We provide intervention services, so this piece argues against something we could easily sell you. It is worth reading for that reason.
The intervention most families picture — everyone assembled, the surprise, the letters read aloud, the ultimatum — has a name. It is broadly the Johnson Institute model, and it is the version that television made famous.
When researchers compared it directly against other approaches, it was not the one that performed best.
The study families should know about
Miller, Meyers and Tonigan (1999) took a straightforward question: if somebody is drinking harmfully and will not seek help, what should the people around them actually do?
They compared three approaches, all delivered through a concerned family member rather than through the drinker:
Al-Anon facilitation, which supports the family member’s own wellbeing and coping, and does not aim primarily to get the drinker into treatment.
The Johnson Institute intervention — preparing the family for a confrontational meeting of the kind most people picture.
CRAFT — Community Reinforcement and Family Training, which coaches the family member in how to change their own responses: how to communicate differently, how to stop inadvertently protecting the person from consequences, how to make sobriety more rewarding, and how to recognise and use the moments when someone is briefly open to help.
One hundred and thirty family members were randomised across the three approaches, in a trial funded by the National Institute on Alcohol Abuse and Alcoholism. The results were not close:
| Approach | Drinkers who entered treatment |
|---|---|
| CRAFT | 64% |
| Johnson Institute intervention | about 3 in 10 |
| Al-Anon based support | barely more than 1 in 10 |
CRAFT engaged roughly two to six times as many people as the comparisons.
The Johnson approach had a further problem that the headline number hides: a large share of families prepared for it never went through with the meeting at all. It is frightening, and many people cannot bring themselves to do it — so its real-world performance is reduced not only by refusals on the day but by interventions that never happen.
Meyers and colleagues (2002) ran the same comparison for drug use rather than alcohol, with twelve hours of individual sessions in each condition. CRAFT engaged 67%, more than double the Nar-Anon and Al-Anon based comparison. Roozen and colleagues (2010) pooled the CRAFT studies in a meta-analysis and found the advantage held across them.
There is a second finding in that first trial which deserves as much attention as the engagement rates. The family members themselves were measured too — on depression, anger, family cohesion, family conflict and relationship happiness. At three and six months, those who had learned CRAFT were less depressed, less angry and reported more cohesive, less conflicted households. Those gains held whether or not their loved one ever entered treatment.
That matters because it changes what a family is deciding. CRAFT is not a bet that only pays out if the person agrees to go.
Why the confrontational version underperforms
It makes intuitive sense that it should work. Overwhelming evidence, delivered by everyone who matters, at once. How could someone argue with that?
Several reasons, and they are worth understanding because they apply to ordinary family conversations too.
It puts the person in a defensive position, and defensiveness produces resistance rather than reflection. People do not generally change their mind while being outnumbered.
It is a single event. One meeting, one answer. If the answer is no, the family has spent all its leverage in one afternoon and often has no plan for the following morning.
It frightens the family out of doing it. An approach that works well when performed but is too daunting to perform has a much lower expected value than its success rate suggests.
It does nothing about the environment. After the meeting, everyone returns to exactly the same household, where the same responses are still in place.
What CRAFT does instead
The premise is different. Rather than trying to force a decision in a single conversation, it changes what happens around the person continuously.
In practice that means the family member learns to stop shielding the person from consequences that would otherwise land naturally — while explicitly not punishing, which is a distinction most people get wrong. They learn a different way of raising the subject, one that does not reliably trigger a row. They learn to make the sober version of life more rewarding and more present. They learn to spot the brief windows — after an incident, during a bad morning, in a quiet moment — when someone is genuinely reachable, and to have a concrete option ready for that moment rather than a general appeal.
And they get support themselves, immediately, whether or not the person ever accepts help. That matters, because the family member is usually in worse shape than anyone has acknowledged.
There is no single high-stakes day. There is no ambush. There is no all-or-nothing answer. It works through accumulation.
What this does not mean
It does not mean professional intervention is useless. A well-run intervention is not the televised version — it is planned over weeks, it incorporates much of the same thinking about communication and consequences, it is small and calm rather than confrontational, and it includes a plan for what changes regardless of the answer. That is a long way from the model that was tested and found wanting.
It does not mean a family meeting is never right. For some situations — particularly where risk is high and time is short — a direct conversation with everyone present is the correct call.
And it does not mean CRAFT always works. It does not. It improves the odds substantially compared to the alternatives, which is a different claim.
What to do with this
If you are choosing between “stage an intervention” and “do nothing”, the research suggests there is a third option that outperforms both, and it starts with the family rather than with the person you are worried about.
If you are being sold a confrontational intervention by someone who has not mentioned that alternative, that is worth asking about directly.
Working out which approach fits your situation — and they are not mutually exclusive — is the conversation our intervention service starts with. Often the answer is a period of CRAFT-style work first, with a planned family conversation held in reserve rather than led with.
References
- Miller WR, Meyers RJ, Tonigan JS (1999). Engaging the unmotivated in treatment for alcohol problems: a comparison of three strategies for intervention through family members. Journal of Consulting and Clinical Psychology, 67, 688-697. 10.1037/0022-006x.67.5.688
- Meyers RJ, Miller WR, Smith JE, Tonigan JS (2002). A randomized trial of two methods for engaging treatment-refusing drug users through concerned significant others. Journal of Consulting and Clinical Psychology, 70, 1182-1185. 10.1037/0022-006x.70.5.1182
- Roozen HG, de Waart R, van der Kroft P (2010). Community reinforcement and family training: an effective option to engage treatment-resistant substance-abusing individuals in treatment. Addiction, 105, 1729-1738. 10.1111/j.1360-0443.2010.03016.x
- Meyers RJ, Wolfe BL (2004). Get Your Loved One Sober: Alternatives to Nagging, Pleading, and Threatening. Hazelden. ISBN 978-1-59285-081-5. — the trials’ co-author describing his own results, and the source of the figures quoted above.
Read more about Intervention.
