Transcript Winston F. Wong - UW Population Health Institute

Knowledge
Into Care… and
Care into
Knowledge
The Wisconsin Council on
Children
Madison, Wisconsin
October 28, 2005
“Lessons from
L. Frank Baum”
Winston F. Wong, MD
Clinical Director, Community
Benefit, Natl. Program Off.
Care Management Institute
Kaiser Permanente
Healthcare’s “Middle Space”…
An Innovation Mother Lode
2
Healthcare’s “Middle Space”…
An Innovation Mother Lode
3
CMI Networks – Distributed Learning
and Knowledge Exchange

Implementation Network
• Regionally based Physician and
Operations Oriented
Implementation Experts

Analytic Network
• Regionally based analysts with
local and national accountabilities

Regular Inter-regional calls
• Competency and Skill Focus
• Clinical Topic Focus
• Improvement Accountability to
each other and to the Program


Visits, Exchanges, Collaborations
Annual Network Retreat
4
Kaiser Permanente







America’s oldest and largest private, nonprofit,
integrated health care delivery and financing system —
Founded in 1945
Multi-specialty group practice prepayment program —
Headquartered in Oakland, CA
8.2 million members — 6.1 million members in California
Over 12,000 physicians representing all specialties and
130,000+ additional employees
Operations in 9 states and Washington, D.C. with 29
Medical Centers and 423 Clinics
KP Research Centers - $100,000,000 in external
funding in 2003 for Health Systems Research
All employees and their families are KP members
5
I’ve got a feeling
we’re not in Kansas
anymore…
An estimated 37% of Kaiser Permanente’s membership is
culturally diverse, compared to 31% for the U.S. population as a
whole.
Asian Americans
5.5%
Other
4.9%
KP Membership
Demographics
2003
Latinos
14.4%
Sources:
KP demographics -- estimates by KP National
Diversity Council based on 2003 data.;
U.S. demographics – U.S. Census Bureau
Estimates as cited in “Key Facts: Race, Ethnicity
& Medical Care,” Henry J. Kaiser Family
Foundation, 2003.
African
Americans
11.8%
Caucasians
63.4%
7
KP Priority Conditions
Clinical Area
KP Members
with this Condition
Asthma
141,000
Coronary Artery Disease 256,000
Depression
411,000
Diabetes
577,000
Heart Failure
94,000
(1 or more of the above
Cancer
Chronic Pain
Elder Care
Obesity
Self Care &
Shared Decision Making
(2.1% of members)
(3.8%)
(6.2%)
(8.7%)
(1.4%)
1,120,000 or 16.1% of members)
25,000 new cases/yr
~1,000,000 (?)
917,000
~ 25% of adults
8.2 MM
8
Additional Health Care Costs of Members
with Chronic Conditions in “CMI Portfolio”
Prevalence
Within KP
Membership
(2002)
Estimated
Members
Affected
(2005)
Annual Incremental Cost
($/member/year)
(assumes 7% cost escalation rate)
2002 estimate
$2,418
$9,811
$5,102
$4,639
$16,134
2.7.%
162,843
Asthma
3.3%
206,234
CAD
6.7%
557,712
Depression
9.3%
584,227
Diabetes
1.6%
100,839
Heart
Failure
16.1%
1,821,443
One or
More
Conditions
Total Incremental Cost of Chronic Conditions in “CMI Portfolio”
Total
Incremental
Cost
($ 2005
millions)
2005
$2,962
$12,019
$6,250
$5,683
$19,765
$482.4
$2,478.7
$3,485.8
$3,320.1
$1,993.1
$11,760.1
Source: Extrapolated from KP Northern California Division of Research estimates
9
Delivering Care…
Then…
Going Forward…
• Process and
experience oriented
• Outcome and
knowledge oriented
• Local and tribal
• National and global
• Access: to Clinicians
and Visits
• Access: to what you need,
whenever you need it
• Knowledge Management
— Paper and Recall
• Knowledge Management —
Electrons and Judgment
• Clinician treating
patients and curing
acute conditions
• Teams — including
members — managing
chronic conditions
10
Lines Between Research, Knowledge
Dissemination and Implementation
Knowledge
Dissemination
Research
Implementation
11
Lines Between Research, Knowledge
Dissemination and Implementation
Knowledge
Dissemination
Research
Implementation
12
If I only had a brain…
13
Population Management &
Levels of Care
Under the principles of population management, the first step in developing proactive strategies for the chronic conditions populations is
to define their service needs. These needs generally fall into 3 service levels. Within these 3 levels, services can further be customized, at
the point of care, to meet the needs of the individual member. Our goal is for the member to achieve and maintain self-management of their
condition (Level 1). Members who require more assistance and monitoring would be potential candidates for Level 2 or 3 programs.

Prevention is
part of every
member’s
care
Intensive or
Case
Management
Assisted Care or
Care Management
Self Care Support
LEVEL 3
Intensive or Case Management
Leverage available resources (both
Kaiser and community-based) to
optimize health status and
coordination of care.
LEVEL 2
Assisted Care or Care Management
Enhance self-care skills and abilities;
provide clinical management using
care paths and protocols.
LEVEL 1
Routine care delivered by APC
Team, as well as self-management
education, support for coping needs,
training in the use of Health-wise
Handbook, etc.
14
Level 1 Care: Achieving and Maintaining Member Self-Management
•Helps the member
achieve and maintain
improved health status
•Five separate, yet
interlocking
components:
•Inreach
•Outreach
•Education
•Psychosocial
support
Clinical Management
•Clinical
management
The components of Level 1 care
15
Asthma
Population Management Program
Asthma
Population
Management Program
Asthma
AsthmaSpecialist
Specialist
Level 3 Intensive Care
• Complex medical issues
• Psych o-social b arriers to
self-management
Level 2
• ED visits
• Hospitalization
• Beta-ago nist
over-use
• Asthma in
poor con trol
Assisted Care
EMER GENCY
Level 1
Self Care
• Asthma is well-controlled
• Member practices
effective self- care
Respiratory
Respiratory Case
CaseManag
Manager
er
•• Con
Confifirms
rmsddiag
iagnosi
nosiss
•• Id
entifie
s
co
-morbi
Id entifie s co -morbiditie
ditiess
•• Op
ti
mizes
med
icatio
Op ti mizes med icationn
reg
regimen
imen
•• Mentors
Mentors Ca
Case
se &
&
C
Care
are Man
Manage
agers
rs
•• Coa
Coaches
ches memb
members
ers in
in crisi
crisiss
•• Mana
Manages
ges acce
access
ssto
to
sp
specia
ecialty
lty && ED
ED care
care
•• Coo
rdin
ates
Coo rdin ates care
care acro
across
ss
con
ti
nuu
m
con ti nuu m
Care Management
• Proa cti ve outrea ch & tria ge
• Assessme nt
• Optimize cl ini cal man age ment,
in clud ing me dica ti on ad justmen t
acco rdin g to e sta bli shed
pro to cols
• Mon itor symptoms & p eak flow s
• Re inforce se lf-mana geme nt ski lls
• Beh avio r ch ang e motivatio n
• Patie nt re tu rns to L evel 1 ca re
wh en asthma i s stab ili zed &
pa ti ent demo nstrates selfman age ment skill s.
Week 1
Week 2
Months 2-6
Assessment
Intake Visit
Telepho ne
Follow-Up
Visits
Asthma cli nic w/
Spe cial ist & Ca re
Man age r
One-to-One
Office Vi sit
(I-to -1 Office Visi t
as ne ede d)
As needed
Primary
Primary Care
CareTeam
Team
Breathe
Breathe Easier
Easier
Class
Class
and
and others
others
Ed
Educational
ucational Resources
Resources
•• Healthwise
HealthwiseHandbook
Handbook
•• KP
KP Online
Online
•• Revi
Reviews,
ews, adj
adjusts
usts
med
medicatio
ications
ns
•• Che
Checks
cksppeak
eak flow
flowss
•• Rei
nforces
Rei nforces
se
self-mana
lf-manageme
gement
nt
16
Trends in cost ratios for members with selected chronic
conditions compared to members without those
conditions, KP Northern California Region
5
Cost Ratio
4
3
2
1
1996 1997 1998 1999 2000 2001 2002 2003
HF
CAD
Depression
Diabetes
Asthma
17
KPNC Adult Asthma
Population Trend Data
Northern California Asthma Monitoring Indicators, 1998-2003
100.0%
80.0
Inhaled
Medication
70.0
Percent of providers with AI ratio > 0.3
80.0%
60.0
70.0%
ED Visits
60.0%
50.0
50.0%
40.0
40.0%
30.0
30.0%
20.0
20.0%
Hospitalization
10.0
10.0%
0.0%

0.0
1998 Q4
1999 Q4
2000 Q4
2001 Q4
2002 Q4
2003 Q4
93.40%
51.9%
63.3%
77.6%
84.5%
90.6%
ED Visits
70.0
56.3
42.9
41.1
39.4
39.2
Hospitalization
10.4
7.5
5.1
5.4
5.3
6.4
Inhaled Medication
ED Visits or Hospitalizations per 1,000 Asthma registry members
.
90.0%
This chart illustrates trends in the monitoring
reports since 1998. The denominator for these
measures is the asthma registry. An increase
in the inhaled medication ratio
correlates well with the decrease in Asthmaspecific ED visits and hospitalizations during
this period.

A variety of factors, including program
interventions with high risk members, may be
involved in the decline in the ED visit rate.
18
Trends in Cost Ratios
The ratio of cost of care for members with asthma is compared to
members without.
Children

All costs of treating members with asthma are
higher than the costs of treating members without
asthma
Adults

Ratio of cost has remained the same 1996-2002
19
Does Care Management Save
Money?





Substantial increases in clinical process and
outcome measures have been achieved for
diabetes, heart failure, coronary artery disease,
asthma and depression
In 2003, these programs “saved” ~$600M relative
to cost trend
These programs did not produce absolute savings
– we spent more on the care of members with
diabetes, heart failure, coronary artery disease,
asthma and depression in 2003 than in 2002.
(Doing more and more things that are costeffective, but not cost saving, does not save money)
These programs continue to produce absolute
value
20
Is this all about chronic care? No!


Hawaii region’s Medicaid immunization rates were
92% in 2004, the 4th straight year over 90%
In 1999, the Medicaid immunization rate was 68%
• RNs and allied staff review medical records and databases
• Telephone outreach, then home visits
• Develop patient centered messages on the importance of
immunizations, keeping appointments, and medications

KP Hawaii Medicaid pediatric immunization rates
have exceeded commercial population rate by 3%
since 1999…most Medicaid populations are
approximately 12% lower than the commercial cohort
21
Prenatal Smoking Cessation



KP Colorado (Denver); Self reported prenatal
smoking rate: 12% among commercial patients, 25%
in Medicaid population
Smoking is the #1 preventable cause of perinatal
morbidity and mortality, mean avg. excess direct
medical cost is $511 for each prenatal pt. (live birth)
Brief cessation counseling session, followed by
directed distribution of specific self help materials
increases smoking cessation two fold: from 10% to
20%
22
If I only had a heart…
23
NOTE:
Identical row/column headings
members with ONLY that condition. The total
Co-morbidities
aredenote
Common
across all cells in the table exceeds the number of members wit h these conditions because
members with more than two chronic conditions are counted more than once.
What percentage of
all members
those with CAD
those with Depression
those with Diabetes
those with HF
Were also in this cohort?
CAD
Depression
3.2%
7.1%
11.9%
5.3%
13.6%
10.4%
46.3%
14.9%
Diabetes
7.9%
33.9%
11.6%
Heart Failure
1.6%
22.8%
3.3%
8.2%
41.6%
14
24
Co-morbidities… impact
Hospital Day Rates Among KP Members, 2001
Among KP Members
with Diabetes
and Depression
Days per 1000 members
2500
2000
Among KP Members
with Diabetes
without
Depression
1500
1000
500
Among Overall
KP Membership
Source: CMI 2002 Diabetes Outcomes Report
25
Many people fail to choose healthy
behaviors because they lack information





One study: 76% of patients with type 2 diabetes
received limited or no diabetes education
50% of patients leave the medical visit without
understanding what happened
Minority patients receive less information than white
patients
Clement, Diab Care 1995;18:1204. Roter and Hall, Annu Rev Publ
Health 1989;10:163. Stewart et al. Milbank Q 1999;77:305.
26
Many people fail to choose healthy
behaviors because they aren’t involved in
decisions



Study of 1000 physician visits, the patient did not
participate in decisions 91% of the time
Multiple studies show that when patients are involved
in decisions, health-related behavior is improved and
clinical outcomes (for example HbA1c levels) are better
than if patients are not involved
Braddock et al. JAMA 1999;282;2313. Heisler et al. J Gen Intern Med
2002;17:243. Greenfield et al. J Gen Intern Med 1988;3:448. Golin et al. Diab Care
1996;19:1153. Piette et al. J Gen Intern Med 2003;18:624. Roter. Health Educ
Monographs 1997;5:281
27
A Partnership with Measurable Outcomes
A 2002 study of results
at the Pediatric Asthma
Clinic of San Francisco
General Hospital, a
demonstration site for
the “Yes We Can”
clinical model, showed
changes
28
High Utilizing Populations
High Utilizing Populations breakdown into 4 buckets:
• Frail Elderly – many diseases, many drugs, support issues, costs
issues (Medicare caps), End of Life issues, different trajectories
•Substance Abuse – Alcohol and Drugs, drug seeking behavior for
prescription drugs
• Psychiatric and Complex Mental Health issues (often mixed with
Substance abuse and chronic pain)
• Chronic Pain – pain medication issues
We need programs other than traditional medical model for acute
and episodic care – CDRP, Chronic Pain, Outpatient Psych
programs, Geriatric programs, Case management (KFH and CCC
programs)
IOM report of 1/03 lists Care Coordination as one of top health care
issues
29
How to get a Population Under Control
Traditional: Target providers and system:
Feedback, reminders, reports, guidelines, champions,
academic detailing, incentives, list management
Provider gives the right med to the right patient:
Patient takes it 50% of the time
Provider gives the right self-management behavior change
message (i.e. – you need to exercise, stop smoking ,
and lose weight)
Patient does this 10% of the time and it will probably
not be sustained
It’s about adherence and concordance – how to help
patient’s to succeed and sustain change not about
creating dependence
30
Strengthening Member Self-Management of Chronic Conditions
Five questions critical to strengthening self-management
practices:
1.
What essential information, beliefs and behaviors do
members need to effectively self-manage their chronic
condition(s)?
2.
3.
4.
5.
What are the key elements and strategies to use in chronic
condition self-management interventions, regardless of type
of condition?
What are effective ways to structure the delivery of chronic
condition self-management interventions in order to maximize
member enrollment?
What are effective approaches to strengthen chronic
condition self-management during the outpatient clinical
encounter?
What are effective approaches to increase adherence to
prescription medication regimens of patients with chronic
conditions?
31
Associating High Performance with
Operational Practices- Examples
Glycemic Screening x Action Plans
90%
85%
80%
75%
70%
0
2
4
6
8
10
Performance (Adjusted) - Eye Exams
Performance (Adjusted) Glycemic Control
Eye Exams x AMR
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
Practice Score - Action Plans
0%
0
2
4
6
8
10
Practice Score - Automated Medical Record
Performance values shown are adjusted for all
other Practices, based on model estimates
32
Practices included in the analysis

Organizational Support
•
•
•
•
•
•
•


•
•
•
•
•
•
•
•
Leadership
Accountability
Champions
Resources
Provider Feedback
Financial Incentives
Program Evaluation
Self-Management
•
•
•
Action Plans
Patient Education
Integration with Care
Delivery System Design

Stratified Services
Risk Stratification
Registry
Outreach and Follow-Up
Inreach
Care Coordination
Team-Based Care
Cultural Competence
Decision Support
•
•
•
Guideline Distribution
and Training
Provider Alerts
Clinical Information
System
33
Associating High Performance
with Operational Practices

Practices most associated with high performance
•
•
•
•
•

KP
HealthConnect
Practices sometimes associated with performance,
but with less strength and/or consistency
•
•
•
34
Patient action plans
Provider financial incentives
Automated medical record
Outreach and follow-up
Provider alerts and Reminders
Registry
Guideline distribution & training
Care coordination
34
Stronger implementation was associated
with significant performance improvement
Average Performance of Locations in Lowest and Highest Quartile of Practice
Implementation, 8 Diabetes Performance Measures Pooled, 2001-2002
Performance on All Measures,
as Percentile
80
67
70
61
60
60
50
55
45
44
60
52
45
46
38
40
37
30
20
10
0
Financial
Incentives
Action Plans
Outreach and
Follow-up
Locations in Lowest Quartile
Provider Alerts
Automated
and Reminders Medical Record
All Practices
(Model)
Locations in Highest Quartile
35
The major findings:
By comparing the level of implementation of diabetes care
practices with eight diabetes performance measures, we
identified five practices that were associated with better
performance:
•
•
•
•
•
Financial incentives
Action plans (patient-specific or personal)
Automated medical record
Outreach and follow-up
Provider alerts and reminders
36
If I only had courage. . .
37
Quality assertions …
“Poor patients don’t deserve poor care”
 Same care does not mean same
outcomes
 Quality outcomes are achieved in years,
not months
 Not what you do, but what you
accomplish
 Medicaid is about care, not payment

38
Courage to confront challenges



Faced with unprecedented financial challenges, can
we implement innovative, population management
approaches to improve outcomes for Medicaid
populations?
Can we develop incentives for patients, providers and
plans that result in improved clinical outcomes?
Can we demonstrate models of care that address the
diverse cultural, linguistic, and literacy characteristics
of Medicaid populations?
39
Healthcare’s “Middle Space”…
An Innovation Mother Lode
40
We always had the answers
41
…we just didn’t
know they were in
our own backyard.
Thank you for your leadership!
Contact: [email protected]
43