Assessment is an Intervention (We Just Seem to Have Forgotten That)
I don't know exactly when it happened, but somewhere over the years I've found myself wondering whether assessment has quietly become something different to what it was ever intended to be.
Working in a CMHT and as an Approved Mental Health Professional, I've carried out hundreds of assessments. Some have lasted twenty minutes, some several hours, and some have carried consequences that will stay with me for a long time. Assessment is an essential part of mental health practice. We have legal duties, ethical responsibilities and, quite rightly, we're expected to make difficult decisions that are evidence-based and defensible.
None of what follows is an argument against thorough assessment. Quite the opposite.
What I've started questioning is whether we've unintentionally confused assessment with eligibility.
Many of the tools and templates we use weren't really designed to help us understand people. They were designed to help organisations decide whether somebody meets criteria for a particular service, intervention or legal framework. That's a perfectly legitimate purpose, but it's a different one. Somewhere along the way those two things seem to have become interchangeable.
The result is that assessments can start to feel transactional. People arrive asking for help and, before we've really established any sort of relationship, we begin working through a fairly predictable sequence of questions. Suicidal thoughts, intent, planning, access to means, psychotic symptoms, substance use, previous admissions, safeguarding concerns... none of those questions are inappropriate. Most of them are absolutely necessary. I've asked every one of them countless times and I'd be worried if they weren't being asked.
What does strike me, though, is that they're all questions about the problem.
We often describe ourselves as recovery-orientated services, yet the first conversation somebody has with us is frequently an hour-long exploration of everything that's gone wrong. If you were designing a process from scratch to help somebody feel defined by their illness, you could probably do worse.
I don't think that's because practitioners have become less compassionate. If anything, most people I work with go to extraordinary lengths to maintain humanity in systems that don't always make it easy. We've all had those moments where you're having a genuinely useful conversation only to remember the electronic record has been sitting there patiently waiting for you to complete the twenty-seven mandatory fields you've managed to ignore. The computer, understandably, isn't interested in rapport. It wants to know whether the risk formulation has been updated.
It was a recent visit that really made me stop and think about this.
I was asked to review a woman in her late fifties with a long history of paranoid schizophrenia. Following a medication change her mental state had deteriorated. She believed there was an imminent threat to her life and was finding meaning in almost everything around her. Television presenters were communicating directly with her. Song lyrics had become coded messages. She had developed a fixed belief that Ant and Dec were planning a Day of the Dead street party outside her house as part of a wider conspiracy against her. Alongside this she was experiencing increasingly distressing command hallucinations telling her to end her own life.
The team had responded exactly as they should have done. She'd been stepped up to our highest level of community monitoring, medication was under review, colleagues had updated crisis plans and risk documentation had expanded at roughly the same rate as it always does when somebody becomes acutely unwell. Bureaucracy, to give it credit, can move with remarkable speed once enough red boxes appear on a screen.
I completed what I would describe as a good assessment. We explored suicidal intent, planning, access to means, previous attempts, protective factors, mental state, insight and the impact of the psychotic symptoms. We discussed the recent medication changes and what had changed over the preceding weeks. Everything that needed covering was covered.
I'd learnt an enormous amount about how unwell she was and surprisingly little about who she actually was.
Towards the end of the visit her cat wandered into the room and immediately appointed itself supervisor of proceedings. I made a comment about it and asked how long she'd had him.
For the next ten or fifteen minutes we talked almost entirely about the cat. She laughed while describing his habit of demanding food despite a full bowl sitting a metre away. Apparently opening cupboard doors was well within his capabilities, but the cat flap remained an unsolved engineering problem. It wasn't a therapeutic technique. I wasn't employing advanced feline-assisted social work. We were simply having an ordinary conversation.
It was also probably the most useful part of the assessment.
The psychosis hadn't gone anywhere. Neither had the suicidal thoughts. We weren't pretending those things didn't exist. What had changed was that the conversation had widened enough for something else to exist alongside them. We started talking about her routines, what mattered to her, who she trusted, what she worried about losing and what she wanted life to look like once this relapse settled. From there, a genuinely collaborative safety plan started to emerge. Not because I'd followed the template particularly well, but because we'd reached the point where we were planning with a person rather than managing a collection of risks.
She remained at home with enhanced support from the CMHT and a medication review. I'm not suggesting the cat conversation was responsible for that outcome. It wasn't. Good psychiatric care, close community follow-up and experienced clinical decision-making all played their part.
But I do think that conversation allowed me to understand something the assessment template hadn't.
It reminded me that she wasn't simply a woman with persecutory delusions, command hallucinations and a significant risk of suicide. She was also somebody who loved a slightly ridiculous cat and worried about who would feed him if she ended up in hospital. Oddly enough, that turned out to be far more useful when we started talking about staying safe than asking the same question about suicidal intent for the third different way.
The more I've reflected on it, the more I think social work has been nudging us towards this understanding for years. Relationship-based practice has long argued that understanding develops through human connection rather than procedure. Steve de Shazer's work on Solution Focused Brief Therapy deliberately shifts attention towards strengths, exceptions and what people want their lives to look like, rather than assuming the problem deserves all of the airtime. Jaakko Seikkula's Open Dialogue approach similarly treats conversation itself as the intervention rather than merely a way of collecting information. More recently, Professor Stefan Priebe's work developing DIALOG+ has shown that relatively small changes to the structure of routine clinical conversations can improve outcomes without creating expensive new services or asking clinicians to find another hour in diaries that are already full.
None of those approaches suggest we stop asking difficult questions. They don't ask us to ignore suicide risk or minimise psychosis. They simply remind us that understanding the problem and understanding the person are not the same thing.
Perhaps that's where social work has something distinctive to offer within community mental health. We naturally tend to think about people's relationships, identity, family, work, housing and the ordinary things that give life meaning. Those things aren't separate from assessment; they're often the context that allows the assessment to make sense in the first place.
I still think we need robust assessments. We need careful formulations and good documentation. Those things protect people, protect practitioners and support sound decision-making.
I just wonder whether we've become so focused on completing assessments that we've occasionally forgotten they are also opportunities to intervene. Every assessment is the beginning of a therapeutic relationship, whether we intend it to be or not. Every question we ask tells somebody something about what we think is important.
Maybe assessment doesn't need redesigning.
Maybe we simply need to spend a little less time asking people to convince us how ill they are and a little more time finding out who they are while we're doing it.
If someone designing electronic patient records happens to read this, I have one modest suggestion. I'd happily sacrifice a couple of mandatory drop-down boxes in exchange for a single question:
"Tell me about your cat."
I have a suspicion that question would tell us rather more about the person sitting in front of us than some of the boxes we're currently required to complete. And, occasionally, it might even tell us something that genuinely changes the course of the assessment.