Imagine you’re in hospital and you’ve been telling the ward counsellor that you are scared that the excruciating pain in your head may be a progression of your cancer. You feel the counsellor listens carefully and seems to understand, and you feel a bit calmer by the time she leaves. Later the consultant comes round and confirms your worst fears: the cancer has spread to your brain. How, then, does it affect your feelings towards the counsellor when a chance remark makes it clear that she already knew the results when you were both speaking earlier? 

This may appear to be extreme, but it can happen. Most ward-based referrals come from attending ward hand-over meetings. As a result the counsellor becomes privy to personal information without the person choosing to share it, and they themselves may not even be aware of some of it. This raises all sorts of dilemmas, ranging from how to respond when the client’s story differs from other versions you have heard, to anticipating the delivery of bad news. Perhaps you decide to ignore the additional information and meet the client with the material they bring. Or maybe subtly you choose to encourage reflection on the possibility of bad news. Perhaps you delay your visit until the bad news has been given. Whatever choices you make within the therapeutic context, it will affect the relationship. 

As an oncology counsellor in a large teaching hospital, I work alone in an office close to the main oncology ward. I provide short-term counselling for outpatients, as well as providing emotional support to patients on the wards. It was a situation similar to the one above that led me to reflect on the strange environment in which I work and the many ways in which boundaries and ethics can become live issues. These issues are not unique to the hospital setting; nor is it possible in this article to provide definitive answers to the many dilemmas that can arise. What I hope to achieve is to increase awareness around some of them, so that all of us who find ourselves in similar situations can make conscious ethical choices. There are many lone counsellors working in isolation in hospital units, with no one but their clinical supervisor to understand their work fully, and this is directed in part to them and to anyone considering a job or placement in a hospital. The issues are also likely to have resonances for people working in other settings.

Information and confidentiality

To some extent information contamination – that is, information that has not come direct from the client – is inevitable. Some people may suggest avoiding the professional meetings and minimising additional information. However, that risks a lower professional profile on the ward and fewer referrals. Nor does it fully avoid exposure: the wards operate in a culture of information exchange within the multidisciplinary team and, while it is accepted that the counsellor will not share the content of his or her sessions, it would be a big challenge for the counsellor to avoid all exposure to information from other professionals. 

Another area of flexibility can be confidentiality, depending on where people are seen. In the counselling room the environment can be controlled: a contract lays out the ground rules; conversations are private (other than those where there is risk of harm to self or others); notes are kept separately and can only be accessed by the counsellor (or the client, with notice), and any statistical data are anonymised. In other words, it is very similar to the standard community-based counselling service, although it is expected that any referrer is notified that a service is being provided. 

In contrast, work on the ward is more fluid. If someone can walk and they are well enough, it may be possible to access a quiet room for a counselling session. However most visits are carried out at the bedside, often surrounded by curtains that give an illusion of privacy but through which conversations can be heard. There can be interruptions, which are often not controllable, by people or activities in the immediate or adjacent areas. There is no contract and the extent to which people share experiences under these conditions is left to them (albeit with the occasional reminder about the absence of privacy). Add in that people on wards are frequently asked personal questions by other staff so they can become disinhibited and subsequently feel exposed. It is important to ensure that people have real choice about whether to speak with the counsellor or not. Most people are ill, vulnerable and scared, and many are keen to be seen as a ‘good’ patient. In those situations they could comply when they would really rather not. 


Thinking back to my early days when I was training to become a counsellor, I remember certain ‘givens’: always start and end sessions on time; expect regular attendance, and explore the meaning of absences with the client as part of their process; be circumspect about the use of touch; keep things consistent – eg time and day of meeting, and where the meeting takes place. In fact I have always aimed to maintain the first of these, keeping the allocated sessions to time. Anyone who works or has been a patient in a hospital will know just how different this is to the prevailing culture. To see how the rest of the given rules can be affected, consider the following case.

D was a 46-year-old black British woman who came for one-to-one counselling midway through receiving chemotherapy for the second time, following a relapse of her bowel cancer. She was understandably frightened, both of the potential side effects of the new treatment and of the implications of relapse. Sessions were arranged to avoid her worst symptom days within the three-weekly cycle of treatment. We did not always meet on the same day each week, although we managed to keep to the same time of day. When she rang to cancel her second session due to ill health, I accepted it at face value, and we confirmed the next appointment. This is not unusual for people receiving chemotherapy. Over the following three weeks she turned up late twice; the reasons could all have been legitimate but I took the decision to explore them. What emerged was her fear of facing the ‘inevitable’ discussion about dying. I reassured her that we would only move at her pace and discuss what she brought. On the strength of that she felt comfortable to extend her sessions. Some time later she announced, ‘I’m ready to talk about dying now,’ and we explored this over several weeks. Towards the end she realised that she had probably explored as far as she could, mainly because she could now see that ‘there’s still life to be lived’ and that at least one more form of treatment was available to her should she relapse again.

That may have been the end of the story but some years later I noticed that D was an inpatient on the ward. This time, I heard, she was dying. I was unsure whether she would want to see me on the ward, since no one would have known she had seen me previously. I asked one of the senior nurses to let her know that a counsellor was on the ward (giving my name), and to ask her if she would like to talk. She was very happy to see me and had no qualms about the confidentiality of her previous contact with me, but said she tired easily. She was candid about her current predicament, but seemed accepting of it until she started having a coughing fit and struggled to catch her breath. Having alerted the nursing staff I stayed with her. She was clearly scared so tentatively I put a supporting arm around her shoulders, unsure if physical contact was appropriate, although it felt the right thing to do. At this point she leant into me and both of us remained there while the staff gave her the medication she needed and the fit passed. She died peacefully a few days later. 

Talking this over in clinical supervision, what made most sense was that, at that point of placing an arm around her shoulders, I stopped being her counsellor and became one human being caring about another.

The experience with D illustrates not only how boundaries can be affected but the importance of checking things out. The changes in days could have caused confusion, but they did not. The lateness could have been genuine, but by looking at it together we uncovered some very potent material and prepared the way for its exploration once she was ready. In hospital it is common to see someone for counselling and then meet them out of context. Some people can be upset by recognition, others by a lack of it. It was possible to check with D but in other contexts of course you can only do your best, and perhaps wait for them to make the first move.

Medical culture

Perhaps the greatest challenge to a lone counsellor working in a hospital is the medical setting itself. The professional paradigm around you is one of being the expert and problem solving. This can be a seductive trap for the therapist, and yet clients feed back that they really value the counsellor being outside the medical model – an independent ear who nonetheless knows the territory. Being seriously ill can leave people feeling extremely vulnerable and powerless; being accompanied by someone who encourages them to reconnect with their own agency can be transforming. In this context counselling is a space where people with serious illness can stay with whatever difficult emotional rollercoaster they find themselves on without it being seen as another problem to be solved. This is not to say that the specialist expertise of the counsellor is redundant. Many people need help to develop tools for managing anxiety and depression. Similarly there are several points at which it is recognised that people facing cancer, for example, can become particularly affected psychologically by predictable concerns.1 This is about knowing the territory. The challenge is to accompany someone through the terrain while respecting their unique experiences and abilities. It can be difficult to stay in contact with that approach in a ‘problem/fix’ culture. 

Another issue where I found myself drawn into the dominant culture for a while was access to and use of patient notes. In the hospital access to information is always on a ‘need to know’ basis, but in the case of a lone worker such as the counsellor this is left to the judgment of that professional. I have access to personal patient records in two ways: electronic data and ward files. On the wards, all members of the multidisciplinary team read through the notes of the patients they are about to meet and I followed suit. It told me their diagnosis, who else was involved in their care, occasionally some comment on their emotional state, and what treatments they might be receiving. I would also write a single line, simply to say that I had visited and whether or not those visits would continue. One day in supervision I mentioned reading information in someone’s notes and was shocked by my supervisor’s response. She felt very strongly that she would object to that, were she the patient, and challenged me to think deeply about my ‘need to know’. This expanded to a reflection on the electronic data to which I had access. The whole process had become so ‘normal’ to me by this stage that it took this shock to make me look at it afresh and question my practice. Initially I wanted to defend my actions by listing the perceived benefits. That led me on to explore how else I might gain that information, whether I needed it and what I might need to know beyond the information provided by the person themselves. I decided to experiment, and refrain from reading any notes for a period. The surprise is that it made very little difference to my actual work with clients, and by dealing only with the information that they chose to share I felt closer to their experience. This episode was a huge learning curve for me, and now I only access the hospital patient records to collect the non-identifying demographic data on which I base my quarterly service reports.

Support and exchange

People’s views on the various boundary issues outlined in this article are likely to vary according to their counselling modality, the levels of support that are available, the pressures on the service, and any number of other factors. It is important to be aware of the choices you make, though, as your decisions are bound to impact on the client. For isolated practitioners, regular clinical supervision and reflective practice are invaluable, provided they are used effectively. This can be augmented through networking with other practitioners in similar situations, such as local counsellor networks. These may bring together counsellors from different specialisms within one hospital, or counsellors working in one specialism across several hospitals. I have facilitated both, formally and informally. I hope this article will also spark debate and ideas more widely.


In the context of hospital-based counselling, many boundary and ethical issues can arise. This article highlights a few: information contamination; confidentiality, and the challenges to maintaining it; the need for flexibility that is based on a foundation of established good practice, and maintaining a helpful distance from the dominant culture of the medical model. There are no absolute answers to many of the issues raised here, and I would suggest resisting the urge to legislate too tightly on any of them. Conditions and workplaces vary too greatly for that to be anything other than a misguided effort to establish safety and certainty in an uncertain world. Instead I would suggest continuing dialogues that are guided by the principles of the Ethical Framework. 

As someone who came into the NHS from the outside, I had little preparation as to the kinds of challenges that this particular way of working would pose. It would have been helpful to have been aware of some of the cultural and practical issues that I was likely to face. I hope this article both offers food for thought for others who may be considering a similar career move and provides a reminder to those already working in this and similar settings of the issues that surround us daily and to which we can become blind through familiarity.

Caroline Armstrong MBACP has been an accredited counsellor since 2002. She is a trained supervisor and worked in a variety of counselling settings prior to her eight years as an oncology counsellor in a London teaching hospital. She has a special interest in mindfulness.


1. Burton M, Watson M. Counselling people with cancer. Chichester: John Wiley & Sons Ltd; 1988.