Why great recovery programmes still produce broken outcomes

There's a moment in this work that never gets easier.

Someone completes a programme.

They've done the work. Showed up. Engaged in the hard conversation. Made breakthroughs they didn’t think were possible. You can feel the shift in them. The steadier eye contact, the calmer presence, the first flicker of “maybe I can do this.”

The intervention has been delivered well. With care. With skill. With fidelity.

The outcome measures look promising.

And then, a few months later, they're back where they started.

Not always with a dramatic relapse. Sometimes it’s quieter than that. Missed appointments. A cancelled script. A return to the same relationship that breaks them down. A housing move that drops them back into the heart of the old environment. Debt. Shame. Loneliness. The slow drift back to survival mode.

It’s tempting to call that “non-compliance.” Or “not ready.” Or to assume the programme didn’t work.

But often, the programme wasn’t the problem.

The problem was that the programme was just… all they had.

We've spent decades and millions of pounds perfecting our interventions. We've refined CBT protocols, strengthened motivational enhancement approaches, built twelve-step facilitation models that genuinely help people find community and purpose.

These aren't small achievements. When delivered well they can be genuinely transformative.

But transformation inside a programme doesn’t automatically translate into stability outside it.

That’s why I’m writing this week’s newsletter.

Because the real question isn’t “do programmes work?”

It’s: what have we built around people once the programme ends and what would it take to stop the predictable drop-off we’ve normalised as inevitable?

If you’ve ever watched someone do brilliantly in treatment and then quietly unravel afterwards, the article is for you.

The lesson we keep learning

There is something we've learned, sometimes the hard way, without really doing much about it: the quality of the intervention matters far less than the quality of the system that holds it afterwards.

Project MATCH taught us this almost 20 years ago, though we didn't fully hear it at the time. Delivered between 1989 and 1997, this was the largest addiction treatment trial ever. Eight years, twenty-seven million dollars, multiple sites across the United States, all designed to answer what seemed like a sensible question: which types of people respond best to which types of treatment? The researchers compared cognitive behavioural therapy, motivational enhancement, and twelve-step facilitation. They matched carefully. They measured meticulously.

And what they found surprised everyone: matching people to specific programmes made almost no difference. Not because the programmes weren't good, they were. The overall impact of treatment was positive and the improvements for most people were impressive. And the treatment manuals developed during the programmes are still used today in our field. But they key hypothesis of the programmes failed gloriously. The programme type wasn't the determining factor. Only 33 out of 952 people going through the programme were somehow matched appropriately in a way that matter for the result of their treatment. Something larger was at play though, something the study design couldn't quite capture: the systems context in which those programmes existed… or didn't.

We've seen similar patterns emerge in our own work across the UK. Brilliant interventions that thrive in one city and wither in another. Not because the intervention model changed, but because the ecosystem did. Because in one place there were peer networks that extended beyond clinic walls, housing support that understood recovery timelines, GPs who saw the person rather than the diagnosis. And in the other place, there was just the programme. Excellent, evidence-based, but ultimately insufficient.

What systems actually do

A recovery-oriented system of care isn't just a fancy term for "joined-up services," though coordination matters enormously. It's something more fundamental: it's the recognition that recovery happens in the context of a life, not a treatment episode.

There is so much research over the last years highlighting that recovery-oriented systems of care have proved to have a transformative impact for individuals, communities, and society at large, which is why investing in such models is vital.

Think about what this means in practice. Recovery unfolds over years, not weeks. It happens in housing situations, in job centres, in family kitchens, in moments of crisis at three in the morning when no service is open. It requires different things at different times - clinical support one month, peer connection the next, practical help with benefits or childcare arrangements the month after that. It needs people who can see you whole, not just your diagnosis or your treatment pathway.

When we build systems instead of just programmes, we create something that can respond to this reality. We create:

Continuity across transitions, because recovery doesn't pause when someone moves from residential to community care, or from one borough to another, or from being "in treatment" to "in recovery."

Coordination across domains, because housing instability undermines the best clinical work, and employment discrimination can undo months of progress, and family estrangement creates vulnerabilities that no amount of therapy alone can address.

Capacity to learn and adapt, because what works shifts over time, for individuals and for communities, and systems that can't evaluate their own impact or respond to changing needs become rigid containers rather than living supports.

Enhanced social inclusion, because recovery-oriented care emphasizes the importance of social connections and community integration, recognizing that these factors are crucial for long-term well-being.

Multiple entry points and pathways, because people don't experience problems in the order we've designed services to address them, and sometimes the door marked "peer support" is the only one someone can walk through today.

The core principles

If we're serious about moving from programmes to systems, we need to be clear about what we're committing to. Based on two decades of designing, delivering, and evaluating recovery-oriented approaches, I've seen certain principles prove themselves essential in building recovery-oriented systems:

The person directs their own care. Not in a tokenistic "tell us your goals" way, but genuinely. This means shared decision-making, care plans that reflect someone's actual life, being honest when we can't offer what someone needs, rather than trying to fit them into what we happen to provide.

Hope is a clinical intervention. Recovery-oriented systems are built on the evidence that belief in the possibility of recovery changes outcomes. This isn't positive thinking or minimising struggle. It's the deliberate creation of environments where people encounter others who have travelled similar roads and found their way through.

Peer voices aren't add-ons; they're essential. People with lived experience must be embedded throughout the system, in governance, in service design, in delivery, in evaluation. Not as service users who graduate to helper roles, but as equal stakeholders with expertise that complements professional knowledge.

We measure what matters to people's lives. Not just abstinence or symptom reduction, though these may matter too. But also: Are you connected to people who care about you? Do you have somewhere stable to live? Are you doing something meaningful with your days? Can you imagine a future you want to move toward?

Flexibility is a feature, not a bug. Systems must be able to respond when someone's needs change, when a family crisis erupts, when a different approach becomes necessary. Rigid pathways, however well-designed, will always fail some people at some times.

The economics of getting this right

There's a persistent myth that systems approaches are too expensive, too complex, too idealistic for our constrained resources. The evidence tells a different story.

Numerous studies have demonstrated the positive impact of recovery-oriented care on individual outcomes. Research conducted by the Substance Abuse and Mental Health Services Administration in the USA revealed that individuals engaged in recovery-oriented services reported higher levels of self-esteem, improved quality of life, and greater satisfaction with their overall care experience.

Beyond the individual level, implementing a recovery-oriented system of care has far-reaching economic advantages. The Mental Health Commission of Canada found that for every dollar invested in services following a recovery-oriented model, there is a return of up seven dollars in improved health and productivity. These are returns measured in improved health markers, increased workforce participation, reduced demand on emergency services and so on.

Here in the UK, we've seen similar patterns. Recovery-oriented approaches reduce hospitalisation rates, shorten crisis episodes, decrease revolving-door patterns of reengagement. They don't achieve this through magic, but through the simple logic of meeting people where they are and providing what they actually need, when they need it.

Building the future we need

None of this is abstract theory. Across the UK, there are pockets of practice where recovery-oriented systems are already working. Places where someone leaving residential treatment is met by a peer who helps them navigate the benefits system. Where GPs and addiction services and housing teams actually talk to each other and share care plans. Where commissioners fund outcomes that matter to people's lives, not just clinical measures.

These aren't perfect systems. No system ever is. But they're learning systems. They create feedback loops. They adapt. They keep the person at the centre, not the programme.

The work ahead of us isn't to abandon the interventions we've developed. CBT is the psychotherapy approach most effective in some many areas. Motivational interviewing is key in starting people’s recovery journey. The right medication is fundamental is some cases. Mutual aid is the sole recovery world for so many people. The work is to stop pretending that delivering these well, in isolation, is enough. The work is to build the systems that allow good programmes to actually do what they're designed to do - support people in reclaiming and rebuilding their lives free from addiction.

This is possible. The evidence is clear. The principles are known.

The shift now is practical: from programmes to systems, from discharge to continuity, from “good treatment” to lives that actually hold.

If you’re commissioning, leading, or redesigning services and you’re ready to build that scaffolding properly, I’d love to stay in the conversation, because this is the work that  actually changes lives.

The people we serve deserve nothing less.