Translational Research on Reinforcement as a Determinant of

Substance Use Disorders

Stephen T. Higgins, Ph.D.University of Vermont

Introduction:Introduction:• Substance use disorders (SUDs) represent a

tremendous public health problem in the U.S., Europe, and most contemporary societies.

• A great deal more needs to be learned regarding the determinants of SUDs and how to more effectively prevent and treat these disorders.

• Also important is that we fully explore and, when possible, translate scientific information that is already available into clinical interventions for SUDs.

Introduction (cont’d)Introduction (cont’d) Extensive evidence indicates that reinforcement

plays an important role in the genesis, maintenance, and recovery from SUDs.

Reinforcement is a process wherein a stimulus event increases the future probability of a response (drug use) by acting as a consequence.

The neurochemical and neuroanatomical systems involved in drug-produced reinforcement have been relatively well characterized (i.e. mesolimbic dopamine system).

Introduction (cont’d)Introduction (cont’d) This is among the most important observations in a

scientific analysis of SUDs because it places them within an existing body of knowledge in the behavioral and biological sciences.

Stated in its most parsimonious form, SUDs are

controlled, at least in part, by the Law of Effect.

This knowledge has been utilized well in pre-clinical research on SUDS, but less consistently so in clinical research and application.

Clinical Research on the Role of Clinical Research on the Role of Reinforcement in Substance Use DisordersReinforcement in Substance Use Disorders

There’s a resurgence of interest in the clinical implications of reinforcement as a determinant of SUDs

The research illustrates two main empirical generalizations with important conceptual and practical implications:

1. Drug use is a form of operant behavior (i.e., sensitive to environmental consequences)

2. The reinforcing effects of drugs are malleable and dependent on environmental context

Evidence to Support Those Evidence to Support Those Empirical GeneralizationsEmpirical Generalizations

• Laboratory settings

• Treatment settings

Drug self-Administration StudiesDrug self-Administration StudiesTwo experiments examining the influence of alternative, non-drug reinforcers on preference for cocaine use illustrate the context-dependent nature of the reinforcing effects of drugs.

• Higgins et al. (1994): recreational users of cocaine hydrochloride made exclusive

choices between (1) cocaine vs. placebo and (2) cocaine vs. varying amounts of money

• Hatsukami et al. (1994): crack cocaine abusers chose between smoking cocaine vs. earning tokens worth varying monetary values

SummarySummary

• Same functional relationships demonstrated with all routes of cocaine use, a wide-range of drug-use histories and doses, as well as other drugs (e.g., heroin, cigarette smoking).

• Contingent availability of non-drug, alternative reinforcers can significantly alter the probability of drug use.

Another Contextual Factor:Another Contextual Factor:Temporal DelaysTemporal Delays In naturalistic settings individuals often

choose between using drugs and enjoying their effects in the present versus abstaining now and reaping delayed positive consequences for doing so in the future.

Roll et al. (2000) illustrates the influence of delays: Regular cigarette smokers abstained from smoking for several hours and then chose between puffs on a cigarette vs. varying amounts of money available after varying temporal delays.

Constraints on Drug ConsumptionConstraints on Drug Consumption Studies emphasized the influence of

alternative reinforcers, but relationships can be considered more generally in terms of constraints on drug use (behavioral economics).

As constraints on drug use increase, in terms of price, effort, or opportunity cost, consumption decreases; as constraints decrease (low price or little opportunity cost), consumption increases (i.e., Law of Demand).

Discounting of Delayed ConsequencesDiscounting of Delayed Consequences

Emerging area of behavioral economics research suggests that those with substance use disorders discount the value of delayed reinforcement more than those without substance use disorders (Bickel & Marsch, 2001)

More specifically, abusers show bias for (a) more immediate, smaller reinforcers over more delayed larger reinforcers and (b) more delayed, larger losses over more immediate, smaller losses.

Seminal StudySeminal Study Study by Madden et al. (1997) illustrates the

procedure commonly used. Opiate-dependent outpatients and community

volunteers made a series of choices between two hypothetical monetary options or drug options (patients only).- Monetary values: $1 to $1,000- Drug doses: comparable # and units to money- Delay intervals: 7 intervals ( 1 week-25 yrs)- 1st choice between highest value available immediately

or in 1 week

SummarySummary At least 14 original studies demonstrating

increased discounting of delayed reinforcement in abusers, and similar outcomes were obtained using real vs. hypothetical consequences.

Co-morbid problems are associated with increased discounting.

Whether the greater discounting represents a cause or consequence of substance use disorders is unknown.

Treatment SettingsTreatment SettingsTreatments based on reinforcement principles

seek to:• increase the availability and reduce

constraints on reinforcement from non-drug options;

• increase constraints on reinforcement derived from substance use and related lifestyle;

• configure these efforts to accommodate the relatively short temporal horizons characteristic of substance abusers.

Contingency ManagementContingency Management CM involves systematic delivery of

reinforcing or punishing consequences contingent on occurrence of a target response.

Most common CM in SUDs literature is voucher-based reinforcement of recent drug abstinence.

Illustrate with studies with cocaine-dependent outpatients and pregnant cigarette smokers and then summarize with results from recent meta-analysis.

TREATMENT PROCEDURES

Weeks 1-12: Twice weekly counseling and thrice weekly urinalysis, Weeks 13-24: once weekly counseling and twice weekly urinalysis, Months 7-12: Aftercare--once monthly check-in with counselor and random urinalysis.

Vouchers, reciprocal relationship counseling, functional analysis, vocational assistance, after recreational/social practices, monitored antabuse therapy

Voucher Program(weeks 1-12 only)

• Specimens that were negative for benzoylecgoine earned pts recorded on vouchers. Pts were worth equiv. of $.25 each.

• 1st neg. test = 10 pts @ $.25/pt = $2.50. Value of vouchers for each subsequent consecutive negative test increased by 5 pts; e.g., 2nd = 15 pts, 3rd = 20 pts, etc.

• Equivalent of $10 bonus earned for every 3 consecutive negative tests.

Voucher Program (cont’d)(weeks 1-12 only)

• Cocaine-positive tests or failure to give a specimen reset vouchers back to initial value

• 5 consecutive negative tests returned vouchers back to the value preceding reset

• Vouchers were exchangeable for retail items in community; max. earnings possible = $997.50

• No cash was ever given to patients

• All purchases were made by staff and had to be deemed by therapists to be consistent w/CRA treatment goals

COMMUNITY REINFORCEMENT APPROACH (CRA)

• Practical needs

• Social & recreational counseling

• Skills training

• Relationship counseling

• Vocational counseling

• Other drug use (monitored disulfiram therapy)

Seminal Randomized Seminal Randomized Clinical TrialClinical Trial

Achieving Cocaine Abstinence With a Behavioral Approach

Stephen T. Higgins, Ph.D, Alan J. Budney, Ph.D, Warren K. Bickel, Ph.D, John R. Hughes, M.D., Florian Foerg, B.A., and Gary Badger, M.S.

Objective: The authors compared the efficacy of a multicomponent behavioral treatment and drug abuse counseling for cocaine-dependent individuals. Method: The 38 patients were enrolled in outpatient treatment and were randomly assigned to the two treatments. Counseling in the behavioral treatment was based on the disease model of dependence and recovery. Patients in the behavioral, but not the drug counseling, treatment also received incentives contingent on submitting cocaine-free urine specimens. Results: Of the 19 patients who received behavioral treatment, 58% completed 24 weeks of treatment, versus 11% of the patients achieved at least 8 and 16 weeks of documented continuous cocaine abstinence, respectively, versus 11% and 5% in the drug abuse counseling group. Conclusions: This multicomponent behavioral treatment appears to be an effective intervention for retaining outpatients in treatment and establishing cocaine abstinence.

(Am J Psychiatry 1993; 150:763-769)

TRIAL COMPARING BEHAVIORAL VS STANDARD CARE

• Standard drug abuse counseling based on disease model and 12-steps

• 38 cocaine-dependent subjects randomized to two treatments (19/gp.)

• 6 months treatment and 6 months of follow-up

Treatment RetentionTreatment Retention

0

20

40

60

80

Treatment Condition

Per

cen

t R

etai

ned

Contingent

Standard

Incentives Improve Outcome in Outpatient Behavioral Treatment of Cocaine

Dependence

Stephen T. Higgins, Ph.D; Alan J. Budney, Ph.D; Warren K. Bickel, Ph.D; Florian E. Foerg; Robert Donham, MA; Gary J. Badger, MS

Reprinted from the Archives of General Psychiatry

July, 1994 Volume 51

Copyright 1994, American Medical Association

METHOD• 40 pts. Randomly assigned to one of two treatment

groups (20/gp).

• Treatment duration was 25 wks for both groups. 2x weekly counseling and 3x weekly urinalysis in weeks 1-12; reduced to 1x weekly counseling and 2x weekly urinalysis in weeks 13-24.

• Counseling in both groups was identical and based on the Community Reinforcement Approach (CRA)Only difference between groups was one group received vouchers contingent on cocaine-negative urinalysis results in weeks 1-12. During weeks 13-24, both groups received only a single $1 Vermont State Lottery ticket/cocaine-negative test.

Treatment RetentionTreatment Retention

0

20

40

60

80

100

12 24

% R

etai

ned

Vouchers

No Vouchers

Weeks of Treatment

Continuous AbstinenceContinuous Abstinence

0

10

20

Treatment Condition

Mea

n #

Wee

ks

Vouchers

No Vouchers

Contingent Reinforcement Increases Cocaine Abstinence During Outpatient Treatment and 1 Year of Follow-up

Journal of Consulting and Clinical Psychology

2000, Vol. 68, No. 1, 64-72Copyright 2000 by the American Psychological Association, Inc

0022-006X/00/$5.00 DOI: 10.1037//0022-006X.68.1.64

Stephen T. Higgins, Conrad J. Wong, Gary J. Badger, Doris E. Huag Ogden, and Robert L. Dantona

University of Vermont

This study assessed whether contingent incentives can be used to reinforce cocaine abstinence in dependent outpatients. Seventy cocaine-dependent outpatients were randomized into 2 conditions. All participants received 24 weeks of treatment and 1 year follow-up. The treatment provided to all participants combined counseling based on the community reinforcement approach with incentives in the form of vouchers exchangeable for retail items. In 1 condition, incentives were delivered contingent on cocaine-free urinalysis results, whereas in the other condition incentives were delivered independent of urinalysis results. Abstinence-contingent incentives significantly increased cocaine abstinence during treatment and 1 year of follow-up compared with noncontingent incentives.

METHOD• 70 cocaine-dependent adults

• Randomized to two treatmentsAll get CRA + vouchers

• Only difference is one gp. gets vouchers contingent on cocaine abstinence, while other gp. gets them non-contingently (yoked control)

• 24 weeks treatment and 1 year of follow-up

Continuous AbstinenceContinuous Abstinence

0

20

40

60

80

>8 weeks >12 weeks > 16 weeks

Durations of Continuous Abstinence

% A

bst

inen

t

Contingent

Noncontingent

Abstinence Post-TreatmentAbstinence Post-Treatment

0

20

40

60

80

6 9 12 15 18

Follow-up Months

% A

bst

inen

t

Contingent

Noncontingent

Voucher-Based Incentives to Treat Pregnant Smokers• Maternal smoking is the most important

preventable cause of poor pregnancy outcomes in the U.S.

• Approx. 30% of women of childbearing age in U.S. are smokers. Among smokers, 20% discontinue smoking upon learning of pregnancy, but vast majority smoke through the pregnancy.

• Effective cessation interventions are available, but quit rates are low (15-18%), and have been at that level for almost 2 decades.

Voucher-Based CM With Voucher-Based CM With Pregnant SmokersPregnant Smokers

Seminal study in application to pregnant smokers: Donatelle et al. (2000). Tobacco Control, 9, iii67-iii69.

Our group followed up: Higgins et al. (2004) Nic & Tob Res, 6, 1015-1020; Heil et al., in preparation.

Donatelle et al., 2000.Donatelle et al., 2000. Women randomized to usual-care (n=108) or

voucher condition (n=112). Abstinent-contingent $50 voucher available

monthly for pregnant smoker, and $50 for 1st month and $25 thereafter for S0.

Contingency in effect through pregnancy and 2 months postpartum.

Quit rates: End-of-pregnancy 32% vs. 9% and 8 wks postpartum 21% vs. 6%.

A Pilot Study on Voucher-Based Incentives to Promote Abstinence from Cigarette Smoking During Pregnancy

and PostpartumStephen T. Higgins, Ph.D., Sarah H. Heil, Ph.D.,

Laura J. Solomon, Ph.D., Jennifer Plebani Lussier, Ph.D., Rebecca L. Abel, B.S., Mary Ellen Lynch, R.N.

Gary J. Badger, M.S.University of Vermont, Burlington VT, USA

Procedure• During initial five days (M-F) of quit attempt

women submit daily breath COs (abstinence criterion < 6 ppm).

• Beginning on day 8, contingency moves to urine cotinine (< 80 mg/ml), thereby requiring less frequent monitoring.

• Contingency operates through the pregnancy and for 3 mos. postpartum . Total possible earning $1,146 ($786 pregnancy, $360 postpartum).

• Pilot data from 58 consecutive admissions.

Abstinence RatesAbstinence Rates(S-R + biochemical verification)(S-R + biochemical verification)

0

5

10

15

20

25

30

35

40

45

End of AP 3 mo PP 6 mo PP

% A

bsti

nen

t

Contingent (N=30)

Noncontingent (N=23)

(Higgins et al., 2004).

Fully-Randomized TrialFully-Randomized Trial Methods largely same as pilot study. 82 women entered; 5 excluded due to fetal

demise; 37 in contingent and 40 in noncontingent conditions.

Only differences from pilot are in random assignment and exclusive reliance on urine cotinine (some use of salivary cotinine in pilot study).

Abstinence RatesAbstinence Rates(S-R + biochemical verification)(S-R + biochemical verification)

0

5

10

15

20

25

30

35

40

45

End of AP 3 mo PP 6 mo PP

% A

bsti

nen

t

Contingent (N=37)

Noncontingent (N=40)

Voucher-Based Reinforcement Therapy for Substance Use

Disorders: A Quantitative Review

J.P. Lussier, S.T. Higgins, S.H. Heil, G.E. Badger, and J. Mongeon

University of Vermont

Methods• Effect sizes & 95% CI calculated for controlled

studies on voucher-based reinforcement therapy (VBRT) published between 1/91 & 3/04.

• 65 studies reviewed, including 40 where effects of contingent vouchers were experimentally isolated

• Focus was on period when vouchers were available.

• Studies grouped by moderator variables: type of drug targeted, control, duration of VBRT, voucher $ value, voucher delivery immediate or delayed, study setting.

Summary / ConclusionsSummary / Conclusions Extensive evidence that reinforcement plays a

central role in the genesis, maintenance, and recovery from SUDs.

Clinical lab research outlines contextual factors critical to understanding vulnerability to and treatment of SUDs.

Observations are framed in terms that have continuity across species and support a common conceptual framework from the lab to the clinic.

Treatment outcome studies illustrate how reinforcement principles can be translated into efficacious interventions, while also providing opportunities for conceptual considerations.

Summary/Conclusions (cont’d)Summary/Conclusions (cont’d) Consider the women who continues smoking

despite being pregnant, but quits with a voucher-based intervention.

Perplexing until you consider that:- Pregnant smokers In U.S. predominately lower SES (fewer

alternatives)- Material feedback on health of fetus is delayed- Women are drug dependent, which is associated with steeper

discounting of delayed rewards- Cessation increases nicotine withdrawal- Smoking produces relatively immediate reinforcing effects- Vouchers bridge the temporal gap to healthy fetus and change

the ratio of reinforcement from drug vs. non-drug activities.

Summary/Conclusions (cont’d)Summary/Conclusions (cont’d) Continuing to examine how reinforcement principles

operate in all aspects of substance use and, where possible, translating that information into clinical interventions has much to offer an evidence-based approach to treatment and prevention of SUDS.