Abstract
Contingency Management is an evidence-based treatment for substance use disorders with strong potential for measurement-based customization. Previous work has examined individual difference factors in Contingency Management treatment response of potential relevance to treatment targeting and adaptive implementation; however, a systematic review of such factors has not yet been conducted. Here, we summarize and evaluate the existing literature on patient-level predictors, mediators, and moderators of Contingency Management treatment response in stimulant and/or opioid using outpatients – clinical populations most frequently targeted in Contingency Management research and clinical practice. Our search strategy identified 648 unique, peer-reviewed publications, of which 39 met full inclusion criteria for the current review. These publications considered a variety of individual difference factors, including (1) motivation to change and substance use before and during treatment (8/39 publications), (2) substance use comorbidity and chronicity (8/39 publications), (3) psychiatric comorbidity and severity (8/39 publications), (4) medical, legal, and sociodemographic considerations (15/39 publications), and (5) cognitive-behavioral variables (1/39 publications). Contingency Management was generally associated with improved treatment outcomes (e.g., longer periods of continuous abstinence, better retention), regardless of individual difference factors; however, specific patient-level characteristics were associated with either an enhanced (e.g., more previous treatment attempts, history of sexual abuse, diagnosis of antisocial personality disorder) or diminished (e.g., complex post-traumatic stress symptoms, pretreatment benzodiazepine use) response to Contingency Management. Overall, the current literature is limited but existing evidence generally supports greater benefits of Contingency Management in patients who would otherwise have a poorer prognosis in standard outpatient care. It was also identified that the majority of previous work represents a posteriori analysis of pre-existing clinical samples and has therefore rarely considered pre-specified, hypothesis-driven individual difference factors. We therefore additionally highlight patient-level factors that are currently understudied, as well as promising future directions for measurement-based treatment adaptations that may directly respond to patient traits and states to improve Contingency Management effectiveness across individuals and over time.
Keywords: Contingency Management, Individual Differences, Substance Use Disorders, Behavioral Interventions, Motivational Incentives, Measurement-Based Care
1. Introduction
Contingency Management (CM) has accumulated substantial empirical support as a treatment option for substance use disorders during the past three decades. As an adjunct to standard outpatient care, CM consistently improves treatment attendance and abstinence (Benishek et al., 2014; Davis et al., 2016; Dutra et al., 2008; Lussier, Heil, Mongeon, Badger, & Higgins, 2006; Milward, Lynskey, & Strang, 2014) by offering incentives to patients (e.g., prizes, vouchers) for abstinence (generally verified though objective drug and/or alcohol testing) or other recovery-consistent behavior change (Petry & Stitzer, 2002). Importantly, CM is a standardized behavioral intervention that can be readily integrated into existing treatment programming (DePhilippis, Petry, Bonn-Miller, Rosenbach, & McKay, 2018) and effectively administered by a variety of behavioral health professionals (e.g., bachelor’s level psychology technicians, nursing staff, social workers, psychiatrists; Hartzler, Beadnell, & Donovan, 2017). However, while CM has gained significant popularity in recent years, it remains underutilized – due in part to concerns regarding available funding for CM treatment programs (Benishek, Kirby, Dugosh, & Padovano, 2010). Importantly, the cost-effectiveness and long-term sustainability of CM programming may further be limited by variable benefits of CM at the individual patient level, including evidence that, as with any evidence-based adjunctive intervention for substance use disorders, some patients do not achieve improved abstinence relative to standard outpatient care (Hagedorn et al., 2013; Petry et al., 2004). Identifying patient-level characteristics associated with greater or lesser benefits of CM will therefore be important to enable more effective and efficient targeting of the intervention to lower costs and support widespread access.
CM provides reliable, tangible incentives to reinforce abstinence from one or more targeted substances over a time-limited treatment interval and is typically delivered in combination with standard outpatient programming (e.g., medication-assisted treatment, group psychotherapy, and/or individual substance use counseling). Maximal earnings generally range from $200 to $1,200 per patient over a 12 week period (Davis et al., 2016) and programmatic costs also include provider time and substance use testing supplies and/or services (J. L. Sindelar & Ball, 2010). Most commonly, patients will undergo objective drug and/or alcohol testing on two or more separate occasions per week and will be eligible for program-based incentives such as monetary vouchers (voucher-based CM) or opportunities to win prizes (prize-based CM) when objective testing is negative for targeted substances. An escalating reinforcement schedule, whereby the magnitude and/or probability of program-based incentives increases with consecutive substance-negative results, is also commonly used to reinforce longer periods of continuous abstinence (Roll & Higgins, 2000; Roll et al., 2000). While the cost-effectiveness of these methods is generally supported by the existing literature on CM (Olmstead & Petry, 2009; J. Sindelar, Elbel, & Petry, 2007; J. L. Sindelar, Olmstead, & Peirce, 2007), relatively few clinical settings have been evaluated in this respect – making the generalizability of these findings unclear (Shearer, Tie, & Byford, 2015). In addition, considerable variability in CM cost-effectiveness has also been noted across clinical contexts, suggesting a potential need for improved treatment targeting (Olmstead, Sindelar, & Petry, 2007).
Evidence regarding the longevity of CM treatment effects at post-treatment is similarly mixed. While a recent meta-analysis by Benishek and colleagues (2014) identified no sustained benefit of CM at 6 months post-treatment, a small-to-medium treatment effect size was identified for more proximal follow-up intervals. Similarly, a meta-analysis by Davis et al. (2016) identified a small post-treatment effect size when a range of post-treatment follow-up latencies was included. Results of previous work further suggest that the longevity of CM effects may be bolstered through combination with other treatment modalities (e.g., cognitive behavioral therapy; Davis et al., 2016; Epstein, Hawkins, Covi, Umbricht, & Preston, 2003; Rawson et al., 2002) or by prolonging the active CM treatment interval (Silverman, Robles, Mudric, Bigelow, & Stitzer, 2004). While individual difference factors could potentially inform strategies to prolong CM treatment effects (e.g. deliberately pairing CM with other treatment options or adjusting the duration of active CM treatment), measurement-based approaches to these aspects of CM-related treatment planning have not yet been developed.
As an adjunctive intervention, the potential benefit of adding CM must also be carefully considered relative to standard care or other active control conditions. While existing evidence suggests that patients achieve improved abstinence and attendance in CM, the degree to which such outcomes are improved relative to standard care is likely to vary considerably from patient-to-patient (as is typically the case for any evidence-based treatment). In particular, evidence of a “ceiling effect” has been noted in some patient samples, whereby both standard care and CM (as an adjunct to standard care) are associated with similarly favorable outcomes (Hagedorn et al., 2013; Petry et al., 2004). Other patients, by contrast, may exhibit similarly poor outcomes in standard outpatient care and CM – potentially indicating that a higher level of care (or higher magnitude version of CM) is necessary (Petry, Barry, Alessi, Rounsaville, & Carroll, 2012; Silverman, Chutuape, Bigelow, & Stitzer, 1999). Indeed, recent work by Petry and colleagues (2012) suggests it may be possible to strategically adapt CM to individual patient needs; for example, by offering higher magnitude CM reinforcement to patients with a poorer prognosis at pre-treatment. By identifying individual difference factors that reliably predict CM treatment response it may therefore be possible to allocate CM resources more efficiently; for example, omitting CM when standard care is likely to be sufficient or strategically offering enhanced versions of CM, when indicated.
Patient characteristics associated with enhanced or diminished benefits of CM may further reveal opportunities to strategically pair CM with other interventions; for example, to remediate specific cognitive or coping skill deficits associated with poorer outcomes. In addition, CM is readily customizable and a variety of treatment parameters (e.g., program duration, reinforcement magnitude and frequency, schedule of escalating reinforcement) may be adapted based on patient-level features to better meet individual needs. However, despite the clear potential for individual difference factors to inform precision implementation of CM, relatively little research has yet been conducted to advance this aim. While many previous studies have examined patient-level factors in CM treatment response, relatively few have been considered as a basis for measurement-based implementation. In addition, a systematic review of individual difference factors in CM has not yet been conducted but would serve to clarify promising directions for this work. To address this need, the current review aims to evaluate and summarize the existing literature regarding individual difference factors in CM, focusing on stimulant and/or opioid using populations, for whom the evidence base for CM is especially well-established (Prendergast, Podus, Finney, Greenwell, & Roll, 2006). We specifically aim to (1) describe the state of the existing literature, (2) determine what types of patient-level data have been considered with respect to prospective treatment response (e.g., demographic variables, comorbid conditions, cognitive performance measures), and (3) identify variables that have demonstrated potential for consideration in future precision implementation of CM. Areas that are currently under-studied (e.g., biomarkers) with respect to individual differences in treatment response will also be highlighted as a potential focus for future research.
2. Material and Methods
A search of PubMed/MEDLINE was conducted to identify manuscripts published between January 1, 1995 and December 1, 2018 wherein the search term “Contingency Management” appeared in conjunction with any of the following terms: “individual differences,” “prediction or predictor,” “moderation analysis or moderator,” “mediation analysis or mediator,” “treatment outcome,” “recovery,” “relapse,” “stimulant,” “cocaine,” “methamphetamine,” “opiate,” “opioid,” or “heroin.” A filter was applied to exclude research on non-human animals. This query yielded a total of 648 unique peer-reviewed publications which were subsequently evaluated for inclusion. Individual difference factors investigated in relation to substance use during the CM treatment interval were the primary focus of the current review; however, factors relevant to treatment retention are also summarized, when available. Publications and analyses exclusively targeting individual differences in longer term outcomes (e.g., six months post-treatment) were not included.
A schematic representation of the evaluation process and outcomes is provided in Figure 1. All publications considered for inclusion were English-language, peer-reviewed, original research articles. Publications selected for inclusion: (1) reported a primary or secondary analysis targeting individual difference factors as predictors, mediators, or moderators of CM treatment response, (2) reported a total sample comprised of ≥ 20 adult stimulant and/or opioid users treated in an outpatient setting, (3) evaluated a standard voucher or prize-based version of CM in which stimulant and/or opioid abstinence was reinforced as a primary target behavior, and (4) evaluated individual difference factors in CM outcome against a non-CM (e.g., “standard care” or noncontingent reinforcement) or low-dose CM (e.g., low magnitude CM) control condition. A minimum total sample size of 20 was specified due to our interest in individual difference factors and approximately corresponds to that necessary to detect a large effect size correlation (r = 0.60, α = 0.05) at 80% power. This threshold also ensures that studies with larger sample sizes, powered to detect smaller effect sizes, are included. Publications describing CM in the context of the Community Reinforcement Approach (CRA; a cognitive-behavioral intervention aiming to reinforce sobriety by restructuring naturalistic environmental contingencies) were included if CM-specific effects were evaluated against a CRA-only control condition. An appropriate control condition was considered necessary to distinguish individual difference factors specifically related to CM treatment response from those associated with substance use treatment outcomes more generally. However, to highlight promising new directions, several additional candidate predictors from studies without such a control condition, are also considered in the Discussion.
Figure 1.
Summary of literature search and review procedure.
With respect to individual differences, we were specifically interested in baseline patient characteristics and/or treatment-related behaviors that have been examined in relation to CM treatment response. Studies investigating differential CM outcomes as a product of treatment modification were excluded unless the design additionally accommodated investigation of individual differences in response to conventional prize- or voucher-based CM. For example, studies wherein CM was experimentally combined with other interventions (e.g., cognitive behavioral therapy or pharmacotherapy) were only included if CM-specific individual difference factors were considered (for example, through comparison against the other intervention alone). Other nonstandard, experimental implementations of CM (such as computerized-, online-, or group-based delivery of the intervention1) were excluded from the current review, as were programs exclusively offering non-monetary/prize-based reinforcement (e.g., housing, employment, or treatment-related privileges). CM programs reinforcing additional target substances or behaviors (i.e., beyond stimulant and/or opioid abstinence) were, however, included – provided that stimulant and/or opioid abstinence was a primary focus of reported results.
We chose to focus on populations that have been a primary emphasis for CM research and dissemination efforts: stimulant users, opioid users, and individuals who use both stimulants and opioids. Samples including patients receiving medication-assisted treatment were considered eligible if CM-specific effects were investigated within (e.g., methadone maintenance with versus without CM) or across (CM in combination with either methadone or buprenorphine versus methadone- or buprenorphine-maintenance only) medication-related treatment conditions. Research samples comprised of patients in medication-assisted treatment programs (e.g., patients engaged in methadone maintenance therapy) are identified as such in tables and text summarizing the existing literature. Research focusing on individual difference factors in special populations (e.g., patients with serious mental illness, pregnant women, men who have sex with men) was also included but is described in reference to the specific population of interest in these cases.
Out of the 648 unique publications identified in our original search, 392 publications were eliminated following review of article titles and abstracts during preliminary screening procedures. The remaining 256 articles were subject to secondary evaluation procedures, involving review of methods, results, and references. An additional 219 articles were excluded at this stage and two publications were added (not identified in our search results), resulting in a final sample of 39 eligible publications. Each of these publications was subsequently re-reviewed and information regarding the sample, design, treatment conditions, and relevant findings was extracted; this information is summarized in Table 1. It is noted that information included in Table 1 focuses on primary findings with respect to individual difference factors. In many cases additional individual difference factors were included as covariates but these methodological details have been omitted for the sake of brevity and clarity.
Table 1.
Summary of literature examining individual difference factors in Contingency Management treatment response
| Study | Sample | Treatment | Study Design | Relevant Findings |
|---|---|---|---|---|
| Epstein & Preston, 2003 | 408 methadone-maintained outpatients with continued heroin or cocaine use
|
Voucher-based CM targeting opiates or cocaine (vs. non-contingent voucher control or modified CM) ± experimental treatment condition (e.g., methadone dose increase, CBT)
|
Secondary analysis of 3 clinical trials examining treatment outcome by cannabis use frequency during treatment (nonuser, infrequent user, frequent user), as well as lifetime history of cannabis use diagnosis. |
|
| Gonzalez, et al., 2003 | 149 buprenorphine-maintained outpatients with recent cocaine use
|
Voucher-based CM targeting opiates and cocaine or non-contingent voucher control ± desipramine
|
Secondary analysis of a clinical trial of desipramine and/or CM, comparing treatment outcomes in patients with and without a lifetime history of major depressive disorder. |
|
| Messina, Farabee, & Rawson, 2003 | 108 methadone-maintained outpatients with cocaine use disorder
|
Voucher-based CM targeting stimulant use and/or CBT + methadone maintenance vs. methadone maintenance only.
|
Secondary analysis of a clinical trial of CM and/or CBT, comparing treatment outcomes in patients with and without antisocial personality disorder (ASPD). |
|
| Sofuoglu, et al., 2003 | 151 buprenorphine-maintained outpatients with recent cocaine use
|
Voucher-based CM targeting opiates and cocaine or non-contingent voucher control ± desipramine
|
Secondary analysis of a clinical trial of desipramine and/or CM, examining treatment outcomes in relation to initial cocaine urine toxicology and self-reported past-month cocaine use frequency at baseline |
|
| Petry, et al., 2004 | 120 outpatients with recent cocaine use
|
Prize-based CM with high (max: $240; n = 38) versus low magnitude (max: $80; n = 45) reinforcement for cocaine, alcohol, and opiate abstinence, and/or completion of goal-related target behaviors and/or standard care
|
Comparative effectiveness trial of high (max: $240) versus low (max: $80) magnitude prize-based CM, wherein initial urine toxicology (cocaine positive versus cocaine negative) were examined as a factor in treatment response. |
|
| Ledgerwood & Petry, 2006 | 142 outpatients with cocaine and/or opioid use disorder
|
Prize- or Voucher-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors and/or standard care
|
Secondary analysis of CM clinical trial in which pretreatment motivation (i.e., readiness to change) was investigated as a factor in treatment outcome. |
|
| Ford, et al., 2007 | 130 outpatients with cocaine and/or opioid use disorder
|
Prize- or Voucher-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors and/or standard care
|
Secondary analysis of CM clinical trial in which trauma exposure / number and complexity of PTSD symptoms were investigated as factors in treatment outcome. The interaction between complex PTSD symptoms and treatment condition was specifically considered. |
|
| Ghitza, Epstein, & Preston, 2007 | 690 methadone-maintained outpatients with continued heroin or cocaine use
|
Voucher-based CM targeting opiates or cocaine (vs. non-contingent voucher control or modified CM) ± experimental treatment condition (e.g., methadone dose increase, CBT)
|
Secondary analysis of 3 CM clinical trials examining treatment outcome by cannabis use disclosure (reporter versus nonreporter) amongst patients who tested positive for cannabis during the treatment interval. |
|
| Stitzer, Peirce, et al., 2007 | 386 methadone-maintained outpatients with continued stimulant use
|
Prize-based CM targeting stimulants and/or usual care (psychosocial counseling)
|
Secondary analysis of CM clinical trial comparing patient outcomes based on initial stimulant urine toxicology. |
|
| Stitzer, Petry, et al., 2007 | 414 stimulant users seeking outpatient treatment for recent or continued stimulant use
|
Prize-based CM targeting stimulants and/or usual care (psychosocial counseling)
|
Secondary analysis of CM clinical trial comparing patient outcomes based on initial stimulant urine toxicology. |
|
| Weinstock, et al., 2007 | 393 outpatients with cocaine and/or opioid use disorder
|
Voucher- or Prize-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors and/or standard care
|
Secondary analysis of 3 CM clinical trials, investigating psychiatric symptom severity (low, moderate, high; based on ASI psychiatric composite score) as a factor in treatment outcome. |
|
| Ghitza, Epstein & Preston, 2008 | 361 methadone-maintained outpatients with continued heroin or cocaine use
|
Voucher- or Prize-based CM targeting opiates and/or cocaine versus non-contingent voucher or prize control
|
Secondary analysis of 2 CM clinical trials examining treatment outcome by self-reported illicit benzodiazepine use during the previous 30 days at baseline (user versus nonuser). |
|
| Rash, Alessi & Petry, 2008a | 393 outpatients with cocaine and/or opioid use disorder
|
Prize- or Voucher-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors and/or standard care
|
Secondary analysis of 3 CM clinical trials in which alcohol dependence (diagnostic criteria met vs. diagnostic criteria not met) was investigated as a factor in treatment outcome. |
|
| Rash, Alessi & Petry, 2008b | 393 outpatients with cocaine and/or opioid use disorder
|
Prize- or Voucher-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors and/or standard care
|
Secondary analysis of 3 CM clinical trials in which prior substance use treatment history (0-1 versus ≥2 prior treatment attempts) was investigated as a factor in treatment outcome. |
|
| Barry, Sullivan & Petry, 2009 | 193 methadone-maintained cocaine dependent outpatients
|
Prize- or Voucher-based CM targeting cocaine (group attendance and opiate abstinence also reinforced in one or more included study) and/or standard care.
|
Secondary analysis of 3 CM clinical trials comparing patient outcomes based on self-identified race/ethnicity (African American, Hispanic, or White). |
|
| Rash, Olmstead & Petry, 2009 | 393 outpatients with cocaine and/or opioid use disorder
|
Prize- or Voucher-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors and/or standard care
|
Secondary analysis of 3 CM clinical trials in which past-year income and income type (earned, unstable, illegal) were investigated as factors in treatment outcome. |
|
| Mancino et al., 2010 | 137 outpatients with cocaine and opioid use disorder in medication-assisted treatment
|
Voucher-based CM targeting cocaine and opiates vs. non-contingent voucher control +/− high- or low-dose levoalpha-acetyl methadyl (LAAM) maintenance
|
Secondary analysis of a clinical trial of LAAM and CM in which current PTSD diagnosis was considered as a factor in treatment outcome. |
|
| Petry & Alessi, 2010 | 393 outpatients with cocaine and/or opioid use disorder
|
Prize- or Voucher-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors versus standard care
|
Secondary analysis of 3 CM clinical trials in which past month gambling participation (self-identified gambler versus non-gambler) was investigated as a factor in treatment outcome. |
|
| Alessi, Rash, & Petry, 2011 | 393 outpatients with cocaine and/or opioid use disorder
|
Prize- or Voucher-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors versus standard care
|
Secondary analysis of 3 CM clinical trials in which past month cannabis use (self-identified user versus non-user) was investigated as a factor in treatment outcome. |
|
| Byrne & Petry, 2011 | 193 methadone-maintained cocaine dependent outpatients
|
Prize- or Voucher-based CM targeting cocaine (group attendance and opiate abstinence also reinforced in one or more included study) and/or standard care.
|
Secondary analysis of 3 CM clinical trials in which past year alcohol dependence (present versus absence) was investigated as a factor in treatment outcome. |
|
| Petry, Ford, & Barry, 2011 | 393 outpatients with cocaine and/or opioid use disorder
|
Prize- or Voucher-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors and/or standard care
|
Secondary analysis of 3 CM clinical trials in which history of sexual abuse (present versus absent) was investigated as a factor in treatment outcome. |
|
| Petry, Rash, & Easton, 2011 | 393 outpatients with cocaine and/or opioid use disorder
|
Prize- or Voucher-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors and/or standard care.
|
Secondary analysis of 3 CM clinical trials in which legal problems (present versus absent) were investigated as a factor in treatment outcome. |
|
| Schottenfeld, Moore, & Pantalon, 2011 | 145 pregnant women and/or custodial mothers of young children with cocaine use disorder
|
Voucher-based CM targeting cocaine abstinence or non-contingent voucher control paired with either 12-Step Facilitation or Community Reinforcement Approach (+ Voucher-based reinforcement for attendance across all conditions)
|
Randomized trial of CM as an adjunct to either of two primary behavioral treatment modalities (12 Step Facilitation, Community Reinforcement Approach) in pregnant women and young mothers; pregnancy status considered as a factor in treatment outcome. |
|
| Washio, et al, 2011 | 36 outpatients with cocaine use disorder
|
High (n = 18) versus low magnitude (n = 18) Voucher-based CM targeting cocaine abstinence in the context of the Community Reinforcement Approach
|
Secondary analysis of CM clinical trial utilizing high- versus low-magnitude voucher reward; delay discounting rate (measured during treatment) was investigated as a factor in treatment outcome. |
|
| Weiss & Petry, 2011 | 393 outpatients with cocaine and/or opioid use disorder
|
Prize- or Voucher-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors and/or standard care
|
Secondary analysis of 3 randomized trials of CM efficacy in which age group (under 30, 30-40, or over 40) was investigated as a factor in treatment outcome. |
|
| DeFulio. et al., 2013 | 415 stimulant users seeking outpatient treatment for recent or continued stimulant use
|
Prize-based CM targeting stimulants and/or usual care (psychosocial counseling)
|
Secondary analysis of CM clinical trial comparing patient outcomes based on whether or not treatment was preceded by a referral from the criminal justice system. |
|
| García-Fernández, et al., 2013 | 108 outpatients with cocaine use disorder
|
Voucher-based CM targeting cocaine abstinence and/or Community Reinforcement Approach
|
Randomized trial of the Community Reinforcement Approach with and without vouchers wherein presence of depressive symptoms at baseline was investigated as a factor in treatment outcome. |
|
| Rash, Andrade & Petry, 2013 | 418 outpatients with cocaine use disorder
|
High or low magnitude prize-based CM targeting alcohol, cocaine, and opiate abstinence or attendance and/or standard outpatient treatment
|
Secondary analysis of a clinical trial of CM variants, in which income amount and source during the treatment interval were investigated as a factor in treatment outcome. |
|
| Secades-Villa, et al., 2013 | 118 outpatients with cocaine use disorder
|
Voucher-based CM targeting cocaine abstinence and/or Community Reinforcement Approach
|
Randomized trial of the Community Reinforcement Approach with and without vouchers wherein self-reported income during the month preceding treatment was investigated as a factor in treatment outcome. |
|
| Weiss & Petry, 2013 | 189 methadone-maintained outpatients with continued cocaine use
|
Prize- or Voucher-based CM targeting cocaine (group attendance and opiate abstinence also reinforced in one or more included study) and/or standard care.
|
Secondary analysis of 3 CM clinical trials in which age group (under 30, 30-40, or over 40) was investigated as a factor in treatment outcome. |
|
| Weiss, et al., 2014 | 428 outpatients with cocaine use disorder
|
High or low magnitude prize-based CM targeting alcohol, cocaine, and opiate abstinence or attendance and/or standard outpatient treatment
|
Secondary analysis of a clinical trial of CM variants, in which age of first cocaine use (≤14 versus ≥15) was investigated as a factor in treatment outcome. |
|
| Burch, Rash, & Petry, 2015 | 323 methadone-maintained outpatients with cocaine use disorder
|
Prize- or Voucher-based CM targeting cocaine (+/− opiates and/or alcohol), completion of goal-related activities, and/or treatment attendance versus standard outpatient treatment
|
Secondary analysis of 4 CM clinical trials in which patient sex was investigated as a factor in treatment outcome. |
|
| Montgomery, Carroll, & Petry, 2015 | 847 outpatients with past year cocaine use disorder
|
Prize- or Voucher-based CM targeting cocaine, opiates, and alcohol, completion of goal-related activities, and/or treatment attendance versus standard outpatient treatment
|
Secondary analysis of 6 CM clinical trials in which race (White versus African American) and initial urine toxicology were investigated as factors in treatment outcome. |
|
| Blanken, et al., 2016 | 214 outpatients with opioid use disorder and continued cocaine use in medication-assisted treatment
|
Voucher-based CM targeting cocaine abstinence and/or medication-assisted treatment for opioid use disorder (methadone and diacetylmorphine)
|
Randomized trial of CM as an adjunct to medication-assisted treatment in heroin users, in which two potential moderators of treatment response (intention to achieve cocaine abstinence and previous abstinence-oriented treatment at baseline) were investigated. |
|
| Rash, Burki, Montezuma-Rusca, & Petry, 2016 | 493 female outpatients with cocaine use disorder
|
Prize- or Voucher-based CM targeting cocaine, opiates, and alcohol, completion of goal-related activities, and/or treatment attendance versus standard outpatient treatment
|
Secondary analysis of 5 CM clinical trials in which history of trading sex for drugs or money (present vs absent) was investigated as a factor in treatment outcome. |
|
| Burch, Rash, & Petry, 2017 | 432 outpatients with cocaine use disorder
|
Prize- or Voucher-based CM targeting cocaine, opiates, and alcohol, completion of goal-related activities, and/or treatment attendance versus standard outpatient treatment
|
Secondary analysis of 4 CM clinical trials in which patient HIV status was investigated as a factor in treatment outcome. |
|
| Ginley et al., 2017 | 323 methadone-maintained outpatients with cocaine use disorders
|
Prize-based CM targeting cocaine (alcohol abstinence also reinforced in one included study) and/or standard care.
|
Secondary analysis of 4 CM clinical trials in which legal problems (present versus absent) were investigated as a factor in treatment outcome. |
|
| Rash, Alessi, & Petry, 2017 | 355 outpatients with cocaine use disorder
|
Prize- or Voucher-based CM targeting alcohol, cocaine, opiates, and/or completion of goal-related target behaviors and/or standard outpatient treatment
|
Secondary analysis of 3 CM clinical trials in which utilization of homeless- or recovery-focused housing programs was investigated as a factor in treatment outcome. |
|
| Oluwoye, et al., 2018 | 176 outpatients with stimulant use disorder and serious mental illness
|
Prize-based CM targeting stimulants (primary) and other substances (secondary) or non-contingent prize draw control
|
Secondary analysis of a CM clinical trial in patients with serious mental illness (SMI) in which SMI diagnosis and initial urine toxicology were considered as factors in treatment outcome. |
|
3. Results
Of the 39 total publications selected for inclusion, nearly all (34/39) reported a secondary analysis of one or more existing datasets and the majority of publications (25/39) represented the work of a single research group. It was additionally identified that nearly half of the selected publications (16/39) included data from three early trials of CM (specifically, Petry et al., 2006; Petry, Alessi, Marx, Austin, & Tardif, 2005; Petry et al., 2004), representing a combined sample of 393 treatment-seeking cocaine and/or opioid using outpatients. Consistent with the prevalence of secondary analyses in the existing literature, data on individual difference factors were primarily derived from clinical inventories and psychodiagnostic assessment tools, commonly used in clinical trial intake procedures. For example, 30.8% of publications utilized data collected in the context of the Addiction Severity Index and 12.8% utilized data collected in the context of the Semi-Structured Clinical Interview for DSM-IV. Another 15.4% of publications examined the predictive utility of substance use data collected as part of the research and/or treatment protocol (e.g., baseline urine toxicology results, frequency of cannabinoid-positive urine specimens during treatment). The majority of previous research therefore appears to reflect an a posteriori approach to investigating individual difference factors in CM treatment response and relatively few studies have included hypothesis-driven measures for the express purpose of outcome prediction.
Review of publications selected for inclusion revealed a variety of individual difference factors, previously investigated with respect to Contingency Management treatment outcomes. These factors generally represented one of the following five categories: 1) pre- and/or in-treatment motivation or substance use, 2) addiction chronicity and comorbidity, 3) psychiatric comorbidity and severity, 4) medical, legal, and sociodemographic considerations, and 5) cognitive-behavioral characteristics. The most studied category (reported in 15/39 or 38% of total publications) was medical, legal, and sociodemographic factors. By comparison, cognitive-behavioral characteristics were only considered in a single study selected for inclusion. All other findings were similarly distributed across the remaining three categories. Individual difference factors investigated in relation to CM treatment response are further summarized by category below. Findings regarding individual difference factors in CM are additionally summarized in Figure 2, which provides a snapshot overview of the current literature.
Figure 2.
Overview of current evidence regarding individual difference factors in Contingency Management treatment response.
3.1. Motivation and Use Prior to and/or During Treatment
Two publications examined pre-treatment motivation to change substance use as a factor in treatment outcome. Ledgerwood and Petry (2006) found that the proportion of patients achieving eight or more weeks of continuous abstinence during treatment did not significantly differ between subgroups with high versus low “readiness to change” scores. While this was the case for both CM and standard care, CM patients achieved longer durations of continuous abstinence overall, suggesting CM effectively prolonged abstinence regardless of pretreatment motivation to change. Blanken, Hendriks, Huijsman, van Ree, and van den Brink (2016), however, determined that self-reported intent to stop using cocaine at pretreatment was a significant factor in CM treatment outcomes in heroin dependent individuals receiving medication-assisted treatment. While intent to stop using cocaine was not associated with differential outcomes in medication-assisted treatment alone, those with intent to stop who received CM in combination with medication-assisted treatment achieved significantly longer periods of continuous cocaine abstinence.
Several additional publications examined treatment outcomes in relation to pretreatment use – most commonly indexed by urine toxicology results at treatment outset as a proxy for use recency and/or severity. Consistent with findings from the broader substance use treatment literature (Alterman et al., 1997; Alterman, McKay, Mulvaney, & McLellan, 1996; Ehrman, Robbins, & Cornish, 2001), individuals testing positive for targeted substances at baseline were generally found to demonstrate poorer outcomes in the CM treatment studies selected for inclusion (Petry et al., 2004; Sofuoglu, Gonzalez, Poling, & Kosten, 2003; Stitzer, Peirce, et al., 2007; Stitzer, Petry, et al., 2007). However, the degree to which patients with initially substance-positive versus -negative urine toxicology differentially benefited from CM versus non-CM treatment was variable and may depend upon treatment-, population-, and methods-related factors. For example, Sofuoglu, et al. (2003) identified poorer outcomes in initially cocaine-positive patients in medication-assisted treatment for opioid dependence but no main effect of CM (relative to non-contingent voucher control) or interaction between initial urine toxicology and CM treatment condition. Similar results were noted for self-reported frequency of past month cocaine use, suggesting poorer outcomes in heavy users, irrespective of CM involvement. However, these results may also reflect the choice of outcome measure, control condition, or reinforcement schedule used in the study, as suggested by the authors. Of note, while Stitzer, Peirce, and colleagues (2007) also found no interaction between treatment condition and initial urine toxicology in stimulant users receiving methadone maintenance, this result reflected comparable improvement in abstinence outcomes for initially stimulant-positive and -negative patients in CM versus standard care. This study, however, utilized a different version of CM (prize-based rather than voucher-based), as well as a different reinforcement schedule, control group, and outcome measure, which may contribute to divergent results.
Interactions between initial urine toxicology and CM treatment condition have also been described elsewhere in the literature. For example, Petry et al. (2004) identified differential effects of CM treatment by initial urine toxicology, with initially cocaine-positive patients demonstrating improved abstinence outcomes in prize-based CM relative to standard care and initially cocaine-negative patients demonstrating comparable outcomes across treatment conditions. Importantly, this effect was further noted to be specific to a higher magnitude version of CM (i.e., $240 in average maximal earnings), while abstinence outcomes in a low magnitude version of CM ($80 in average maximal earnings) did not differ from standard care, for either positive or negative initial urine toxicology subgroups. Stitzer, Petry, et al. (2007), however, identified a contrasting interaction in a similar sample of outpatient stimulant users, wherein initially stimulant-negative patients exhibited improved abstinence outcomes in prize-based CM versus standard care but initially stimulant-positive patients demonstrated similar outcomes across treatment conditions. Of note, this interaction was only significant when missing urine samples were coded as positive for patient-level treatment outcomes. Improved early retention of initially stimulant-positive patients in the Petry and colleagues (2004) study has been suggested as a potential explanation for contrasting results.
Importantly, Petry and colleagues (2004) also demonstrated that patients with cocaine-negative initial urine toxicology achieved similar outcomes in CM and standard care conditions irrespective of CM magnitude, suggesting no benefit of CM in this subgroup. At odds with this finding, however, both Stitzer, Petry, et al. (2007) and Stitzer, Peirce, et al. (2007) identified that initially stimulant-negative patients did achieve improved abstinence in CM (relative to standard care). Of note, Petry et al. (2004) reported a high occurrence of substance-negative specimens during treatment (≥90%) for cocaine using outpatients with negative urine toxicology at baseline, irrespective of treatment condition. By comparison, lower rates of negative specimens in the initially negative subsample were identified by both Stitzer, Peirce, et al. (82% stimulant and alcohol negative) and Stitzer, Petry, et al. (54% stimulant and alcohol negative) and significant CM treatment effects were identified in both these studies. Overall, the absence of significant CM treatment effects in initially cocaine-negative patients in the Petry et al. (2004) study would therefore appear consistent with a “ceiling effect,” whereby high rates of overall abstinence precluded detection of further treatment-related gains in CM. However, it is also possible that differences in the abstinence contingencies employed in these studies may contribute to discordant findings. Specifically, Petry et al. awarded prize draws for abstinence from cocaine, alcohol, and opiates, while both Stitzer et al. publications report data from a parent trial in which abstinence from only stimulants and alcohol was necessary to earn prize draws (with additional bonus draws awarded for opiate- and/or cannabinoid-negative urines).
Use of non-targeted substances during treatment was also considered as a factor in treatment outcome in two publications selected for inclusion. Specifically, Epstein & Preston (2003) identified that in-treatment frequency of cannabinoid-positive urines in methadone-maintained polydrug abusers was not significantly associated with abstinence from targeted substances (cocaine or opiates) or retention in CM or non-CM treatment conditions. Ghitza and colleagues (2007), however, identified that methadone-maintained patients who used cannabis during treatment and were not assigned to CM, exhibited elevated rates of cocaine and opiate use if they did not disclose their use when compared against cannabis users who did disclose use. Importantly, this effect was absent in CM wherein use of targeted substances was comparable for both disclosing and non-disclosing cannabis users. Taken together, these results suggest that treatments targeting stimulant and opioid use can be effective in patients who continue to actively use cannabis during treatment and further that CM may be particularly beneficial in those who fail to disclose ongoing cannabis use.2
3.2. Addiction Chronicity and Comorbidity
Co-occurring substance use disorders may also contribute to differential treatment outcomes (Dutra et al., 2008). Consistent with the established effectiveness of CM in patients with co-occurring opioid and cocaine use disorders (Petry & Martin, 2002; Schottenfeld et al., 2005), CM is also supported for patients with comorbid alcohol dependence. Specifically, CM has been associated with longer durations of abstinence from cocaine, opioids, and alcohol regardless of current alcohol dependence (Rash, Alessi, & Petry, 2008a) and has similarly been shown to support longer durations of cocaine abstinence in methadone-maintained patients with and without past year alcohol dependence (Byrne & Petry, 2011). In addition, Petry & Alessi (2010) demonstrated that benefits of CM (i.e., reduced cocaine, opioid, and alcohol use and improved treatment retention) were comparable for patients with and without past month gambling behavior and a significant reduction in gambling behavior at post-treatment was identified in CM recipients only.
Pretreatment use of other non-targeted substances has also been associated with a differential response to CM in some cases. Patients with self-reported cannabis use during the month preceding treatment, for example, exhibited improved benefits of CM with respect to treatment retention (relative to non-cannabis users), as well a comparable response to CM with respect to improved abstinence from targeted substances (Alessi, Rash, & Petry, 2011). Past month benzodiazepine use was, by contrast, associated with lower rates of cocaine abstinence amongst methadone-maintained patients (across conditions) and those without past month benzodiazepine use exclusively demonstrated increased rates of opioid use disorder remission in CM (relative to a non-contingent control condition; Ghitza, Epstein, & Preston, 2008).
Number of previous treatment episodes may also reflect patient-level substance use severity and chronicity, relevant to substance use treatment outcomes. Consistent with greater CM benefits in more severe clinical presentations, Rash, Alessi, and Petry (2008b) identified improved retention and longer durations of abstinence in CM for patients with two or more previous treatment episodes. Those with fewer previous treatment episodes, by comparison, achieved similar outcomes in CM and standard care with respect to retention and only modest benefits of CM with respect to improved abstinence. Blanken et al. (2016), however, found no significant moderating effect of previous abstinence-oriented treatment experience (i.e., any versus none) on treatment outcomes in opioid dependent patients receiving CM for co-occurring cocaine use in conjunction with medication-assisted treatment. Benefits of CM with respect to duration of cocaine abstinence and treatment retention were also comparable for cocaine users with an early (prior to age 15) versus later (age 15 or later) onset of cocaine use (L. M. Weiss & Petry, 2014) – the former generally being associated with a more severe and chronic course of substance use problems (Jordan & Andersen, 2017; Robins & Przybeck, 1985).
3.3. Psychiatric Severity and Comorbidity
Greater severity of comorbid psychiatric symptoms has previously been associated with poorer outcomes in substance use treatment (McLellan, Luborsky, Woody, O’Brien, & Druley, 1983) and may similarly impede treatment response in CM. Weinstock, Alessi, and Petry (2007), however, demonstrated that patients consistently achieved improved outcomes (i.e., higher rates of prolonged abstinence) in CM relative to standard care, regardless of psychiatric symptom severity. Moreover, while patients with higher levels of psychiatric symptom severity exhibited poorer retention in standard treatment, somewhat improved retention with greater symptom severity was identified for patients in CM.
Unique benefits of CM have similarly been identified in patients with specific psychiatric diagnoses and risk factors. For example, while Petry, Ford, & Barry (2011) found that all patients exhibited improved outcomes in CM relative to standard care, patients with a history of sexual abuse demonstrated greater benefits of CM with respect to longer durations of abstinence than patients without such a history; an effect absent in standard care alone. Patients with a history of major depressive disorder have also demonstrated improved abstinence outcomes in CM (specifically, evidence of a more rapid reduction in use); although, poorer retention was also noted for these patients in CM, relative to other treatment conditions (Gonzalez, Feingold, Oliveto, Gonsai, & Kosten, 2003). With respect to current depressive symptoms, however, work by Garcia-Fernandez, Secades-Villa, Garcia-Rodriguez, Pena-Suarez, and Sanchez-Hervas (2013) supports both improved abstinence and retention outcomes in CM relative to non-CM treatment, regardless of depressive symptoms at treatment onset. Similarly, while a current diagnosis of PTSD was associated with poorer outcomes in cocaine and opioid dependent patients receiving medication-assisted treatment, patients with PTSD had similar abstinence and retention outcomes to those without PTSD when assigned to CM – suggesting unique benefits in this population (Mancino, McGaugh, Feldman, Poling, & Oliveto, 2010). This may additionally be the case for methadone-maintained patients with antisocial personality disorder, for whom abstinence outcomes (i.e., total substance-negative urines during treatment) were significantly improved, relative to patients without antisocial personality disorder, in CM but not other treatment conditions (i.e., CM + CBT, CBT, or methadone maintenance only; Messina, Farabee, & Rawson, 2003).
By comparison, only a single publication reported evidence of diminished CM outcomes in relation to a psychiatric comorbidity factor. Specifically, Ford, Hawke, Alessi, Ledgerwood, and Petry (2007) identified that more complex PTSD symptoms (e.g., dissociation, somatization, alienation/distrust, risk-taking) were associated with poorer abstinence and retention outcomes in CM, but were not clearly related to these outcomes in standard care alone. It is, however, noted that patients with both high and low complexity PTSD presentations achieved longer periods of abstinence and treatment retention on average when assigned to CM. While a statistical comparison of outcomes in CM versus standard care was not conducted, reported findings are consistent with benefits of CM, regardless of PTSD symptom complexity, but with more pronounced benefits in patients without complex symptoms.
Interestingly, psychiatric comorbidity may also impact other proposed predictors of treatment outcome as demonstrated by Oluwoye and colleagues in patients with serious mental illnesses (2018). These authors identified that, initially stimulant-positive urine toxicology was associated with a higher likelihood of stimulant-positive specimens during treatment in patients with bipolar, major depressive, and schizophrenia spectrum disorders when controlling for treatment condition. When considering only patients receiving CM, by contrast, the relationship between initial urine toxicology and in-treatment urine results was no longer significant for patients with schizophrenia spectrum disorders. Indeed, initially stimulant-positive patients with schizophrenia spectrum disorders were somewhat less likely to submit positive samples during CM than initially stimulant-positive patients with bipolar disorders. Initially stimulant-negative patients with schizophrenia spectrum disorders, however, were over 11 times more likely to submit stimulant-positive samples during CM than patients with bipolar disorders. These results suggest the predictive utility of initial urine toxicology in CM may differ by diagnosis – at least among patients with serious mental illness.
3.4. Medical, Sociodemographic, and Legal Considerations
While medical, legal, and sociodemographic individual difference factors were most prevalent in the included literature, few were found to meaningfully impact CM treatment response. Indeed, despite the use of monetary and/or prize-based incentives in CM, income level – both during treatment and in the month and year preceding treatment – did not significantly interact with CM versus non-CM treatment conditions with respect to substance use outcomes (Rash, Andrade, & Petry, 2013; Rash, Olmstead, & Petry, 2009; Secades-Villa et al., 2013). A null effect of past month income on retention was also demonstrated (Secades-Villa et al., 2013); however, retention was not explicitly considered in studies targeting past year (Rash et al., 2009) and in-treatment (Rash et al., 2013) income as a factor in treatment response. Other factors including race/ethnicity (Barry, Sullivan, & Petry, 2009), HIV status (Burch, Rash, & Petry, 2017), history of trading sex for drugs or money (in a sample of female outpatients; Rash, Burki, Montezuma-Rusca, & Petry, 2016), and current pregnancy (within a sample of women who were pregnant and/or caring for young children; Schottenfeld, Moore, & Pantalon, 2011) similarly had no significant impact on treatment outcomes either within or across treatment conditions. Female sex (Burch, Rash, & Petry, 2015), absence of current legal problems (Ginley, Rash, Olmstead, & Petry, 2017; Petry, Rash, & Easton, 2011), and utilization of homeless or recovery-focused housing programs during treatment (Rash, Alessi, & Petry, 2017), by contrast, were associated with improved outcomes in CM, as well as other treatment conditions. Importantly, however, none of these variables interacted with treatment condition – consistent with a generalized impact on treatment outcome rather than a CM-specific effect.
Out of the 15 publications addressing medical, legal, and sociodemographic individual difference factors, only four identified evidence of differential treatment outcomes in CM relative to a comparison condition. First, DeFulio and colleagues (2013) identified that, while patients referred to treatment by the criminal justice system demonstrated improved outcomes overall, benefits of CM with respect to abstinence and retention were most pronounced in patients who did not endorse such a referral. Two additional publications focused on age-related effects. Specifically, Weiss and Petry (2011) found that while, patients in older age groups generally remained in treatment longer, benefits of CM with respect to retention, longest duration of abstinence, and proportion of negative urine samples provided during treatment were greatest in younger patients (specifically, those under the age of 40). However, subsequent work by Weiss and Petry (2013) in methadone-maintained patients identified evidence of the opposite; specifically, greater CM-related improvement in abstinence outcomes (i.e., longest duration of abstinence) in patients aged 40 or older. Of note, these two publications targeted different clinical populations (primarily cocaine dependent outpatients versus methadone-maintained patients with concurrent cocaine use) and in both cases the age group demonstrating greater benefits of CM had poorer abstinence outcomes in standard care alone.
A fourth publication by Montgomery, Carroll, and Petry (2015) identified a significant three-way interaction between race (African American, White), initial urine toxicology results (positive, negative), and treatment condition (CM, standard care). Specifically, while differences in abstinence and retention associated with CM versus standard care were comparable for African American and White patients who tested negative for cocaine at baseline, outcomes differed by race in patients who tested positive. White patients with cocaine-positive urine toxicology at baseline demonstrated significantly improved retention and abstinence outcomes in CM relative to standard care. However, African American patients exhibited no significant benefit of CM with respect to retention and only modest benefits with respect to abstinence outcomes. While African American and White patients with positive initial urine toxicology did not significantly differ with respect to self-reported use frequency during the month preceding treatment, a trend toward greater use frequency in African American patients was noted as a possible explanation for the observed interaction.
3.5. Cognitive-behavioral Markers
A single publication examined individual differences in cognitive-behavioral performance in reference to CM treatment response. In view of research and theory linking delay discounting behavior to pathological substance use (Bickel & Marsch, 2001; Bickel et al., 2007), Washio and colleagues (2011) examined decision-making in an intertemporal choice paradigm, wherein patients could elect to receive a smaller, immediately-available monetary reward or a delayed reward of greater value. Consistent with previous work demonstrating steeper devaluation of delayed reward in favor of more immediate pay-outs in chronic cocaine users (Heil, Johnson, Higgins, & Bickel, 2006), individuals exhibiting greater delay discounting behavior generally achieved shorter periods of continuous cocaine abstinence across CM treatment conditions offering high and low magnitude voucher rewards. However, while the interaction between delay discounting and treatment condition was nonsignificant, it was identified that greater delay discounting was associated with shorter abstinence duration in low magnitude (maximum: $499 over 12 weeks) CM only. The relationship between delay discounting and abstinence duration was not significant for a high magnitude version of CM (maximum: $1,995 over 12 weeks), which was generally associated with improved outcomes. These results suggest that patients with steeper delay discounting functions may require higher magnitude CM reinforcement to achieve comparable outcomes to those with less extreme discounting behavior.
4. Discussion
A variety of individual difference factors have been examined with respect to primary treatment outcomes (i.e., abstinence, retention) in CM. While most individual difference factors investigated had no significant impact on CM treatment response, several factors were associated with improved or diminished benefits of CM as an adjunctive intervention. Several factors associated with more severe and/or complex clinical presentations (i.e., testing positive for target substances at treatment entry, more previous treatment attempts, history of sexual abuse, diagnosis of antisocial personality disorder, failure to disclose ongoing cannabis use, history of major depressive disorder, current PTSD) have been associated with increased benefits of CM. Correspondingly, several factors associated with less severe/complex presentations (e.g., testing negative for target substances at treatment entry, fewer previous treatment attempts, no history of sexual abuse, etc.) have been associated with lesser benefits. However, evidence for factors studied repeatedly and across different contexts (e.g., initial urine toxicology) suggests results are variable and may depend upon specific treatment parameters (e.g., reward magnitude, reinforcement schedule) and patient characteristics (e.g., comorbidity, frequency of use). While study inclusion criteria specified herein were intended to limit some such methodological heterogeneity, the current review indicates that variation in population, setting, and intervention parameters may still impact results.
Other factors did not follow the general trend of improved CM responsivity in patients with more severe or complex presentations. For example, patients with higher PTSD symptom complexity had poorer outcomes in CM than patients with less complex symptoms (Ford et al., 2007) – although still appearing to benefit from CM relative to standard care alone. Patients without intent to stop use also had poorer outcomes in CM targeting cocaine but further exhibited no benefit of CM as an adjunct to medication-assisted treatment for opioid use disorder, while those with intent to stop using cocaine did significantly benefit (Blanken et al., 2016). On the other hand, patients with significant legal problems (specifically, those referred into treatment by the criminal justice system) demonstrated lesser benefits of CM but had better outcomes across treatment condition – likely reflecting additional contingencies in place due to patient legal status (DeFulio et al., 2013). Taken together, the existing literature on individual difference factors in Contingency Management treatment response in stimulant and/or opioid users does not yet support firm conclusions for immediate translation into clinical practice. However, the foundational work summarized in the current review highlights promising future directions for this developing area of research. With respect to clinical implications, CM was only rarely associated with no clinical benefit and rarer still with poorer outcomes. CM would therefore appear to be generally indicated for the populations and settings included in the current review. While there are also indications that CM may be especially beneficial in some cases, continued research will be necessary to confirm and clarify these findings; in particular, because previous results may reflect specific methodological and clinical conditions and may therefore fail to generalize to novel circumstances.
Some individual difference factors were, however, better studied than others. Initial urine toxicology was most frequently investigated as a potential prognostic marker of CM treatment response – directly evaluated in six separate publications (Montgomery et al., 2015; Oluwoye et al., 2018; Petry et al., 2004; Sofuoglu et al., 2003; Stitzer, Peirce, et al., 2007; Stitzer, Petry, et al., 2007) and included as a covariate in several others (e.g., Barry et al., 2009; Burch et al., 2017; Rash et al., 2017; Rash et al., 2009; L. Weiss & Petry, 2013). To our knowledge, initial urine toxicology is also the only factor that has previously been considered as the basis for prospective, measurement-based adaptation of CM in a clinical trial (N. M. Petry, D. Barry, et al., 2012). Based on previous work (Petry et al., 2004), patients who initially test negative for target substances may have a diminished response to CM because they achieve high rates of in-treatment abstinence in both standard care and CM (resulting in a “ceiling effect”). Patients who test positive for target substances at treatment entry, by comparison, may demonstrate enhanced benefits of CM as an adjunct to standard care; although, perhaps, only in higher magnitude versions of the intervention (e.g., offering a maximum of $240 in CM rewards, rather than $80). In the first effort to prospectively modulate reinforcement magnitude based on a candidate outcome predictor, Petry and colleagues (2012) identified unique benefits of a higher magnitude CM condition (i.e., $560 versus $250 in average maximal winnings) in initially cocaine-positive patients. Work by Washio and colleagues (2011) further suggests that individual differences in delay discounting behavior could similarly inform decision-making with respect to CM magnitude. While delay discounting has not yet been considered in the context of a prospective adaptive design, these authors identified poorer overall outcomes in patients who exhibited steeper discounting of future reward but this relationship vanished in a high magnitude version of CM, wherein high and low discounters achieved comparable outcomes.
Importantly, existing work suggests that features such as initial urine toxicology and delay discounting behavior could inform precision implementation of CM (for example, identifying individuals for whom high magnitude CM is most appropriate). However, as demonstrated by Montgomery, et al. (2015) and Oluwoye, et al., (2018), interactions between individual difference factors may complicate the use of such information in clinical decision-making. These authors found that factors including race and psychiatric comorbidity can affect the relationship between initial urine toxicology and later treatment outcomes – suggesting additional caution in the use of such information. Identifying and understanding clinically-relevant interactions between individual difference factors will ultimately be necessary to refine and optimize precision delivery of CM and should be prioritized in future work. In addition, it will also be necessary to consider and address ethical implications of precision CM delivery– particularly with respect to personalized adjustment of treatment parameters that are expected to impact patient preference (e.g., reward magnitude). While an assumption of equal services for equal need is compatible with precision implementation (wherein, personalization reflects differential need), the potential perception of such treatment as inequitable, unjust, and/or discriminatory is nevertheless problematic. Factors used as the basis for measurement-based customization (for example, substance-positive initial urine toxicology) could also potentially promote unwanted behaviors (e.g., pretreatment substance use) if known to the patient and under his/her own volition. Given these concerns, more implicit and/or immutable markers (e.g., cognitive or physiological measures, genetic markers) may be favorable as the basis for adaptation. Similarly, treatment adaptations with less expected impact on patient preference (e.g., different reinforcement schedules with equivalent overall reward value) may also have stronger potential for translation into clinical practice.
Overall, we reviewed individual difference factors in the following five categories: (1) motivation and substance use prior to and/or during treatment, (2) addiction comorbidity and chronicity, (3) psychiatric severity and comorbidity, (4) medical, legal, and sociodemographic factors, and (5) cognitive-behavioral characteristics. The majority of reported findings resulted from secondary analysis of previously collected datasets. This a posteriori approach to the examination of individual difference factors has likely shaped the current literature by constraining the breadth of patient-level data available for consideration. For example, while most publications focused on variables that are commonly evaluated in clinical trials (e.g., demographics, clinical diagnoses, symptom severity), only one study involved a hypothesis-driven cognitive-behavioral measure. Furthermore, the majority of previous work (comprising 64% of included publications) was conducted by a single research group and nearly half included data pooled from the same three early trials of CM in opioid and/or cocaine users. For these reasons, generalizability remains a potential concern, despite previous findings predominantly reflecting large, appropriately powered clinical samples.
Our review supports the need for more studies investigating hypothesis-driven predictors of CM treatment response, such as theoretically-informed psychometric, genetic, and/or neurobiological markers. While only a single study included herein investigated individual differences in a putative cognitive-behavioral trait related to addiction (delay discounting behavior; Washio et al., 2011), other potential psychometric and/or neural markers have been considered elsewhere. For example, Dean and colleagues (2009) investigated a variety of cognitive-behavioral performance measures in methamphetamine-dependent individuals receiving both CM and CBT and identified that baseline working memory and reaction time measures were associated with treatment completion. While the predictive utility of these measures was low and variables related to pre-treatment substance use were more strongly associated with both completion and abstinence, CM-specific treatment effects were not explicitly considered. Stotts and colleagues (2015) similarly examined patterns of subjective craving/withdrawal symptoms, affective experiences, coping style, and impulsivity in CM responders and non-responders in the absence of a control group. While no differences in craving/withdrawal, negative affect, or impulsivity were identified, CM non-responders were characterized by greater experiential avoidance and inflexibility in response to unpleasant drug-related thoughts, feelings, and sensations – suggesting these patients may benefit from supplementary coping skills training. Additional work (excluded herein due to absence of a comparison condition) further suggests that CM may especially improve early retention in high novelty seekers (Helmus et al. 2001) but that patients with significant anhedonia may exhibit a diminished treatment response (Wardle, et al., 2017). These findings may, however, reflect individual difference factors of more general relevance to substance use treatment response and their specificity to CM remains to be systematically evaluated.
Potential neuromarkers of substance use treatment response have also demonstrated promise in previous work but have not yet been considered in relation to CM-specific outcomes (e.g., in-treatment abstinence during CM relative to standard care). Xu and colleagues (2014), for example, identified that hippocampal volume mediates the positive relationship between pretreatment use frequency and use frequency during treatment for cocaine use disorder (using CM and/or CBT) but did not consider CM treatment outcomes, specifically. Similarly, Luo and colleagues (Luo, Martinez, Carpenter, Slifstein, & Nunes, 2014) examined dopamine binding potential and release capacity in the striatum as a potential prognostic indicator of treatment outcomes in CM, delivered in combination with the Community Reinforcement Approach. Consistent with reduced motivation and/or reward sensitivity in individuals with lower dopamine activity, patients with reduced dopamine release capacity in the ventral striatum were more likely to be “non-responders” (achieving less than a month of abstinence during treatment). Importantly, these authors also demonstrated that neuromarker data (i.e., ventral striatal dopamine release capacity) outperformed demographic and pre-treatment substance use severity variables in predicting treatment outcome, supporting the utility of such metrics in measurement-based care. CM treatment response was not, however, considered relative to a comparison condition in this work and it is therefore unclear to what degree striatal dopamine measures relate to outcomes in CM and/or the Community Reinforcement Approach, specifically, or treatment outcomes in cocaine users, more generally.
Factors associated with differential CM treatment response may additionally inform treatment modifications and/or combinations which may result in improved effectiveness. For example, consistent with evidence of improved CM outcomes in patients with increased striatal dopamine activity, pairing CM with the dopamine-enhancing medication, levodopa, has been demonstrated to increase in-treatment abstinence in cocaine users in previous work (Schmitz, Lindsay, Stotts, Green, & Moeller, 2010; Schmitz et al., 2008; although also see Wardle et al., 2017). While levodopa should be particularly efficacious in patients with deficient dopamine tone, associated biomarkers have not yet been considered as a basis for strategically pairing CM with levodopa therapy. Individual difference factors could similarly inform selection of other medication-assisted treatment options to boost CM benefits. CM has already acquired support as an adjunct to several established and experimental pharmacotherapy options for opioid and stimulant use disorders. However, efforts to optimize medication selection and dosage in the context of CM are relatively new and rarely leverage a measurement-based approach.
Benefits of CM may also be improved and prolonged when delivered in combination with specific psychotherapeutic approaches such as cognitive-behavioral, relapse prevention, and/or mindfulness-based skills training. Importantly, individual difference factors may also guide informed selection between such options, as well as strategic adaptation of interventions to better meet individual needs. For example, patients with specific clinical features (e.g., complex PTSD symptoms) or coping styles (e.g., experiential avoidance) may require targeted skills training in order to more fully benefit from motivational incentives offered in CM. Cognitive abilities including executive working memory and inhibitory control have also been implicated in theoretical models of CM treatment response (Bickel, Snider, Quisenberry, Stein, & Hanlon, 2016; Regier & Redish, 2015) and deficits in these domains may similarly reflect a need for targeted adjunctive services. For example, modest benefits of cognitive remediation have already been demonstrated in the context of CM (Rass et al., 2015) and may be more pronounced if specifically targeted toward patients with identified cognitive deficits. It is also possible that behavioral interventions like CM could be further enhanced through the addition of noninvasive brain stimulation (Koffarnus, Jarmolowicz, Mueller, & Bickel, 2013). This approach allows for strategic up- or down-regulation of neural processes underlying key treatment mechanisms (e.g., representation/valuation of future abstinence-contingent reward) but would first require identification of appropriate targets for neuromodulation. Evidence supporting delivery of CM in combination with other specific interventions will, however, need to outweigh the cost of adding such complementary treatments.
Fortunately, CM treatment parameters can also be readily adjusted, affording additional opportunities for personalized delivery. As previously described, the magnitude and probability of CM reinforcement has been manipulated in previous work (Ghitza, Epstein, Schmittner, et al., 2008; N. M. Petry, D. Barry, et al., 2012; Petry et al., 2004; Washio et al., 2011) and can be adjusted across individuals, as well as over time, to flexibly respond to traits and states that may impact treatment response. Shaping strategies, whereby requirements for contingent reward change over time, may be particularly well-suited to dynamic, personalized adaptation but have not yet been considered as the focus of measurement-based treatment delivery. By the same token, the frequency of CM sessions and overall duration of treatment can also be adjusted in response to individual patient needs and goals, representing another target for precision implementation. For example, it may be possible to extend benefits of CM by transitioning patients with longer term treatment needs to a modified intermittent schedule of reinforcement (e.g., a variable interval schedule with reinforcement typically occurring once or twice per month), providing for longer term treatment with a more gradual step-down preceding termination. Systems for remote abstinence verification and reward delivery may also be developed to facilitate CM delivery during such a late, transitional phase of treatment. Finally, both the type (e.g., money, vouchers, goods, privileges) and timing (e.g., immediate versus delayed) of CM reinforcement may significantly interact with patient traits and preferences to determine motivational salience. While, for example, similar outcomes have been noted for money- versus voucher-based reinforcement (Festinger, Dugosh, Kirby, & Seymour, 2014), patient preferences have been demonstrated to be roughly evenly split between these two options and may also change depending on specific contextual factors (for example, reward amount; Reilly, Roll, & Downey, 2000).
Interestingly, the in-treatment behavior of redeeming monetary incentives for more tangible rewards (e.g., food, clothing, electronics) has previously been associated with improved outcomes in research comparing CM voucher “savers” versus “spenders” (Ling Murtaugh, Krishnamurti, Davis, Reback, & Shoptaw, 2013; although see Fletcher, Dierst-Davies, & Reback, 2014, for evidence that more frequent spending may be associated with use in substance dependent men who have sex with men). Consistent with evidence that CM is strengthened when incentives are more reliably translated into tangible reward, patients are encouraged to articulate reward goals (e.g., desired prizes, items to be purchased with vouchers) as part of competent delivery of CM (N. M. Petry, S. M. Alessi, & D. M. Ledgerwood, 2012; Nancy M. Petry, Sheila M. Alessi, & David M. Ledgerwood, 2012; Petry, Alessi, Ledgerwood, & Sierra, 2010). In patients who struggle with this task, it may be possible to better support and incentivize goal-oriented thinking and planning in CM; for example, through formalized planning activities in session, incorporating a goal setting group, or directly reinforcing goal-related behaviors. Indeed, CM has already been successfully used to reinforce a variety of non-drug-related, recovery-oriented behaviors (e.g., making a “to do” list, conducting a job search, problem-solving a legal problem, attending 12 step programming, exercising), in addition to abstinence (Petry et al., 2006; Petry, Tedford, & Martin, 2001). This may be particularly important for patients who have difficulty setting and working toward complex goals or for those with specific needs in early recovery; such as securing employment, housing, or sober social support. Specific recovery-oriented behaviors, such as those related to physical exercise and religious activities, have also been associated with improved CM outcomes in previous work (Petry, Lewis, & Ostvik-White, 2008; Weinstock, Barry, & Petry, 2008) and may therefore be particularly beneficial to target through CM programming.
5. Conclusion
A variety of individual difference factors have previously been examined in relation to CM treatment response, including motivation and pre-treatment use, use of non-targeted substances during treatment, substance use chronicity and comorbidity, psychiatric comorbidity and severity, medical and legal factors, sociodemographic characteristics, and cognitive-behavioral variables. While CM was generally associated with improved abstinence and retention in stimulant and/or opioid using outpatients, some patient-level features were associated with either an enhanced or diminished response to CM as an adjunct to other outpatient services. On the whole, existing evidence suggests that CM may be especially beneficial in patients who might otherwise have poorer outcomes (for example, those who test positive for targeted substances at treatment entry, patients with more previous treatment attempts, and those with antisocial personality disorder). However, specific treatment parameters may be necessary to achieve these benefits (e.g., higher reinforcement magnitude and reinforcement schedules that encourage early retention) and these requirements may also meaningfully intersect with patient-level characteristics (e.g., age, race, psychiatric comorbidity). Patients expected to have better outcomes may, correspondingly, exhibit less pronounced benefits of CM due to a “ceiling effect,” whereby further improvement in measured outcomes is not possible due to high levels of abstinence in standard care. Other results were not, however, consistent with this general framework. For example, patients without intent to stop using (who would therefore be expected to have poorer outcomes), exhibited no benefit of CM and other factors intuitively associated with a poorer prognosis (e.g., greater psychiatric symptom severity, co-occurring substance use disorders, earlier age of first use) were not associated with CM treatment response.
We also identified that the current literature is limited in several respects and additional work will be necessary to definitively characterize individual difference factors with the strongest potential to inform precision implementation of CM in the future. First, we identified that nearly all previous work reflects secondary analysis of existing datasets and, accordingly, that patient-level data were predominantly derived from intake assessments commonly used in clinical trials (for example, psychodiagnostic interviews such as the Addiction Severity Index and SCID). Consequently, very little previous work has examined hypothesis-driven predictors of CM treatment response, such as theoretically informed genetic, cognitive, or biomarker variables. We also found that the majority of previous research in this area has been conducted by a single research group and has often utilized the same primary data sources for secondary analyses. To address these limitations, future work should focus on establishing the replicability and generalizability of previously reported individual difference factors in CM, as well as examining promising, hypothesis-driven factors that have not yet been considered. Such work will be necessary to rapidly advance precision implementation of CM – an emerging area of research which has already yielded promising results (N. M. Petry, D. Barry, et al., 2012). Ultimately, efforts in this area have strong potential to further improve upon the effectiveness of this important intervention by (1) enabling improved targeting and customization of treatment, informed by individual patient needs and (2) revealing novel treatment modifications and combinations with promise to enhance and extend CM benefits.
Highlights.
Contingency Management (CM) is an evidence-based treatment for addictive disorders.
Individual differences in treatment response may inform precision adaptation of CM.
A systematic review was conducted to identify individual difference factors in CM.
Evidence suggests CM may be especially beneficial in patients with poorer prognosis.
Limitations of the extant literature also offer promising directions for future work.
ACKNOWLEDGMENTS
Dr. Forster was responsible for the concept of this review article and performed the literature search, screening, and evaluation procedures. Drs. Forster, Forman, and DePhilippis contributed to the interpretation of results. Dr. Forster drafted the manuscript and all authors participated in revision of the manuscript. All authors critically reviewed content and approved the final version for publication. Dr. Forster was supported by funding from the VISN 4 Mental Illness Research, Education and Clinical Center (MIRECC, Director: D. Oslin; Pittsburgh Site Director: G. Haas), VA Pittsburgh Healthcare System. The contents do not represent the views of the Department of Veterans Affairs, Department of Defense, or the United States Government. None of the authors have any conflicts of interest to declare.
Footnotes
It is noted that data from group-based Contingency Management represent a small proportion of data included in three secondary analyses included in the current review. Specifically, these data represent 17% of the total sample analyzed by Montgomery, et al., 2015, 15% of the total sample analyzed by Rash, et al., 2016, and 34% of the total sample analyzed by Burch, et al., 2017.
It is noted that the current review focuses on abstinence from targeted substances and that the extant literature generally offers mixed findings regarding the effect of CM on non-targeted substance use. While several studies have found that effects of CM targeting a single drug or drug class generalize to abstinence from other substances (Jones, Haug, Silverman, Stitzer, & Svikis, 2001; M. McDonell et al., 2014; M. G. McDonell et al., 2013; Shoptaw, Jarvik, Ling, & Rawson, 1996; Silverman et al., 1999; Silverman et al., 1998), others found no CM effect on non-targeted substances (Epstein et al., 2003; Petry, Peirce, et al., 2005; Rawson et al., 2002; Silverman et al., 1999). Nevertheless, there is no evidence that targeting a single substance or substance class increases use of non-targeted substances (Lussier et al., 2006). Rather, a generalized clinical benefit of stimulant-targeted CM has previously been described, including fewer psychiatric symptoms and hospitalizations, as well as fewer problems in major life areas including medical, legal, employment, and social domains (Kiluk, Nich, Witkiewitz, Babuscio, & Carroll, 2014; M. McDonell et al., 2014; M. G. McDonell et al., 2013). Such outcomes were not, however, specifically considered herein.
REFERENCES
- Alessi SM, Rash C, & Petry NM (2011). Contingency management is efficacious and improves outcomes in cocaine patients with pretreatment marijuana use. Drug Alcohol Depend, 118(1), 62–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Alterman AI, Kampman K, Boardman CR, Cacciola JS, Rutherford MJ, McKay JR, & Maany I (1997). A cocaine-positive baseline urine predicts outpatient treatment attrition and failure to attain initial abstinence. Drug Alcohol Depend, 46(1-2), 79–85. [DOI] [PubMed] [Google Scholar]
- Alterman AI, McKay JR, Mulvaney FD, & McLellan AT (1996). Prediction of attrition from day hospital treatment in lower socioeconomic cocaine-dependent men. Drug Alcohol Depend, 40(3), 227–233. [DOI] [PubMed] [Google Scholar]
- Barry D, Sullivan B, & Petry NM (2009). Comparable efficacy of contingency management for cocaine dependence among African American, Hispanic, and White methadone maintenance clients. Psychol Addict Behav, 23(1), 168–174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Benishek LA, Dugosh KL, Kirby KC, Matejkowski J, Clements NT, Seymour BL, & Festinger DS (2014). Prize-based contingency management for the treatment of substance abusers: a meta-analysis. Addiction, 109(9), 1426–1436. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Benishek LA, Kirby KC, Dugosh KL, & Padovano A (2010). Beliefs about the empirical support of drug abuse treatment interventions: a survey of outpatient treatment providers. Drug Alcohol Depend, 107(2-3), 202–208. [DOI] [PubMed] [Google Scholar]
- Bickel WK, & Marsch LA (2001). Toward a behavioral economic understanding of drug dependence: delay discounting processes. Addiction, 96(1), 73–86. [DOI] [PubMed] [Google Scholar]
- Bickel WK, Miller ML, Yi R, Kowal BP, Lindquist DM, & Pitcock JA (2007). Behavioral and neuroeconomics of drug addiction: competing neural systems and temporal discounting processes. Drug Alcohol Depend, 90 Suppl 1, S85–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bickel WK, Snider SE, Quisenberry AJ, Stein JS, & Hanlon CA (2016). Competing neurobehavioral decision systems theory of cocaine addiction: From mechanisms to therapeutic opportunities. Prog Brain Res, 223, 269–293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Blanken P, Hendriks VM, Huijsman IA, van Ree JM, & van den Brink W (2016). Efficacy of cocaine contingency management in heroin-assisted treatment: Results of a randomized controlled trial. Drug Alcohol Depend, 164, 55–63. [DOI] [PubMed] [Google Scholar]
- Burch AE, Rash CJ, & Petry NM (2015). Sex effects in cocaine-using methadone patients randomized to contingency management interventions. Exp Clin Psychopharmacol, 23(4), 284–290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Burch AE, Rash CJ, & Petry NM (2017). Cocaine-using substance abuse treatment patients with and without HIV respond well to contingency management treatment. J Subst Abuse Treat, 77, 21–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Byrne SA, & Petry NM (2011). Concurrent alcohol dependence among methadone-maintained cocaine abusers is associated with greater abstinence. Exp Clin Psychopharmacol, 19(2), 116–122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Davis DR, Kurti AN, Skelly JM, Redner R, White TJ, & Higgins ST (2016). A review of the literature on contingency management in the treatment of substance use disorders, 2009-2014. Prev Med, 92, 36–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dean AC, London ED, Sugar CA, Kitchen CM, Swanson AN, Heinzerling KG, … Shoptaw S (2009). Predicting adherence to treatment for methamphetamine dependence from neuropsychological and drug use variables. Drug Alcohol Depend, 105(1-2), 48–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- DeFulio A, Stitzer M, Roll J, Petry N, Nuzzo P, Schwartz RP, & Stabile P (2013). Criminal justice referral and incentives in outpatient substance abuse treatment. J Subst Abuse Treat, 45(1), 70–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- DePhilippis D, Petry NM, Bonn-Miller MO, Rosenbach SB, & McKay JR (2018). The national implementation of Contingency Management (CM) in the Department of Veterans Affairs: Attendance at CM sessions and substance use outcomes. Drug Alcohol Depend, 185, 367–373. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dutra L, Stathopoulou G, Basden SL, Leyro TM, Powers MB, & Otto MW (2008). A meta-analytic review of psychosocial interventions for substance use disorders. Am J Psychiatry, 165(2), 179–187. [DOI] [PubMed] [Google Scholar]
- Ehrman RN, Robbins SJ, & Cornish JW (2001). Results of a baseline urine test predict levels of cocaine use during treatment. Drug Alcohol Depend, 62(1), 1–7. [DOI] [PubMed] [Google Scholar]
- Epstein DH, Hawkins WE, Covi L, Umbricht A, & Preston KL (2003). Cognitive-behavioral therapy plus contingency management for cocaine use: findings during treatment and across 12-month follow-up. Psychol Addict Behav, 17(1), 73–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Epstein DH, & Preston KL (2003). Does cannabis use predict poor outcome for heroin-dependent patients on maintenance treatment? Past findings and more evidence against. Addiction, 98(3), 269–279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Festinger DS, Dugosh KL, Kirby KC, & Seymour BL (2014). Contingency Management for Cocaine Treatment: Cash vs. Vouchers. J Subst Abuse Treat, 47(2), 168–174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fletcher JB, Dierst-Davies R, & Reback CJ (2014). Contingency management voucher redemption as an indicator of delayed gratification. J Subst Abuse Treat, 47(1), 73–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ford JD, Hawke J, Alessi S, Ledgerwood D, & Petry N (2007). Psychological trauma and PTSD symptoms as predictors of substance dependence treatment outcomes. Behav Res Ther, 45(10), 2417–2431. [DOI] [PubMed] [Google Scholar]
- Garcia-Fernandez G, Secades-Villa R, Garcia-Rodriguez O, Pena-Suarez E, & Sanchez-Hervas E (2013). Contingency management improves outcomes in cocaine-dependent outpatients with depressive symptoms. Exp Clin Psychopharmacol, 21(6), 482–489. [DOI] [PubMed] [Google Scholar]
- Ghitza UE, Epstein DH, & Preston KL (2007). Nonreporting of cannabis use: Predictors and relationship to treatment outcome in methadone maintained patients. Addict Behav, 32(5), 938–949. [DOI] [PubMed] [Google Scholar]
- Ghitza UE, Epstein DH, & Preston KL (2008). Self-report of illicit benzodiazepine use on the Addiction Severity Index predicts treatment outcome. Drug Alcohol Depend, 97(1-2), 150–157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ghitza UE, Epstein DH, Schmittner J, Vahabzadeh M, Lin JL, & Preston KL (2008). Effect of reinforcement probability and prize size on cocaine and heroin abstinence in prize-based contingency management. J Appl Behav Anal, 41(4), 539–549. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ginley MK, Rash CJ, Olmstead TA, & Petry NM (2017). Contingency management treatment in cocaine using methadone maintained patients with and without legal problems. Drug Alcohol Depend, 180, 208–214. [DOI] [PubMed] [Google Scholar]
- Gonzalez G, Feingold A, Oliveto A, Gonsai K, & Kosten TR (2003). Comorbid major depressive disorder as a prognostic factor in cocaine-abusing buprenorphine-maintained patients treated with desipramine and contingency management. Am J Drug Alcohol Abuse, 29(3), 497–514. [DOI] [PubMed] [Google Scholar]
- Hagedorn HJ, Noorbaloochi S, Simon AB, Bangerter A, Stitzer ML, Stetler CB, & Kivlahan D (2013). Rewarding early abstinence in Veterans Health Administration addiction clinics. J Subst Abuse Treat, 45(1), 109–117. [DOI] [PubMed] [Google Scholar]
- Hartzler B, Beadnell B, & Donovan D (2017). Predictive Validity of Addiction Treatment Clinicians’ Post-Training Contingency Management Skills for Subsequent Clinical Outcomes. J Subst Abuse Treat, 72, 126–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Heil SH, Johnson MW, Higgins ST, & Bickel WK (2006). Delay discounting in currently using and currently abstinent cocaine-dependent outpatients and non-drug-using matched controls. Addict Behav, 31(7), 1290–1294. [DOI] [PubMed] [Google Scholar]
- Jones HE, Haug N, Silverman K, Stitzer M, & Svikis D (2001). The effectiveness of incentives in enhancing treatment attendance and drug abstinence in methadone-maintained pregnant women. Drug Alcohol Depend, 61(3), 297–306. [DOI] [PubMed] [Google Scholar]
- Jordan CJ, & Andersen SL (2017). Sensitive periods of substance abuse: Early risk for the transition to dependence. Dev Cogn Neurosci, 25, 29–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kiluk BD, Nich C, Witkiewitz K, Babuscio TA, & Carroll KM (2014). What happens in treatment doesn’t stay in treatment: cocaine abstinence during treatment is associated with fewer problems at follow-up. J Consult Clin Psychol, 82(4), 619–627. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Koffarnus MN, Jarmolowicz DP, Mueller ET, & Bickel WK (2013). Changing delay discounting in the light of the competing neurobehavioral decision systems theory: a review. J Exp Anal Behav, 99(1), 32–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ledgerwood DM, & Petry NM (2006). Does contingency management affect motivation to change substance use? Drug Alcohol Depend, 83(1), 65–72. [DOI] [PubMed] [Google Scholar]
- Ling Murtaugh K, Krishnamurti T, Davis AL, Reback CJ, & Shoptaw S (2013). Spend today, clean tomorrow: predicting methamphetamine abstinence in a randomized controlled trial. Health Psychol, 32(9), 958–966. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Luo SX, Martinez D, Carpenter KM, Slifstein M, & Nunes EV (2014). Multimodal predictive modeling of individual treatment outcome in cocaine dependence with combined neuroimaging and behavioral predictors. Drug Alcohol Depend, 143, 29–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lussier JP, Heil SH, Mongeon JA, Badger GJ, & Higgins ST (2006). A meta-analysis of voucher-based reinforcement therapy for substance use disorders. Addiction, 101(2), 192–203. [DOI] [PubMed] [Google Scholar]
- Mancino MJ, McGaugh J, Feldman Z, Poling J, & Oliveto A (2010). Effect of PTSD diagnosis and contingency management procedures on cocaine use in dually cocaine- and opioid-dependent individuals maintained on LAAM: a retrospective analysis. Am J Addict, 19(2), 169–177. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McDonell M, McPherson S, Vilardaga R, Srebnik D, Angelo FN, Leickly E, … Ries R (2014). Preliminary findings: Contingency management targeting psycho-stimulant use results in secondary decreases in smoking for severely mentally ill adults. Am J Addict, 23(4), 407–410. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McDonell MG, Srebnik D, Angelo F, McPherson S, Lowe JM, Sugar A, … Ries RK (2013). Randomized controlled trial of contingency management for stimulant use in community mental health patients with serious mental illness. Am J Psychiatry, 170(1), 94–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McLellan AT, Luborsky L, Woody GE, O’Brien CP, & Druley KA (1983). Predicting response to alcohol and drug abuse treatments. Role of psychiatric severity. Arch Gen Psychiatry, 40(6), 620–625. [DOI] [PubMed] [Google Scholar]
- Messina N, Farabee D, & Rawson R (2003). Treatment responsivity of cocaine-dependent patients with antisocial personality disorder to cognitive-behavioral and contingency management interventions. J Consult Clin Psychol, 71(2), 320–329. [DOI] [PubMed] [Google Scholar]
- Milward J, Lynskey M, & Strang J (2014). Solving the problem of non-attendance in substance abuse services. Drug Alcohol Rev, 33(6), 625–636. [DOI] [PubMed] [Google Scholar]
- Montgomery L, Carroll KM, & Petry NM (2015). Initial abstinence status and contingency management treatment outcomes: does race matter? J Consult Clin Psychol, 83(3), 473–481. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Olmstead TA, & Petry NM (2009). The cost-effectiveness of prize-based and voucher-based contingency management in a population of cocaine- or opioid-dependent outpatients. Drug Alcohol Depend, 102(1-3), 108–115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Olmstead TA, Sindelar JL, & Petry NM (2007). Clinic variation in the cost-effectiveness of contingency management. Am J Addict, 16(6), 457–460. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Oluwoye O, Hirchak K, Leickly E, Skalisky J, McPherson S, Srebnik D, … McDonell MG (2018). Interaction between pre-treatment drug use and heterogeneity of psychiatric diagnosis predicts outcomes in outpatients with co-occurring disorders. Psychiatry Res, 260, 233–235. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Petry NM, & Alessi SM (2010). Prize-based contingency management is efficacious in cocaine-abusing patients with and without recent gambling participation. J Subst Abuse Treat, 39(3), 282–288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Petry NM, Alessi SM, Carroll KM, Hanson T, MacKinnon S, Rounsaville B, & Sierra S (2006). Contingency management treatments: Reinforcing abstinence versus adherence with goal-related activities. J Consult Clin Psychol, 74(3), 592–601. [DOI] [PubMed] [Google Scholar]
- Petry NM, Alessi SM, & Ledgerwood DM (2012). Contingency management delivered by community therapists in outpatient settings. Drug Alcohol Depend, 122(1-2), 86–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Petry NM, Alessi SM, & Ledgerwood DM (2012). A randomized trial of contingency management delivered by community therapists. Journal of Consulting and Clinical Psychology, 80(2), 286–298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Petry NM, Alessi SM, Ledgerwood DM, & Sierra S (2010). Psychometric properties of the contingency management competence scale. Drug Alcohol Depend, 109(1-3), 167–174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Petry NM, Alessi SM, Marx J, Austin M, & Tardif M (2005). Vouchers versus prizes: contingency management treatment of substance abusers in community settings. J Consult Clin Psychol, 73(6), 1005–1014. [DOI] [PubMed] [Google Scholar]
- Petry NM, Barry D, Alessi SM, Rounsaville BJ, & Carroll KM (2012). A randomized trial adapting contingency management targets based on initial abstinence status of cocaine-dependent patients. J Consult Clin Psychol, 80(2), 276–285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Petry NM, Ford JD, & Barry D (2011). Contingency management is especially efficacious in engendering long durations of abstinence in patients with sexual abuse histories. Psychol Addict Behav, 25(2), 293–300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Petry NM, Lewis MW, & Ostvik-White EM (2008). Participation in religious activities during contingency management interventions is associated with substance use treatment outcomes. Am J Addict, 17(5), 408–413. [DOI] [PubMed] [Google Scholar]
- Petry NM, & Martin B (2002). Low-cost contingency management for treating cocaine- and opioid-abusing methadone patients. J Consult Clin Psychol, 70(2), 398–405. [DOI] [PubMed] [Google Scholar]
- Petry NM, Peirce JM, Stitzer ML, Blaine J, Roll JM, Cohen A, … Li R (2005). Effect of prize-based incentives on outcomes in stimulant abusers in outpatient psychosocial treatment programs: a national drug abuse treatment clinical trials network study. Arch Gen Psychiatry, 62(10), 1148–1156. [DOI] [PubMed] [Google Scholar]
- Petry NM, Rash CJ, & Easton CJ (2011). Contingency management treatment in substance abusers with and without legal problems. J Am Acad Psychiatry Law, 39(3), 370–378. [PMC free article] [PubMed] [Google Scholar]
- Petry NM, & Stitzer ML (2002). Contingency Management: Using motivational incentives to improve drug abuse treatment. New Haven, CT: Yale University. [Google Scholar]
- Petry NM, Tedford J, Austin M, Nich C, Carroll KM, & Rounsaville BJ (2004). Prize reinforcement contingency management for treating cocaine users: how low can we go, and with whom? Addiction, 99(3), 349–360. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Petry NM, Tedford J, & Martin B (2001). Reinforcing compliance with non-drug-related activities. J Subst Abuse Treat, 20(1), 33–44. [DOI] [PubMed] [Google Scholar]
- Prendergast M, Podus D, Finney J, Greenwell L, & Roll J (2006). Contingency management for treatment of substance use disorders: a meta-analysis. Addiction, 101(11), 1546–1560. [DOI] [PubMed] [Google Scholar]
- Rash CJ, Alessi SM, & Petry NM (2008a). Cocaine abusers with and without alcohol dependence respond equally well to contingency management treatments. Exp Clin Psychopharmacol, 16(4), 275–281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rash CJ, Alessi SM, & Petry NM (2008b). Contingency management is efficacious for cocaine abusers with prior treatment attempts. Exp Clin Psychopharmacol, 16(6), 547–554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rash CJ, Alessi SM, & Petry NM (2017). Substance Abuse Treatment Patients in Housing Programs Respond to Contingency Management Interventions. J Subst Abuse Treat, 72, 97–102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rash CJ, Andrade LF, & Petry NM (2013). Income received during treatment does not affect response to contingency management treatments in cocaine-dependent outpatients. Drug Alcohol Depend, 132(3), 528–534. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rash CJ, Burki M, Montezuma-Rusca JM, & Petry NM (2016). A retrospective and prospective analysis of trading sex for drugs or money in women substance abuse treatment patients. Drug Alcohol Depend, 162, 182–189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rash CJ, Olmstead TA, & Petry NM (2009). Income does not affect response to contingency management treatments among community substance abuse treatment-seekers. Drug Alcohol Depend, 104(3), 249–253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rass O, Schacht RL, Buckheit K, Johnson MW, Strain EC, & Mintzer MZ (2015). A randomized controlled trial of the effects of working memory training in methadone maintenance patients. Drug Alcohol Depend, 156, 38–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rawson RA, Huber A, McCann M, Shoptaw S, Farabee D, Reiber C, & Ling W (2002). A comparison of contingency management and cognitive-behavioral approaches during methadone maintenance treatment for cocaine dependence. Arch Gen Psychiatry, 59(9), 817–824. [DOI] [PubMed] [Google Scholar]
- Regier PS, & Redish AD (2015). Contingency Management and Deliberative Decision-Making Processes. Front Psychiatry, 6, 76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Reilly MP, Roll JM, & Downey KK (2000). Impulsivity and voucher versus money preference in polydrug-dependent participants enrolled in a contingency-management-based substance abuse treatment program. J Subst Abuse Treat, 19(3), 253–257. [DOI] [PubMed] [Google Scholar]
- Robins LN, & Przybeck TR (1985). Age of onset of drug use as a factor in drug and other disorders. NIDA Res Monogr, 56, 178–192. [PubMed] [Google Scholar]
- Roll JM, & Higgins ST (2000). A within-subject comparison of three different schedules of reinforcement of drug abstinence using cigarette smoking as an exemplar. Drug Alcohol Depend, 58(1-2), 103–109. [DOI] [PubMed] [Google Scholar]
- Roll JM, Huber A, Sodano R, Chudzynski JE, Moynier E, & Shoptaw S (2000). A Comparison of Five Reinforcement Schedules for use in Contingency Management-Based Treatment of Methamphetamine Abuse. Psychologic, 56, 67–81. [Google Scholar]
- Schmitz JM, Lindsay JA, Stotts AL, Green CE, & Moeller FG (2010). Contingency management and levodopa-carbidopa for cocaine treatment: a comparison of three behavioral targets. Exp Clin Psychopharmacol, 18(3), 238–244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schmitz JM, Mooney ME, Moeller FG, Stotts AL, Green C, & Grabowski J (2008). Levodopa pharmacotherapy for cocaine dependence: choosing the optimal behavioral therapy platform. Drug Alcohol Depend, 94(1-3), 142–150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schottenfeld RS, Chawarski MC, Pakes JR, Pantalon MV, Carroll KM, & Kosten TR (2005). Methadone versus buprenorphine with contingency management or performance feedback for cocaine and opioid dependence. Am J Psychiatry, 162(2), 340–349. [DOI] [PubMed] [Google Scholar]
- Schottenfeld RS, Moore B, & Pantalon MV (2011). Contingency management with community reinforcement approach or twelve-step facilitation drug counseling for cocaine dependent pregnant women or women with young children. Drug Alcohol Depend, 118(1), 48–55. [DOI] [PubMed] [Google Scholar]
- Secades-Villa R, Garcia-Fernandez G, Pena-Suarez E, Garcia-Rodriguez O, Sanchez-Hervas E, & Fernandez-Hermida JR (2013). Contingency management is effective across cocaine-dependent outpatients with different socioeconomic status. J Subst Abuse Treat, 44(3), 349–354. [DOI] [PubMed] [Google Scholar]
- Shearer J, Tie H, & Byford S (2015). Economic evaluations of contingency management in illicit drug misuse programmes: A systematic review. Drug Alcohol Rev, 34(3), 289–298. [DOI] [PubMed] [Google Scholar]
- Shoptaw S, Jarvik ME, Ling W, & Rawson RA (1996). Contingency management for tobacco smoking in methadone-maintained opiate addicts. Addict Behav, 21(3), 409–412. [DOI] [PubMed] [Google Scholar]
- Silverman K, Chutuape MA, Bigelow GE, & Stitzer ML (1999). Voucher-based reinforcement of cocaine abstinence in treatment-resistant methadone patients: effects of reinforcement magnitude. Psychopharmacology (Berl), 146(2), 128–138. [DOI] [PubMed] [Google Scholar]
- Silverman K, Robles E, Mudric T, Bigelow GE, & Stitzer ML (2004). A randomized trial of long-term reinforcement of cocaine abstinence in methadone-maintained patients who inject drugs. J Consult Clin Psychol, 72(5), 839–854. [DOI] [PubMed] [Google Scholar]
- Silverman K, Wong CJ, Umbricht-Schneiter A, Montoya ID, Schuster CR, & Preston KL (1998). Broad beneficial effects of cocaine abstinence reinforcement among methadone patients. J Consult Clin Psychol, 66(5), 811–824. [DOI] [PubMed] [Google Scholar]
- Sindelar J, Elbel B, & Petry NM (2007). What do we get for our money? Cost-effectiveness of adding contingency management. Addiction, 102(2), 309–316. [DOI] [PubMed] [Google Scholar]
- Sindelar JL, & Ball SA (2010). Cost evaluation of evidence-based treatments. Addict Sci Clin Pract, 5(2), 44–51. [PMC free article] [PubMed] [Google Scholar]
- Sindelar JL, Olmstead TA, & Peirce JM (2007). Cost-effectiveness of prize-based contingency management in methadone maintenance treatment programs. Addiction, 102(9), 1463–1471. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sofuoglu M, Gonzalez G, Poling J, & Kosten TR (2003). Prediction of treatment outcome by baseline urine cocaine results and self-reported cocaine use for cocaine and opioid dependence. Am J Drug Alcohol Abuse, 29(4), 713–727. [DOI] [PubMed] [Google Scholar]
- Stitzer ML, Peirce J, Petry NM, Kirby K, Roll J, Krasnansky J, … Li R (2007). Abstinence-based incentives in methadone maintenance: interaction with intake stimulant test results. Exp Clin Psychopharmacol, 15(4), 344–350. [DOI] [PubMed] [Google Scholar]
- Stitzer ML, Petry N, Peirce J, Kirby K, Killeen T, Roll J, … Li R (2007). Effectiveness of abstinence-based incentives: interaction with intake stimulant test results. J Consult Clin Psychol, 75(5), 805–811. [DOI] [PubMed] [Google Scholar]
- Stotts AL, Vujanovic A, Heads A, Suchting R, Green CE, & Schmitz JM (2015). The role of avoidance and inflexibility in characterizing response to contingency management for cocaine use disorders: A secondary profile analysis. Psychol Addict Behav, 29(2), 408–413. [DOI] [PubMed] [Google Scholar]
- Wardle MC, Vincent JN, Suchting R, Green CE, Lane SD, & Schmitz JM (2017). Anhedonia Is Associated with Poorer Outcomes in Contingency Management for Cocaine Use Disorder. J Subst Abuse Treat, 72, 32–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Washio Y, Higgins ST, Heil SH, McKerchar TL, Badger GJ, Skelly JM, & Dantona RL (2011). Delay discounting is associated with treatment response among cocaine-dependent outpatients. Exp Clin Psychopharmacol, 19(3), 243–248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weinstock J, Alessi SM, & Petry NM (2007). Regardless of psychiatric severity the addition of contingency management to standard treatment improves retention and drug use outcomes. Drug Alcohol Depend, 87(2-3), 288–296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weinstock J, Barry D, & Petry NM (2008). Exercise-related activities are associated with positive outcome in contingency management treatment for substance use disorders. Addict Behav, 33(8), 1072–1075. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weiss L, & Petry NM (2013). Older methadone patients achieve greater durations of cocaine abstinence with contingency management than younger patients. Am J Addict, 22(2), 119–126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weiss LM, & Petry NM (2011). Interaction effects of age and contingency management treatments in cocaine-dependent outpatients. Exp Clin Psychopharmacol, 19(2), 173–181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weiss LM, & Petry NM (2014). Substance abuse treatment patients with early onset cocaine use respond as well to contingency management interventions as those with later onset cocaine use. J Subst Abuse Treat, 47(2), 146–150. [DOI] [PMC free article] [PubMed] [Google Scholar]