CATR: Define contingency management for us. 

Dr. DePhilippis: Contingency management (CM) is an evidence-based treatment with particular effectiveness for stimulant use disorder (StimUD). I like to say that everything you need to know about CM you already learned in Psychology 101. It’s based on fundamental learning principles described long ago by B.F. Skinner. At its core, CM is the therapeutic application of operant conditioning.

CATR: Remind us of the basics of operant conditioning and tell us how they apply to CM.

Dr. DePhilippis: In operant conditioning, behaviors are shaped by the application of reinforcements and punishment. In CM, we are taking advantage of four features of operant conditioning. First, we use positive reinforcement, meaning we provide a reward, or incentive, when a patient engages in a desired behavior. In our case, that behavior is abstaining from drug use. And the longer they abstain, the bigger the reward. Second, if they use drugs—that is, they don’t engage in recovery behavior—they get no reward. That takes advantage of the second feature of operant conditioning: extinction. This term describes how behaviors tend to die out if they are not reinforced. Third, we have a mild punishment in place. If the patient uses drugs, not only do they receive no incentive that day, but they also have their incentive amount reset to the beginning. And fourth is negative reinforcement. Negative reinforcement is when behaviors are encouraged through the avoidance of an unpleasant consequence—in this case, the reset of the reward.

CATR: How do these programs work logistically?

Dr. DePhilippis: First, we must decide what behavior we want to reinforce. For our purposes, that behavior is abstinence. Next, we must be able to tell whether a patient has engaged in that behavior. Fortunately, urine drug screens (UDS) with immediately available results accomplish this easily (Editor’s note: For more on UDS, see CATR May/June 2022.) As an aside, I can’t overstate the importance of immediately available results; learning is best when consequences are immediate. That’s why we utilize immunoassay-based drug testing that gives results right away. So, we have our behavior and we have our means of monitoring it; now we apply the operant conditioning principles. If the patient is abstinent, meaning they test negative for stimulants, they receive an incentive immediately. 

CATR: What kind of incentives are used?

Dr. DePhilippis: There are various possibilities, but the incentive must compete with the reinforcement associated with substance use. And that can be a tough sell. Substance use says, “I can make you feel really good. There may be consequences down the road, but right now you’re gonna feel good.” Recovery has an inherent disadvantage. Not only can it be uncomfortable in the moment, but it can also give you a more lucid, clear-eyed view of the devastation in your life brought on by substance use. So, the incentive we provide must be immediately reinforcing. It also needs to be broadly appealing. In the network of Veterans Administration (VA) hospitals, we have access to coupons for the Veterans Canteen Service. These coupons are legal tender that can be used to purchase all sorts of merchandise available through VA outlets—anything from coffee or a candy bar to a laptop computer or flat-screen TV. 

CATR: And what about providers who don’t have access to these coupons?

Dr. DePhilippis: Some research protocols use cash as an incentive, and that certainly meets our criteria of being broadly appealing. The concern with cash payments, of course, is that cash can be used to buy drugs or other materials that may complicate recovery, like tobacco or alcohol products, or it can be used for gambling. And while that is a reasonable concern, evidence suggests that it happens less than you might think. In fact, cash and noncash reinforcement conditions show similar rates of stimulant use when compared head-to-head (Festinger DS et al, J Subst Abuse Treat 2014;47(2):168–174). 

CATR: Most clinics don’t have the resources to provide cash to patients, either.

Dr. DePhilippis: That’s very true. Gift cards are the most common alternative to cash. But sourcing gift cards can be a challenge as well. Funding meaningful incentives remains one of the greatest challenges to implementing CM more broadly. But even so, I would encourage your readers to investigate local resources because there are state and county grants out there that fund CM programs. For example, California has a particularly robust CM system in place (www.tinyurl.com/n6sucka2).

CATR: Critics of CM might say, “Why should we pay someone not to use drugs? They should be doing that anyway.” How do you respond to that?

Dr. DePhilippis: I don’t think this is a proper application of the word “pay.” When I pay someone, I am compensating them for a task or a good that serves my best interest. I pay a mechanic to fix my car; I pay the store for groceries. But in CM, we’re providing reinforcement to strengthen a behavior that’s in the patient’s best interest, not our own. So, referring to it as payment is not an accurate description. Furthermore, we know that managed contingencies, positive and negative reinforcement, extinction, and punishment are how we learn. In fact, that’s a large part of how substance use disorders develop in the first place. We are simply leveraging these same contingencies in the interest of treatment.

CATR: You mentioned that CM is particularly effective for StimUD. Can you walk us through that evidence?

Dr. DePhilippis: The strongest evidence supporting the application of CM comes from a series of meta-analyses (Bentzley BS et al, JAMA Netw Open 2021;4(5):e218049). These studies have very reliable and consistent findings that CM is the most effective treatment modality for StimUD. In fact, the evidence is so robust that the American Academy of Addiction Psychiatry (AAAP) and the American Society of Addiction Medicine (ASAM) jointly published a set of clinical practice guidelines for the treatment of StimUD in which they identify CM as the very best treatment (www.tinyurl.com/mr2rxuun). In addition, an expert panel from the VA and the Department of Defense developed treatment guidelines for substance use disorder (SUD) and named CM as the treatment of choice for StimUD (www.tinyurl.com/78nwyam6).

CATR: What is it about StimUD that makes it so conducive to CM?

Dr. DePhilippis: That’s a good question. Because CM is based on operant conditioning, you might expect it to be effective across the board: for alcohol use disorder (AUD), opioid use disorder (OUD), and so forth. However, one challenge is in the mechanics—namely, the importance of immediate incentives and reliably detecting whether someone has used drugs. A UDS remains positive for roughly 24–72 hours after use of stimulants. That’s a convenient length of time. We can test a patient twice a week, which is a typical schedule, and we’re surveilling a full week’s worth of behavior. In other words, if a patient tests negative, we can be fairly confident that the patient has abstained for the past week. But let’s look at alcohol. A breathalyzer will only remain positive for six to 12 hours after the last drink. So, if we’re truly going to surveil for alcohol abstinence, we’d have to test the patient twice a day, which is impractical. There is some work being done with the biomarker ethyl glucuronide, but we don’t have a reliable immediate test for it (McDonell MG et al, Am J Psychiatry 2017;174(4):370–377). On the flip side, a UDS for cannabis can remain positive for a month or more, so in that case, we’re unable to get results about recent behaviors. 

CATR: What about opioids?

Dr. DePhilippis: Implementing CM with OUD is complicated for a few reasons. First, opioids bought and sold on the street are a highly heterogeneous group of drugs: prescription pills, heroin, fentanyl, and fentanyl derivatives, in addition to other compounds like xylazine, kratom, and tianeptine. An effective CM program would need to surveil this entire spectrum, which would be a huge challenge. Second, CM targets abstinence as its recovery behavior. If a patient abstains, their tolerance drops. We could end up inadvertently increasing the risk of fatal opioid exposure if the patient ever uses again, and addiction is a relapsing-remitting illness by its nature. Finally, the evidence is clear that the most effective, life-saving treatments for OUD are the three FDA-approved medications: methadone, buprenorphine, and extended-release naltrexone. Not only are those treatments effective, but they also offer a measure of safety because they protect against overdose. CM can’t offer that. 

CATR: And StimUD doesn’t have the same medication options that OUD and AUD have.

Dr. DePhilippis: That’s right. In fact, there are no FDA-approved medication options at all. There are some signals in the literature for pharmacotherapy, primarily for specific patient subgroups, but none of them are as effective as CM (Chan B et al, Drug Alcohol Depend 2020;216:108193).

CATR: What about false positive UDS results for stimulants?

Dr. DePhilippis: Yes, false results are challenges one must be prepared to address. With cocaine, false positives are rare, but with methamphetamine and amphetamine, the likelihood is greater. Trazodone and bupropion can cause false positives, and so can some patients’ amphetamine medication for ADHD that is indistinguishable from other amphetamine use on the UDS. The consent process should make clear that self-report and clinician judgment don’t have bearing on CM’s reinforcement schedule; the UDS result is the sole determinant of reinforcement. Before starting, providers should review prescriptions and, if possible, make adjustments to minimize the risk of false results. If such changes are not possible or are unsuccessful, then the patient might not be a good candidate for CM. An alternative treatment, like cognitive behavioral therapy (CBT) for SUD, might be a better option.

CATR: What if you are working with a patient who isn’t ready for abstinence? Is there still a place for CM in their treatment?

Dr. DePhilippis: What I’m hearing in that question is: “What about harm reduction? What about patients who don’t embrace an abstinence model?” Well, the beauty of CM is that it does not compel a patient to embrace an abstinence goal. Instead, it suggests, “Dip your toe in the water; if you were to try abstinence for even a brief period of time, CM is there to provide reinforcement that could tip the motivational balance.” The goal is for the patient to realize not only that they can do this, but that it’s worth doing.

CATR: What are some remaining open questions in the field?

Dr. DePhilippis: One of the biggest challenges is knowing what level of incentive is necessary. I mentioned the state of California earlier. Well, California received a waiver to offer CM funded through Medicaid, and they established their reinforcement level at $599 maximum per patient over the course of the study. This is because at $600 you trigger IRS reporting requirements, which obviously creates implementation challenges. But is $599 enough? We don’t know. Some studies found that $500 is an amount that could effectively reinforce abstinence, but those studies are decades old (Petry NM. Contingency Management for Substance Abuse Treatment: A Guide to Implementing This Evidence-Based Practice. New York, NY: Routledge/Taylor & Francis Group; 2012:141). We likely need to increase the amount to accommodate for increases in cost of living, but we don’t know how much is needed. 

CATR: Over what time period is the money distributed?

Dr. DePhilippis: It varies. The $500 figure emerged from studies of 12 weeks in duration. The California program is $599 over 24 weeks. Veterans in our CM program typically earn $250–$300 over the course of a 12-week protocol. 

CATR: These programs are all time limited. What happens afterward?

Dr. DePhilippis: Your question speaks to a common critique of CM: “Once you withdraw the reinforcement, you’re going to have recurrence of symptoms.” First, we don’t level this criticism toward other chronic conditions. Imagine saying, “As soon you stop antihypertensive medication, blood pressure goes back up, so I guess antihypertensives aren’t effective.” No, that’s not how it works. In fact, the recurrence of symptoms is evidence that the treatment did work. Am I suggesting that CM be unlimited? For some, maybe. What about for others? Surely the answer depends on the individual. Right now, we don’t have the precision to identify who might benefit from what duration of treatment. But there is another concern here: Might we be decreasing our patients’ own internal motivation toward abstinence? Well, I don’t think so, and I’d argue that we are enhancing it. In fact, we have evidence that CM effects are durable beyond the protocol itself. Studies that follow patients after they have completed CM have found enduring benefits for up to a year after incentives are discontinued (Ginley MK et al, J Consult Clin ­Psychol 2021;89(1):58–71). And when a course of CM is completed, the patient can receive other treatments like CBT.

CATR: What resources are available for those interested in learning more? 

Dr. DePhilippis: I already mentioned the clinical practice guidelines put out by the VA and the Department of Defense, as well as the guidelines put out by the AAAP and the ASAM. Both are great resources and very comprehensive. I’d also point readers to educational materials put out by the Substance Abuse and Mental Health Services Administration through their Addiction Technology Transfer Center (www.tinyurl.com/6hd2vz89). Finally, the Northwest Addiction Technology Transfer Center has a set of asynchronous CM learning materials with modules designed for clinicians and program leaders (www.tinyurl.com/5a22yyva). These materials are not intended to be stand-alone training resources, but they can help a program do a deeper dive into what CM implementation entails.

CATR: Thank you for your time, Dr. DePhilippis.