I was recently invited to take part in a roundtable exploring some pretty fundamental questions about our mental health system. What makes care humane? What gets in the way of it? And what would need to change for people experiencing profound distress to receive care that is not only safe, but actually helpful?
I keep coming back to fear. Mental health services are frightened places. We are frightened of suicide. Frightened of dependency. Frightened of complaints, coroners, regulators, litigation and scrutiny.
When systems are frightened, they become defensive. And fear fundamentally changes the way we relate to people. It can turn care into surveillance. It can make ordinary human closeness feel professionally dangerous. It can make practitioners more concerned with demonstrating that they followed a process than asking whether that process actually helped the person in front of them.
Defensive practice and defensible practice
There is an important distinction between these practices. Defensible practice does not mean being reckless with risk. It means knowing the person you are working with. Understanding their patterns of distress and risk. Making thoughtful decisions that can be explained, supported by evidence and reviewed when circumstances change.
Sometimes that means taking therapeutic risks because autonomy, power and choice matter too.
Defensive practice is different. Defensive practice asks: What do I need to do so that nobody can blame me if something goes wrong.
Staff cannot contain what their organisations cannot contain
We talk a lot about the capacity of mental health staff to contain distress. We ask them to sit alongside unbearable pain, uncertainty, suicidality, anger, hopelessness, dependency and risk without immediately acting to make those feelings disappear. But who is containing them? Working relationally requires more than telling individual practitioners to be compassionate. It requires organisations to create the conditions in which compassion is actually possible.
Staff need time and space to think about the people they work with, not simply process them through increasingly pressured systems. They need reflective spaces where they can talk about the emotional experience of the work itself. They need access to experienced colleagues and senior leaders when something feels frightening or overwhelming.
And those spaces have to allow people to say the things that are difficult to say.
I am frightened by this person’s risk.
I am really angry about what happened yesterday.
I don’t like working with this person at the moment.
I don’t know what to do.
These feelings do not disappear because we decide they are unprofessional. When there is nowhere safe to think about them, they are enacted instead.
A practitioner who cannot acknowledge that they are frightened may become increasingly restrictive. Someone who cannot acknowledge that they are angry may become distant. Teams who cannot say that they feel overwhelmed by somebody may begin excluding them, avoiding them or passing responsibility elsewhere.
The feeling remains. It just moves into the relationship. And often the patient is acutely aware. They know when the nurse who previously spent time with them suddenly avoids them. They know when staff no longer volunteer to be on their observations. They know when the tone in the room has changed.
Then something particularly damaging can happen. The person asks, Are we okay? Have I done something? Why are you treating me differently? And because acknowledging the rupture feels professionally unacceptable, they are told:
Nothing is different.
I’m just busy.
It’s not about you.
Their perception of the relationship is denied at precisely the point at which they are trying to understand it. If rupture cannot be acknowledged, it cannot be repaired.
I have seen the consequences of this in inpatient care. I have worked with someone who spent months in segregation on high level observations. Regular staff had become exhausted and distressed and increasingly felt unable to work with them. Their observations were consequently often undertaken by agency staff they did not know.
The patient knew something had changed. They repeatedly asked why regular staff would not work with them. Why people did not want to talk to them. Why it felt as though people hated them. They were repeatedly reassured that it wasn’t about them.
Elsewhere, staff were openly talking about how burnt out they were, how upsetting they had found incidents and how difficult they were finding working with this patient. This protected staff from an incredibly difficult conversation. It did not protect the patient. It also denied them the opportunity to learn.
If we never sensitively tell people about the impact they have on relationships, how can they do anything differently? They experienced the consequence — people withdrawing from them — without being given any meaningful way of understanding what had happened or how the relationship might be repaired.
That isn’t relational care. It is punishment without explanation.
Containment is also about where responsibility sits
Organisational containment is not just reflective practice or supervision. It is also about whether frontline clinicians genuinely believe their organisation will stand beside them when they make difficult decisions.
I have seen the difference this can make. I have worked with several people who had spent many years in hospital in extremely complex and high-risk situations. There were understandably very different views about how care and risk should be managed. Over time, senior leaders became directly involved in the work. They attended meetings, understood what we were proposing and engaged with the complexity rather than expecting the frontline team to carry it alone.
At one meeting, after a plan had been agreed, a deputy director made their position explicit: they had heard the plan, understood it and supported it. The plan also received legal scrutiny and reassurance that the approach was defensible and consistent with relevant guidance.
That mattered enormously because it redistributed the anxiety.
The system was saying: We understand the risk. We have thought about this together. This is a defensible plan. You are not carrying it alone. When people feel supported, they can make better decisions. Too often systems encourage clinicians to defend themselves, not their patients.
Relational care requires relationships that are real
There is another uncomfortable part of this conversation.
Somewhere within clinical systems we seem to have created an idea that a therapeutic relationship is one in which the professional is not affected by anything.
I don’t think that is humanly possible. People affect each other. There have been times in my own work when I have needed to have incredibly difficult conversations with people I care deeply about. That can be done without blaming someone for being distressed.
Sometimes relational work sounds like:
I care about you. What happened was really difficult for me. I wanted to be able to help you before things got to that point, and I couldn’t. We need to think together about what happened and what we might do differently next time.
And it has to work both ways. The people we work with need to be able to tell us when we have messed up. Because sometimes we do.
The answer cannot be to hide behind professional authority and insist that because our intention was therapeutic, the impact must have been therapeutic too.
We acknowledge it. We hold our hands up. We apologise. We try to understand what happened. We try to make it right.
That is repair. The basis of relationships everywhere else in human life.
Yet something happens within clinical systems where these ordinary cornerstones of relationships suddenly seem too dangerous.
Instead, people can be offered an artificial version of a relationship. Warm enough to be called therapeutic, but distant enough that nobody is supposed to genuinely affect anybody else.
What exactly are people supposed to learn about relationships from that?
Compassion is emotional labour. And emotional labour comes at a cost.
We cannot endlessly ask individual staff to absorb distress created by profoundly distressed people, frightened families and overstretched systems without somewhere for that emotional load to go.
If organisations want relational care, they have to build relational organisations. That means psychologically containing staff rather than simply training them to contain patients. It means senior leaders being available when risk becomes frightening. It means genuine reflective practice and good clinical supervision. It means giving teams permission to talk honestly about the emotional impact of their work without being shamed for having ordinary human responses.
It means recognising that risk belongs to systems, not individual clinicians. And it means creating cultures where uncertainty can be tolerated rather than immediately converted into restriction. None of this means abandoning boundaries, accepting abuse or pretending risk does not exist.
Quite the opposite. Good relational practice requires boundaries. But boundaries are not the same as emotional withdrawal. Safety is not synonymous with restriction. Professionalism does not require pretending that we are unaffected by one another. And containment does not mean making distress disappear.
Being human should not be a specialist intervention
It has to be system wide, top down and be role modelled in how organisations treat their own staff.
Staff need systems that make compassion possible. Patients need staff who are allowed to be human.
And all of us need services capable of tolerating the uncomfortable reality that meaningful care involves uncertainty, emotional impact and relationships in which things sometimes go wrong and need to be repaired.
Funding matters enormously. But a better-funded inhuman system is still an inhuman system.
If we genuinely want humane mental health care, the cultural change has to be every bit as ambitious as the financial investment.
Hollie Berrigan is lived experience lead at Beam Consultancy
