Article Text

Download PDFPDF

Haematologists’ perceptions of palliative care and specialist palliative care referral: a qualitative study
  1. Bethany Wright1 and
  2. Karen Forbes2
  1. 1St Peter's Hospice, Bristol, UK
  2. 2University of Bristol and University Hospitals Bristol NHS Trust, Bristol, UK
  1. Correspondence to Dr Bethany Wright, St Peter's Hospice, Charlton Road, Brentry, Bristol BS10 6NL, UK; bethanywright{at}


Objectives This study explored the views and perceptions of haematologists towards palliative care, based on their own clinical experiences, focusing on those factors that helped or hindered referral to specialist palliative care (SPC) services.

Method We conducted in-depth face-to-face interviews with a purposive sample of eight trainee and consultant haematologists working in tertiary referral centres in the West Midlands. All interviews were recorded, transcribed and subsequently analysed using the principles of the grounded theory approach.

Results Data analysis revealed a core category around boundaries alongside four emerging categories: perceptions of palliative care; philosophy of approach; determining roles; and responsibility and control. Positive attitudes towards SPC involvement were expressed by most participants. The role of SPC services was sometimes difficult for participants to define, with timing of referral often determined by their level of confidence in providing end-of-life care. Almost all highlighted a lack of inpatient palliative care unit provision; this was viewed negatively and impacted on referral.

Conclusions While positive attitudes towards palliative care were expressed, barriers to collaboration between haematology and SPC services were identified and provided opportunities for improved interdisciplinary working. Suggestions for improvement are focused around: improved training; enhanced mutual respect and understanding; clearer definition of the role of SPC services; and consistency and flexibility in service provision.

  • Haematological disease

Statistics from


Haematological malignancies form a diverse group of diseases collectively representing the fifth most common type of cancer in the UK. While mortality figures are comparable to those of solid malignancies, they are distinct in retaining the potential to respond to curative treatment even in advanced disease.1 Whether intent of treatment is palliative or curative, patients with haematological malignancies experience a similar symptom burden to patients with solid tumours.2 In 2003, the National Institute of Health and Care Excellence (NICE) concluded that access to and uptake of SPC services was lower in patients with haematological malignancies than in solid tumours.3

Despite a recommendation from NICE for greater integration between haemato-oncology and SPC services,1 the uptake of SPC services remains lower in patients with haematological malignancy than patients with solid tumours. Fewer patients with haematological malignancies are referred to SPC services and the interval between referral and death appears shorter in this group.4 ,5 In addition, qualitative data from interviews with bereaved carers and healthcare professionals demonstrate end-of-life care for haematology patients is often inadequate and an unrealistic and continuing focus on cure blocks referral of these patients to SPC services.6 Available evidence suggests little consensus between haematologists on when and why they should refer patients to SPC services.7

In this study, we explored the views and perceptions of haematologists towards palliative care, based on their own experiences of caring for patients dying of haematological malignancy.


Participant selection

Research governance was granted by participating National Health Service (NHS) trusts. Purposive sampling was used to identify participants from two tertiary referral centres in the West Midlands of England, one of which had a bone marrow transplant centre. For inclusion in the study, participants were required to specialise in haematology and have personal experience of managing patients with haematological malignancy. Potential participants were emailed an invitation letter and participant information leaflet. Those interested then contacted the research team to arrange an interview.

Data collection

Data were collected using face-to-face, in-depth interviews as this provided the best means of probing and gathering information about each participant's individual experience. Interviews were conducted between November 2009 and February 2010 by the principal investigator (BW), at the time a specialty registrar in palliative medicine. All interviews were digitally recorded and transcribed verbatim. An interview guide was developed and then modified throughout data collection, enabling flexibility to explore new areas of importance that emerged as interviews progressed.8

Data analysis

Given that the research subject was relatively undiscovered, a grounded theory approach was used.9 An iterative process of constant comparison-guided data collection and analysis: open coding generated initial concepts through analysis of each transcript at a micro level (phrase by phrase, word by word); axial coding sorted and resorted these open codes into groups of concepts and common themes; selective coding grouped overarching themes into defined categories and central concepts. Testing and verification (or omission) of these concepts took place in subsequent interviews.

All transcribed interviews were systematically analysed by BW. A proportion of transcripts were coded and grouped into categories independently by the local supervisor (HJ). One further researcher (KF) reviewed the final presentation of emergent themes and the corresponding verbatim quotes supporting them. KF is an experienced researcher who is familiar with grounded theory and BW undertook the research as part of a taught Masters degree in Palliative Medicine.


Participant characteristics

Eighteen potential participants were invited by email to take part in the study. Of these, eight failed to respond with no reason given for non-participation. Data saturation was reached after eight interviews. The characteristics of participants are summarised in table 1, below. The average duration of interview was 42 min (range 35–70 min).

Table 1

Characteristics of participants

The data collected in this study provided ‘insider’ accounts of the experiences of haematologists caring for patients with haematological malignancy. Four categories emerged through the process of constant comparison: philosophy of approach; perceptions of palliative care; determining roles; and responsibility and control. The relationship between these categories and their lower order codes is summarised in table 2 and described sequentially below.

Table 2

Summary of relationship between categories and codes

Philosophy of approach

Aims of care

Participants were universally committed to their choice of specialising in haematology, all describing their motivation to work at the ‘cutting edge’ of science (see box 1). A sense of shared identity, purpose and philosophy, with focus on treatment and cure, were contrasted against the more ‘abstract’ ethos of palliative care. While participants valued a holistic approach, switching between this and the ‘biomedical model’ was challenging.

Box 1 Accounts of participants relating to the theme ‘Philosophy of Approach’

Dr D: What is unique though, in haematology, is that no other specialty has the lab component in it and no other specialty has so much links with pure science, so it's kind of unique.

Dr F: And yet what attracts all the trainees now, they all want to do haem-onc. They don't want to do haemophilia, they don't want to do general haematology, they don't want to do labs. It's all haem-onc. So, there's a price that is paid for doing haem-onc and that is that you're gonna lose a lot of your patients. You've got to be comfortable managing that aspect of care.

Dr B: Palliative care is more of an art than a science, you know, with a chemotherapy protocol you give what the protocol says … I sometimes joke that haematology is intensive care and oncology is palliative care. And um, obviously that's not true but um …… (Haematology) is very intensivist … And if our mind-set is like that with the first 12 patients on the ward then, you know, it's quite difficult then to … switch.

Dr F: Now, we all like to think we're special and different. I mean, there's a couple of us that have quite a strong leaning towards that group of (palliative care) patients, whereas others are a bit more surgical in their approach. And, you know, we're all different.

Influence of training

Several participants were concerned that haematology specialist training was biased towards laboratory medicine, leaving trainees naïve to the fact that most haemato-oncology involved management of patients with terminal illness (see box 1). Lack of palliative care training within haematology specialist training was a theme highlighted by all participants. Its relative absence from the specialty curriculum was seen as an indicator that haematology, as a specialty, placed little value on palliative care.

Perceptions of palliative care

Meaning of palliative care

Participants’ perceptions of palliative care were universally favourable, with no suggestion that palliative care involvement caused harm (see box 2). However, participants reported some of their patients still equated the words ‘palliative care’ and ‘Macmillan nurse’ with worsening prognosis, death and dying. The word ‘hospice’ held particularly negative connotations and participants reported that conversations about the benefits of hospice referral required sensitive and proactive explanation.

Box 2 Accounts of participants relating to the theme ‘Perceptions of Palliative Care’

Dr B: I have worked for a team where a Macmillan nurse wasn't allowed on the ward round, because the consultants felt that that meant that the patients might think that some of them were dying (laughs) …

Dr H: I think (palliative care) is integral to everybody's care with haematological malignancy um, at all stages really. But I don't necessarily think we always think about it at the right stage. I think we still think of it as end of life.

Dr F: Um, most people have an aversion to hospices, ‘cause they think it's just a charnel house and … if you're offering hospice support in the non-terminal phase, the look of horror, you know, ‘Oh, this is it! It's the end isn't it doc!’ But you know, if you've got the time and you elaborate on why you think hospice support at home might be useful … then usually people are comfortable with it.

Dr H: And (the hospice) were, I would say, a little bit forceful in saying, ‘Well if she comes here we're not going to give her any blood products’… But it was partly their attitude about it: ‘Well with haematology patients you often want us to give blood products and that's not part of our end-of-life care. We don't do blood tests … That's why we don't have many haematology patients’.

Dr F: (When referring patients for hospice inpatient care), now we've got this stricture of ‘expected to die within the next 2 weeks’. None of us has a crystal ball … They might survive—it might be 2 days, it might be 2 weeks, it might be a month, we don't know. So I think that makes life very difficult because they're asking for a definitive statement.

Purpose of palliative care

In patients with haematological malignancy, symptom control was cited as the most significant palliative care need and the commonest reason for referral to SPC services (see box 2). However, participants described significant variation in symptom prevalence, with myeloma singled out for its high symptom burden and other diseases viewed as being relatively asymptomatic. In terms of symptom severity, several participants asserted patients with solid malignancies had greater palliative care need.

Perceptions of SPC services

Although participants reported positive experiences when engaging with palliative care, variation in provision of SPC services, both between and within regions, made services hard to navigate (see box 2). Experience of referring patients to local hospices for inpatient care was generally negative, with almost all participants reporting some difficulty. Reasons for this included inadequate bed resource, inability to provide blood product support and waiting list times.

With regards to referral for terminal care, participants reported inflexible referral criteria, in which a specific prognosis was required. Several participants suggested haematology patients were considered by hospices as ‘difficult’ and were in some way prejudiced against. This engendered a sense of negativism towards referral, as if it wasn't ‘worth trying’. One participant summed it up thus, ‘I don't think the hospice sees itself as a place for people to die anyway’. (Dr A)

Determining roles

Timing of palliative care

The relapsing and remitting nature of haematological malignancy made the ‘timing of palliative care’ a significant challenge for all participants (see box 3). For certain patients, transition from curative to palliative care may preclude referral altogether due to its rapidity or complexity. Prognosis was not a hindrance to referral in itself, since patients with myeloma were frequently referred with even ‘years of life expectancy’. Despite appreciating the benefits of ‘early referral’, participants were uncomfortable referring patients with curable disease, even when the chance of cure was relatively low. Referral at the time of diagnosis was also avoided, due to fears patients would be overwhelmed and confused by a barrage of conflicting information.

Box 3 Accounts of participants relating to the theme ‘Determining Roles’

Dr B: Um, I mean I think sometimes you can say to a patient, the treatment hasn't worked, and now we are going to give you palliative care. But my experience is that's what happens on the television, that's rarely what happens in real life. I think the thing is, for haematology, a lot of the time it's a phase that's not a single step.

Dr H: I think … there's often a lot going on when they're first diagnosed. I mean we do try and introduce (palliative care) but I'm not sure that the patients necessarily take on the concept of it. Um, you know, palliative care nurses often do come and see them when they're in but I'm not sure they necessarily always associate that with them, if that makes sense.

Dr F: If you're looking after a group of patients where two-thirds are gonna die, then if you haven't got any of those generic skills then I think you're in the wrong trade, to be honest. That's not to say … it's a land grab, that you feel you should do everything. But if you're not comfortable looking after that group of patients, and by definition that must involve a degree of palliative care, then you shouldn't be doing it.

Dr B: I have worked in places where um… basically no palliative care goes on on the ward … And that's partly a lead from the haematologists who don't feel it's something that they do.

Other factors, less to do with disease trajectory and more to do with relationships between professionals and patients, were also relevant. An individual's approach to their illness and treatment influenced acceptability of referral. For some patients, early referral was welcomed while others resisted this right up until death. The referrer's own experience in palliative care also altered the threshold for referral, depending on their scope of expertise.

Overlap in roles

One of the challenges identified by participants related to the distinction between ‘specialist’ and ‘generalist’ palliative care. All participants in this study described palliative care as integral to their role as haemato-oncologists, with SPC referral triggered by ‘inability’, rather than unwillingness to meet their patients’ needs (see box 3). A ‘joint care’ approach was viewed as the preferred model, with collaboration rather than abrogation of responsibility being the goal. However, several participants recalled working with colleagues who considered palliative care a separate entity, for which they had no responsibility or interest.

Responsibility and control


All participants described the rewards of supporting a patient through the course of their illness (see box 4). Establishing an open, honest relationship with patients was considered fundamental to their care. Participants believed that taking time to nurture this relationship paid dividends when inevitable challenges were faced later down the line.

Box 4 Accounts of participants relating to the theme ‘Responsibility and Control’

Dr C: There are some groups of patients that … both family and themselves, they want like a full intensive treatment, but on the other hand I'm not entirely sure how well they understand that by giving the full intensive treatment that they can actually become worse than the kind of state of disease.

Dr A: I think that transition from trying to cure you … to one where we are basically into symptom control and quality of life … that's probably one of the things that haematologists have been least good at.

Dr G: Making the wrong decision. You're constantly scared … So I guess it's always the fear that you've got to get everything right and do everything you possibly can but then I think once someone's got a cancer it's a lot easier and we're so control freaky in haematology.

Dr E: If somebody mentioned a patient of mine, you know I would know them by name and what's been happening etc. yeah because there's that relationship and that's one of the most rewarding things about haematology. And often by then the patients have built such a relationship with us and I think if we say, ‘Oh well we'll no longer give you any blood products up on the ward’, they do feel that we've written them off.

Dr B: I am in the middle of a case just at the moment we've gone from a patient who I thought, in all honesty, not kidding myself, 12 days ago was doing very well and five days ago it's become clear that she is not. And that was a brick wall for me, let alone the patient and her family … I think most of us are reasonably good at having robust conversations with each other and I think you need that because sometimes when you are emotionally involved with the patient you feel you're um, abandoning them or you feel … you've failed them and all of us to some extent feel like that. Rationally you know that's not true.

Dr E: In terms of a patient relapsing after a transplant … You see, a transplant evokes a high degree of expectation in the patient, they expect to be cured, so of course it's been brought up from there (points up) right down here (points down) and it's managing that as well.

Many felt a need to be ‘in control’ of all aspects of their patients’ care. Therapeutic continuity was not, however, without difficulty; it risked creating ‘over-dependence’ in patients and a sense of ‘abandonment’ when a change in focus of care was discussed. Handing over the care of long-term patients also took its emotional toll on medical and nursing staff, where bonds and attachments had developed with patients and their families over a number of years; one participant remarked that palliative care support was not the same as ‘somebody who has seen them through the thick and thin of it’. While accepting external advice and support, some participants seemed to view handing over to palliative care as professional failure.

Control over disease outcome

Participants acknowledged taking a paternalistic approach towards certain treatment decisions, believing their experience and expert knowledge left them better informed than patients to make objective decisions about risks and consequences. The weight of this responsibility hung heavily, with concerns voiced frequently about ‘letting down’, ‘abandoning’ or ‘failing’ patients (see box 4). Participants explained haematologists’ reputation for therapeutic optimism by their need to bolster patients undergoing toxic treatment. Withdrawing treatment required certainty that ‘all options had been exhausted’ and the assurance that ‘no more could be done’, since stopping treatment signified failure and ‘the point of no return’.

Participants recognised how differences in the value placed on quantity and quality of life influenced decision-making, describing situations in which the decision to discontinue treatment had been clear cut in their own minds but their colleagues had pressed ahead regardless. SPC services were not considered in supporting such decision-making.

Hope and expectation

Patient expectation was considered to be linked closely to a societal belief that modern medicine can cure all disease (see box 4). Offering repeated courses of treatment risked perpetuating this fallacy and the idea of there always being ‘something more’ to offer. Extending this view, participants suggested patients were selective in the information they retained when being counselled about treatment options, dismissing the potential dangers of treatment in an effort to receive more intensive therapies. No matter how explicit the discussion, younger patients in particular found it difficult to acknowledge the possibility of treatment failure. Perhaps most challenging were the situations in which deterioration was so rapid and unexpected that all plans had to be abandoned. Participants found adjusting to this extreme change in events led them to raise fundamental questions of themselves and their professional abilities.

Despite these challenges, ‘striving to do the right thing’ emerged as the driving pattern of behaviour, with participants’ only concern being to deliver the best care for patients.

Core category

During the process of data analysis, a core category emerged entitled ‘boundaries’, describing the sense of division between three areas; be these services, disease states or professional disciplines. Issues around ‘boundaries’ were mentioned with both high frequency and connectedness in all four categories. For instance, the contrast between the biomedical and holistic approach to care; the transition from active treatment to palliation; the provision of services between settings and the overlap in role between haematologist and palliative care professional. All of these describe boundaries, whether conscious or subconscious, real or imagined.


In the UK, perceptions of palliative care being of ‘low status’ and poor scientific merit have been reinforced for haematologists by its absence from specialty training. This situation may be changing, with the 2010 Haematology Specialty Curriculum requiring competency in palliative care to be demonstrated in the FRCPath examination.10 Although the results of this study confirmed haematologists are rooted in a biomedical model of care, they also suggested the priorities of individual clinicians and their interest in palliative care vary widely. It follows that differences in approach, focus of interest and values need to be understood and respected if greater collaboration between haematology and palliative care services is to be achieved.

Recent models of SPC provision advocate a phased transition, with palliative care provided concurrently alongside active treatment from diagnosis of a life-limiting illness.7 ,9 The impetus to integrate SPC services upstream in the disease trajectory has been strengthened by recent evidence suggesting this approach offers prognostic benefit.11 ,12 Participants struggled to see how such integration would work in practice, particularly as this might shift focus away from terminal care provision, the area they perceived to be in greatest need. Doubts relating to the symptom burden of patients with haematological malignancy, especially early on, were in conflict with data from retrospective studies which suggest that the symptom burden of patients with haematological malignancy is not dissimilar to those with solid tumours.2 ,3

A transition has been described as a ‘change of focus in the goals of care from ‘active treatment’, where focus is on cure to a ‘palliative care’ approach, where the focus is on maximising quality of life’.13 The speed and unpredictability of disease transition, alongside the necessity to treat iatrogenic illness aggressively, has led to haematologists being accused of delaying referral to palliative care.9 ,14 ,15 The results of our study suggest that in haematological disease this transition period is complex, non-linear and often ill-defined, with the dying trajectory being very different from that of patients with solid tumours. Participants did not consciously delay SPC referral but had reservations about early referral, with concerns that, in the absence of specific need, this would risk duplication of effort and pull SPC services away from meeting the needs of haematology patients in the terminal phase.

The pattern of illness in haematological malignancy predisposes to a relationship between haematologist and patient which is more intimate, protected and intense than in many other specialties. The benefits of such close care include the development of expertise in, and comfort with palliation, making transfer to palliative care more disruptive than beneficial to patients.16 This view is supported by our findings, which emphasise the value placed in therapeutic continuity and the forging of a trusting relationship with patient and family.

Previous authors have suggested that haematologists are death-denying with a therapeutic optimism that prevents patients from accessing opportunities offered by the palliative care system.4 In our study, the relationship between ‘responsibility for’ and ‘control over’ disease outcome were inextricably linked, with a desire to ‘do something’ mutually reinforced by the intellectual challenge of overcoming the disease process. The expectations of patients and carers, based on a widespread belief in society that modern medicine can resist, postpone or prevent death,17 added further pressure. This shared intention not to ‘give up’ was strengthened by the intense physician–patient relationship and by the unpredictable nature of the treatment itself.

Study limitations

The scope of this study was small, with sampling limited to two NHS Trusts within the West Midlands. While data saturation was reached after eight interviews, the views expressed by participants may not be representative of other haematologists. Although conclusions drawn from this study cannot be generalised, the emergence of new themes would be worthy of future study in a wider population.

Using the grounded theory approach requires reflexivity (ie, sensitivity to the ways the researcher and research process have shaped the data collected).18 In this study, active awareness of potential for responder and researcher bias, alongside techniques to limit these (including member checking, interviewer triangulation and memowriting), ensured that analysis remained grounded in data and further enhanced its rigour.


The results of this study demonstrate the rich and diverse range of views and perceptions held by haematologists towards palliative care. In general, we found that participants held positive attitudes about palliative care involvement although some barriers to collaboration were also alluded to. These included difficulties in determining boundaries, specifically with regards to clarity of role and responsibilities; disease transition points; and the relationship between services. The role of SPC services was sometimes difficult to define, with timing of referral often determined by the participant's level of confidence in end-of-life care. Almost all highlighted a lack of inpatient palliative care unit provision, resulting in negativity towards referral.

Perhaps the most significant conclusion to be drawn from the study is the complex interplay between personal and professional experience, which underpinned participants’ understanding of palliative care and their referral behaviour. Participants’ desire to ‘do the right thing’ and preserve their relationship with patients lay at the core of many decisions but a natural predilection for science channelled their energies more towards disease and treatment than holistic care. Although the theoretical value of palliative care was recognised, there seemed to be some resistance to the reality and practicalities associated with referral to SPC services.

While this study has gone some way towards defining ‘palliative care need’ from the perspective of haematologists, this needs to be considered more broadly in the context of the perception of patients, carers and healthcare professionals. Although some evidence exists regarding the palliative care needs of patients with haematological malignancies,2–6 this is mainly retrospective and based on subgroup analysis. Further research to better establish the symptom burden, end-of-life issues and psychosocial needs of patients with haematological malignancy would assist in determining the extent to which haematologists’ views are representative of their patients’ views. In future, studies demonstrating benefits from SPC involvement might provide haematologists with the incentive to offer palliative care referral as a proactive alternative to the fatalistic resignation; ‘there is nothing more we can do’.


The authors would like to thank the participants for their honesty and responsiveness, and Helen Johnson (HJ) for her assistance with coding interview transcripts.



  • Contributors BW conceived the study and was responsible for data collection. BW and KF were responsible for data analysis and contributed towards writing of the manuscript. BW and KF are guarantors.

  • Competing interests None.

  • Ethics approval The study was conducted with ethical approval from Coventry Local Research Ethics Committee (reference 09/H1210/81).

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Unpublished data would be available by emailing the corresponding author.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.