Transcript Effectiveness and mediators of 12

A professional perspective on
mutual help organisations for
addictions
Scotland’s Futures Forum, Edinburgh, Scotland UK
23 March 2009
Keith Humphreys
Professor of Psychiatry
Veterans Affairs and Stanford University Medical Centers
Palo Alto, California USA
Acknowledgement
• U.S. National Institute of Alcohol Abuse and
Alcoholism
• U.S. Department of Veterans Affairs
• Rudy Moos, Christine Timko, John Finney
• Research participants
Overview
•
•
•
•
Definition of mutual help groups
Effectiveness and Cost-effectiveness
Mediators of Change
So What?: Implications for Designing
Better Services
Characteristics Shared by All SelfHelp/Mutual Aid Groups
• Members share some problem or status that
results in suffering/distress
• Groups are organized and facilitated by
members themselves
• Experiential knowledge is the basis of
expertise
• All members are both “helpers” and
“helpees”
• No fees are charged, save “pass the hat”
contributions
Characteristics of Only Some SelfHelp/Mutual Aid Groups
• A codified world view/program for change
• Residential setting
• Internet presence
• Connection to a larger organization
• Political Advocacy
Note: Based on MIDUS survey (N=3032)
Source: Kessler, R.C. et al., 1997, Patterns and correlates of self-help group
membership in the United States. Social Policy, 27, 27-46.
Any
Other
Family
Physical
Illness/Disease
Emotional
Crisis
Eating
Problems
20
18
16
14
12
10
8
6
4
2
0
Substance
Problems
Lifetime and past 12 months participation
in self-help groups, 1995
Past 12 mos.
Lifetime
Help-seeking visits in U.S. for psychiatric and
substance abuse problems by sector
40.1%
Self-help
8.1%
General Medical
Human Services
16.5%
Mental Health
Specialty
Source: Kessler, R.C. et al.
(1997). Differences in the use if
psychiatric outpatient services
between the U.S. and Ontario.
NEJM. 336. 551-557.
35.3%
Alcohol and drug-related self-help/mutual
aid organizations (12 step in red)
Estimated Number of Groups Worldwide
Alcoholics Anonymous
Al-Anon
Narcotics Anonymous
Cocaine Anonymous
Adult Children of Alcoholics
LifeRing/Secular Organization for Sobriety
Marijuana Anonymous
Women for Sobriety
SMART Recovery
Moderation Management
95,000
32,000
21,000
2,000
1,500
1,200
1,000
350
300++
50
Source: White and Madara (1998). Self-help sourcebook. Denville, NJ: American Self-help
clearinghouse and Humphreys, K. (2004) Circles of Recovery.
12-step groups have established themselves
in the once-impenetrable Middle East
Note: NA is for all drugs not just narcotics
Selected data on clinical and
cost-effectiveness of 12-step
mutual help organizations
Clinical trial of Oxford House
• Oxford House is a 12-step influenced, peermanaged residential setting
• 150 Patients randomized after inpatient
treatment to Oxford House or TAU
• 77% African American; 62% Female
• Follow-ups every 6 months for 2 years,
90% of subjects re-contacted
At 24-months, Oxford House (OH)
produced 1.5 to 2 times better outcomes
80
70
60
50
OH
TAU
40
30
20
10
0
Abstinent
Employed
Incarc
Jason et al. (2006). Communal housing settings enhance substance abuse recovery. American J Public Health, 96, 1727-1729.
Veterans Affairs RCT on AA/NA
referral for outpatients
• 345 VA outpatients randomized to standard
or intensive 12-step group referral
• 81.4% FU at 6 months
• Higher rates of 12-step involvement in
intensive condition
• Over 60% greater improvement in ASI
alcohol and drug composite scores in
intensive referral condition
Source: Timko, C. (2006). Intensive referral to 12-step self-help groups and 6-month substance use disorder
outcomes. Addiction, 101, 678-688.
Study of cost consequences
Comparable baseline demographic and alcohol use
characteristics of 201 alcoholic individuals who initially
chose AA or outpatient treatment
Characteristic
Total sample AA
(n=201)
(n=135)
%
%
Outpatient
(n=66)
%
Caucasian Race
Female
Married
Employed
86.6
49.3
25.9
52.2
88.9
54.1
23.7
48.1
81.8
39.4
30.3
60.6
Age (years)
ETOH (oz)
ADS Score
Days intox
Mean
35.3
11.5
10.1
11.7
Mean
34.7
12.3
10.9
11.9
Mean
36.4
10.0
8.6
11.4
Total alcohol-related health care costs over three years by
comparable alcoholic individuals who initially chose
Alcoholics Anonymous or professional outpatient treatment
AA group
(n=135)
mean SD
Outpatient group
(n=66)
F
mean SD (df=1,199)
$1,115 $2,386
$1,136 $4,062
$2,251 $5,075
$3,129 $4,355
$948 $2,852
$4,077 $5, 371
Per person costs
Year 1
Years 2 and 3
Total
Note *p<.05
5.52*
Alcohol-related outcomes of individuals initially
selecting AA or outpatient treatment (OP)
14
12
AA
AA OP
AA
OP
OP
10
8
6
Baseline
1 Year
3 Year
4
2
0
Ozs. of Ethanol
per day
Days
Intoxicated in
past month
Alcohol Dependence
Symptoms
Source: Humphreys, K., Moos, R.H. (1995). Reduced substance abuse related health-care costs among voluntary participants in
Alcoholics Anonymous. Psychiatric Services, 47, 709-713.
Replication of cost offset
findings in Department of
Veterans Affairs Sample
Source: This study appeared in Alcoholism:
Clinical and Experimental Research, 25, 711-716.
Design
• Follow-up study of over 1700 patients
receiving one of two types of care:
• 5 programs were based on 12-step
principles and placed heavy emphasis on
self-help activities
• 5 programs were based on cognitivebehavioral principles and placed little
emphasis on self-help activities
Participants
• N=
1,774
• Sex = 100% Male
• Race= 49% Caucasian, 46% AfricanAmerican
• SES=
• Age=
• Dx=
100% low-income
42.9 Years (Sd = 9.8)
16% Drug, 46% Alc, 38% Both
Self-help group participation at 1year follow-up was higher after selfhelp oriented treatment
• 36% of 12-step program patients had a
sponsor, over double the rate of cognitivebehavioral program patients
• 60% of 12-step program patients were
attending self-help groups, compared with
slightly less than half of cognitive-behavioral
program patients
1-Year Treatment Costs, Inpatient
Days and Outpatient visits
$1000 cost
Cog-Beh
12-step
IP Days
OP Visits
0
5
10
15
Note: All differences significant at p <.001
20
25
1-Year Clinical Outcomes (%)
90
80
70
60
50
12-step
Cog-Beh
40
30
20
10
0
Abstinent
No SA Prob
Note: Abstinence higher in 12-step, p< .001
Pos MH
2-year follow-up of same sample
• 50% to 100% higher self-help group
involvement measures favoring 12-step
• Abstinence difference increased: 49.5% in
12-step versus 37.0% in CB
• A further $2,440 health care cost reduction
(total for two years = $8,175 in 2006USD)
What mediates these benefits?
B “mediates” the relationship
between A and C
A>>>>>>>>>B>>>>>>>>C
Structural equation modeling results from over 2,000
patients assessed at intake, 1-year, 2-year
Active Coping
Self-Help
Group
Involvement
Motivation to
change
General
Friendship Quality
Friends’ Support
For Abstinence
Note
All paths significant at p<.05. Goodness of Fit Index = .950.
Reduced
Substance
Use
Partial mediators of 12-step groups’ effect
on substance use identified in research
•
•
•
•
•
•
Increased self-efficacy
Strengthened commitment to abstinence
More active coping
Enhanced social support
Greater spiritual and altruistic behavior
Replacement of substance-using friends
with abstinent friends
12-step vs. non-12 step based
friendship networks of 1,932 treated
SUD patients
100
90
80
70
60
50
40
30
20
10
0
12-step
Non
%friends in 12S
%Abstainers
Source: Humphreys, K., & Noke, J. (1997). The influence of posttreatment mutual help group participation on the
friendship networks of substance abuse patients. American J of Community Psychology, 25, 1-16.
So what?:Clinical implications
12-step group involvement of 2,045 substance-dependent
veterans after 12-step or cognitive-behavioral treatment
100
75
66%
60%
%
50
58%
48%
45%
48%
36%
Cog Beh
25
18%
12-Step
0
attended meetings
had sponsor
read 12-step
literature
had a friend who
Attends AA/NA
12-step self-help group involvement
Note: Involvement was measured one year after discharge by patient reports of activities in the past 3 months. Data in this
table were drawn from Humphreys et al. (1999), Alcoholism: Clinical and Experimental Research, 23, 558-563.
Abstinence from illicit drugs and alcohol as a function of
self-help involvement and treatment type in 3,018 patients
Self-help group
Involvement
-1SD +1SD
Proportional
Difference
19%
75%
1:4.0
Non-12-step 25%
65%
1:2.6
Treatment
Orientation
12-step
How can referrals to self-help groups
be more effective?
Sample:
20 alcohol outpatients
Design:
Outpatients randomly assigned to standard 12-step
self-help group referral (list of meetings and therapist
encouragement to attend) or intensive referral (in-session
phone call to active 12-step group member)
Results:
Attendance rate after intensive referral: 100%
Attendance rate after standard referral: 0%
Source:
Sisson, P.W., & Mallams, J.H. (1981). The use of systematic encouragement and community
access procedures to increase attendance at AA meetings. Am J Drug Alc Abuse, 8, 371-376.
Self-help referral can be beneficial
in non-specialty settings
Control
BI
BI+Peer
6-month abstinence
36%
51%
64%
TX/AA Initiation
9%
15%
49%
Source: Study by Rick Blondell, M.D. of 140 patients hospitalized For alcohol-related injuries, J Fam Practice, 50
Implications of mediational
results for clinicians
• Promoting involvement as important as
promoting attendance
• May help to focus treatment on mediators
even for non-12-step involved patients
• Help anticipate and reinforce mediational
changes likely to occur in self-help groups
Conclusions
• 12-step group participation significantly
reduces drug and alcohol use.
• 12-step group involvement reduces surplus
health care utilization.
• Benefits of 12-step groups mediated both by
psychological and social changes.
• Applying these findings in treatment settings
should improve outcomes and reduce costs.
Collaboration and competition between
self-help groups and professionals
Keith Humphreys
Veterans Affairs and Stanford
University School of Medicine
Bases of collaboration
• Shared commitment to a stigmatized activity
• Overlap of some goals, e.g., recovery
• Shared value orientation
• Overlap of personnel
Self-reported global attitudes are positive
in German, U.S. and Canadian studies
Items w/95-99% endorsement:
• “Self-help groups are an important resource”
• “Professionals and self-help groups should
work together to help people in need”
• “As a self-helper/professional, I want to
collaborate with self-helpers/professionals”
“We love humanity in general,
but we don’t much like anyone in
particular
The code of the Minnesota Liberal, as
described by Garrison Keillor
Beneath apple pie sentiments
• Fear of lost legitimacy, status and income
• Different conceptions of “collaboration”
• Different norms, values and language
• Lack of faith in the other fellow
Project MESH: A Collaboration
that Worked
• Form self-help promoting coalitions
• Give them money and staff support
• Ask them to use media and education to
expand self-help groups
• Ask them to emphasize people of color and
Spanish-language speakers
• Get out of their way
Coalition work
•
•
•
•
•
•
Adopted “get a grip” motto
Designed bus posters with tear offs
Recorded public service announcements
About 20% Espaňol, 80% English
No direct appeals to physicians
Los Angeles modifications
Example poster with tear off
Example poster with tear off
Number of visits to self-help groups in Oakland and Los Angeles
in 3 months of MESH Project vs. in same 3 months of prior year
Visits to self-help groups
2500
2250
Oakland (prior to
intervention)
Oakland (during
intervention)
Los Angeles (prior
to intervention)
Los Angeles
(during intervention)
2000
1750
1500
1250
1000
1
2
3
Why did the collaboration in
Project MESH work?
•
•
•
•
Independent bases of authority and legitimacy
Independent resources
Shared commitment to self-help groups
Socially skilled group leaders and project
facilitator
• Shared commitment to an outcome rather than
a process