Prevention-Research Centers Health Aging Research Network (PRC-HAN)
Webinar Series
Evidence-Based Depression
Care Management:
Improving Mood-Promoting
Access to Collaborative
Treatment (IMPACT)
Tuesday, October 16th 2008
2-3:30 PM EST
Moderated by: Cate Clegg
Jürgen Unützer, MD, MPH, MA
Virna Little, PsyD, LCSW-R
Sponsors:
Prevention Research CentersHealthy Aging Research Network
http://www.prc-han.org/
Retirement Research Foundation
http://www.rrf.org/
National Council on Aging
http://ncoa.org/index.cfm
2
IMPACT
Primary Care Based
Team Care for
Late-Life Depression
Jürgen Unützer, MD, MPH, MA
Professor & Vice Chair
Psychiatry & Behavioral Sciences
University of Washington
Virna Little, PsyD, LCSW-R
Vice President for Psychosocial Services and
Community Affairs
Institute for Family Health
3
Depression
Common
10% in primary care
Disabling
#2 cause of disability (WHO)
Expensive
50-100% higher health care costs
Deadly
Over 30,000 suicides / year
4
Depression is deadly
Older men have the highest rate of suicide.
5
Depression is often not
the only health problem
Cancer
Chronic
Pain
10-20%
40-60%
Depression
Geriatric
Syndromes
20-40%
Heart
Disease
20-40%
Neurologic
Disorders
10-20%
Diabetes
10-20%
6
Efficacious treatments
for depression
Antidepressant Medications
– Over 20 FDA approved
Psychotherapy
– CBT, IPT, PST, brief dynamic, etc.
Other somatic treatments
– ECT
Physical activity / exercise
Unutzer et al, NEJM 2008.
7
But: few older adults
get effective treatment
Only half are ‘recognized’
a particular problem for older men & minorities
– “I didn’t know what hit me …”
– “I am not crazy”
– “Isn’t depression just a part of ‘normal aging?”
Fewer than 10 % seek care from a mental
health specialist. Most prefer their primary care
physician.
8
Depression Treatment
in Primary Care
50 % are recognized and started on treatment or
referred
Limited access to evidence-based psychosocial
treatments (psychotherapy)
Increasing use of antidepressants
•
•
•
PCPs prescribe 70 – 90 % of antidepressants
10 - 30 % of older adults are on antidepressants
MAJOR OPPORTUNITIES for Quality Improvement – even for
nonprescribing providers
But treatment is often not effective
– Only 20 – 40 % improve substantially over 12
months
9
Why integrate care?
Home &
Community based
social services?
Primary
Care
PC
Alcohol &
substance
abuse care?
CM
HC
Community
Mental Health
Center
10
Depression Care Management
in Primary Care
Limited access to / use of
mental health specialists
Treat mental health disorders
where the patients are
- Established
provider-patient
relationship
- Less stigma
- Better coordination with
medical care
11
Components of evidence
based integrated care programs
Screening / case finding
Patient education / self-management support
Support medication treatment prescribed in primary care
–
Monitor adherence, side effects, effectiveness
[Nonprescribing providers function as the ‘eyes and ears of the doctor’]
Proactive outcome measurement / tracking
–
e.g., PHQ-9, GDS, CES-D
Brief counseling (e.g., Behavioral Activation, PST-PC, IPT, CBT)
Stepped care (initial treatments often are not enough)
–
–
increase treatment intensity as needed
mental health consultation to help guide or provide care for patients not
responding as expected
12
IMPACT Study
Funded by
John A. Hartford Foundation
California Healthcare Foundation
Robert Wood Johnson Foundation
Hogg Foundation
13
IMPACT Team
“None of us is as smart as all of us”
Study coordinating center
Jürgen Unützer (PI), Sabine Oishi, Diane Powers, Michael Schoenbaum, Tom Belin,
Linqui Tang, Ian Cook. PST-PC experts: Patricia Arean, Mark Hegel
Study sites
University of Washington / Group Health Cooperative
Wayne Katon (PI), Elizabeth Lin (Co-PI), Paul Ciechanowski
Duke University
Linda Harpole (PI), Eugene Oddone (Co-PI), David Steffens
Kaiser Permanente, Southern CA (La Mesa, CA)
Richard Della Penna (Co-PI), Lydia Grypma (Co-PI), Mark Zweifach, MD,
Rita Haverkamp, RN, MSN, CNS
Indiana University
Christopher Callahan (PI), Kurt. Kroenke, Hugh. Hendrie (Co-PI)
UT Health Sciences Center at San Antonio
John Williams (PI), Polly Hitchcock-Noel (Co-PI), Jason Worchel
Kaiser Permanente, Northern CA
Enid Hunkeler (PI), Patricia Arean (Co-PI)
Desert Medical Group
Marc Hoffing (PI); Stuart Levine (Co-PI)
Study advisory board
Lisa Goodale (NDMDA), Rick Birkel (NAMI), Thomas Oxman, Kenneth Wells,
Cathy Sherbourne, Lisa Rubenstein, Howard Goldman
14
Study Methods
1998 – 2003
Randomized controlled trial
8 health care organizations in 5 states
– 18 primary care clinics
1,801 older adults with major depression or chronic
depression
– 450 primary care providers
– Patients randomly assigned to IMPACT or usual care
– Usual care = antidepressant Rx in primary care (~ 70
%) and / or referral to mental health specialists (20 %)
– All followed with independent assessments for 2 years
15
IMPACT Team Care Model
Prepared, Pro-active Practice Team
Informed, Activated Patient
Effective
Collaboration
Photo credit: J. Lott, Seattle Times
Photo: Courtesy D. Battershall & John A. Hartford
Foundation
Practice Support
16
Evidence-based
‘team care’ for depression
TWO NEW ‘TEAM MEMBERS’
TWO PROCESSES
Care Manager
1. Systematic diagnosis and
outcomes tracking
- Patient education / self
management support
e.g., PHQ-9 to facilitate diagnosis
and track depression outcomes
- Close follow-up to
make sure pts don’t ‘fall
through the cracks’
2. Stepped Care
- Support antidepressant Rx by PCP
a) Change treatment according
to evidence-based algorithm if
patient is not improving
- Brief counseling
(behavioral activation,
PST-PC, CBT, IPT)
b) Relapse prevention once
patient is improved
- Facilitate treatment
change / referral to
mental health
Consulting
Psychiatrist
- Caseload consultation for
care manager and PCP
(population-based)
- Diagnostic consultation on
difficult cases
- Consultation focused on
patients not improving as
expected
- Recommendations for
additional treatment / referral
according to evidence-based
guidelines
17
- Relapse prevention
Treatment Protocol
Assessment and education,
Behavioral Activation / Pleasant Events Scheduling
AND
(3) a) Antidepressant medication
usually an SSRI or other newer antidepressant
OR
b) Problem Solving Treatment in Primary Care
(PST-PC)
6-8 individual sessions followed by monthly group
maintenance sessions
(4) Maintenance and Relapse Prevention Plan for patients in
remission
18
Stepped Care
Systematic follow-up & outcomes tracking
Patient Health Questionnaire (PHQ-9)
The “cheap suit”
Treatment adjustment as needed
- based on clinical outcomes
- according to evidence-based algorithm
- in consultation with team psychiatrist
Relapse prevention
19
20
Greater Satisfaction with
Depression Care
(% Excellent, Very Good)
Usual Care
100
P=.2375
Intervention
P<.0001
P<.0001
percent
80
60
40
20
0
0
3
month
12
21
Unützer et al. JAMA. 2002; 288: 2836-2845.
IMPACT Doubles Effectiveness
of Depression Care
50 % or greater improvement in depression at 12 months
Usual Care
70
IMPACT
60
50
%
40
30
20
10
0
1
2
3
4
5
6
7
8
22
Participating Organizations
Evidence-based Care Benefits
Disadvantaged Populations
50 % or greater improvement in depression at 12 months
60%
50%
54%
43%
42%
IMPACT Care
40%
30%
23%
Care as Usual
19%
14%
20%
10%
0%
White
Black
Latino
Areán et al. Medical Care, 2005
23
Improved Physical
Functioning
SF-12 Physical Function Component Summary Score (PCS-12)
P<0.01
41
P<0.01
P<0.01
P=0.35
40.5
40
Usual Care
IMPACT
39.5
39
38.5
38
Baseline
3 mos
6 mos
12 mos
24
Callahan et al. JAGS. 2005; 53:367-373.
Fewer thoughts of suicide
18
% patients with suicidal
thoughts
16
14
12
10
IMPACT
Usual Care
8
6
4
2
0
Baseline
6 months
12 months
25
Unützer et al, JAGS 2006
IMPACT Saves Money
Intervention
group cost
in $
Usual care
group cost in
$
Difference in
$
522
0
522
661
558
767
-210
7,284
6,942
7,636
-694
Other outpatient costs
14,306
14,160
14,456
-296
Inpatient medical costs
8,452
7,179
9,757
-2578
Inpatient mental health /
substance abuse costs
114
61
169
-108
31,082
29,422
32,785
-$3363
Cost Category
4-year
costs in
$
IMPACT program cost
Outpatient mental health costs
Pharmacy costs
Total health care cost
Savings
26
Unutzer et al. Am J Managed Care 2008.
IMPACT Summary
- Less depression
IMPACT doubles effectiveness of
usual care
- Less physical pain
- Better functioning
- Higher quality of life
- Greater patient and
provider satisfaction
- More cost-effective
Photo credit: J. Lott, Seattle Times
“I got my life back”
27
IMPACT Endorsements
–President’s New Freedom
Commission on Mental
Health
–National Business Group
on Health
–Institute of Medicine
(Retooling for An Aging America)
–POGOe
–CDC Consensus Panel
–Annapolis Coalition
–Partnership to Fight
Chronic Disease
–SAMHSA NREPP
–Commonwealth Fund
–Integrated Behavioral
Health Partnership
28
Taking IMPACT
from Research to Practice
Support from JAHF (2004-2009)
Over 170 clinics have implemented core
components of the program to date
– DIAMOND program in Minnesota implementing
the program state-wide in partnership with 25
medical groups and 9 health plans
Several large health plans and disease
management organizations are
incorporating core components of IMPACT
29
IMPACT Implementation
Trained over 3000
Over 3,000 clinicians
Providers intrained
over 150
practices to date
2004
2005
2006
2007
30
2008
Kaiser Permanente of
Southern California
Pilot Study
- Compare 284 clients in ‘adapted program’ with 140
usual care patients and 140 intervention patients in
the IMPACT study (Grypma et al, 2006)
Dissemination
- Implemented core components of program in 10
regional medical centers
31
KPSC – San Diego
‘After IMPACT’
Fewer care manager contacts
IMPACT Study
Post-Study
18.9
10.2
8.7
7.9
5.1
2.8
Total contacts
Clinic visits
Phone calls
32
Grypma et al, General Hospital Psychiatry, 2006.
IMPACT Remains Effective
>= 50 % drop in PHQ-9 depression scores
66%
68%
64%
At 3 months
68%
At 6 months
IMPACT
Post-Study
33
Grypma et al, General Hospital Psychiatry, 2006.
Lower Total
Health Care Costs
$8,800
$8,400
$ / year
$8,588
Study Usual
Care
$8,000
Study
IMPACT
$7,949
$7,600
$7,200
$7,471
Post Study
IMPACT
$6,800
34
Grypma, et al; General Hospital Psychiatry, 2006
Institute for
Urban Family Health
Number
Age at enrollment:
Percent
Mean
Range
71.6 years
60 – 99 years
Female
Male
165
74
69.0%
31.0%
90
70
56
23
37.7%
29.3%
23.4%
9.6%
44
48
47.8%
52.2%
Gender:
Ethnicity:
Hispanic
African American
Caucasian
Other
Marital Status:
Married
Single, Widowed,
Divorced/separated
35
IMPACT Effective
for Depression
Mean PHQ-9 Depression Scores
20
18
Mean Depression Scores
16
14.03
14
12
10
8.14
7.91
8
6
4
2
Initial
3 Months
6 months
0
Time
36
Change in Depression
Initial to 6 months
Initial PHQ-9 Depression Scores
6 Month PHQ-9 Depression Scores
(Mean Score of 7.91)
160
160
63%
140
120
100
80
60
28%
40
20
0
9%
Number of Patients
Number of Patients
140
120
100
80
40
20
0
Under 10: 10-14:
15-19:
Mild
Moderate Mod.
Severe
PHQ-9 Score
20+:
Severe
65%
60
24%
Under 10: 10-14:
Mild
Moderate
5%
15-19:
Mod.
Severe
6%
20+:
Severe
PHQ-9 Sore
37
A word from providers…
“It is good to see that mental health is
once again becoming part of the medical
Interview, as so much of our patient's
health depends on their mental well being.”
- Dr. Eric Gayle
“Project IMPACT has allowed me to incorporate
a new tool (PHQ-9)into my primary care practice,
which has improved the accuracy of my diagnosis
while increasing my efficiency and productivity as well.
It helped me identify patients I initially overlooked.”
38
-Dr. Joseph Lurio (68th Street)
Depression Is Associated With a Higher
> 3 Cardiac Risk Factors (%)
Number of Cardiac Risk Factors
100
90
80
70
60
50
40
30
20
10
0
Non Depressed
Depressed
62.5
38.4
61.3
35
Diabetic Patients With CVD Diabetic Patients Without CVD
N=3010
N=1215
Katon et al, J Gen Intern Med, 2004
Depression Increases Mortality Rate
in Patients With Diabetes by 2-Fold
Katon et al. Diabetes Care, 2005
Depression and Diabetes: More
Depression Free Days over 2 Years
500
300
200
100
359
Increment
Days
400
331
215.5
115.5
53
0
Pathways
Increment
412
IMPACT
Intervention
Usual Care
Increment
Two Collaborative Care Trials Demonstrate
Improved Depression Care in Diabetes
Lowers Total Health Care Costs Over 2 Years
$22,258
$21,148
$18,932
$20,000
Savings
$1,110
Usual Care
$5,000
Savings
$10,000
Usual Care
$15,000
Intervention
$18,035
Intervention
$25,000
$897
$0
Pathways
IMPACT
Katon et al. Diabetes Care 2006, Simon et al Arch Gen Psychiatry 2007
Project Dulce + IMPACT
Principal Investigator: Todd Gilmer, UCSD
Combined diabetes and depression care
management program targeting low-income and
primarily Spanish speaking Latinos in San Diego
community clinics
Added a depression care manager to an existing
diabetes team (RN/CDE, promotoras)
Translation for Cultural Competency
– DCM bilingual with experience serving Latino
pop.
– PST-PC adapted to low-literacy population
43
Project Dulce + IMPACT
Results
Screened 499 patients with PHQ9
31% with scores of 10+
75% Latino, 70% Spanish speaking
65% had depressive symptoms for 2+ years
26% interested in pharmacological treatment
74% interested in psychological treatment
48% reported financial stressors
44
Depressive Symptoms at Baseline and SixMonth Follow-Up As Measure with PHQ-9
.
Inter-Quartile Range (box)
Highest and Lowest (whiskers)
Outlier (dots)
Median
45
Gilmer et al. Diabetes Care 2008
Collaborative Care
for Alzheimer’s Disease
Collaborative Care for Alzheimer’s Disease
Christopher M. Callahan, MD
Cornelius and Yvonne Pettinga Professor
Director, Indiana University Center for Aging Research
Research Scientist, Regenstrief Institute, Inc.
Improvement in Dementia-related
Problem Behaviors
Patient NPI Score
20
15
10
5
0
baseline
6 months
12 months
Augmented Usual Care
IU Center for Aging
Research
18 months
Intervention
Callahan et al. JAMA 2006
Improvement in Caregiver Stress
Caregiver NPI Score
10
5
0
baseline
6 months
12 months
Augmented Usual Care
IU Center for Aging
Research
18 months
Intervention
Callahan et al. JAMA 2006
Implementing
Collaborative Care
Shared vision
– How will we know success?
– Shared, measurable outcomes
• (e.g., # and % of population screened, treated, improved)
Engaged leaders & stakeholders
– Clinic leaders & administration
– PCPs, care managers, psychiatry, other mental health providers
Clinical & operational integration
– Functioning teams, communication, and handoffs
– Clear about ‘shared workflow’ & roles of various team members
Adequate resources
• Personnel, IT support, funding
Proactive problem solving re barriers & competing demands
• Minimize complexity, PDCA
49
http://impact-uw.org
50