Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Dec 1.
Published in final edited form as: J Subst Use Addict Treat. 2024 Sep 12;167:209522. doi: 10.1016/j.josat.2024.209522

Examining the Impact of Low Magnitude Incentives in Contingency Management Protocols:Non-engagement in Petry et al. 2004

Carla J Rash 1, Sheila M Alessi 1, Kristyn Zajac 1
PMCID: PMC11527562  NIHMSID: NIHMS2026263  PMID: 39277143

Abstract

Introduction/Method:

Current federal regulations limit the use of incentives in contingency management (CM) interventions to a nominal total value (i.e., up to $75/patient/year in aggregate of federal funds). This limit represents a striking divergence from the magnitudes used in evidence-based CM protocols. In the present report, we re-analyze data from the Petry et al. (2004) study, which was designed to test the efficacy of two different magnitude CM protocols ($80 and $240 in 2004 dollars) relative to usual intensive outpatient services for treatment-seeking patients with cocaine use. Petry et al. (2004) found that the $240 condition [~$405 in 2024 dollars], but not the $80 condition [~$135 in 2024 dollars], improved abstinence outcomes relative to usual care. The lower-cost $80 condition is the closest condition to the current federal $75 limit that permits a head-to-head comparison of magnitudes. A re-analysis offers an opportunity to examine the impact of low magnitude protocols in more detail, specifically in terms of non-engagement with treatment (defined as absence of negative samples and thus not encountering incentives for abstinence).

Results:

We found moderate to large effects favoring the $240 condition over both usual care (ds ranging 0.33 to 0.97) and the $80 condition (ds ranging 0.39 to 0.83) across various thresholds of non-engagement with the incentives/reinforcers for abstinence. Importantly, the $80 condition evidenced higher (worse) rates of non-engagement compared to the usual care condition (i.e., small and negative effect sizes ranging −0.30 to 0.14), though not reaching statistical significance).

Conclusions:

These results suggest that CM protocols designed to stay within the federal limitation of $75 should be discouraged, and evidence-based protocols should be recommended along with the regulatory reforms necessary to support their implementation.

Keywords: motivational incentives, vouchers, prizes, cocaine use disorder, incentives

1. Introduction

Contingency management (CM) has robust evidence supporting its efficacy as a substance use disorder (SUD) treatment (Benishek et al., 2014; Bolivar et al., 2021; Davis et al., 2016; Dutra et al., 2008; Ginley et al., 2021; Griffith et al., 2000; Lussier et al., 2006; Pfund et al., 2021, 2022; Prendergast et al., 2006). CM is effective across a wide range demographic and clinical characteristics (Foster et al., 2019; Rash & Petry, 2015; Rash et al., 2008a,b, 2009, 2013, 2017; Zajac et al., 2020) and across various substance use disorders (e.g., Lussier et al., 2006). This large foundation of scientific support has led federal agencies to recommend CM as an evidence-based treatment for SUD, including the National Institutes of Health (Volkow, 2020), the Substance Abuse and Mental Health Services Administration (2020), the Department of Health and Human Services (2023), the Department of Veterans Affairs (2021), and the Department of Defense (2021). In addition, the White House’s (2022) National Drug Control Strategy recommends CM be accessible to patients with SUD.

Design parameters such as magnitude (Bolivar et al., 2021; Businelle et al., 2012; Dallery et al., 2001; Higgins et al., 2007; Lussier et al., 2006; Petry et al., 2004, 2012; Regnier et al., 2022; Silverman et al., 1999), immediacy (Griffith et al., 2000; Lussier et al., 2006; Packer et al., 2012), and frequency (Griffith et al., 2006; Pfund et al., 2021) of reinforcement opportunities impact CM’s effect on patient outcomes. However, these core features of effective CM design diverge from evidence-based practices in real-world implementation (Rash et al., 2012; 2020), and this deviation is particularly evident in reinforcement magnitude. In a sample of 617 substance use disorder treatment providers, about half of the providers reported using reinforcers (colloquially called “rewards”) clinically (Rash et al., 2012). Of those, 90% used rewards totaling less than $100 per patient, and 43% used strictly social reinforcement with no monetary value. A subsequent study (Rash et al., 2020) similarly found that nearly three-fourths of providers reported using low (≤ $100) rewards. Nationally, 54% of treatment programs reported that they offer contingency management to patients as one of their clinical/therapeutic approaches (Substance Abuse and Mental Health Services Administration, 2021). It is likely that the magnitudes used in these CM programs are well below effective amounts (Rash et al., 2012; 2020).

Programs involving such low per patient magnitude incentives stand in stark contrast to most research protocols for CM. The median average maximum expected incentive magnitude1 per patient across 12-week prize-based CM protocols targeting alcohol or non-tobacco drug abstinence in Benishek et al. (2014; includes 17 protocols) was $420 (interquartile range $250-$497). For voucher-based protocols, the median maximum magnitude per patient for 12-week voucher protocols in Ainscough et al. (2017; includes 7 protocols) was $900 (inter-quartile range $743-$1121); and in the Bolivar et al. (2021; includes 21 protocols) meta-analysis, the median maximum for 12-week voucher protocols was $738 (inter-quartile range $407-$1155). The chasm between research protocols and real-world use of incentives is concerning given the strong evidence that CM’s efficacy is tied to its magnitude (Bolivar et al., 2021; Businelle et al., 2012; Dallery et al., 2001; Higgins et al., 2007; Lussier et al., 2006; Petry et al., 2004, 2012; Regnier et al., 2022; Silverman et al., 1999).

1.1. Federal Policy Challenges

Policy and regulatory barriers pose challenges to moving CM forward from research to practice with fidelity and at effective magnitudes. Clark and Davis (2023) outline several federal laws related to the use of incentives in patient care (see Andraka-Christou et al., 2023 for a discussion of US state laws). Of relevance to the issue of non-evidence-based magnitudes in CM protocols, the beneficiary inducement statute prohibits providers from offering rewards or giveaways to patients that are likely to influence the patient’s choice of provider or for meeting milestones or activities, unless the value of the reward was deemed to be of nominal value. In 2016, the Office of Inspector General (OIG; 2016) set the limits of nominal value gifts or services to be a retail value of no more than $15 per item and no more than $75 per patient per year in aggregate. The OIG affirmed these limits in the 2020 Final Rule (OIG, 2020). States can request a 1115 Medicaid Demonstration Waiver to allow for higher incentive CM to be provided as a covered benefit and reduce the potential for CM incentives to be considered Medicaid fraud. Only four states (California, Washington, Delaware, Montana) have successfully received approval of waivers for CM (McDonell et al., 2024). Another process involves seeking an OIG advisory opinion to review risks related to the Anti-Kickback Statute, but this case-by-case process is lengthy and expensive (~$10,000 per opinion, see McDonell et al., 2024). Not surprisingly, few programs have pursued this route.

In 2018, the Substance Abuse and Mental Health Service Administration (SAMHSA) announced incentives could be used for CM programs, but using the OIG definition of nominal value, SAMHSA limited the incentive to $75/patient/year. As of the most recent release in May 2024 of the SAMHSA State Opioid Response and Tribal Opioid Response Grant notice of funding opportunities, the $75/patient/year limit remains unchanged despite feedback from the scientific community that this limit remains well below evidence-based CM magnitudes (Rash et al., 2024). The Center for Substance Abuse Treatment Director, Dr. Yngvild Olsen, has acknowledged the $75 limit is not compatible with evidence-based CM magnitudes on its own (2023). Dr. Olsen has noted that grantees can braid funding streams together using other resources to reach an evidence-based, effective “dose” of CM; however, this barrier increases complexity and burden for organizations. Further, many organizations may not be aware that CM magnitude is a critical parameter of efficacy and neglect to take the extra steps to align funding of their CM protocols with evidence-based magnitudes. While SAMHSA’s support of CM is a positive step toward making effective CM treatment available to patients with substance use disorders, the current state of affairs encourages CM implementation but with restrictions likely to hinder effectiveness and its potential to help patients.

The studies by Rash et al. (2012, 2020) suggest that incentive delivery in practice will not align with the science of CM unless clear guidelines and structure to support fidelity are in place. With the restriction on incentive magnitude per patient and the complexities around pursuing additional funding streams to supplement the base $75 SAMHSA dollars, a possible result will be use of CM protocols that by design are unlikely to be effective. Not only would this practice represent missed opportunities for patients to benefit from state-of-the-science care, but it may stymy growing interest in integrating CM into clinical practice when implementation efforts fail to reproduce patient outcomes regularly documented in research trials.

1.2. Purpose of this paper

Of relevance, the Petry et al. (2004) study, aptly titled “How Low Can We Go?” examined a low-cost CM protocol for patients with cocaine use disorder initiating community intensive outpatient treatment. The study compared three conditions: a low-cost prize-based CM with an average expected maximum value of $80 in total incentives, a standard-for-the-time prize-based CM condition with an average expected maximum of $240 in total incentives, and a usual care comparison condition. Petry’s $80 condition provides an avenue to examine how a protocol using only the $75/patient of federal funds for incentives might perform compared to typical magnitude CM protocols in the scientific literature.

Petry et al. (2004) reported two abstinence outcomes: continuous abstinence defined as the number of consecutive weeks of uninterrupted negative samples, and percent negative samples out of total samples submitted. To be considered a negative sample, the samples had to test negative for all substances targeted by the CM protocol: alcohol, opioids, and cocaine. The $240 condition was superior to the usual care condition for both abstinence outcomes, and the $240 condition was superior to the $80 condition for percent negative samples and trended (p = .09) toward a significantly better outcome for abstinence duration. In contrast, the $80 did not statistically differ from the usual care group for either of the abstinence outcomes. See Figure 1 panels A and B.

Figure 1.

Figure 1.

Open in a new tab

Abstinence outcomes from Petry et al. (2004) by condition: Longest duration of abstinence in weeks and percent negative samples submitted. Panel A presents average weeks of continuous abstinence achieved by each of the three treatment conditions. Panel B presents the percent of negative samples for each condition. Posthoc comparisons (p-value) are noted in both panels.

These results, as reported in Petry et al. (2004), suggest a clear advantage of the higher magnitude condition. However, some may interpret the small numerical advantage in the $80 condition compared to usual care as support that “something is better than nothing” even though testing indicates no statistical advantage.

While the original paper (Petry et al., 2004) calculated percent negative samples out of total samples submitted, we evaluated treatment non-engagement with the active component of CM (the incentives) by calculating the percent of participants who submitted zero or few negative samples during the study across various thresholds. For CM to be effective (and to shape the probability of future negative samples), patients must access the reinforcement. Patients who provide zero or few negative samples are not accessing the active ingredient of CM. Thus, we are assessing patients’ lack of engagement with the abstinence incentives/reinforcers by analyzing the number of negative samples submitted across conditions. In the present report, we re-examine Petry et al. (2004) data to evaluate whether the low magnitude condition is associated with non-engagement with the incentives/reinforcers. We hypothesize the higher magnitude condition would be more likely to engage individuals than either the usual care or low-cost condition, and that participants in the low magnitude CM condition would have similar rates of non-engagement with the active ingredients of CM (i.e., incentives/reinforcers) as those in the usual care condition. Examining lack of engagement with incentives/reinforcers in CM treatment may provide additional perspective on why the $80 condition was not superior to usual care, and this timely re-analysis might be useful in informing federal policy related to CM incentives and may guide real-world implementation efforts to fully align with best practices.

2. Materials and methods

2.1. Participants

Petry et al. (2004) included 120 persons with cocaine use disorder who initiated intensive outpatient treatment in two community clinics in the Northeast. Participants were 18+ years of age, had recent use of cocaine (via urinalysis or self-report), and were English-speaking. Exclusion criteria were: 1) cognitive impairment, 2) diagnosis of current DSM-IV opioid dependence, 3) active, uncontrolled psychosis or bipolar disorder, and 4) gambling disorder (i.e., DSM-IV pathological gambling).

Participants were 56% female, 64% African American, 10% Hispanic, and 35 years old (SD = 7.18) on average. Participants did not differ across the three treatment conditions in terms of demographics (i.e., age, sex, race/ethnicity, marital status, past year income, or employment status) and clinical characteristics (i.e., percent cocaine positive at baseline, prior treatments, years of cocaine use, recent cocaine use, recent alcohol use, percent meeting criteria for cocaine dependence, percent with alcohol abuse/dependence).

2.2. Procedure

In the original study (Petry et al., 2004), research assistants randomized eligible participants to one of the three study conditions: usual care at the clinic (n = 37), usual care plus low cost ($80) CM (n = 45), and usual care plus “standard” ($240) CM (n = 38). Usual care at the clinic included group counseling sessions focused on 12-step topics, cognitive-behavioral therapy, health education and AIDS prevention, and life skills training. Treatment duration was 12-weeks post-randomization. During the treatment period, participants in all three groups were expected to submit 21 breath and urine samples arranged to have higher collection frequency (3 times per week) early in treatment and decreasing frequency thereafter. Research staff tested samples onsite and provided results immediately.

Both CM conditions used prize-based protocols, providing draws from a fishbowl for negative samples and completion of weekly treatment goal-related activities (e.g., attending a medical appointment, attending a job skills training class). A total of 165 draws (81 for negative samples, 84 for verified activities) was possible over the 12-week treatment period. In both CM conditions, participants who had earned draws selected slips from a fishbowl containing 250 total slips; half of the slips were non-winning and half were winning slips with a monetary prize. In the low cost ($80 condition), 109 slips represented “mini” prizes worth about $0.33 in value, 15 slips represented medium prizes worth about $5 in value, and 1 slip was a jumbo prize worth up to $100. The $240 condition used the same probabilities, but small prizes worth up to $1 replaced “minis”, and large prizes worth up to $20 replaced mediums. The other categories remained the same as in the $80 condition. Research staff delivered the CM intervention sessions and returned slips to the fishbowl after draws so that probabilities remained constant across treatment and patients.

2.3. Data Analysis

Rates of CM non-engagement were determined for the three treatment groups using during-treatment alcohol and substance samples testing negative (0 to 21 samples possible). We explored multiple thresholds for non-engagement with the incentives/reinforcers beginning with those who provided no negative samples during treatment (thus never earning reinforcement for abstinence). Because no single cut-off threshold singularly captures all patterns of engagement/non-engagement, we also examined those who submitted less than 2 negative samples, less than 3 negative samples, less than 4 negative samples, less than 6 negative samples, and less than 8 negative samples. Frequency and percent of non-responders were determined for each condition using the various thresholds, with significant differences at p < .05 noted. Cohen’s d effect size was calculated individually for each CM condition compared to usual care and for the two CM conditions compared to one another (Wilson, 2023). Total samples submitted (regardless of result) are also reported for each of the three conditions.

3. Results

Across all thresholds considered, rates of non-engagement with the incentives/reinforcers were lower (better) in the $240 condition compared to either the usual care or the $80 condition. Participants randomized to the $80 condition had higher (worse) non-engagement rates compared to not only the $240 condition as hypothesized, but also had higher (worse) rates of non-engagement compared to those randomized to usual care alone. See Table 1.

Table 1.

Non-engagement with the incentives/reinforcers across various thresholds for Petry et al.’s (2004) usual care, $80 CM, and $240 CM conditions

Thresholds of Non-engagement Usual Care (n = 37) Prize CM $80 (n = 45) Prize CM $240 (n = 38) $80 vs. UC d $240 vs. UC d $80 vs. $240 d
No Negative Samples 24%, 9 20%, 9 5%, 2 0.14 0.97* 0.83*
<2 Negative Samples 27%, 10 30%, 13 13%, 5 −0.05 0.49 0.54
<3 Negative Samples 35%, 13 42%, 19 18%, 7 −0.17 0.48 0.65*
<4 Negative Samples 38%, 14 51%, 23 24%, 9 −0.30 0.37 0.67*
<6 Negative Samples 54%, 20 58%, 26 40%, 15 −0.08 0.33 0.42
<8 Negative Samples 60%, 22 62%, 28 45%, 17 −0.06 0.33 0.39
Open in a new tab

Notes. Percent of non-engagers, n are provided for each condition by various thresholds. Effect size (Cohen’s d) are provided comparing each of the CM conditions to usual care and comparing the two CM conditions to one another (last column).

*

Indicates chi-square p-value < .05.

Comparing the $240 condition to usual care, effect sizes ranged from moderate to large (ds range 0.33 to 0.97), indicating a benefit of $240 CM over usual care (significant for the 0 versus any negative samples provided threshold). Effect sizes comparing the $240 to the $80 CM condition showed similar moderate to large effect sizes supporting the $240 condition (ds range 0.39 to 0.83), with 3 of the 6 thresholds reaching significance. The comparison of the $80 condition to usual care, however, suggested minimal effects for one threshold (i.e., those who provided no negative samples during the treatment period, d = 0.14) and negative effect sizes for all other thresholds indicating worse performance in the $80 condition compared to usual care. However, none of these comparisons reached statistical significance.

We also examined overall number of samples submitted (regardless of whether positive or negative for targeted substances). See Figure 2. As reported in the original report, the three conditions did not differ significantly in total samples submitted: 8.9 (5.3), 8.7 (6.3), and 10.5 (6.6) for usual care, $80, and $240 conditions, respectively (p = .37). An examination of Figure 2 suggests that the frequency of patients submitting low overall numbers of samples across treatment was denser for the usual care and $80 conditions compared to the $240 condition.

Figure 2.

Figure 2.

Open in a new tab

Total number of samples collected per patient (regardless of whether sample was positive or negative) across the three treatment conditions in Petry et al. (2004)

As reported in the original report, study retention (0-12 weeks) did not differ across conditions, M (SD): 6.2 (4.1), 6.2 (4.1), and 6.7 (4.1) for usual care, $80, and $240 conditions, respectively. Completion rates were: 13.5%, 20.0%, and 31.6%, respectively (p = .16). To mirror our focus in the present manuscript on non-engagement/non-response, we categorized patients according to whether they dropped out in the first two weeks versus those who stayed involved longer across the 3 conditions (27.0%, 24.4%, and 18.4%, usual care, $80, and $240 conditions respectively, ns). We also examined drop-out in the first month (43.2%, 40.0%, and 31.6% respectively, ns) versus those who were retained longer.

4. Discussion

In this re-analysis of Petry et al. (2004), we found clear indications that rates of non-engagement with the incentives/reinforcers for abstinence differed across low and higher magnitude CM protocols. Higher magnitude protocols generated more engagement (i.e., more access to the active component of CM: the reinforcers) among participants compared to usual care (with moderate to large effect sizes and statistical significance in one threshold) and compared to the lower magnitude CM condition (with moderate to large effect sizes and statistical significance in 3 of 6 thresholds). Importantly, the poor response in the lower magnitude protocol compared to treatment as usual (small to negative effect sizes) suggests a possible iatrogenic effect. These findings raise concern over the current SAMHSA limits on CM incentives and may indicate that implementing non-evidence-based CM is not innocuous in terms of patient impact in addition to being a poor use of resources.

In the primary study, the three conditions did not differ statistically with respect to mean total samples submitted regardless of results. However, inspection of the histogram (Figure 2) suggests that submissions may have skewed toward fewer samples submitted in the usual care and $80 condition compared to the $240 condition. Thus, some of the differences in efficacy between the $80 and $240 condition may be explained by slightly lower participation in sample submissions regardless of results. Early drop-out rates favored (were lower) for the $240 condition compared to either usual care or the $80 condition. However, neither total samples submitted, nor early drop-out rates, reached statistical significance, suggesting that these patterns do not fully explain why the $240 condition fared better (and the $80 worse) than usual care.

CM protocol recommendations (Petry, 2012) note that incentive magnitude should match the expected magnitude of effort required. It is possible that the lower magnitude $80 condition was under funded relative to the effort required and thus did not produce the expected results. Low magnitude incentives may not provide enough leverage to overcome the complexity of personal and contextual barriers relevant to behavior change (Carrera et al., 2018; Middleton et al., 2013). Another consideration is loss aversion, which refers to the tendency for individuals to have heightened sensitivity toward losses relative to gains. If the perception of low magnitude payouts is that they are inadequate compared to the loss of activities forfeited in lieu of treatment engagement, the incentives program may be rendered ineffective (Halpern et al., 2012; Mehrotra et al., 2010). Another factor may be that low payouts can be demoralizing (Eijkenaar, 2013), which could also be expected to generate dis/non-engagement. Additional studies of patient perspectives of CM may be needed to ascertain the key reasons for non-engagement. Indeed, both hypothetical scenarios and studies of participants who disengaged during CM treatment would provide novel information for the CM field and potentially inform future design considerations.

Programs interested in implementing CM using these SAMHSA dollars are faced with a decision to use a non-evidence-based magnitude in order to stay within the $75 limit, delay implementing until another funding source is identified, or supplement with alternative funds up to an effective magnitude. This restriction unnecessarily delays or restricts an evidence-based treatment from reaching patients. In a recent consensus statement (Rash et al., 2024), 26 CM experts recommended federal regulatory reforms that address barriers to effective CM delivery, including this SAMHSA restriction on incentives used for CM. In addition, the consensus recommends development of specific, approved protocols that adhere to best practices in CM delivery, including effective magnitude, immediacy, duration, and frequency of CM reinforcers.

Rash (2023) provides one such example using a 12-week prize-based protocol to reduce stimulant use that details best practices and acceptable modifications to CM parameters that avoid undermining CM’s efficacy (see Petry, 2012 for additional recommended protocols). Rash (2023) adjusted recommended expected maximum magnitudes for 12-week prize protocols for inflation, resulting in suggested magnitudes of $385 to $533. Similar adjustments to maximum magnitude possible for a 12-week voucher protocol would be $949 to $1,158 using the medians reported in Ainscough et al. (2017) and Bolivar et al. (2021). The prize CM conditions in the Petry et al. (2004) study had arranged expected maximums of $80 and $240. Adjusted for inflation, these conditions would be about $135 and $405 in value today. These inflation adjustments widen the gap between evidence-based CM magnitudes and the current SAMHSA $75 restriction.

Protocol design does have impact on patient outcomes. The Petry et al. (2004) protocol included a tapered frequency of visits with three sample submissions required in early weeks and decreasing to once weekly sample submissions in the later weeks of the protocol. The ideal approach to CM session frequency would be three visits per week (ideal) or twice weekly but well-spaced (e.g., Monday/Thursday, Tuesday/Friday) sample submissions (acceptable modification; Rash, 2023). Session frequency of less than twice weekly in abstinence-based CM protocols is discouraged (Rash, 2023) in part because this low frequency would likely miss some substance use should it occur. Petry et al. (2004) also chose to reinforce abstinence from multiple substances (alcohol, opioids, cocaine) and to reinforce multiple behaviors (abstinence, activity completion). After several decades of CM research, we now know that single drug or single drug class reinforcement results in at least double the impact on patient outcomes compared to multi-drug reinforcement targets (see Rash, 2023). The inclusion of these features (tapered schedule, multi-drug focus, dual abstinence and activity completion targets) may have suppressed outcomes based on current understanding of CM parameter design; however, any impact would have likely affected both the $240 and $80 conditions similarly as these features were consistent across CM conditions.

5. Recommendations

Federal agencies should provide “approved” protocols developed with fidelity to best practices and strictly limit programs from developing their own CM protocols given that deviations from evidence-based practices, especially in magnitude, are common (Rash et al., 2012, 2020). Major modifications or protocols designed from scratch should only be done in consultation with a CM expert who understands the impact of protocol parameter changes on efficacy. A prescribed protocol approach has been successful in implementation of CM in Department of Veteran’s Affairs (Petry et al., 2013; Rash et al., 2013) and in the Recovery Incentive Program in California (Freese et al. 2023), the two largest clinical implementation efforts in the United States to date. It is also important to note that training and ongoing coaching have been critical for these implementation initiatives (DePhilippis et al., 2018; DePhilippis et al., 2023; Freese et al., 2023; Rash & DePhilippis, 2019), as well as other state-wide (Parent et al., 2023) and regional (Becker et al., 2023) efforts.

We acknowledge the reasons behind the federal statutes (see Clark & Davis, 2023). However, in the midst of an overdose crisis, serious ethical concerns also present when restricting an evidence-based treatment from reaching patients in crisis. The need, arguably a highly pressing need, is for federal agencies to provide explicit guidance to providers on how CM can be implemented with fidelity, including to evidence-based magnitudes, while avoiding running afoul of federal laws. Without explicit guidance, providers are likely to stay within the nominal value recommendations (currently $15 per item or $75 per patient per year in aggregate) or decline to implement CM altogether, restricting access to a treatment that works. Importantly, this re-analysis suggests that offering CM designed to stay within the $75 limit may have possible harmful effects. In addition to easing the restrictions, we recommend a requirement that any federal funds used for CM be permitted only if all parameters of CM protocols adhere to best practices, including magnitude, or, alternately, the development of federally approved prize- and voucher-based CM protocols that have been well vetted by CM experts as evidence-based and likely to positively impact patient outcomes.

6. Conclusions

With strong support and guidance for evidence-based CM protocols, this much needed intervention can be more widely implemented in clinical settings with fidelity and in a manner that will benefit patients. Without requirements that programs adhere to best practices, prior research suggests deviations, especially in maximum incentive magnitude, are likely. This report highlights possible negative impacts of low magnitude CM protocols. Beyond the broader ethical concerns in offering a less effective version of a treatment, providers implementing CM protocols within the current $75 SAMHSA limit may introduce harm to some patients by increasing non-engagement rates. We recommend strict adherence to best practices in CM protocols, federal reforms to ease CM implementation, and the development of recommended evidence-based prize and voucher protocols.

Highlights.

  • Low incentive magnitude may have iatrogenic effects on patient engagement.

  • The $75/patient/year limit on incentives is a barrier to evidence-based contingency management.

  • Regulatory reforms are needed to promote best practices in contingency management.

Funding:

Preparation of this report was supported by National Institutes of Health grants: P30-DA023918, P50-DA09241, P50-AA027055, and General Clinical Research Center Grant M01-RR06192. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

CRediT statement: CJR conceptualized the study, conducted data analyses, and wrote the first draft. SMA was an investigator on the original 2004 study and contributed to the writing of the manuscript. KZ contributed to the writing of the manuscript.

1

In contrast to voucher CM protocols, in which the maximum possible magnitude is a set value and the same for all patients, for prize CM, the expected magnitude per patient is calculated based on the probabilities of prize categories in the fishbowl. The resulting value is termed ‘average expected maximum magnitude’, representing a typical or ’average’ maximum because individual patients may earn more or less based on the results of their draws. This maximum value differs from earnings, which are typically about half of the expected maximum in clinical trials. Statements about the magnitude of CM protocols typically refer the maximum value rather than earnings.

Declaration of interests: CJR discloses past consulting relationships with Affect Therapeutics and RealWorks, and collaboration with Dynamicare Health. SMA discloses collaboration with Q2i, LLC and Contingency Management Innovations, LLC.

References

  1. Ainscough TS, McNeill A, Strang J, Calder R, & Brose LS (2017). Contingency Management interventions for non-prescribed drug use during treatment for opiate addiction: A systematic review and meta-analysis. Drug and Alcohol Dependence, 178(June), 318–339. 10.1016/j.drugalcdep.2017.05.028 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Andraka-Christou B, Williams M, Buksbaum S, Karkkainen J, Stein BD, Batchelder SR, et al. (2023). US state laws relevant to incentives for health behavior: A qualitative analysis. Preventive Medicine, 176(Nov), 107645. doi: 10.1016/j.ypmed.2023.107645 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Becker S. (2023). Effective, but underused: Lessons learned implementing contingency management in real-world practice settings in the United States. Preventive Medicine, 176(October), 107594. doi: 10.1016/j.ypmed.2023.107594 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. 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. 10.1111/add.12589 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bolívar HA, Klemperer EM, Coleman SRM, DeSarno M, Skelly JM, & Higgins ST (2021). Contingency management for patients receiving medication for opioid use disorder: A systematic review and meta-analysis. JAMA Psychiatry, 78(10), 1092–1102. 10.1001/jamapsychiatry.2021.1969 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Businelle MS, Rash CJ, Burke RS, & Parker JD (2009). Using vouchers to increase continuing care participation in veterans: Does magnitude matter? The American Journal on Addictions, 18, 122–129. 10.1080/10550490802545125 [DOI] [PubMed] [Google Scholar]
  7. Carrera M, Royer H, Stehr M, & Sydnor J (2018). Can financial incentives help people trying to establish new habits? Experimental evidence with new gym members. Journal of Health Economics, 58, 202–214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Clark HW, & Davis M (2023). Federal legal and regulatory aspects of contingency management incentives. Preventive Medicine, 176(Nov), 107726. 10.1016/j.ypmed.2023.107726 [DOI] [PubMed] [Google Scholar]
  9. Dallery J, Silverman K, Chutuape MA, Bigelow GE, & Stitzer ML (2001). Vocuher-based reinforcement of opiate plus cocaine abstinence in treatment-resistant methadone patients: Effects of reinforcer magnitude. Experimental and Clinical Psychopharmacology, 9(3), 317–325. 10.1037//1064-1297.9.3.317 [DOI] [PubMed] [Google Scholar]
  10. 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. Preventive Medicine, 92, 36–46. 10.1016/j.ypmed.2016.08.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Department of Veterans Affairs & Department of Defense (2021). VA/DoD Clinical Practice Guideline for the Management of Substance Use Disorders. Prepared by the Management of Substance Use Disorders Work Group with support from the Office of Quality and Patient Safety, Veterans Health Administration, & Clinical Quality Improvement Program, Defense Health Agency. https://www.healthquality.va.gov/guidelines/MH/sud/VADoDSUDCPG.pdf [Google Scholar]
  12. 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 and Alcohol Dependence, 185(December 2017), 367–373. 10.1016/j.drugalcdep.2017.12.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. DePhilippis D, Khazanov G, Christofferson DE, Wesley CW, Burden JL, Liberto J, & McKay JR (2023). History and current status of contingency management programs in the Department of Veterans Affairs. Preventive Medicine, 176(Nov), 107704. doi: 10.1016/j.ypmed.2023.107704 [DOI] [PubMed] [Google Scholar]
  14. Dutra L, Stathopoulou G, Basden SL, Leyro TM, Powers MB, & Otto MW (2008). A meta-analytic review of psychosocial interventions for substance use disorders. The American Journal of Psychiatry, 165(2), 179–187. 10.1176/appi.ajp.2007.06111851 [DOI] [PubMed] [Google Scholar]
  15. Eijkenaar F. (2013). Key issues in the design of pay for performance programs. The European Journal of Health Economics, 14, 117–131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Forster SE, DePhilippis D, & Forman SD (2019). “I’s” on the prize: A systematic review of individual differences in Contingency Management treatment response. Journal of Substance Abuse Treatment, 100, 64–83. 10.1016/j.jsat.2019.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Freese TE, Rutkowski BA, Peck JA, Urada D, Clark HW, Nigusse A, Friedman J, & Rawson RA (2023). Recovery incentives program: California’s contingency management benefit. Preventive Medicine, 176(September), 107703. 10.1016/j.ypmed.2023.107703 [DOI] [PubMed] [Google Scholar]
  18. Ginley ΜK, Pfund RA, Rash CJ, & Zajac K (2021). Long-term efficacy of contingency management treatment based on objective indicators of abstinence from illicit substance use up to 1 year following treatment: A meta-analysis. Journal of Consulting and Clinical Psychology, 89(1), 58–71. 10.1037/ccp0000552 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Griffith JD, Rowan-Szal GA, Roark RR, & Simpson DD (2000). Contingency management in outpatient methadone treatment: A meta-analysis. Drug and Alcohol Dependence, 58(1–2), 55–66. 10.1016/S0376-8716(99)00068-X [DOI] [PubMed] [Google Scholar]
  20. Halpern SD, Asch DA, & Volpp KG (2012). Commitment contracts as a way to health. BMJ, 344. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Higgins ST, Heil SH, Dantona R, Donham R, Matthews M, & Badger GJ (2007). Effects of varying the monetary value of voucher-based incentives on abstinence achieved during and following treatment among cocaine-dependent outpatients. Addiction, 102(2), 271–281. 10.1111/j.1360-0443.2006.01664.x [DOI] [PubMed] [Google Scholar]
  22. 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. 10.1111/j.1360-0443.2006.01311.x [DOI] [PubMed] [Google Scholar]
  23. McDonell MG, Hirchak K, McCrady E, Clark HW, & Rawson R (2023). Federal policy changes must occur to provide a lifesaving stimulant drug intervention. Health Aff Forefr [Internet cited 2024 July 18]; Available from: https://www.healthaffairs.org/do/10.1377/forefront.20240209.728693/full/
  24. Mehrotra A, Sorbero ME, & Damberg CL (2010). Using the lessons of behavioral economics to design more effective pay-for-performance programs. The American Journal of Managed Care, 16(7), 497. [PMC free article] [PubMed] [Google Scholar]
  25. Middleton KR, Anton SD, & Perri MG (2013). Long-term adherence to health behavior change. American Journal of Lifestyle Medicine, 7(6), 395–404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Office of the Inspector General, Department of Health and Human Services (2020). Medicare and State Health Care Programs; Fraud and Abuse; Revisions to Safe Harbors Under the Anti-Kickback Statute, and Civil Monetary Penalty Rules Regarding Beneficiary Inducements. Federal Register, 85(232). [PubMed] [Google Scholar]
  27. Olsen Y. (2023). Dr. Olsen Opening Remarks 032223 [video]. Substance Abuse and Mental Health Administration, Center for Substance Abuse Treatment, Contingency Management Taskforce Kick-off Meeting. [Google Scholar]
  28. Packer RR, Howell DN, McPherson S, & Roll JM (2012). Investigating reinforcer magnitude and reinforcer delay: A contingency management analog study. Experimental and Clinical Psychopharmacology, 20(4), 287–292. 10.1037/a0027802 [DOI] [PubMed] [Google Scholar]
  29. Parent SC, Peavy KM, Tyutyunnyk D, Hirchak KA, Nauts T, Dura A, Weed L, Barker L, & Mcdonell MG (2023). Lessons learned from statewide contingency management rollouts addressing stimulant use in the Northwestern United States. Preventive Medicine, 176(November), 107614. 10.1016/j.ypmed.2023.107614 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Petry NM (2012). Contingency management for substance abuse treatment: A guide to implementing this evidence-based practice. Routledge. [Google Scholar]
  31. 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. Journal of Consulting and Clinical Psychology, 80(2), 276–285. 10.1037/a0026883 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Petry NM, Dephilippis D, Rash CJ, Drapkin M, & McKay JR (2014). Nationwide dissemination of contingency management: The Veterans Administration initiative. American Journal on Addictions, 23(3), 205–210. 10.1111/j.1521-0391.2014.12092.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. 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. 10.1111/j.1360-0443.2003.00642.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Pfund RA, Ginley MK, Boness CL, Rash CJ, Zajac K, Witkiewitz K, Pfund RA, Ginley MK, Boness CL, Rash CJ, Zajac K, & Witkiewitz K (2022). Contingency Management for Drug Use Disorders: Meta-Analysis and Application of Tolin’s Criteria Contingency Management for Drug Use Disorders. Clinical Psychology: Science and Practice. 10.1037/cps0000121 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Pfund RA, Ginley MK, Rash CJ, & Zajac K (2021). Contingency management for treatment attendance: A meta-analysis. Journal of Substance Abuse Treatment, March, 108556. 10.1016/j.jsat.2021.108556 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. 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. 10.1111/j.1360-0443.2006.01581.x [DOI] [PubMed] [Google Scholar]
  37. Rash CJ, Alessi SM, & Zajac K (2020). Examining implementation of contingency management in real-world settings. Psychology of Addictive Behaviors, 34(1), 89–98. 10.1037/adb0000496 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Rash CJCJ, Alessi SMSM, & Petry NMNM (2017). Substance abuse treatment patients in housing programs respond to contingency management interventions. Journal of Substance Abuse Treatment, 72, 97–102. 10.1016/j.jsat.2016.07.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Rash CJ, Petry NM, Kirby KC, Martino S, Roll J, & Stitzer ML (2012). Identifying provider beliefs related to contingency management adoption using the Contingency Management Beliefs Questionnaire. Drug and Alcohol Dependence, 121(3), 205–212. 10.1016/j.drugalcdep.2011.08.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Rash CJ, & DePhilippis D (2019). Considerations for implementing contingency management in substance abuse treatment clinics: The Veterans Affairs initiative as a model. Perspectives on Behavior Science, 42(3). 10.1007/s40614-019-00204-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Rash Carla J. (2023). Implementing evidence-based prize contingency management for stimulant use. Journal of Substance Use and Addiction Treatment. Aug:151:209079. doi: 10.1016/j.josat.2023.209079 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Rash Carla J., Andrade LF, & Petry NM (2013). Income received during treatment does not affect response to contingency management treatments in cocaine-dependent outpatients. Drug and Alcohol Dependence, 132(3), 528–534. 10.1016/j.drugalcdep.2013.03.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Rash Carla J., McDonell MG, Erath TG, Parent SC, Rawson RA, & Wattenberg S. (2024). A call to action: Evidence-based contingency management. The American Journal of Psychiatry, in press. [Google Scholar]
  44. Rash Carla J., Olmstead TA, & Petry NM (2009). Income does not affect response to contingency management treatments among community substance abuse treatment-seekers. Drug and Alcohol Dependence, 104(3), 249–253. 10.1016/j.drugalcdep.2009.05.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Rash Carla J, Alessi SM, & Petry NM (2008a). Cocaine abusers with and without alcohol dependence respond equally well to contingency management treatments. Experimental and Clinical Psychopharmacology, 16(4), 275–281. 10.1037/a0012787 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Rash Carla J, Alessi SM, & Petry NM (2008b). Contingency management is efficacious for cocaine abusers with prior treatment attempts. Experimental and Clinical Psychopharmacology, 16(6), 547–554. 10.1037/a0014042 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Rash Carla J, DePhilippis D, McKay JR, Drapkin M, & Petry NM (2013). Training workshops positively impact beliefs about contingency management in a nationwide dissemination effort. Journal of Substance Abuse Treatment, 45(3), 306–312. 10.1016/j.jsat.2013.03.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Rash Carla J, & Petry NM (2015). Contingency management treatments are equally efficacious for both sexes in intensive outpatient settings. Experimental and Clinical Psychopharmacology, 23(5), 369–376. 10.1037/pha0000035 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Register F. (2020). Medicare and State Health Care Programs: Fraud and Abuse. Revisions to Safe Harbors Under the Anti-Kickback Statute, and Civil Monetary Penalty Rules Regarding Beneficiary Inducements. https://www.federalregister.gov/documents/2020/12/02/2020-26072/medicare-and-state-health-care-programs-fraud-and-abuse-revisions-to-safe-harbors-under-the [PubMed] [Google Scholar]
  50. Regnier SD, Strickland JC, & Stoops WW (2022). A preliminary investigation of schedule parameters on cocaine abstinence in contingency management. Journal of the Experimental Analysis of Behavior, 118(1), 83–95. 10.1002/jeab.770 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. 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, 146(2), 128–138. 10.1007/s002130051098 [DOI] [PubMed] [Google Scholar]
  52. Substance Abuse and Mental Health Services Administration. (2021). Treatment of Stimulant Use Disorders: Treatment Improvement Protocol (TIP) Series 33. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2021. [Google Scholar]
  53. Substance Abuse and Mental Health Services Administration, National Survey of Substance Abuse Treatment Services (N-SSATS): 2020. Data on Substance Abuse Treatment Facilities. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2021. [Google Scholar]
  54. U.S. Department of Health and Human Services. (2023). Contingency management for the treatment of substance use disorders: Enhancing access, quality, and program integrity for an evidence-based intervention. Office of the Assistant Secretary for Planning and Evaluation. [PubMed] [Google Scholar]
  55. Volkow ND (2020). Rising Stimulant Deaths Show that We Face More than Just an Opioid Crisis. National Institute on Drug Abuse (NIDA). https://www.drugabuse.gov/about-nida/noras-blog/2020/11/rising-stimulant-deaths-show-we-face-more-than-just-opioid-crisis [Google Scholar]
  56. White House, Executive Office of the President, Office of National Drug Control Policy (2022). National Drug Control Strategy. https://www.whitehouse.gov/wp-content/uploads/2022/04/National-Drug-Control-2022Strategy.pdf [Google Scholar]
  57. Wilson DB (2023). Practical meta-analysis effect size calculator (Version 2023.11.27) https://www.campbellcollaboration.org/escalc/html/EffectSizeCalculator-Home.php [Google Scholar]
  58. Zajac K, Rash CJCJCJ, Ginley MKMK, & Heck NC (2020). Sexual orientation and substance use treatment outcomes across five clinical trials of contingency management. Psychology of Addictive Behaviors, 34(1), 128–135. 10.1037/adb0000494 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES