Alcoholism Treatment Matching: Methodological and Clinical Issues Dennis M. Donovan, Ph.D. Alcohol & Drug Abuse Institute and Department of Psychiatry & Behavioral Sciences University of Washington CONJ 556: Addiction: Mechanisms, Prevention, Treatment
73
Embed
Alcoholism Treatment Matching: Methodological and Clinical Issues Dennis M. Donovan, Ph.D. Alcohol & Drug Abuse Institute and Department of Psychiatry.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Alcoholism Treatment Matching: Methodological and Clinical Issues
• It may no longer be appropriate to ask if treatment works, which is the best available treatment, or which treatment is more effective than another.
• The question needs to be reframed:
“Which kinds of individuals, with what kinds of alcohol problems, are likely to respond to what kind of treatments by achieving what kinds of goals when delivered by which kinds of practitioners?”
What Is Treatment Matching?
“Deliberate and consistent attempt to select a specific candidate for a specific method of intervention in order to achieve specific goals.”
Glaser & Skinner, 1982
“I utilize the best from Freud, the best from Jung, and the best from my Uncle Marty, a very smart fellow.”
The “Science” of Treatment Selection and Matching
Key Conceptual and Methodological Issues in Patient-Treatment Matching Conceptual issues: • Selecting effective matching variables
• Specifying the end result that matching is to enhance
• Determining the stage(s) in the treatment process at which matching decisions are to be made
Methodological issues deal with the type of patient-treatment match or interaction effect and include:
• Nonlinear interaction effects
• Higher-order interaction effects
• Multilevel interaction effects
Finney & Moos, 1986
Conditions Enhancing Probability of Identifying Interaction Effects Testing the Treatment Matching Hypothesis
• Large clinical sample, allowing subject heterogeneity for meaningful subtypes
• Controlled trial with random assignment to treatment conditions
• Use of reliable and valid instruments to assess prognostic characteristics
• Well-defined treatments that are clearly differentiable
• Use of a multidimensional assessment approach
• Inclusion of psychological variables (including degree of alcohol dependence) and demographic characteristics
• Use of data analytic strategies that are appropriate for the detection of complex interactions while also controlling for prognostic indicators1992
Methods of Treatment Matching
• Client preference/self-selection
• Client attribute by type of treatment (e.g., Project MATCH)
• Identified problem by services provided (e.g., McLellan / ASI)
• Severity of substance abuse problems / consequences by treatment intensity or setting (e.g., ASAM Criteria)
Degree of Empirical Evaluation Relative to Implementation of Matching Approaches
Em
pirical Evaluation
Implementation
Severity- Setting
Client-Treatment
Problem-Services
Client Preference
Low High
Low
High
Methods of Treatment Matching:
Client Preference/Self-Selection
Role of Client Choice in the Therapy Process
• A common clinical process following assessment is that the client is advised of the “appropriate” form of therapy.
• This is seen as undermining the client’s sense of responsibility with regard to the therapeutic process and may lead to dissatisfaction, reduced compliance and dropout.
Van Audenhove & Vertommen, 2000
Role of Client Choice in the Therapy Process
Treatment goals and approaches that have been chosen by the client, either independently or through negotiation with the clinician, are likely to capitalize on the client’s motivation and to increase compliance.
Miller, 1989
“try to make me go to rehab, i won't go, no, no, no…!”
Does Self-Selection of Treatments Lead to Better Outcomes than Random Assignment?
• Receiving treatment of preference had no measurable impact on treatment outcome, either for drinking behavior or general functioning.
• Receiving treatment of preference had no measurable impact on treatment process, utilizing client-rated (satisfaction and effectiveness), clinician rated (rapport and engagement) and objective (number of sessions attended) measures.
Adamson, Sellman, & Dore, 2005
Self-Directed Care
• A system that is “intended to allow informed consumers to assess their own needs . . . determine how and by whom these needs should be met, and monitor the quality of services they receive” (Dougherty, 2003).
• A system “in which funds that would ordinarily be paid to service provider agencies are transferred to consumers, using various formulas to account for direct, administrative, and other costs.” (Cook et al., 2004).
SAMHSA, 2004
Patient-Centered Care
“Patient-centered care” is care that is “respectful of and responsive to individual patient preferences, needs, and values and ensures that patient values guide all clinical decisions.”
• Patient access to and receipt of information that permits informed health care decisions
• Supporting the client through disagreements about treatment decisions
• Asking the patient’s goals for recovery
• Factoring these into shared decision making for the recovery process
• Assessing and supplementing education/information
Source: Institute of Medicine (2006), Improving the Quality of Health Care for Mental and Substance-Use Conditions.
Methods of Treatment Matching:
Client Attribute by Type of Treatment
(e.g., Project MATCH)
Matching Alcoholism Treatment to Client Heterogeneity
Purpose of Project MATCH
To determine if various types of alcoholics respond differentially to
different treatment approaches
Why a Multi-Site Matching Trial?
• There is considerable heterogeneity among alcoholics
• No single treatment helps all alcoholics
• Promising results in small-scale matching studies
• Matching is a dominant question on the research agenda
Clinical Research UnitsProject MATCH
SeattleVAMC
Albuquerque
HoustonVAMC
CharlestonVAMC
Milwaukee
Buffalo
Providence
WHaven VAMC*
**
* ***
**
Farmington
Two Parallel Study Arms
To test the generalizability of matching in different client populations and
treatment settings
Outpatient Aftercare
Project MATCH Therapy Manuals
To evaluate matching clients to distinct, manual-driven, theoretically-based treatments that are widely applicable
to a range of settings and providers
Research Design: Outpatient Study
M o tiva tio n a lE n ha n ce m e nt
T h e ra py
C o gn it iveB e h av io ra l
T ra in ing
T w e lveS tep
F a c ilita tion
R a nd omA ss ig nm e nt
R e c ru itm e nt fromC o m m u n ity a t 5 s ites
Assessment and Follow-up Protocol
• Baseline Pre-Treatment
• End of Treatment (Month 3)
• Post-Treatment Follow-up: Mos 6, 9, 12, 15
• All contacts were in person except month 12
• Self reports corroborated by blood/urine samples and collateral reports
Client Attributes Examined in Project MATCH
• Gender
• Alcohol involvement
• Cognitive impairment
• Meaning seeking (spirituality)
• Motivation
• Sociopathy
• Social network support for drinking
• Alcohol dependence
• Level of anger
• Interpersonal dependency
• Prior AA involvement
• Self-efficacy
• Social functioning
• Antisocial personality disorder
• Type and severity of psychiatric disorder
• Religiosity
• Alcoholism type
• Readiness to change.
Outcome Variables
Primary Measures• Percent of Days Abstinent (frequency)• Drinks per Drinking Day (intensity)
Secondary Measures (Partial list)• Other measures of drinking• Negative consequences of drinking• Other substance use• Social functioning• Psychological functioning
Compliance in Project MATCH:Treatment Attendance
Percent of Prescribed Sessions
0102030405060708090
100
Outpatient Aftercare
CB
MET
TS
The Crushing Weight of the Data
Mean Percent Days Abstinent as a Function of Time (Outpatient)
0102030405060708090
100
-2 -1 0 4 5 6 7 8 9 10 11 12 13 14 15
CBT
MET
TSF
Project MATCH Research Group, 1997
Proportion of Patients Maintaining Total Abstinence as a Function of Time (Outpatient)
0
0.2
0.4
0.6
0.8
1
1.2
1 3 7 14 30 60 90 120 150 180 240 300 360
Time in Days
Pro
po
rtio
n o
f C
lien
ts
CBT
MET
TSF
Mean Drinks per Drinking Day as a Function of Time (Outpatient)
0
2
4
6
8
10
12
14
-2 -1 0 4 5 6 7 8 9 10 11 12 13 14 15
Time in Months
Mean N
um
ber of D
rinks
CBTMETTSF
Project MATCH Research Group, 1997
Proportion of Outpatients Avoiding a Heavy Drinking Day as a Function of Time
0
0.2
0.4
0.6
0.8
1
1.2
1 3 7 14 30 60 90 120 150 180 240 300 360
Time in Days
Pro
po
rtio
n o
f C
lien
ts
CBT
MET
TSF
Project MATCH Research Group, 1997
Only 4 of 21Possible Treatment - Attribute Matches Found in Project MATCH
Alcohol Dependence: In the aftercare group, individuals with high levels of alcohol dependence benefited more from TSF than from CBT, whereas the reverse was true for patients low in dependence.
Psychopathology: In the outpatient group, those without psychopathology were found to benefit more from TSF than from CBT.
Anger: Also in the outpatient arm of the trial, patients high in anger had more successful outcomes with the MET than with the other two approaches.
Social Network Support for Abstinence: Patients whose social networks offered less support for abstinence had better outcomes in TSF than in MET.
Project MATCH Findings
• Treatment attendance was high across all three treatments
• Excellent overall outcomes, with substantial reductions in frequency and intensity of drinking following treatment
• Few differences among treatments
• Outcomes similar for MET vs. CBT+TSF
• Observed main effects generally favored TSF
• Outcomes are not substantially improved by client-treatment matching.
“In sum, Project MATCH’s findings challenged the notion that patient-treatment matching is a prerequisite for optimal alcoholism treatment. Other than the four relationships, the findings did not show that matches between patient characteristics and treatments produced substantially better outcomes.”
Interaction between self-efficacy and treatment assignment on percentage of drinking days for infrequent drinking class (top) and frequent drinking class (bottom)
Original Project MATCH Hypothesis: Individuals lower in baseline self-efficacy would have better outcomes if they were randomly assigned to CBT rather than to MET.
Original Project MATCH Finding:No interaction effect obtained.
Witkiewitz, et al., Finding:Provided support for the original self-efficacy matching hypotheses. Baseline self-efficacy was related to different outcomes depending on whether individual was randomly assigned to CBT or to MET, and this relationship was moderated by levels of drinking frequency.
Subsequent Therapy Component by Client Attribute Interactions
• A focus on emotional material in therapy was associated with more frequent alcohol consumption among patients high in depressive symptoms
• Therapist use of confrontation was associated with more frequent alcohol consumption among patients at medium and high levels of trait anger while it was associated with less frequent alcohol consumption among patients low in anger
• Confrontation was associated with more frequent alcohol consumption among patients high in interpersonal reactance (extent to which an individual generally resists being influenced by others)
• Therapy structure was associated with more frequent alcohol consumption among patients at medium or high levels of reactance
Karno & Longabaugh, Addiction, 102, 587–596, 2007
Effects of Match between Patient Depressive Symptoms and Therapy Emotion Focus
• Mismatches between patient attributes and treatment appear to have serious consequences, and this effect is magnified with multiple mismatches.
• Matches, on the other hand, while beneficial, may not be necessary to achieve good outcomes
Karno & Longabaugh, Addiction, 102, 587–596, 2007
Still left with variable response…..
• Even when treatment delivery is standardized and high adherence to manual is achieved, some patients do well and others do not.
• Very hard to predict who will do well in a particular treatment
• Nonresponse is often blamed on the patient, but that is likely not the whole story.
Another Possible Approach?
Adaptive Treatment
In Adaptive Treatment Protocols…
• One of the conceptual issues identified by Finney & Moos (1986): Determining the stage(s) in the treatment process at which matching decisions are to be made
• Treatment is tailored or modified on the basis of measures of response (e.g., symptoms, status, or functioning) obtained at regular intervals during treatment
• Goal is to deliver the least burdensome treatment that is effective, to promote better compliance over time
• Rules for changing treatment are clearly operationalized and described…..
Combined Pharmacological and Behavioral Interventions for Alcohol Dependence
Ways in Which Psychopharmacology Is Used to Treat Alcohol or Other Drug Dependencies
Purpose Treatment Goal Examples
Relapse Prevention
• Make drinking alcohol aversive
• Reduce alcohol craving
Disulfiram (Antabuse)
Naltrexone (ReVia, Vivitrol)
Acamprosate (Campral)
Three Medications Approved by the FDA for Use inthe Treatment of Alcohol Dependence
Rationale for Project COMBINE
• Recent advances have occurred in the development of pharmacological and behavioral treatments for alcohol dependence
• The hypothesis that pharmacological and behavioral treatments may enhance each other and yield optimal improvement rates requires investigation
• COMBINE will evaluate the efficacy of naltrexone, acamprosate, and psychotherapy individually and in combination
80
70
60
50
40
30
20
10
0
Baseline Month 1 Month 2 Month 3 Month 4
PercentHeavyDrinking Days
Lines: Top to Bottom at 4 monthsType A Placebo Type B Naltrexone Type B Placebo Type A Naltrexone
FIGURE 1.
Percent Heavy Drinking Days by Typology and Medication ConditionPercent Heavy Drinking Days by Typology and Medication ConditionBabor’s Type A Alcoholism
• Later age of alcoholism onset
• Weaker family history (i.e.,
• Fewer first-degree relatives who are alcoholics)
• Less severe dependence
• Fewer symptoms of co-occurring psychiatric disorders
• Less psychosocial impairment (i.e., negative familial, social, legal, or occupational consequences of drinking).
Effect of Alcoholism Typology on Response to Naltrexone in the COMBINE Study
Bogenschutz, Tonigan, Pettinati, under review
Among those receiving medication management without CBI, Type A alcoholics had better drinking outcomes with naltrexone than placebo, whereas medication conditiondid not influence outcomes significantly in the Type Bs. For those who received CBI, there were no significant effects of A/B typology.
Methods of Treatment Matching:
Identified Problem by Services Provided
(e.g., McLellan / ASI)
Drug Abuse Treatment Core Components and Comprehensive Services
Medical Services Mental
Health
Vocational
Legal Services
Educational
AIDS/HIV Services
Financial Services
Housing & Transportation
Child Care
Family Services
Substance Use Monitoring
Self-Help/Peer Support Groups
Continuing Care
Pharmaco-therapy
Intake Processing/ Assessment
Treatment Plan
Clinical and Case
Management
Behavioral Therapy and Counseling
CoreTreatment
An investigation that randomized 94 patients to problem–service matching versus standard services found that those who received three or more service sessions matched to mental health, family or employment problems had better treatment completion and post-treatment outcomes.
McLellan et al., 1997
Change in “Wrap-Around” Services 1980s (TOPS) vs. 1990s (DATOS)
77
50
60
1520
5
LTR ODF OMT
1980s 1990s
Percent of Sample Receiving 2 or More Services
Etheridge, Craddock, Dunteman, & Hubbard, 1995
Change in “Wrap-Around” Services 1980s (TOPS) vs. 1990s (DATOS)
26 23
68
10
43
10
21
10
MedicalServices
PsycholServices
FamilyServices
EducationalServices
1980s 1990s
Percent of Outpatient Psychosocial Treatment Sample
• Examined five areas of need: medical, mental, vocational, family, and housing
• Most participants indicated needing at least one service (an average of 3.22 services needed per person)
• Only about 1/3 of service needs were met/matched
– (1.11 met, 2.22 unmet)
• Overall, matching of needs with services was associated with significant reductions in drug use
• Of the five areas, only matching on mental health failed to be associated with improved drug use outcomes
• The effects of matching are greater for clients with high needs (having needs in 4-5 areas) than those with low needs
Effects of Comprehensive Matching of Service Needs on Drug Use Outcomes
Friedmann, et al., 2004
Computer-Assisted System for Patient Assessment and Referral: CASPAR
Compared to clients of counselors conducting standard assessments, those of counselors using CASPAR
• Had treatment plans that were better matched to their needs
• Received significantly more and better-matched services
• Were less likely to leave treatment against medical advice
• Were more likely to complete the full course of treatment than patients of counselors in the SA group.
Carise, et al, 2005
Methods of Treatment Matching:
Severity of Substance Abuse Problems
and Consequences
by Treatment Intensity or Setting
(e.g., ASAM Patient Placement Criteria)
• Degree of direct medical management provided
• Degree of structure, safety, and security provided
• Degree of treatment intensity provided
ASAM Criteria Describe Levels of Treatment Differentiated by Three Characteristics
ASAM PLACEMENT CRITERIA LEVELS OF I. OUTPT II. INTENSIVE III. MED IV. MED
OF CARE OUTPT MON INPT MGD INPT CRITERIA
Withdrawal no risk minimal some risk severe risk
Medical Complications no risk manageable
medical monitoring
required
24-hr acute med. care required
Psych/Behav Complications no risk mild severity moderate
24-hr psych. & addiction Tx required
Readiness For Change cooperative
cooperative but requires
structure
high resist., needs 24-hr motivating
Relapse Potential
maintains abstinence
more symptoms, needs close monitoring
unable to control use in
outpt care
Recovery Environment
supportive
less support, w/ structure
can cope
danger to recovery, logistical
incapacity for outpt
Finding from Research on ASAM Patient Placement Criteria
• Individuals who receive a less intensive level of care than recommended have poorer outcomes than those who receive the recommended level
• Receiving a more intensive level of care than recommended did not improve outcomes over those for appropriately matched level of care
• There is substantial disagreement between clinicians’ recommended level of care and that recommended by a computer algorithm (which recommended more intensive levels than did counselors)
Practical Barriers to Implementing Matching Procedures
• Lack of true alternative treatments either within a given program or across programs in a community
• Need to specify the treatments long a number of theoretically or therapeutically relevant dimensions
• Need to specify the individual client characteristics or problems at which the treatment is targeted
• Practical impact of the increased workload in order to provide the necessary assessments and treatments
Donovan & Mattson, 1994
Practical Barriers to Implementing Matching Procedures
• Need to account for influence of therapist characteristics either within or across treatment modalities
• Potential requirement of “staged” matching across different levels of motivation or readiness to change and phases of treatment and recovery
• Need to determine the best methods of and criteria for matching clients to treatment
Donovan & Mattson, 1994
Suggestions for Conceptual Changes in Allocation Research
• Extend the scope of relevant factors in change processes
– Clarify the relevance of social support for maintaining or reducing problematic substance use
– Study patient–therapist interactions
– Study site effects
• Clarify the relevance of patient decision making for treatment allocation
– Analyze the relevance of motivation/readiness for change
– Clarify the role of patients’ treatment choices
• Search for mediator and moderator effects
Bühringer, 2006
Suggestions for Conceptual Changes in Allocation Research
• Consider treatment ‘macro-level’ allocation needs in practice (setting, duration, intensity)– Determine the need for in-patient/residential
interventions– Improve the knowledge on duration versus
intensity of interventions– Determine the choice of relevant problem areas