: providers,
purchasers, payers, patients, policy-makers
Colorado
Connect with local foundations and groups
already doing the work
Washington/ Political leadership involvement can be
Penn.
critical catalyst
Maine/
California
Organizing diffuse efforts is a big but
important job
Contact us at:
www.improvingchroniccare.org