Transcript Implementation of EBP: Who, What, When, Where & How?

Implementation: A
Practical Approach
Mark L. Willenbring, MD
NIAAA/National Institutes of Health
ASAM 2006, San Diego, CA
Case
Manager
OR…
Scientific Inquiry
• Level I: Consensus of experts
• Level II: Small clinical trials
• Level III: Large randomized
controlled trials (RCTs)
• Meta-analysis
Guidelines are more likely to be valid
if they:
• are developed using systematic
reviews
• use national or regional guideline
development groups (including
representatives of key disciplines)
• note explicit links between
recommendations and scientific
evidence.
» Grimshaw J, Eccles M. et al. “Developing clinically valid
QUERI
Inclusion Criteria:
30-50%
1. Has specified disorder
2. Willing to participate
3. Has home, phone, &
transportation
4. Fine upstanding citizen
Exclusion Criteria:
1. Coexisting psychopathology (incl. addictions)
50-70%
2. Multiple prior treatments
3. Serious medical problems
4. Lack of housing, transportation
5. Unmotivated
6. Too busy surviving to bother
7. Taking other medications
8. History of non-compliance
9. Lives too far away
10. Serious personality disorder
Dodo bird:
Friend or enemy?
Looking in the wrong
place?
Percent Days Abstinent
Tx 1
90
80
70
60
50
40
30
20
10
0
Tx 2
Examined change
Unexamined change
0
4
8
Weeks
12
Results remarkably
similar
90
Percent Days Abstinent
80
70
Disulf
RREP
MATCH OP
MATCH AC
MEAN
60
50
40
30
20
10
0
Baseline
12 mo
Miller et al., J Stud Alc 62:211-220, 2001
Project MATCH Main
Outcomes
Immediate
substantial
improvement
Increased
drinking
quantity
before entry
Project MATCH Research Group, J Stud Alc 59:631-639, 1998
Treatment: instigator or result of
change?
Does change occur
prior to tx?
PDA
0
4
8
week
12
etc.
Compared 5 different treatment approaches
for cannabis use disorder in adolescents
No difference across groups
Dennis et al., JSAT 2004
UKATT Trial
• Compared 4 sessions of
MET with 8 sessions of
social and behavior therapy
UKATT findings
50
40
30
PDA
20
DDD
10
0
Baseline
3 mo
12 mo
No difference across groups
Dodo bird strikes again!
A broader view of
change
• Therapeutic techniques
may have small effects
relative to extrinsic
factors
A broader view of
change
• Common mechanisms may
account for change
– Therapeutic alliance and
Empathic listening
– Social support for change
– Therapeutic rituals (cf. Jerome
Frank)
– Help-seeking itself
Implications for
research
There may be multiple
pathways (or mechanisms
of action) with similar
outcomes,
OR
We may have been looking
in the wrong place
There’s many a
slip between the
cup and the lip.
• Implemented CBT (high and low
standardization) in community
program
• Compared to Treatment as Usual
Whoops!
Morgenstern et al., JCCP 2001
Motivational interviewing to improve
treatment engagement and outcome
in individuals seeking treatment for
substance abuse: A multisite
effectiveness study
Kathleen M. Carroll, Samuel A. Ball, Charla Nicha, Steve Martinoa,
Tami L. Frankfortera, Christiane Farentinosb, Lynn E. Kunkelc,
Susan K. Mikulich-Gilbertsond, Jon Morgensterne, Jeanne L.
Obertf, Doug Polcing, Ned Sneadh, George E. Woodyi and for the
National Institute on Drug Abuse Clinical Trials Network
Drug and Alcohol Dependence Volume 81, Issue 3, 28 February 2006,
Pages 301-312
Effectiveness of MI
• Standard intake vs. MI
• Achieved desired differences in
technique
• Improved engagement
• No difference in 28 day or 84 day
substance use outcomes
Carroll et al., Drug and Alcohol Dependence, 2006
Take Home Message
• Multiple treatments have been
shown to produce acceptable
outcomes
• Caution is required
– Interpretation of empirical findings
– Implementation
• Careful attention to implementation
and evaluation is necessary
• Focus on outcomes
73% of primary care physicians
believe that they immunize a greater
percentage of patients than their
colleagues
Noe CA, Markson LJ, Prev Med 1998;27(6):767-72
In a closed panel academic medical
center:
• 65% of the practitioners are more
efficient than their peers in the same
practice
• 65% have sicker patients
• 75% have better outcome
J Perlin, 2000
. . . and all of the children in Lake
Wobegone are above average
J Perlin, 2000
Rates of Coronary
Angiography
HI
LO
Rates of Coronary
Angiography
HI
LO
Rates of Radical
Prostatectomy
HI
LO
Rates of Radical
Prostatectomy
HI
LO
Variation in Inpatient Care
for COPD by VISN (Region)
Hospital Days / 1,000 COPD Patients
25
20
15
10
5
0
Variation in Ambulatory Care
for MDD by VISN (Region)
MDD Clinic Visits / Yr
20
15
10
5
0
Reasonable Expectation
Unfortunate Experience
after DA Burnett, ©UHC, 1995
What you would want for
your family
Should it matter where
you receive care?
What you would not want
after DA Burnett, ©UHC, 1995
How do we capture and
systematize these practices?
How do we drive these practices
to be more like those above?
after DA Burnett, ©UHC, 1995
Implementing Guidelines
Requires Broad Skills
Leadership and teamwork
Negotiation and tact
“Selling” ideas to peers
Link evidence to practice guideline
Educational Strategies
Ineffective:
– Traditional CME
– Printed materials
– ?Audit and feedback
VERDICT Brief, Spring 1998
Educational Strategies
Effective: Intensive conferencing w/
interaction, discussion, & role
playing
VERDICT Brief, Spring 1998
Reminders and
Prompts
Pocket cards & wall charts
Formatted records
Computerized reminders
VERDICT Brief, Spring 1998
Interactive Strategies
Opinion leaders
Academic detailing (outreach)
VERDICT Brief, Spring 1998
Multiple strategies work best.
Organizational
strategies
• “None of the strategies produced
consistent results.”
• Professional performance was
improved by revision of
professional roles and computer
systems
Wenning et al., 2006
Organizational
strategies
• Patient outcomes improved by
multidisciplinary teams, integrated
care services, and computer
systems.
• Cost savings [came from]
integrated care services
Wenning et al., 2006
So what should a program
[leader/provider] do?
•
•
•
•
Technique
Outcome measurement
Quality improvement
Implementation strategies
Implications for
implementation
• Examine current system/outcomes
• Determine where you wish to
improve
• Develop model for how your
program improvements will
translate into process and/or
outcome improvements
Is treatment as usual
ok?
• Examine general quality of care
– Professionalism
– Empathy
– Training and skill
– Supervision
• Determine whether implementing
a new technique will be helpful,
why, and how
Implementation
• Buy in by leadership critical
• Resources must be provided
• Multiple strategies
Implementation
• Monitoring performance of
individuals not just systems
• Supervision
• Plan for improving
performance
Summary
• Research on mechanisms of
change and improvement in
outcomes in the community is
needed
• Until then, paying attention to
monitoring performance and
upgrading the skills of staff are
reasonable strategies
Summary
• Community-research partnerships
hold promise for gradually learning
about how to improve overall
community outcomes
• “Make haste slowly.”
– Caesar Augustus