The Drug Killing More Americans Every Year Has No Approved Medication. The One Thing That Works Is Funded in Five States.

There is a 30-second script that treatment counselors at meth-using populations have had to internalize as their baseline pitch for going on two decades now. It goes something like this: I know you’re struggling. There are things we can do. There is no medication. But if you can get to this clinic, take this urine test, and it comes back negative, we will give you a gift card worth about twenty-five dollars. Come back and do it again next week. We will give you another one, slightly more. The idea is that the small reward helps your brain — which the meth has spent months or years depleting of the ability to feel much of anything — start to re-associate sobriety with something that feels good.

That script is contingency management. It is the best clinical tool the field has for stimulant use disorder. It is not medication. It is not a cure. It works by doing manually, through structured incentives, what a functioning dopamine system would do on its own — reward behaviors that support recovery. Across a substantial clinical evidence base, it works: it reduces stimulant use, improves treatment retention, and does not produce the side effects of a drug because it is not a drug.

In five U.S. states, Medicaid pays for it. In forty-five, it does not.

Why there is no medication — and why that matters

The opioid epidemic transformed addiction medicine in one specific way that does not apply to stimulants: it gave the field medications. Buprenorphine and methadone, both opioid agonists, can directly address the craving and withdrawal that drive opioid use. They work because opioids and their pharmacological substitutes operate on the same receptor system. The chemistry of opioid dependence is, in this sense, clinically tractable.

Methamphetamine and cocaine operate through an entirely different mechanism. Both are stimulants that cause a massive flood of dopamine — the brain’s reward signal — into the synaptic cleft, producing the intense euphoria that drives compulsive use. But the mechanism makes them pharmacologically hard to treat with a substitute. There is no “methamphetamine maintenance” that functions the way methadone maintenance functions for opioids, because the pharmacology doesn’t translate. The rewarding effect of stimulants is not mediated through a single receptor subtype the way opioid effects are. It’s diffuse, plastic, and deeply integrated with the brain’s broader reward-learning system.

What prolonged stimulant use does to the brain is well-documented and genuinely sobering. The dopamine flood is followed by depletion. The brain’s D2 receptor density — the number of receptors available to detect dopamine — decreases significantly in people who use methamphetamine chronically. The result is anhedonia: an inability to experience pleasure from things that previously produced it. Food doesn’t taste as good. Sex is less rewarding. A conversation with someone you love doesn’t produce the warmth it once did. What produces anything at all, at least initially, is the drug.

Anhedonia is not a metaphor or a mood. It is a measurable neurological state with real implications for recovery. When the expected experience of sobriety is a protracted gray period in which almost nothing feels good, the pull of relapse is not a moral failing — it’s a rational response to a terrible choice architecture. This is why the “just quit” framework fails so completely for stimulant use disorder. It asks someone whose reward system has been profoundly disrupted to sustain behavior that offers no neurological return, at least in the short term.

Recovery from stimulant use disorder often involves a period of weeks to months during which dopamine signaling gradually normalizes. The brain does recover, at least partially. But that recovery window is when people are most at risk, and it is precisely the window during which the treatment system currently offers the least.

But that recovery window is when people are most at risk, and it is precisely the window during which the treatment system currently offers the least.

What contingency management does — and what it doesn’t

Contingency management works by providing external reward signals during the period when the brain’s internal reward system is offline. The incentives are small — federal rules cap them at $750 per patient per year, structured as escalating gift cards, never cash — but the effect on behavior and on retention in treatment is clinically significant. A 2024 analysis of California’s Recovery Incentives program, the most mature Medicaid CM program in the country, found meaningful reductions in methamphetamine-positive urinalysis results compared to standard care.

The mechanism is not mysterious. If your dopamine system has been depleted and sobriety produces very little that feels rewarding, an external reward — even a $25 Walmart gift card — can provide enough of a signal to reinforce the behavior until the brain starts to provide its own. It is also possible that the structure of the program itself — the regular clinic visits, the relationship with a counselor, the accountability — provides therapeutic benefit independent of the incentive. The research can’t fully separate these effects, and for clinical purposes, it doesn’t have to.

What CM doesn’t do is address the neurological underpinnings of craving directly the way buprenorphine addresses opioid craving. Someone leaving a CM program faces the same reward-system deficit they entered with, tempered by whatever normalization has occurred during the treatment period. The incentive removes when the program ends. Long-term outcomes data for CM are promising but not definitive, and the research community is clear that CM is most effective when it’s part of a broader treatment infrastructure — counseling, housing stability, social support — not a standalone fix.

But the alternative to CM is not some better treatment that doesn’t exist yet. The alternative to CM, for most people with stimulant use disorder in the United States right now, is nothing evidence-based at all.

Five states. Forty-five that aren’t.

The Medicaid CM landscape in 2026 is: California, Delaware, Hawaii, Montana, and Washington have approved Section 1115 waivers to cover contingency management. Michigan and Rhode Island have applications pending. The other forty-five states and territories have not begun the process.

The reasons are a mix of the structural and the ideological. Structurally, Medicaid waiver applications are slow, resource-intensive, and require states to demonstrate that the program will be cost-effective — which is harder to prove for behavioral interventions than for medications, because behavioral interventions don’t have a pharmaceutical company funding the evidence generation. Ideologically, there is persistent congressional and administrative resistance to what critics frame as “paying people to stop using drugs” — a framing that strips the clinical context from a behavioral health intervention and applies a moral lens to a medical question.

California’s program is under particular pressure. The state’s Medicaid director has until the end of 2026 to demonstrate that the Recovery Incentives program produces enough clinical and economic value to justify continued federal funding. The data is encouraging, but the political climate is not. Under the current administration, the expansion of CM to additional states through waivers has stalled.

Meanwhile, stimulant overdose deaths — the overwhelming majority of them involving fentanyl in the supply chain, methamphetamine users inadvertently consuming it in adulterated product — have reached a level where a Frontiers in Psychiatry review in 2026 called an “urgent” response necessary. More than half of U.S. overdose deaths now involve a stimulant. Stimulant-involved fatalities reached 57,500 in 2022. An estimated 70 percent of those deaths involved fentanyl — meaning many people dying from stimulant use are dying not from the stimulant itself but from what’s in the supply alongside it.

What the treatment field — and platforms like Rize — can do right now

The most immediate practical implication is a navigation one: people seeking help for methamphetamine or cocaine use disorder should be connected explicitly to programs that offer contingency management, in the states where Medicaid covers it, and to sliding-scale options elsewhere. For Arizonans, AHCCCS does not currently have an approved CM waiver — which means this is a gap in Arizona’s treatment infrastructure that advocacy organizations and providers are positioned to address through the state’s opioid settlement spending processes.

Longer term, the research questions that remain open for stimulants — better understanding of which patients benefit most from CM, whether digital delivery platforms can extend CM’s reach, how CM interacts with emerging pharmacotherapies being tested in trials — are the questions that deserve investment. The PEER-CM study, published in Addiction Science & Clinical Practice, is testing peer-delivered CM as a way to reach people who are not in formal treatment — an approach that could dramatically expand access if it proves effective.

For now, the situation is this: stimulants are involved in more than half of overdose deaths, no medication exists to treat the underlying disorder, the best behavioral treatment is funded in five states, and the people most at risk are often precisely the people least able to navigate a fragmented system to find it. That gap is not an accident of science — the science is clear. It’s a gap in political will and infrastructure investment. It is treatable.

That gap is not an accident of science — the science is clear.

Browse stimulant treatment resources at /newsroom/substances/stimulants or explore Rize’s treatment matching for Arizona-specific options.