Palliative care: a foreign language? Nurses’ perspectives on palliative care in a French comprehensive cancer centre

Abstract

Background

Integrated patient-centred palliative care for cancer patients is widely advocated internationally, but promoting it often proves difficult. The literature suggests that one key factor is physicians’ perspectives of palliative care (PC). Nurses’ views, however, from their pivotal stance within the healthcare team between patients and physicians, have been less well researched. This study explores French nurses’ perspectives on PC, how their view frames their role, and how they experience PC implementation. Nurses’ discourse is then explored as a reflection of their experience.

Methods

Semi-structured interviews were audio-recorded of 21 registered nurses, purposively sampled, in a French comprehensive cancer centre. Reflexive thematic analysis was applied by a team including a patients’ rights representative (a PC carer and applied linguist), the quality manager, and three PC physicians. Considerations of speech emphasis and style completed the analysis.

Results

Analysis generated three themes. Nurses perceive PC as a complex and continuous journey, passing through phases to the end of life. It revolves around patients’ physical and psychological well-being, timeliness, patient communication and empowerment. This perception frames their caring role as members of the healthcare team, based on necessary knowledge, training, and working conditions allowing them to fulfil their mission. Consequently, they experience PC implementation as either a virtuous or a vicious circle and suggest ways to improve the latter. Analysis of nurses’ discourse enhances the meaningfulness of the thematic analysis, reflecting that the more difficult the care context, the more in-depth the description and the more evaluative and emphatic the words chosen.

Conclusions

These French nurses see their holistic view of PC as diverging from that of physicians, keeping them (painfully) from fulfilling their caregiving role. While the nurses are able to “speak PC” with patients, they see it as a foreign concept/language for physicians (and management), resulting in a sometimes vicious circle of care. Physicians’ and management’s openness to hearing nurses’ views would be a first step towards patient and professional well-being. As the nurses suggest, a palliative approach and communication skills can be learned, and institutions can commit to rethinking priorities, policies, and resources. PC can become a shared language.

Background

Ever-increasing cancer survival rates have introduced new concepts and new terms: “survivorship”, “cancer as a chronic disease”, “controlled/cancer-free but not cured” [1, 2]. In parallel, the practice and discourse of palliative care (PC) have also evolved: “comfort care”, “enhanced supportive care”, “a palliative culture/approach”, “palliative medicine” [1, 3,4,5,6,7,8,9,10]. Nevertheless, in 2019, the concept and the words used to express it remained problematic: “...stigmatisation of the term [PC] among patients, professionals, and the public continues to counter positive messages about its benefits” [11].

Guidelines, standards, and recommendations for optimal PC have been set out by the World Health Organisation [12] focusing on key concepts that PC specialists are familiar with — among others, a holistic patient-centred approach, advance care planning in a continuum of care, interprofessional teamwork, PC training, and communication. Individual countries approach the question according to their own cultural traditions, with “limitations… seen as a cultural by-product” [13]. In France PC first emerged in the 1980s [3]; guidelines have been issued by various government agencies particularly since 2010 [14,15,16,17,18,19,20,21,22,23,24,25,26,27] as well as by learned societies [28].

While the essence of PC is thus clearly defined, the literature reports that across countries, realities vary [8, 13, 29,30,31]. Aside from specific national policies and financial constraints, difficulties are reported from various perspectives — physicians, nurses (and nurse’s aides), and more rarely, hospital administrations and patients. In France in 2024, despite government investment over many years, the country was only ranked 15th among OECD members in terms of density of specialised PC services [32], meeting only 30% of PC needs [33].

Regardless of the research perspective, however, the key underlying factor in optimal PC rests on a shared “therapeutic ideology” [34, 35], grounded in knowledge and training, framing the practice of all professionals involved. Since physicians (oncologists) are traditionally the ones to announce a cancer diagnosis and supervise care [36, 37], an overwhelming proportion of the literature explores their ideology, professional practice, and therapeutic difficulties: among others, a traditional curative paradigm [1, 5, 8, 10, 35, 38] precluding PC screening [8, 39, 40] where PC rhymes with sequential terminal care [34, 36, 39]; inappropriate or insufficient communication with patients due to a lack of PC training and to patients’/families’ negative attitudes [1, 8, 29, 34, 36, 37, 41,42,43,44,45]; poor interprofessional communication and collaboration [35,36,37, 41, 46]; a lack of shared decision-making [47,48,49] and advance care planning, including early PC referral in a concurrent model of care [1, 5, 6, 8, 29, 30, 34, 36, 39, 40, 50,51,52]. Oncologists’ therapeutic difficulties can also be compounded by external factors such as late or reluctant referral from the community physician [36], institutional barriers (PC resources, physicians’ heavy workload and unrecognised psychological difficulties) [36, 37, 46], along with the innate unpredictable nature of cancer itself [37, personal communication with the hospital’s Mobile Palliative Care Team (MPCT)].

But with interprofessional teamwork also being a cornerstone of optimal PC [53], the research leaves a gap, since far fewer authors focus on nurses’ corresponding ideology or practice, hindering or helping effective PC. In a francophone context (France and French-speaking Belgium and Switzerland), a precursor study in 2001 among PC nurses compared “simplistic” (in the authors’ terms) questionnaire results to earlier results from PC physicians on communication with terminal cancer patients and decision-making processes. Nurses differed from physicians only in the strength of their perception of patients’ information needs and of shared decision-making [53]. A second quantitative study, in 2008, broadened the horizon to include French nurses’ lived experience, practices and representations (key words) associated with PC. Only 34 nurses (35%) actually worked in oncology; influential factors were PC training and working conditions [54]. Lastly, a 2012 study reported French nurses hesitating on MPCT referral for fear of provoking anxiety [55]. More recent French and international work (2015, 2018, 2020) [40, 56,57,58,59] portrays nurses’ mission as both physical and psychological caregiver and pro-active sentinel, advocating for patients by sharing an ethical and timely palliative approach with physicians. Oncologists’ curative, or at best sequential, approach is seen to ignore “the elephant in the room” [60], compromising nurses’ strategic pivotal role and relegating these day-to-day caregivers to a difficult “powerless” position hindering effective patient-centred palliative and end-of-life (P/EOL) care [58].

In 2019, in a Swedish context, Wallerstedt et al. [8] posited that: “[I]t is obvious [across a range of professionals] that the current ambiguous understanding of palliative care has a negative impact on the care that is provided to patients and the emotions of family members and healthcare professionals. We strongly argue that a common conceptualisation would enhance care, interprofessional communication and teamwork.” But sharing a conceptualisation means putting thought into words, inextricably linking concept, language (i.e., a linguistic system available to humans), and speech (i.e., actual utterances, or contextualised language use — “discourse”¹, see F. de Saussure’s linguistic theory, 1916 [61]). Boyd et al. [11] argue from a British PC context that “use of language matters, and getting it right or wrong can promote or prevent an ethos of shared endeavour across multidisciplinary teams”. To fully explore nurses’ views of P/EOL care and their practice, targeting their actual discourse is thus essential: “How people talk about something… can tell us as much about what they think as the content of what they’re actually saying” [62].

Looking into the literature on how P/EOL physicians and nurses express themselves reveals a second research gap. The rich body of Anglo-American and Northern European research, often focusing on metaphor, contrasts with a scarcity of work, at least in the French context, on health professionals’ (and patients’) discourse of any kind. By metaphor, we mean thinking about one thing (often an abstract, complex, or sensitive experience) in terms of another (more often a concrete, tangible and accessible experience) [63]. Its use in emotional, intimate, subjective, if not taboo experiences such as cancer and death, “may help overcome, but also contribute to, communicative problems” [64]. In 1978, Susan Sontag [65] first analysed the harmful impact of metaphors in tuberculosis and cancer. Several authors have further stressed the importance of healthcare professionals being sensitive to patients’ — and their own — metaphorical speech [66,67,68,69]. Starting in 2014, the Metaphor in end-of-life care (MELC) project determined that different metaphors (e.g., “battle” or “journey”) are appropriate for different people at different times [64]. In 2020, this work was followed up with Swedish cancer patients [70]. In the Dutch context, independently of metaphor, as early as 2004 van Kleffens et al. studied the use of the term “palliative” in clinical oncology to distinguish between curative and palliative oncological treatments to clarify decision-making for both physicians and patients [71]. More recently, Busser et al. explored GPs’ language use concerning the existential dimension of PC [72], and van Meurs et al. studied oncologists’ responses to advanced cancer patients [45].

In France, only two major publications have ever addressed the language of P/EOL care. Both focused on words and speech actually used; neither explored metaphor. On the postulate that words surrounding death are left unsaid by patients, professionals and institutions, Mino & Fournier [1] sought to break that silence in 2008. Respectively a medical researcher in PC and public health and a philosopher, they focused at length on the inherent confusion around the word palliative (vs curative). In 2010 Chvetzoff’s doctoral thesis in medicine [37] explored the ethical stakes of decisions to stop chemotherapy. In analysing oncologists’ and patients’ lexis, semantic fields, and attitudes, expressions of emotion — primarily negative — ran high throughout. On a more modest scale, a PC university diploma thesis (2016) investigated the use of the term “palliative” in patient/professional perceptions and interrelations in a context described as “le soin qui-ne-se-dit-pas-palliatif” (care that-must-not-be-called-palliative) [44]. Two quantitative questionnaire studies in 2008 and 2024 [54, 73] explored PC representations and their impact on nurses’ (and other health professionals’) practice through key words from free association.

The gaps in the literature on both nurses’ views of P/EOL care and on how people involved in this context — other than Anglo-Americans and Northern Europeans — put words to their thoughts thus raise several questions: how do (French) nurses interpret PC? How do their views frame their role? How do they experience PC implementation? As a corollary, is PC a “foreign language” (metaphoric or otherwise) for any or all types of (French) health professionals?

The above queries arose indirectly from an in-house (unpublished) assessment in a French comprehensive cancer centre prior to a national certification protocol in 2017. The MPCT identified the risk of “absence of or non-conformity with the palliative approach” and carried out a medical chart review of professional PC practices to target difficulties. Among other points, divergencies in perception were found between physicians and nurses, along with an overall lack of discussion. These results prompted the MPCT, the head of Quality, Safety and Risk Management, along with a patients’ rights representative (a PC carer and applied linguist) to explore the phenomenon. Bringing together these diverse research perspectives offered a privileged opportunity to deepen understanding of the situation from the nurses’ perspective.

Methods

Design and setting

The study was commissioned by the cancer centre’s statutory Patients’ Rights Commission to shed light on the institutional PC context further to the in-house assessment. The above research questions were explored within an interpretivist paradigm to enable a deep understanding of participants’ subjective experience of their reality. Analysis of nurses’ discourse was intended to further stakeholders’ understanding [74, 75].

The research was set within a French comprehensive cancer centre, one of a network of 18 national specialised care and research centres, each one with an MPCT and an integrated dedicated PC care pathway. This particular centre was one of the first to have a full Supportive Care department offering over 20 different specialties (psychologists, social workers, a sexologist, a sophrologist, sports programs, etc.) as well as a PC day hospital. The centre’s PC tumour board — a rare phenomenon in French cancer care — meets weekly. Finally, when the study was carried out, the hospital was one of only two such comprehensive centres accredited by the Organisation of European Cancer Institutes. The hospital has no PC unit, but has 12 designated identified PC beds in two full hospitalisation wards. These beds, allotted by the regional health authority, have a lower staff/patient ratio than those in a PC unit [76]. Our sample of nurses worked in these two wards where care is both palliative and curative.²

Since no patient or professional health data were involved, under French law no prior approval was required by the Institutional Ethics Committee/Review Board. Nevertheless, the project was presented to the Committee to foster discussion on the ethics of PC; a favourable opinion was received. The project was also presented to the Director of Nursing Care and the head nurse of each of the two wards who all approved the study’s rationale and relevance. Since the in-house challenges identified are similar to those reported internationally, we believe that our findings are transferable to other care environments and contexts where the use of the “p-word” [44] is also difficult [75, 77, 78].

Participants

To explore the nature of nurses’ discourse, alongside thematic content, we sought a maximum number of participants for the largest possible dataset. Our sampling was thus purposive — making no selection. Data saturation (seen as a “criterion for discontinuing data collection and/or analysis” [79]) was thus not a pertinent consideration. Participation was voluntary, out of ethical respect for nurses’ freedom of choice and as part of informed consent. Following the project coordinator’s presentation to the head nurse of each ward (A, B), these two professionals introduced the project to their entire teams and posted a sign-up sheet for volunteers to choose their own date and time slot (early morning, daytime, early evening). Twenty-one (out of 28) registered nurses volunteered, equally distributed between the two wards. Table 1 shows complete demographic data, but noteworthy are the facts that all participants were women with widely varying levels of experience in cancer care, many with night shift experience, and most with minimal PC training.

Table 1 Participant characteristics

Members of the research team adopted an interdisciplinary approach to explore PC: GT as project coordinator, a patients’ rights representative, PC carer, and applied linguist; NCH, head of Supportive Care and the MPCT; MB and VMS, MPCT physicians; DV, head of Quality, Safety and Risk Management. As early as 2014 (beginning of the MELC project [63]) GT felt frustrated with the apparent lack of reference made to PC, in metaphorical terms or otherwise. She and NCH exchanged at length from 2015, each one bringing to the other’s attention her respective difficulties. MB and VMS joined in on these exchanges over time, adding professional insight and co-developing the research design. DV was also challenged by the PC question and contributed his analytical sensitivity and experience in handling textual data from both qualitative and quantitative perspectives.

Materials

The semi-structured interview guide was inspired by the MELC project [63] and was designed to be reproducible. Developed collaboratively by the MPCT and GT, it included an ice-breaking demographic introduction, three open questions echoing the research questions (with follow-up probes as necessary to obtain the fullest picture possible), and a final query opening out to further considerations (Fig. 1). The interview guide was piloted for clarity and pertinence by GT with two nurses from services other than those under study; no changes were deemed necessary.

Fig. 1

Interview guide

Procedures

Prior to the interviews, GT liaised with the Patients’ Rights Commission, the Director of Nursing Care, the head nurse of each ward, and the Institutional Ethics Committee/Review Board. The two head nurses had already presented the project to their nursing staff and agreed to volunteers’ taking time off work; interviews thus took place between January and April 2019 at participants’ convenience. A quiet, dedicated room was provided in each service. Following an oral presentation of the study’s rationale, aims and methodology, GT introduced her institutional role as patients’ rights representative and her professional interest in “PC-speak”. Both data security and anonymity were guaranteed. Verbal informed consent was collected and recorded with QuickTime Player on a MacBook Pro computer; GT also took notes. Interviews averaged 46 min (26–86), totalling 16-and-a-quarter hours, and were carried out in the absence of any power dynamics in a climate of confidence, trust, and spontaneity [78, 80]. Indeed, as a probable reflection of the project’s originality and the interview experience itself, several nurses expressed pleasure at being heard “for the first time”. All participants agreed to answer follow-up questions and were assured of feedback on outcome and results.

Analysis

To reflect nurses’ description of their beliefs and lived experience, reflexive thematic analysis was chosen to analyse and interpret the qualitative interview data [80,81,82,83]. This method respects participants’ subjectivity, as well as that of researchers’ individual discipline-specific influence. In answer to our research questions, including a focus on discourse, we thus adopted an essentialist epistemological perspective. In Braun & Clarke’s words, a “largely unidirectional relationship is assumed between meaning and experience and language (language reflects and enables us to articulate meaning and experience)” [81]. An experiential orientation to understanding nurses’ discourse — particularly emotional (specific word choice to communicate experience [84]) and evaluative (expressing attitude towards content [85]) — was thus embraced to mirror the description of their lived experience on the wards [75, 81, 83]. While GT, as an applied linguist, had in mind the gap of language research in the P/EOL context [63], fellow team members stressed the importance from their perspective of a data-driven approach to explore these nurses’ “first time ever” experience. Analysis was thus inductive, and coding, semantic, for a descriptive analysis of content [81, 83].

The focus on expressive discourse was also descriptive. While Braun & Clarke [81] do not prone quantitative reporting in thematic analysis — a high frequency does not mean value, nor does a low percentage of codes/themes in an interview study mean insignificance —, they do state that “occasionally reporting percentages or frequencies is useful” [86]. In the present case, ranking the top ten key words (SphinxQuali word counts) to characterise lexical fields (collaboratively decided with DV) provides an overall view of the nurses’ message. Secondly, tabulating coded qualitative data as frequencies or percentages focusing on evaluative word use (primarily adjectives, adverbs) [41, 85] reveals emphasis (relative volume, repetition [84]) and style (discourse choices for a particular purpose), allowing readers to “hear” nurses’ voices [75, 87].

After recording the interviews and noting demographic data, GT began familiarising herself with the data — the first step of Braun & Clarke’s six-phase reflexive thematic analysis [81, 82]. This corresponded to transcribing the interviews with Macintosh iTunes, using transcription conventions to typographically represent the level of emotions expressed in speech production [84]: emphasis through upper case, boldface, etc. and repetition (Additional file 1). GT then repeatedly read through the data, noting observations and insights within each interview and across the dataset. Initial codes were manually generated before being collated with data extracts. Inspired by Chvetzoff’s early work on language in PC [37], neutral codes were excluded (e.g., cancer, diagnosis, rights, hospital, chemotherapy, oncologist, etc.) to focus on nurses’ feelings [“the commonsense notion that feelings get put into words” 84] and evaluative discourse [85]. The data set was then imported into Excel, identifying nurses numerically from 1 to 21, and tabulating demographic data to obtain descriptive statistics.

Generating initial themes with SphinxQuali was collaborative (GT, DV); reliability was not a methodological criterion since the goal was to obtain a reflexively nuanced reading [74, 80, 82]. Potential themes were then recursively reviewed against coded data from the entire corpus, keeping in mind the definition of “theme” as “patterns of shared meaning underpinned… by a… central organizing concept” [82]. The fifth step, before reporting, was defining and naming each theme and sub-theme in relation to the entire dataset and our research questions, and then identifying illustrative extracts (Thematic map, Fig. 2). Regular meetings were held throughout the process with the three MPCT physicians and GT to discuss findings. GT drafted the report, translating interview extracts. All authors approved the final version. Reporting the study followed the 32-item Consolidated criteria for reporting qualitative research guidelines (COREQ, Additional file 2) [88].

Fig. 2

Thematic map

Finally, during interviewing GT was struck by the nurses’ lengthy, in-depth, and expressive description of difficult PC situations, recalling the above-mentioned in-house assessment of the risk of “absence of or non-conformity with the palliative approach”. In contrast, she also felt it important to explore the relative length of the description of better, or good, PC situations. This entailed distinguishing coded data (collaboratively with DV) by contextually determining the nature of experiential discourse (negative, positive), considering spoken emphasis (volume, repetition), and calculating length (number of words) and percentages. Particularly evaluative terms were then targeted and tabulated (frequency, percentage) to reflect their impact across the entire dataset [1, 41, 84, 85].

Results

Reflexive thematic analysis allowed us to identify three main themes relevant to our research questions. First, these French nurses conceptualise PC as a continuous, patient-centred journey, including notions (sub-themes) of appropriate timing, patient (and family) communication and patient empowerment. Second, this perspective frames their role as one of caregiving, involving sub-themes of necessary interprofessional collaboration, knowledge and training, and working conditions. Third, these two key ideal conceptualisations, confronted with everyday life on the wards, allow nurses to reflect on their professional experience in terms of a virtuous circle, a vicious circle, and of ways to improve patient-centred PC in this French comprehensive cancer centre. Finally, the focus on nurses’ discourse itself furthered understanding of the main research questions, reflecting how these professionals actually and expressively “speak PC”.

PC is a continuous, patient-centred journey

First and foremost, PC is clearly seen by all 21 nurses — regardless of the number of years in cancer care or the extent of PC training — as a continuum passing through phases. It is definitely not a synonym of EOL, even if death is the inevitable end of the journey: “ It’s the end of curative treatment, when you consciously stop treating people FOR a disease” (nurse no. 17), but “it’s continuing to care for people, not abandoning them because they can’t be cured” (6). While the nurses refer to French legislation on PC in terms of “unreasonable obstinacy”, “comfort”, “advance directives”, “sedation”, for them PC means patient-centred physical and psychological well-being. Physical ill-being is readily visible: “discomfort, difficult breathing, you can see it in the patient’s face” (15); “most important, PAIN. It MUST be well managed” (11); “Pain, that’s part of sedation, physical AND moral pain” (11). The latter — patients’ psychological well-being — is described in terms of clarity of mind/spirit, confidence, and serenity: “Patients need to understand what the term ‘PC’ means because when you KNOW what it means, it’s less frightening. You have to take the time to explain to the patient, to the family, what the care will be, and what happens when. They’re relieved” (13); “that patients trust the healthcare team, that we can explain as much as possible” (5).

On the wards, however, the nurses’ ideal conception becomes a complex reality involving other healthcare professionals and intertwining essential issues of appropriate timing, patient (and family) communication and patient empowerment. These three notions are discussed below; teamwork will be discussed thereafter in relation to nurses’ caregiving role.

Appropriate timing

This refers both to particular moments (night, weekends) and to when decisions are made, if indeed they are. The result of inappropriate timing leaves patients — and nurses — in distress: “We have patients in a borderline state for two-three weeks, palliative without being identified as palliative. If there’s an acute problem, do we send them to resuscitation?” (3); “PC happens TOO late, or not at all, — chemotherapy on Monday, EOL on Friday” (15); “The night lasts ten hours! When you call the junior doctor on duty to say that a patient’s really suffering and she replies ‘OK, but we’ll see tomorrow.’ OK, but NO!!! I can’t leave a patient in that condition. Our hands are tied” (21).

Patient communication

This refers to bilateral exchange between nurses and patients (and family). Nurses recognise their own skill in communicating when the cancer progresses, stressing the importance of repeated, knowledgeable exchange as an integral part of PC practice: “Knowing how to find the right words — or to stay silent — can be learned” (11); “It’s our job; we were trained to do this. You learn to take a step back” (7); “She knew there were no more treatments and cried all night. It was quiet, so I could sit with her for several hours, and I asked her point blank what her questions were, promising to answer. They were difficult words to say — sedation, death, dying, burial.” (21).

Only twice do nurses report hearing patients request or merely mention PC. This taboo leaves those directly concerned with neither concepts nor words: “The word ‘palliative care’ REALLY frightens patients and their family, so they don’t say it.” (12); “For them, PC leads directly to EOL” (11). But helping patients put words to the situation is essential for their psychological well-being: “It’s important for us to know where [the patient] is in her head, to speak in words that are easier to hear” (13).

Patient empowerment

Without actually using the term — agency in sharing informed decision-making —, nurses advocate in patients’ favour. The concept is recent in France [1, 32, 36, 42, 43], with the theme being cited by only six nurses: “It’s the patient’s right to know what the situation is” (4); “to be clear from beginning to end so that a patient feels well taken care of, an integral part of her own care pathway, and not someone external to her care who’s the object of procedures or treatments she doesn’t understand” (13).

Nurses’ caregiving role

The nurses’ “therapeutic ideology” [35] of PC frames their professional practice as a caregiving role: empathetic presence, active listening, appropriate words, and respect for patients’ needs and wishes: “Empathy — this is a hard time for them — so they can trust us” (5); “We first have to listen to the patient to know what she wants and needs (9); “respecting her lifestyle, at home? in hospital?” (10); “PC? Information, language, the right words, just being there” (7). On the wards, however, this ideal care involves a complex reality of interprofessional collaboration, knowledge and training, and working conditions. The degree to which these pieces of the puzzle fit together affects nurses’ ability to fulfil their caregiving mission, either helping or hindering effective PC.

The healthcare team

“You can’t be palliative unless there are people around to propose solutions. There must be a pluridisciplinary team” (18). Principal team members, along with nurses in a pivotal position between patients/families and physicians, are physicians and junior doctors, nurses’ aides and the MPCT.

The MPCT is very much appreciated by the nurses for its expertise, resources, and effective communication: “The team has keys that we don’t, for pain, anxiety, and we don’t have the time. They spend a long time with patients, families, taking everything into account. It’s a reflex for us to call on them” (10). “For nights, sometimes the MPCT comes ahead of time to anticipate a prescription. They mark down observations — ‘if deterioration, do this, this, and this’ —; the ward doctor writes out the prescription. It REALLY helps us A LOT” (13). The MPCT makes systematic visits once or twice a week to the two full hospitalisation wards studied but can also be specially called in by the ward physician or junior doctors. Nurses can and do also call them in. Nurses’ aides are also very much appreciated, in close solidarity: “at the forefront of patient comfort, able to exchange during personal care, when we [nurses] have to worry about technical procedures like chemotherapies, transfusions” (13).

Oncologists and junior doctors, however, are seen to diverge from nurses in their therapeutic ideology with a curative paradigm. This lack of a common conceptualisation leads to clearly expressed therapeutic difficulties for nurses: “There are some doctors who want to do, do, do until it’s TOO late. The referring oncologist wanted to manage things until the end, but at a certain point, it went beyond her competence to help the patient, and WE called in the MPCT. But it was just TOO late” (2); “They keep pushing chemotherapies, but then they ask us to manage end-of-life. We can’t do both as we should, unless we ‘forget’ people… Physically and morally, it’s HARD enough without, in addition, constantly fighting for them to hear what’s simply logical to do” (17).

Triangular communication (doctors-patients-nurses) is also difficult. The nurses need physicians to put words to reality but see them as not being able to talk about PC: “The referring oncologists don’t often come by. We wait for them to clearly explain the situation. Often, it’s us, alone, facing the problem. It’s the doctors’ role to say these things, not ours!” (11); “It’s not easy, but how come WE [nurses] can explain things gently and they can’t, or don’t take the time?” (17); “Especially at night, on weekends, we NEED clear labels — PC/curative — and instructions — Do not resuscitate/Resuscitate —, and NOT ‘non-invasive resuscitation’” (10).

Knowledge and training

Among the 21 nurses, only two had diplomas in PC, 11 had minimal in-house or nursing school PC training and eight reported no training. However, their view of PC reflects an open mind on the issue; indeed, 18 of them reported choosing to work on these two specific wards. Physicians, on the other hand, are seen to lack training and knowledge: “They only know the theory they learned and are not WITH the patients to see what’s happening (17); “It’s VERY HARD for the junior doctors, especially at night. They’re not trained in PC and EOLC, and they’re VERY reticent to set up sedation. They’re afraid to hasten the process, and since they don’t know the files, they don’t want the patient to die that night. Result? No prescription” (19). “Interdisciplinarity and communication: managing pain, midazolam, you can learn that. People get scared because they don’t know how to dose it, because they don’t know the consequences.” (20).

Working conditions

Within this third component of nurses’ caregiving role, recognition of their professional identity and role is paramount: “…knowing the value of a nurse, of a nurses’ aide, that — UNFORTUNATELY — shouldn’t have to be learnt; it should be innate” (20). Nurses see denial of their practical PC experience and disregard for their advice: “Often, we see that the patients are uncomfortable… We tell the day shift that the doctor has to see them, and maybe call in the MPCT. But the doctors either don’t listen or see” (21); “Without a doctor who FOLLOWS our opinion, or that we can exchange with, we can’t have good EOLC” (4).

Human and material resources are also necessary for good PC. In the absence of a PC unit — a reality several nurses regret —, they report managing as they can: “Twelve PC beds for the entire cancer centre when most of the time there are 12 needing PC in just this ward!” (3) On night shifts frustration culminates: “At night, we don’t have the resources, and it can be REALLY difficult, really, REALLY (15); “We’re two, with no nurses’ aides, for twenty patients. Sometimes we have FIVE EOL patients and we’re still just two” (12).

Finally, the nurses need an ethical working environment. Physicians “not telling a patient because she couldn’t bear to hear a poor prognosis…” is seen as morally wrong: “…Who are YOU to judge whether a person can’t bear it?” (17). The decision (or not) to stop curative treatment and implement PC “is doctor-dependent, so a gamble for patients” (15); “It depends on whether the oncologist is more curative oriented, whether the junior doctor at night or on the weekend has any PC experience, whether communication with patients and families was satisfactory and how easily they let go, whether nurses were able to accompany them. It’s VERY COMPLICATED and EXTREMELY UNEQUAL” (20).

Living PC implementation

Nurses’ lived experience of PC implementation is the third theme generated by the interviews. Their “solid clinical evaluation” [46] allows them to credibly express satisfaction when PC is well implemented, but also disapproval, anger, and frustration when it is not, putting them in the position of “treating while maltreating” [46]. In this light, they see life on the wards, a complex reality, as a virtuous or a vicious circle: “When you KNOW it’s going to f*** up, and you’ve told the junior doctor several times, and she replies, ‘Don’t worry!’, I’m angry, frustrated, but at the same time proud, knowing I told her what was going to happen, but she didn’t believe me, and I’m tired of feeling ‘I told you so!!’” (20).

A virtuous circle

Nurses speak favourably and with positive emotions of situations where patients (and family) were physically and psychologically comfortable, where communication flowed with patients and among all team members, and where nurses were able to fulfil their professional mission of caregiving and supporting: “It works when everyone’s on the same wavelength with treatment decisions — the patient, her family, the doctors —, respecting the patient’s will” (10); It’s “very, very successful” when you call in the doctor, to please her, and she sees that you increased the treatments, that you titrated [the appropriate sedative], and she leaves, saying ‘You explained everything, the patient had no questions for me’. You know you did your job well; the woman died as well as her disease would allow. Top-notch work; you can be proud” (20).

A vicious circle

More guardedly, nurses describe experiences tending towards a vicious circle: “Sure, we agree, but we’re not on the same wavelength, and we don’t have the same experience. They’re in a curative mind set and aren’t even aware they can break rules at the end of a cancer journey. What difference does it make?” (20); “EOL care, we’re not BAD, given the means we have. But we do mess up, and at the end of life, a failure is always more difficult” (6). Particularly at night, “when you know it’s the end, and patients are all alone, and in-between rounds you realise the person’s gone. That’s really HARD to manage” (15).

Describing a full vicious circle, nurses relate their experience with strong emotion, particularly in EOL care in critical and complex contexts: “The patient suddenly took a serious turn for the worse and needed sedating to stop her feeling oppressed and suffocated. I needed the doctor’s approval, but she wasn’t on duty yet! It took twenty LONG minutes. I’d SAID she was getting worse, and repeated that they should have ANTICIPATED” (19); “The patient is dying and we’re carrying out an invasive treatment because the doctor decided we had to continue, but I know very well she’s going to die! It’s sh***y, she died that night” (20); “It’s not OUR role to tell the doctors ‘this patient needs to be in PC; she needs to be sedated’. Far from it! We don’t have the knowledge, and we don’t have the power. Oncologists need to be sensitised to PC and TRAINED!” (6); “Often we’re walking on eggs; we don’t know what doctors actually SAID to patients when nothing is written down. We have to guess; it’s very uncomfortable!” (18).

Ways forward

In a constructive spirit, nurses’ describing difficult vicious circle situations led to concrete suggestions on improving P/EOL care. PC and communication training and rethinking human resources were the main focus: “[Training], especially for the junior doctors; they’re totally LOST in PC, and in real trouble at night, on weekends” (18). “I see worried patients. Shouldn’t we [nurses] be present to actually hear what’s said? Patients understand the first few words, then nothing. There’s TOO much information” (15); “We have two-three deaths a week here. We could make time for feedback to review how we managed things, if there are points to improve, to share how we felt, if we were on the same wavelength” (5).

“Speaking PC”: focusing on nurses’ discourse

Recalling that Braun & Clarke [81] assume “a simple, largely unidirectional relationship… between meaning and experience and language (language reflects and enables us to articulate meaning and experience)”, we first highlight nurses’ top ten key words characterising lexical fields for an overall view of their thematic message (Table 2).

Table 2 Top 10 key words by number of occurrences across entire corpus^a, number of nurse users (minimum 5/21) and lexical field

Knowledgeable communication, then, must be continuous to allow nurses to properly carry out their complex caregiving role. Recalling that the interviewer was struck by the lengthy and expressive description of difficult PC situations, we now focus on nurses’ evaluative discourse (emphasis, style [84, 85]). Emotion-laden descriptions of virtuous and vicious circle situations account for nearly half of the dataset (49.86% of 83,922 words), with vicious circle descriptions, especially of EOL, being notably longer, e.g., 25,021 words vs. 13,448 for positive P/EOL experience. Particularly evaluative and emotional words were identified and tabulated (Table 3). Typography (upper case, boldface, exclamation points) reflects the degree of speech emphasis (relative volume, repetition) and style (discourse choices for a particular purpose) to reflect the emotional load as data — albeit subjective [85] — in its own right (Transcription conventions, Additional file 1).

Table 3 Expressive items

Explicitly vulgar words are used solely in recounting vicious circle situations. Strong qualifiers (primarily adjectives, adverbs) can be divided into two categories, reasonably and less strong. Reasonably strong qualifiers (boldface) account for a minimum of around two-thirds of all strong qualifiers, whereas less strong qualifiers (upper case) and utterances expressed in a particularly animated tone — “!(!!)” —, make up around 50%. More general words expressing negative emotions like “to be afraid” and “pain/painful”, nonetheless, account for around 40% of total qualifier usage [1, 37].

Finally, unlike the above-mentioned studies reporting on metaphor in EOL care [63,64,65,66,67,68,69,70], our French results can only be seen as exploratory. Indeed, our nurses use this figure of speech only five times in a corpus of nearly 84,000 words to describe particularly difficult situations, reflecting earlier research on a very small scale (Additional file 3). Readily identified by their striking, contrasting contexts, these few occurrences express major difficulties such as interprofessional collaboration, P/EOL care, knowledge and training, and nurses’ pivotal caregiving role. It might be hypothesised that such images are less necessary for nurses, given their caregiving role accompanying patients (“shifting away from the disease to focus on the person”) compared to physicians delivering difficult treatment information (“to continue or stop chemotherapy” [57]). Similarly, unlike patients who may discuss their cancer as a “journey”, a “battle” or otherwise [70], our nurses are probably not called on — other than in an interview context — to describe their view of P/EOL care. A single “positive” metaphor (not included in Additional file 3) brought the corpus-wide total to six occurrences, with one nurse calling for “a new choreography” (2) to favour hospital-wide PC acculturation.

Discussion

Our results shed light on the state of P/EOL care in a French comprehensive cancer centre as seen through nurses’ eyes. How do they interpret PC? How does their perspective frame their role? How do they experience PC implementation? Finally, does PC seem to be a “foreign language” for any or all types of health professionals?

In accordance with WHO guidelines [12], these nurses have a holistic view of PC, seeing it as a continuous patient-centred journey until EOL. Patients’ physical and psychological well-being is paramount and requires appropriately-timed pertinent responses, patient (and family) communication, and patient empowerment. Accepting the assumption that a healthcare professional’s vision of P/EOL care frames her professional practice [34, 35, 44], these nurses describe their role as a pivotal position for caregiving including empathetic presence, active listening, appropriate words, and respect for patients’ needs and wishes. Fulfilling this mission requires interprofessional collaboration, effective triangular communication among all stakeholders, PC knowledge and training, and appropriate working conditions.

Nurses’ ideal conceptualisation of PC and of their own role, however, encounters real-life obstacles on the wards. In their words, the root of the problem is oncologists’ and junior doctors’ therapeutic ideology targeting curing at the expense of caring. Seen primarily as being due to a lack of PC training, along with interprofessional challenges, the result is inappropriate PC behaviour. This, in turn, keeps nurses from doing their job and results in patients’ — and nurses’ — distress. Finally, nurses’ description of experiencing PC implementation suggests a virtuous circle/vicious circle reality, albeit with a middle ground. Emotions run high, and nurses are sometimes satisfied and proud, but more often disapproving, frustrated, and angry. Nevertheless, in a constructive spirit, they make concrete suggestions to improve P/EOL care.

One of the two nurses with a university diploma in PC sums up the need for patient-centred PC, calling for a reality check: “ ‘Fortunately’ for our patients, they have a disease that allows them access to PC — but the decision has to be made to do so! We know that metastatic cancer is palliative. On that basis, three-quarters of our patients [out of around 35,000 in 2020] are already in the palliative phase”.

These findings help bridge the first research gap identified between physicians’ and nurses’ respective PC views and practice. The nurses’ in-depth description confirms what the literature reports: physicians’ therapeutic ideology is perceived as a curative, sequential paradigm equating PC with EOL [1, 8, 10, 34,35,36, 38, 39]. The nurses say that they are not “on the same wavelength” and critically describe what the literature reports: physicians’ inappropriate or insufficient communication with patients [1, 34, 36, 37, 41,42,43,44,45, 51], poor interprofessional communication and collaboration [29, 36, 41, 46], rare shared decision-making [47,48,49] and advance care planning/early PC referral [5, 6, 8, 30, 34, 36, 37, 39,40,41, 50,51,52]. Often with very emotional discourse, they describe how such practices leave patients — and themselves — in distress [84, 85]. The nurses thus clearly confirm the importance of a “common conceptualisation” of PC [8, 11]. The sole recent French study (2024) exploring the question of PC ideology and its impact on professional practice among physicians, nurses and nurses’ aides — albeit quantitatively and in general medicine —, still highlights divergencies on both levels between physicians on the one hand and nurses/nurses’ aides on the other [73].

A first step towards a shared patient-centred PC perspective cited by the nurses and reported in the literature [10, 29, 31, 34, 36, 37] is PC education and communication training for physicians, both as students and as professionals. Two fundamental aspects of such training are physicians’ acceptance of their own professional psychological challenges and appropriate support [36, 37, 46] and their perceived non-recognition of nurses’ professional identity and role. Nurses’ feelings of injustice, frustration, and anger on this last level are echoed in the literature [56, 89, 90]. Moreno-Milan et al. [91] highlight the impact of such denial through the construct of “meaning of work” among palliative care professionals, concluding that job recognition is a key factor in avoiding burnout. Horlait et al. discuss similar questions for oncologists [36]. Oude Engberink et al. [58] posit that the benefit of recognizing nurses’ specific role is an essential concept in understanding delayed PC implementation. Finally, nurses’ own constructive suggestions for PC and communication training (interprofessional, but particularly for physicians) again echo the literature [50, 58, 90], as do their recommendations for the institution to rethink human resources (physician-nurse tandems, sufficient staff to allow for team meetings) [29, 58, 92]. These suggestions are reflected in the single “positive” metaphor cited above calling for “a new choreography” for hospital-wide PC acculturation.

Since the nurses were interviewed in 2019, the cancer centre has taken steps to forestall the identified risk of poor PC: the MPCT provides increased in-house PC training (responding to, in particular, nursing staff turnover) and holds weekly interprofessional meetings for complex cases. The hospital employs an additional day nurse, a night nurses’ aid and a second physician for the two wards involved. A campaign was run by the Institutional Ethics Committee/Review Board to favour the use of advance directives. Nevertheless, the head of Supportive Care and the MPCT (NCH, personal communication 2024) reports that the term “PC” is still primarily used only among professionals and rarely so with patients unless they themselves bring it up (as the nurses reported in 2019). NCH confirmed that, for most patients, “knowing you’re in palliative care means thinking you’re doomed rather than chronically ill”.

Our final focus on nurses’ discourse first highlights through key words their patient-centred PC perspective requiring continuous knowledgeable exchange to carry out their complex caregiving role. The evaluative nature of their discourse, reflecting their lived experience [81, 84, 85], echoes Chvetzoff [37], where the more difficult the care context, the longer and more emotional the discourse and the more evaluative and emphatic the words used. The notion of evaluation is indeed “slippery” [85], with “no single agreed upon scientific definition of the term ‘emotion’” [84], but this finding corroborates Ruszniewski’s psychotherapeutic experience with health professionals accompanying dying patients: the context obliges professionals to process their own intense emotional voyage [46]. In this study, typographics and basic tabulations allow readers to “hear” the intensity of the interviews describing, in particular, difficult EOL situations.

In response to our fourth research question, in the present context, talking about PC does seem difficult for physicians (and the institution). Our findings thus help bridge the second research gap contrasting Anglo-American/Northern European interest in language use in P/EOL care to that reported elsewhere (at least in France). These nurses use very few metaphors to describe complex, difficult situations, but the overall emotion and expressiveness of their discourse add to the international literature calling for PC professionals to be aware of their patients’ use of language [63, 70] as well as their own [45, 72]. Sensitivity to patients’ ability to understand, process and use information — “health literacy” [93] — is also a consideration. As long ago as 2007, Delavigne et al. [94] offered a French dictionary of cancer terms for patients and caretakers. In 2017, Wittenberg et al. [95] reported the beneficial effect of nurses using “plain language” (i.e., sixth-grade level) in palliative cancer care, encouraging other health professionals to follow suit.

Strengths and weaknesses of the study

The origin of this study, commissioned by the hospital’s statutory Patients’ Rights Commission, designed by an interdisciplinary research team, with interviews conducted by a patients’ rights representative, lent bottom-up authenticity to the project. It provided a reflexive framework to interpret subjective qualitative data. This context made it possible, in particular, for nurses to be heard by a researcher outside the organisation chart, seen as a first-ever vector of their experience. The focus on specific discourse as a research tool is also original in a French PC context.

A potential shortcoming was, despite purposive sampling for a maximal dataset, being unable to interview all 28 nurses working in the two services. It was not possible to obtain an explanation for their non-participation.

Limitations to the study may be, first, given its dual focus (what nurses say and how they say it), striking a balance between the two aims to satisfy readers more interested in one aspect or the other. Similarly, for results where metaphor plays a minimal role, raising the metaphorical question of “speaking PC as a foreign language” may not seem appropriate. However, having defined this figure of speech as thinking about a difficult experience in terms of a more accessible one, particularly in the context of cancer and death, we feel that this image captures the innate challenge posed by our study. A further limitation may be the fact that all the nurses who volunteered were women (although among the seven absent, there were no male nurses): while women represent nearly 87% of all nurses in France [96], a mixed sample might have provided other perspectives. Finally, as for other studies of health professionals’ perspectives, our sample reflects specific actors involved in a specific context — one French comprehensive care cancer centre with a certain number of PC designated identified beds, but no PC unit. However, the dual nature of the nurses’ working environment, including both palliative and curative care for patients whose disease tends towards chronicity, may be extrapolated to other care environments and contexts. The fact that international literature raises similar challenges in PC makes it feasible that our findings may well cross borders.

Conclusion

Nurses in a French comprehensive cancer centre see patient-centred PC as a continuous journey rather than a synonym of EOL care and death. Framing their pivotal caregiving role, this perspective allows them to feel reasonably competent in discussing PC with patients and families. However, nurses see oncologists as diverging on the “p-word”, keeping them from fulfilling their caregiving mission and resulting in poorly implemented patient-centred palliative and end-of-life care. Shifting paradigms is challenging in any culture, but if oncologists and institutions could hear nurses’ words, opening their minds to PC and learning to speak its language, a first step would be taken to turn a sometimes vicious circle into a virtuous one for both patient and professional well-being. Future research should replicate this study among junior doctors and young permanent physicians, focusing on the discourse they use as an integral part of the message itself.