Behavior Analysis and Treatment of Drug Addiction (Silverman et al. 2021) PDF

CHAP TER 29 Behavior Analysis and Treatment of Drug Addiction Recent Advances in Research on Abstinence Reinforcement Kenneth Silverman, August F. Holtyn, Brantley P. Jarvis, and Shrinidhi Subramaniam Extensive evidence in animals and in humans, in AN OVERVIEW OF RESEARCH ON ABSTINENCE bas...

CHAP TER 29 Behavior Analysis and Treatment of Drug Addiction Recent Advances in Research on Abstinence Reinforcement Kenneth Silverman, August F. Holtyn, Brantley P. Jarvis, and Shrinidhi Subramaniam Extensive evidence in animals and in humans, in AN OVERVIEW OF RESEARCH ON ABSTINENCE basic laboratory research and randomized clinical REINFORCEMENT TO 2011 trials, suggests that drug addiction is an operant behavior that is maintained and modifiable by For over 40 years, researchers have applied absti- its consequences (Silverman, DeFulio, & Everly, nence reinforcement interventions to treat drug 2011). This body of research serves as a rich foun- addiction in diverse populations of adults and ado- dation for applying the principles of operant con- lescents who have used many different commonly ditioning to the treatment of drug addiction. Re- abused drugs (Silverman, Kaminski, et al., 2011). searchers have applied operant principles to drug addiction treatment in a variety of ways, but they Early Studies have applied it most directly and arguably with the greatest effectiveness in the direct reinforce- Early studies applied abstinence reinforcement in- ment of drug abstinence. Under abstinence rein- terventions to treat so-called “Skid Row alcohol- forcement procedures, patients receive a desirable ics”; adults enrolled in methadone treatment who consequence contingent on providing objective continued to use opiates, benzodiazepines, and evidence of drug abstinence. A chapter in the first alcohol during methadone treatment; health care edition of this handbook provided the context and professionals who abused various drugs; and ciga- overview of about 40 years of research on the de- rette smokers (Silverman, Kaminski, et al., 2011). velopment and evaluation of abstinence reinforce- Although these early studies differed considerably ment interventions for drug addiction (Silverman, in the settings and procedures used to apply ab- Kaminski, Higgins, & Brady, 2011). In the present stinence reinforcement contingencies, they “estab- chapter, we provide a brief overview of the earlier lished a firm scientific foundation for the develop- chapter, and then provide a qualitative and selec- ment of abstinence reinforcement interventions, tive summary and discussion of the research that and illustrated a range of creative and useful ap- investigators have published since we wrote the plications of an abstinence reinforcement technol- original chapter. ogy” (Silverman, Kaminski, et al., 2011, p. 453). 490  Behavior Analysis and Treatment of Drug Addiction 491 Voucher-Based Reinforcement a monetary voucher if the CO level displayed on the CO meter met the criterion for reinforcement. Higgins et al. (1991) developed and tested an abstinence reinforcement intervention to treat adults addicted to cocaine. The intervention of- Reviews and Meta-Analyses fered participants monetary vouchers exchange- Reviews and meta-analyses have shown that ab- able for goods and services for providing routine stinence reinforcement interventions are among urine samples that were negative for cocaine. Im- the most effective psychosocial interventions for portantly, the voucher intervention used a sched- the treatment of drug addiction (e.g., Dutra et al., ule of escalating pay for sustained abstinence, in 2008; Lussier, Heil, Mongeon, Badger, & Higgins, which the value of the vouchers increased as the 2006; Pilling, Strang, Gerada, & National Insti- number of consecutive cocaine-negative urine tute for Clinical Excellence [NICE], 2007). One samples increased. This voucher-based abstinence meta-analysis, for example, examined 34 con- reinforcement intervention proved effective and trolled studies that evaluated the effectiveness of versatile. Over the next 20 years, researchers ap- abstinence reinforcement interventions (called plied and evaluated the effectiveness of the vouch- contingency management in that paper); relapse er intervention in promoting cocaine abstinence prevention; general cognitive-behavioral therapy; in adults with primary cocaine dependence; adults and treatments combining cognitive-behavioral and patients receiving methadone treatment who therapy and contingency management. In that continued to use cocaine during this treatment; meta-analysis, the strongest effect was for contin- opiate abstinence in patients who continued to gency management interventions (Dutra et al., use opiates during methadone treatment; smoking 2008). cessation in diverse populations of cigarette smok- ers; and abstinence from marijuana use (Silver- man, Kaminski, et al., 2011). Improving Outcomes Petry, Martin, Cooney, and Kranzler (2000) Abstinence reinforcement interventions are clear- developed a variation of the voucher-based absti- ly effective in promoting abstinence from most nence reinforcement intervention, in which par- commonly abused drugs and in diverse popula- ticipants earned the opportunity to draw prizes tions. However, the interventions have two main from a fishbowl contingent on alcohol-negative limitations: (1) They are not effective for all pa- breath or drug-negative urine samples. The possi- tients, and (2) many patients resume drug use ble prizes had small, large, jumbo, or no monetary when the abstinence reinforcement intervention values, and a participant had a chance of drawing ends (Silverman, Kaminski, et al., 2011). As de- one prize value on any given occasion. To rein- scribed in the original chapter, the effectiveness of force sustained abstinence, the researchers used a abstinence reinforcement interventions can vary schedule of escalating reinforcement for sustained as a function of familiar parameters of operant abstinence in which the number of draws in- conditioning that affect any operant-reinforce- creased as the number of consecutive alcohol- or ment contingency. drug-negative samples increased. The prize-based abstinence reinforcement procedure was effective, and the National Institute on Drug Abuse’s Clini- Increasing Effectiveness cal Trials Network evaluated the procedure in Conclusions from reviews and meta-analyses sug- two multisite randomized controlled clinical trials gest that parameters such as immediacy and fre- (Peirce et al., 2006; Petry, Alessi, Marx, Austin, & quency of reinforcement and response require- Tardif, 2005). ments (i.e., abstinence from single vs. multiple Dallery and Glenn (2005) developed a novel drugs) alter the effectiveness of abstinence re- Internet-based approach to reinforce smoking ces- inforcement interventions. However, individual sation that proved both effective and convenient. studies show most clearly that the magnitude of Under that system, participants provided breath reinforcement determines the effectiveness of carbon monoxide (CO) samples in front of a video these interventions (Silverman, Kaminski, et al., camera connected to the Internet. The video, 2011). One study, for example, demonstrated that which included the reading on the CO meter, some patients who used cocaine during metha- was time-stamped, transmitted across the Inter- done treatment did not initiate cocaine absti- net, and evaluated by staff. Participants received nence when offered a standard voucher interven- 492 S U B S P E C I A LT IE S I N A P P L IE D B E H AV I O R A N A LY S I S tion in which they could earn up to about $1,150 Dissemination in vouchers for providing cocaine-negative urine At the time we wrote our earlier chapter, profes- samples three times per week for 12 weeks. In a sionals in the community had not used abstinence subsequent within-patient cross-over period, many reinforcement interventions widely (Silverman, of those treatment-refractory patients did initi- Kaminski, et al., 2011). Most efforts to apply such ate sustained cocaine abstinence when offered a interventions sought to integrate those contingen- high-magnitude voucher intervention in which cies into substance abuse treatment clinics. they could earn up to $3,400 in vouchers for pro- Resources available to substance abuse treat- viding cocaine-negative urine samples three times per week for 9 weeks. Importantly, they achieved ment clinics constrained applications of absti- significantly higher rates of cocaine abstinence nence reinforcement interventions in those clinics. than during low- and zero-magnitude reinforce- Researchers had attempted to “use reinforcers that ment conditions (Silverman, Chutuape, Bigelow, [were] available in clinics, to devise ways to pay for & Stitzer, 1999). One study also showed that an reinforcers, and to use low-cost reinforcers” (Sil- abstinence reinforcement intervention could be- verman, Kaminski, et al., 2011, p. 464). Take-home come ineffective if reinforcement magnitude de- methadone doses for patients receiving methadone creased too much (Petry et al., 2004). treatment proved to be one reinforcer that clinics could provide at relatively little additional cost. Researchers had used deposit contracts, in which Preventing Relapse patients deposited money at the start of treatment Relapse to drug use is common after drug abuse that they could earn back by achieving and main- treatment, independent of the type of treatment taining drug abstinence during treatment, since (McLellan, Lewis, O’Brien, & Kleber, 2000). the earliest days of research on abstinence rein- Relatively few studies have evaluated interven- forcement interventions (Elliott & Tighe, 1968). tions that could prevent relapse to drug use after Researchers tried to reduce the magnitude of an abstinence reinforcement intervention ends, reinforcement to make the interventions more even though researchers have observed relapse practical; however, reducing reinforcement mag- reliably since the earliest studies of such interven- nitude had the undesirable effect of reducing and tions (Silverman, Kaminski, et al., 2011). Several possibly eliminating the effectiveness of the in- studies examined whether abstinence reinforce- terventions (e.g., Glasgow, Hollis, Ary, & Boles, ment would produce lasting effects if researchers 1993; Petry et al., 2004). The National Institute combined it with cognitive-behavioral relapse on Drug Abuse Clinical Trials Network conducted prevention therapy, a counseling intervention de- large-scale multisite investigations in which the signed to prevent relapse. Results of those studies researchers effectively used higher-magnitude ab- showed that voucher-based abstinence reinforce- stinence reinforcement that were effective in pre- ment produced higher rates of abstinence during vious controlled studies (e.g., Peirce et al., 2006). treatment than the cognitive-behavioral therapy These studies illustrated both the effectiveness of when researchers presented each alone, and the abstinence reinforcement interventions and the combined treatment did not increase rates of ab- willingness of community treatment programs to stinence compared to the voucher intervention apply these interventions, at least when exter- alone either during or after treatment (Silverman, nal sources funded the reinforcers. The United Kaminski, et al., 2011). Kingdom provided the greatest evidence that Results of a few studies have suggested that re- communities could adopt abstinence reinforce- searchers could use abstinence reinforcement as a ment interventions. Based on a rigorous review maintenance intervention to sustain abstinence of psychosocial treatments for drug addiction, the and prevent relapse, at least while the abstinence NICE recommended routine use of voucher-based- reinforcement intervention continued (e.g., Pres- reinforcement interventions for the treatment of ton, Umbricht, & Epstein, 2002; Silverman, Ro- drug addiction (substance misuse) in the United bles, Mudric, Bigelow, & Stitzer, 2004). One study, Kingdom’s National Health Service (Pilling et al., for example, showed that methadone-maintained 2007). This resulted in the National Health Ser- patients could maintain cocaine abstinence for up vice’s offering voucher-based reinforcement as one to a year if the abstinence reinforcement contin- of its interventions for drug addiction treatment. gency was in place for that period (Silverman et Several researchers sought to use reinforcers al., 2004). available outside of the standard clinic setting  Behavior Analysis and Treatment of Drug Addiction 493 for drug addiction treatment (Silverman, Kamin- ness (Benishek et al., 2014; Davis et al., 2016). ski, et al., 2011). Three research programs on this First, the duration of abstinence reinforcement method stand out. Milby et al. (1996) investigated interventions is relatively short, with a mean of 12 the use of abstinence-contingent housing and weeks in both Davis et al. (2016) and Benishek et work therapy to promote abstinence in home- al. (2014), and a median of 8 and 12 weeks in Davis less cocaine-dependent adults. Ries et al. (2004) et al. and Benishek et al., respectively. Davis et al. used U.S. Social Security Disability benefits in a reviewed voucher- and money-based contingency contingent fashion to promote drug abstinence in management interventions for substance use dis- adults with severe mental illness. Finally, Silver- orders from 2009 through 2014; they reported that man (2004) used abstinence-contingent access to 43 of 51 studies (84%) showed significant effects of employment in a series of studies to initiate and the abstinence reinforcement interventions, and maintain drug abstinence in unemployed adults the rest (16%) did not produce significant effects. with long histories of drug addiction. Of the 22 studies that produced significant effects while the abstinence reinforcement interventions were in effect and assessed posttreatment absti- RECENT ADVANCES IN RESEARCH nence, only 7 (32%) showed a statistically signifi- ON ABSTINENCE REINFORCEMENT cant effect at follow-up. Benishek et al. conducted a meta-analysis of prize-based abstinence rein- At the time the earlier chapter was written, conclu- forcement interventions from 2000 through 2013. sions about research on abstinence reinforcement Results were like those Davis et al. reported. Eigh- interventions were relatively simple (Silverman, teen of the 19 studies (95%) reported a significant Kaminski, et al., 2011): (1) These interventions effect while the prize-based interventions were in promoted abstinence from most commonly abused effect. The effects of these interventions decreased drugs and in diverse populations and settings; (2) in the 6 months after discontinuation of the in- a need existed to develop procedures to increase terventions. Six of the nine studies (66%) with a the proportion of patients who achieved absti- 3-month follow-up assessment showed a significant nence when exposed to these interventions; (3) a effect of the prize-based abstinence reinforcement need existed to develop procedures that promoted interventions. Two of the six studies (33%) with long-term abstinence and prevented relapse; and a 6-month follow-up assessment showed a signifi- (4) a need existed to develop practical applica- cant intervention effect. An analysis that com- tions of abstinence reinforcement interventions bined data from the 6-month follow-up showed that ensured the widespread application of these that the overall effects of the prize-based interven- procedures in society. In the remainder of this tions were not detectable at that time. chapter, we review studies conducted since we The more recent reviews and meta-analyses completed the earlier chapter, with special atten- document the effectiveness of operant-condi- tion to whether recent research has confirmed and tioning principles in the treatment of drug ad- extended our earlier conclusions. diction. Abstinence reinforcement interventions significantly increase drug abstinence while the interventions are in effect and occasionally after General Utility of Abstinence they end (Benishek et al., 2014; Cahill et al., 2015; Reinforcement Interventions Castells et al., 2009; Davis et al., 2016). However, Reviews and meta-analyses published in the last these reviews and meta-analyses summarize the several years have confirmed that abstinence re- literature by statistical significance and effect inforcement interventions are highly effective sizes, but they do not fully highlight the clinical in promoting abstinence from most commonly importance and limitations of abstinence rein- abused drugs and in diverse populations (Benishek forcement interventions: What proportion of pa- et al., 2014; Cahill, Hartmann-Boyce, & Perera, tients fail to respond to such interventions? Do 2015; Castells et al., 2009; Davis et al., 2016). these interventions serve a clinically useful role if These reviews and meta-analyses also confirmed they do not maintain abstinence over time? Has that such interventions are not always effective recent research identified methods of promoting and rarely produce effects that are evident after abstinence in treatment-resistant patients and pro- the interventions end. moting long-term abstinence? In this section, we Two analyses of these interventions highlight summarize the results of selected studies to shed key and consistent findings about their effective- some light on these questions. 494 S U B S P E C I A LT IE S I N A P P L IE D B E H AV I O R A N A LY S I S Cigarette Smoking recent analysis that combined data collected in three studies confirmed that the voucher-based One of the greatest challenges of applying absti- abstinence reinforcement intervention was effec- nence reinforcement interventions to smoking tive in promoting smoking cessation in a larger cessation is the practical problem of collecting sample of pregnant women; it also extended the frequent breath CO samples from participants. findings of the earlier research by showing that The short half-life of breath CO is the reason for the voucher-based intervention produced signifi- this problem. As described in our earlier chapter, Dallery and colleagues (e.g., Dallery & Glenn, cantly increased birth weight for babies and re- 2005) developed a novel intervention including duced the percentage of low-birth-weight babies an Internet-based video technology to solve this (Higgins et al., 2010). Again as in previous absti- practical problem. Participants use that system nence reinforcement studies, the voucher-based to provide breath CO samples in front of a video intervention was not effective for all participants. camera connected to the Internet. The software For example, 34% of women were abstinent at the transmits the video of the participant providing end of treatment. The combined analysis showed the breath sample and the CO level displayed that the effects of the voucher-based abstinence on the CO meter to the investigator. The par- reinforcement intervention were still evident ticipant receives a monetary voucher if the CO and statistically significant after the intervention level displayed on the meter meets the criterion ended, although the rates of abstinence decreased for abstinence reinforcement. At the time of the progressively over the 12- and 24-week period after previous chapter, limited evidence had demon- delivery. By 24 weeks postpartum, only 14% of strated the effectiveness of this approach, but women exposed to the voucher-based intervention two more recent randomized controlled clinical were abstinent from smoking. trials have provided rigorous and real-world evi- Researchers also have used abstinence rein- dence (Dallery, Raiff, & Grabinski, 2013; Dallery forcement to promote smoking cessation in pa- et al., 2017). In one study (Dallery et al., 2013), tients in a residential substance abuse treatment researchers randomly assigned cigarette smokers program (Alessi & Petry, 2014), in opioid-main- who were interested in quitting smoking to a con- tained patients (Dunn, Saulsgiver, & Sigmon, tingent-CO or noncontingent-CO group. Partici- 2011), in smokers in Spain (Secades-Villa, Garcia- pants in the contingent-CO group could earn up Rodriguez, Lopez-Nunez, Alonso-Perez, & Fernan- to $530 in vouchers for providing two breath CO dez-Hermida, 2014), and in pregnant Indigenous samples per day that confirmed recent abstinence women in New Zealand (Glover, Kira, Walker, & from smoking over a 7-week period. Noncontin- Bauld, 2015). Two related studies assessed the ef- gent-CO (control) participants earned vouch- fects of financial incentives in promoting smok- ers regardless of the CO levels of their breath ing cessation in socioeconomically disadvantaged samples. Contingent-CO participants provided adults (Kendzor et al., 2015) and homeless adults significantly more CO-negative breath samples (Businelle et al., 2014). These studies generally than participants in the noncontingent-CO confirmed prior findings that the abstinence re- group. As in previous abstinence reinforcement inforcement interventions (1) were effective in studies, the smoking cessation intervention was promoting smoking cessation, (2) were not effec- not effective for all participants, and participants tive for all participants, and (3) did not prevent did not maintain the effects at 3- and 6-month postintervention relapse reliably. Some studies follow-ups. showed that the abstinence reinforcement inter- As reported in the earlier chapter, voucher- ventions could produce effects that were still evi- based abstinence reinforcement can have pro- dent and statistically significant in the weeks after found effects in pregnant women who smoke ciga- termination (e.g., Kendzor et al., 2015; Secades- rettes during pregnancy. Pregnant women in the Villa et al., 2014). Rates of abstinence after the abstinence reinforcement intervention initially abstinence reinforcement ended, however, were earned vouchers for providing CO-negative breath consistently lower than such rates while the ab- samples and then for providing urine samples that stinence reinforcement was in effect. Thus, even were negative for cotinine. That intervention in the studies with significant postintervention ef- was effective in promoting smoking cessation in fects on smoking cessation, smoking relapse after pregnant women and produced significant in- discontinuation of an abstinence reinforcement creases in fetal growth (Heil et al., 2008). A more intervention remains a problem.  Behavior Analysis and Treatment of Drug Addiction 495 Opioids randomly assigned to receive cognitive-behavioral therapy with or without prize-based reinforcement Three studies evaluated the effects of prize-based of cocaine abstinence over a 24-week period (Pe- reinforcement on opiate abstinence and retention titjean et al., 2014). The study did not show effects in methadone treatment for patients in China. of the prize-based reinforcement intervention on One randomized study showed clear effects of the several measures of cocaine abstinence, but it did prize-based reinforcement on retention of patients show effects on the percentage of cocaine-negative in methadone treatment in one of two clinics, but the prize-based system did not affect opiate absti- urine samples at selected weeks during the study. nence rates either in the clinic or overall (Hser There were no significant differences between et al., 2011). A second study randomly assigned groups at the 6-month follow-up assessment. methadone-treated patients to standard metha- In one study, 127 men who routinely had sex done maintenance with or without prize-based with men and who used methamphetamine were reinforcement of opiate abstinence and retention randomly assigned to a 12-week voucher-based ab- (Jiang et al., 2012). This study showed that the stinence reinforcement intervention or to a con- prize-based reinforcement intervention had no ef- trol condition (Menza et al., 2010). Participants in fect on treatment retention or opiate abstinence. the voucher-based intervention could earn up to A final study randomly assigned clinics to prize- about $450 in vouchers over the 12-week interven- based reinforcement for opiate abstinence and tion for providing methamphetamine-free urine retention in methadone treatment or to standard samples two or three times per week. The study methadone maintenance, and found a small but failed to show a significant effect of the voucher significant effect of prize-based reinforcement on intervention on methamphetamine use during or both treatment retention and opiate abstinence after treatment. (Chen et al., 2013). Thus these studies on the One study randomly assigned 176 outpatients prize-based reinforcement intervention on opiate with serious mental illness and stimulant depen- abstinence in China produced mixed effects on dence to a 12-week intervention in which they opiate abstinence. Only one of the studies showed received either prize-based reinforcement for stim- that prize-based reinforcement increased opiate ulant-negative urine samples or noncontingent abstinence. As with other studies, prize-based re- prizes (McDonell et al., 2013). During treatment, inforcement of opiate abstinence was not effective participants assigned to receive the prize-based in- for all participants, and the only study that showed tervention provided significantly more stimulant- an increase in opiate abstinence during the inter- negative urine samples (over 80% negative) than vention did not assess postintervention effects. the participants assigned to the noncontingent (control) condition (equal to or less than 70% negative). Participants assigned to the contingent Stimulants (Cocaine, Amphetamine, prizes appeared to provide significantly more stim- or Methamphetamine) ulant-negative urine samples (46%) than partici- One study evaluated the effects of prize-based pants assigned to noncontingent prizes (35%) 16, reinforcement on cocaine abstinence in a rela- 20, and 24 weeks after the intervention ended. Re- tively small sample (N = 19) of cocaine-dependent sults of this study were limited, however, because adults receiving treatment in an outpatient mental of the high levels of missing data during the follow- health clinic (Petry, Alessi, & Rash, 2013). The up period, and the significant differences between participants were randomly assigned to receive groups were not maintained across all methods of prize-based reinforcement of cocaine abstinence handling missing urine samples. or not over an 8-week period. The study showed Another study evaluated the effects of flexible mixed effects on cocaine abstinence. Results were methadone dosing and a voucher intervention significant only for most consecutive weeks of that reinforced cocaine abstinence in patients in cocaine abstinence and proportion of expected a methadone treatment program (Kennedy et al., samples that were cocaine-negative. Results were 2013). After a 6-week baseline, participants who not significant for proportion of submitted samples continued to use opiates and cocaine were ran- that were cocaine-negative. The study did not as- domly assigned to one of four conditions: voucher- sess postintervention effects. based reinforcement of cocaine abstinence, flex- In another study, 60 patients in an outpatient ible methadone dosing, flexible methadone dosing unit of a psychiatric hospital in Switzerland were and voucher-based reinforcement of cocaine ab- 496 S U B S P E C I A LT IE S I N A P P L IE D B E H AV I O R A N A LY S I S stinence, or a no-treatment control group. Partici- disorder who were enrolled in an outpatient treat- pants in the voucher intervention could earn up to ment program were randomly assigned to receive $1,418 in vouchers for providing cocaine-free urine usual care or usual care plus prize-based absti- samples every Monday, Wednesday, and Friday nence reinforcement (Killeen, McRae-Clark, Wal- over 16 weeks. Although the study examined the drop, Upadhyaya, & Brady, 2012). After a 2-week effects of flexible methadone dosing and vouch- washout period, participants in the prize-based ab- er-based reinforcement of cocaine abstinence on stinence reinforcement group could earn the op- opiate and cocaine use, we review only the ef- portunity to draw prizes for providing alcohol-free fects of the voucher-based intervention here. The breath samples and urine samples that were nega- group exposed to the voucher-based intervention tive for marijuana, cocaine, opiates, methamphet- had significantly higher rates of cocaine-negative amine, and amphetamine; control participants urine samples during treatment than the group earned two prize draws for providing samples, in- that received neither intervention. Surprisingly, dependent of the breath or urinalysis results. The the group exposed to both flexible dosing and prize intervention was in effect for 10 weeks. The voucher-based cocaine abstinence reinforcement study showed no difference in urinalysis results did not provide significantly more cocaine-neg- between the two groups on any measure of drug ative urine samples than the group that received use. Although this study included other drug use neither intervention. As in other studies, although in the abstinence reinforcement intervention and the voucher-based intervention increased cocaine in the outcome measures, marijuana was the pre- abstinence, not all participants responded to the dominant drug participants used in this study, and intervention. Only about half the urine samples of there were very low levels of drug use other than participants in the group exposed to the voucher- marijuana in both groups. based intervention alone were cocaine-negative. In a second study, 59 adolescents between the This study did not assess posttreatment results. ages of 14 and 18 who met Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM- IV) criteria for cannabis abuse or dependence and Marijuana showed recent evidence of marijuana use, but no Litt, Kadden, and Petry (2013) assigned 215 mar- use of substances other than alcohol and tobacco, ijuana-dependent adults to one of three 9-week were randomly assigned to one of two conditions: outpatient treatments: a case management (con- cognitive-behavioral therapy and voucher-based trol) condition; a condition combining motiva- abstinence reinforcement, or cognitive-behavioral tional-enhancement therapy, cognitive-behavioral therapy and noncontingent vouchers. The voucher therapy, and prize-based reinforcement for home- condition started in Week 3 and continued until work completion; and a condition combining Week 10, and participants could earn up to $242 motivational-enhancement therapy, cognitive- in vouchers for providing drug-free urine samples. behavioral therapy, and prize-based reinforcement Researchers tested urine samples for cannabis, co- for abstinence from marijuana. The study evalu- caine, opioids, benzodiazepines, amphetamines, ated the effects of those interventions on self-re- and methamphetamine. Participants exchanged ported marijuana use in the months after the voucher earnings for gift cards. There were no sig- intervention ended. Neither of the experimental nificant differences between groups on measures interventions that included prize-based reinforce- of marijuana use either during or after treatment. ment increased self-reported marijuana abstinence In a final study, 153 adolescents between the relative to the control condition; however, the ages of 12 and 18 who met DSM-IV criteria for intervention that included prize-based reinforce- cannabis abuse or dependence and showed recent ment for marijuana abstinence produced transient evidence of marijuana use, but were not dependent increases in marijuana abstinence in Months 5–8, on another substance, were randomly assigned compared to prize-based reinforcement of home- to one of three groups (Stanger, Ryan, Scherer, work completion. There were no significant differ- Norton, & Budney, 2015). All groups received ences between groups at the latest follow-up time motivational-enhancement therapy and cogni- at Months 11–14. tive-behavioral therapy for 14 weeks. Two groups Three controlled studies evaluated the effects received an additional contingency management of abstinence reinforcement interventions on intervention, one with a parent-training compo- marijuana use in adolescents. In the first of these nent and one without parent training. After a studies, 31 adolescents with primary marijuana use 2-week washout period, participants in the contin-  Behavior Analysis and Treatment of Drug Addiction 497 gency management groups began the contingency experimental design to evaluate its effectiveness management intervention (Weeks 3–14). During (Barnett, Tidey, Murphy, Swift, & Colby, 2011). this intervention, participants could earn voucher- The second study used a within-participant based reinforcement for urine samples collected cross-over design to evaluate the effectiveness of twice per week that were negative for all drugs a contingency management intervention, includ- assessed (cannabis, cocaine, opioids, benzodiaz- ing use of the SCRAM bracelet, in promoting epines, amphetamine, and methamphetamine). alcohol abstinence (Dougherty et al., 2014). In Participants in the voucher condition could earn that study, participants (N = 26) were exposed up to $590 in vouchers that were exchangeable for to $0 and $25 contingency management condi- gift cards for providing drug-free urine samples. In tions in counterbalanced order, followed by a $50 addition, participants received a prize-based absti- contingency management condition. Each condi- nence reinforcement intervention in which they tion was in effect for 4 weeks. In the $0 condition, could earn up to about $135 in prizes for provid- researchers did not give participants any instruc- ing drug-free urine samples during Weeks 1–4. tions about alcohol consumption. During the $25 Finally, parents could earn prize draws for devel- and $50 conditions, participants could earn $25 oping and using a substance-abuse-monitoring and $50 per week, respectively, if their transder- contract. The contingency management interven- mal alcohol concentration never exceeded 0.03 tion produced significant increases in abstinence grams per decaliter (g/dl; this concentration cor- from marijuana during treatment and at the end responded to light to moderate drinking) on any of treatment. Forty-eight percent of participants in day during the week. The $25 incentive condition the two groups exposed to the contingency man- significantly reduced drinking episodes and heavy agement intervention achieved 4 or more weeks of drinking, relative to the $0 condition. In addition, continuous abstinence during treatment, whereas participants exposed to the $25 condition followed only 30% of participants in the group not exposed by the $0 condition had less frequent drinking and to the contingency management intervention did less heavy drinking in the $0 condition than par- so. Abstinence rates decreased between the end ticipants exposed to the $0 condition followed by of treatment and the 3-month follow-up, and the the $25 condition. This result demonstrated that three groups had similar rates of abstinence at the the $25 condition produced some lasting carry- 3-month follow-up. over effects on drinking that persisted after the $25 condition ended. There were few differences on drinking outcomes between the $25 and $50 Alcohol conditions, except that the $50 condition was Researchers have conducted limited research on more effective in reducing heavy weekend drink- abstinence reinforcement interventions for alco- ing than the $25 condition. hol use, at least in part because measures have A third study using the SCRAM technology not been available to detect alcohol use reliably enrolled 80 alcohol-dependent adults in a three- beyond a relatively brief window of several hours phase study that included an observation phase, since the last drink. Researchers can use breath a contingency management phase, and a follow- alcohol tests, but participants can test alcohol- up phase (Dougherty, Karns, et al., 2015). At each negative on these by remaining abstinent from weekly clinic visit during the observation and con- alcohol for several hours before providing a breath tingency management phases and monthly during sample. the follow-up phase, participants completed Time- Since our earlier chapter appeared, three stud- line Follow-Back interviews about drinking. Dur- ies have evaluated the effectiveness of using a ing the initial observation phase, each participant transdermal alcohol sensor bracelet as a part of wore a SCRAM bracelet with no explicit conse- an abstinence reinforcement intervention for al- quences for drinking. During the contingency cohol use. A participant wears this bracelet, called management phase, a participant earned $50 each the Secure Continuous Remote Alcohol Monitor week if the SCRAM bracelet showed that the (SCRAM), locked on the ankle. The participant participant’s transdermal alcohol concentration can wear the bracelet continuously, including in stayed below 0.03 g/dl every day of the week. Dur- the shower. The bracelet includes features that ing the follow-up phase, participants returned to detect removal and tampering. One of the studies the clinic every month for 3 months and complet- used the bracelet in a contingency management ed Timeline Follow-Back interviews about drink- intervention, but did not include an adequate ing in the past 28 days. Transdermal alcohol con- 498 S U B S P E C I A LT IE S I N A P P L IE D B E H AV I O R A N A LY S I S centration data showed that heavy drinking was cohol,