Relapse Prevention: What Actually Works
Most relapse prevention advice is a list of reasonable-sounding suggestions with no indication of which ones matter. Exercise, journaling, meetings, therapy, medication, a support network, avoiding triggers — all presented as equally weighted, as though the reader’s job is to do all of them and hope.
The research does not treat them equally. Some of these have randomized trials behind them and measurable effect sizes. Some are good ideas with thin evidence. Knowing which is which is the difference between a plan and a wish list.
First, the number that reframes the question
Between 40% and 60% of people treated for a substance use disorder return to use at some point.1 That figure gets quoted as evidence that treatment fails. It is actually the opposite: it is close to the relapse rate for hypertension and asthma, two conditions nobody describes as untreatable.
The comparison is not a consolation prize. It changes what the goal should be. For a chronic condition, the measure of success is not whether symptoms ever return — it is how quickly they are caught, and whether the response is adjustment or abandonment. A person whose blood pressure creeps up does not conclude that medicine does not work. They change the dose.
Relapse is a process, and it starts long before the drink
The most useful single idea in this field is that relapse has three stages, and the substance shows up only in the last one.2
Emotional relapse. The person is not thinking about using. But the maintenance has stopped — sleep is poor, meetings are skipped, meals are irregular, feelings are bottled, isolation is creeping in. This stage can run for weeks. Nothing in it looks like a warning sign to the person living it, because from the inside it feels like being busy or being tired.
Mental relapse. Now there is an argument going on. Part of the mind is bargaining — remembering using fondly, minimizing past consequences, thinking about people and places, planning around supervision, looking for a scenario where it would be fine this time. Some people describe this as lying to themselves and knowing it.
Physical relapse. The drink or the drug. By this point most of the decisions have already been made.
The practical consequence is that relapse prevention aimed at the third stage is aimed too late. The techniques that work at stage three are emergency measures with a low success rate. The ones that work at stage one are unglamorous: sleeping, eating, showing up, telling someone the truth about how the week is actually going.
What the evidence actually supports
Ranked roughly by strength of evidence rather than by popularity.
Medication, where it applies
This is the most under-used effective intervention in the field, and the most likely thing to be missing from a plan.
For alcohol use disorder, naltrexone and acamprosate have the strongest support of the approved options. In the largest meta-analysis, roughly one in twelve people treated with acamprosate avoided a return to drinking who otherwise would not have, and roughly one in twenty on naltrexone.4 Those are not miracle numbers, and they are better than most psychosocial interventions manage on their own.
For opioid use disorder, methadone and buprenorphine are associated with substantially reduced overdose and all-cause mortality.5 The evidence here is strong enough that withholding these medications is itself a risk.
For stimulants, cannabis and several other substances, there is currently no medication with comparable support. That is worth knowing so nobody waits for one.
None of this is a decision to make from an article. It is a conversation to have with a prescriber, and the point of raising it here is that many people never have that conversation at all.
Structured relapse prevention therapy
Cognitive-behavioral relapse prevention — identifying high-risk situations, rehearsing responses, and building coping skills before they are needed — has been the backbone of this field since the 1980s and holds up.3
A randomized trial comparing mindfulness-based relapse prevention, standard relapse prevention and treatment as usual found both structured approaches beat treatment as usual, with the mindfulness-based version showing fewer drug-use days at twelve months.6 The mechanism is straightforward: it trains people to notice a craving and stay put rather than obey it. That particular skill has a name and a technique — urge surfing — and it is worth learning properly rather than in summary.
Contingency management
Small, immediate, tangible rewards for verified abstinence. It has the largest effect sizes of any psychosocial intervention for stimulant use disorder, and it is chronically under-used because it sounds too simple to be serious. It works because it puts a reward on the same short timescale the substance operates on.
Continuing care, and the length of it
Outcomes improve with the duration of contact more reliably than with its intensity.7 Long, light follow-up — regular check-ins over a year or more — tends to outperform a short burst of intensive aftercare that ends. The most common failure mode in recovery is not a dramatic collapse; it is a gradual disconnection from everything that was holding the structure up.
This is also the strongest practical argument for structured housing after treatment. It is not that the house does something clinical. It is that it keeps the connection going through the months when motivation dips and nothing dramatic is happening. What sober living actually is covers how that works.
Mutual-help groups
Twelve-step and other peer groups have better evidence than skeptics assume, particularly for sustained abstinence, and they are free and available everywhere, which no clinical intervention can claim. They are not for everyone and they are not the only option — but the reflex to dismiss them is not supported by the data. How the 12 steps work is the longer version.
Treating what is underneath
If a psychiatric condition is driving the use, relapse prevention that ignores it is building on sand. Untreated depression, anxiety, PTSD and ADHD are among the most reliable predictors of return to use, and the pattern is usually recognizable in hindsight: the substance was doing a job. Take it away without replacing the function and the pressure does not go anywhere.
This is the case for treating both conditions at once rather than in sequence, which is what dual diagnosis treatment means in practice.
What a real plan contains
Most written relapse prevention plans are too abstract to be used in the moment. A usable one is specific enough that a person in trouble at 11pm can follow it without deciding anything.
- Named high-risk situations. Not “stress” — the specific bar, the specific person, the specific Sunday.
- Named early warning signs, in the person’s own words, ideally identified with someone who has watched them before. Sleep, irritability, and withdrawing from contact are the three that show up most often.
- A written response for each, decided in advance.
- Three people, by name, with phone numbers, and a prior agreement with each of them that they can be called.
- What happens after a lapse — written down before there is one. This is the part most plans omit, and it is the part that decides how much damage a slip does.
- A regular review, because the plan that fits month one does not fit month nine.
Boundaries belong in this too, particularly around family and old friendships. Setting boundaries in recovery is a subject in its own right and one of the more common places a plan quietly fails.
The most important twenty minutes: after a lapse
Research on this is unusually clear. What predicts whether one drink becomes a return to daily use is not the drink. It is what the person concludes about themselves in the hour afterwards.
Someone who thinks “that was a lapse, here is what led to it, here is who I am calling” frequently does not go further. Someone who thinks “I have blown it, I am the same as I always was, it does not matter now” very often does. The second reaction is so common it has a clinical name — the abstinence violation effect3 — and knowing the name helps, because it lets a person recognize the thought as a known phenomenon rather than a verdict.
The practical instruction is unromantic: tell someone within twenty-four hours. Shame runs on secrecy, and the gap between the lapse and the disclosure is where a slip turns into a slide. If a lapse has already become more than that, recovering from a relapse is the more relevant read.
What is oversold
- Avoiding triggers as a whole strategy. Useful early, unsustainable long-term. You cannot remove every cue from a life. At some point the work has to shift from avoidance to tolerance.
- Willpower. Not because people lack it, but because it is a finite resource that fails predictably under fatigue, hunger, conflict and stress — exactly the conditions relapse selects for. Plans that rely on it are relying on the thing that stops working first.
- Keeping busy. Genuinely helpful. Not a substitute for anything above.
- One-off intensive treatment with no follow-through. The evidence for duration is one of the most consistent findings in the field, and the standard failure is stopping too early rather than choosing wrong.
For a fuller list of the day-to-day practices that support all of this, relapse prevention activities covers the practical end.
The short version
Relapse is common, predictable and staged. It starts with maintenance quietly stopping, moves through an internal argument, and only then reaches a substance. The interventions with the best evidence are medication where it applies, structured relapse prevention therapy, contingency management for stimulants, and continuing care that lasts longer than it feels necessary. The plan is only useful if it is specific, and the most valuable line in it is the one describing what to do after a slip — written before there is one.
If you’re struggling to stay sober, or you have already slipped and don’t know what comes next, you don’t have to work it out alone. Choice House offers residential treatment and transitional sober living for men in Boulder, Colorado, with care for co-occurring mental health conditions alongside substance use. Call 720-577-4422 or reach out through our website to talk it through.
References
- National Institute on Drug Abuse — Treatment and Recovery. Relapse rates of 40–60% and the chronic-illness comparison.
- Melemis SM. Relapse Prevention and the Five Rules of Recovery. Yale J Biol Med, 2015 — the three stages of relapse.
- Marlatt GA, Gordon JR. Relapse Prevention (1985); Witkiewitz K, Marlatt GA (2004) — the abstinence violation effect and the dynamic model.
- Jonas DE et al. Pharmacotherapy for Adults With Alcohol Use Disorders in Outpatient Settings. JAMA, 2014 — naltrexone and acamprosate NNTs.
- Sordo L et al. Mortality risk during and after opioid substitution treatment. BMJ, 2017.
- Bowen S et al. Relative Efficacy of Mindfulness-Based Relapse Prevention… JAMA Psychiatry, 2014.
- McKay JR. Continuing-care research program — duration versus intensity.