Death denial: obstacle or instrument for palliative care? An analysis of clinical literature


  • Camilla Zimmermann

    1. Division of Medical Oncology and Haematology, University of Toronto and Department of Psychosocial Oncology and Palliative Care, Princess Margaret Hospital, University Health Network.
    Search for more papers by this author

Address for correspondence: Camilla Zimmermann, MD, MSc, FRCPC, 610 University Avenue, Suite 16-744, Toronto, Ontario, Canada, M5G 2M9


As a society and as individuals, we have come to recognize ourselves as ‘death-denying’, a self-characterisation particularly prominent in palliative care discourse and practice. As part of a larger project examining death attitudes in the palliative care setting, a Medline search (1971 to 2001) was performed combining the text words ‘deny’ and ‘denial’ with the subject headings ‘terminal care’, ‘palliative care’ and ‘hospice care’. The 30 articles were analysed using a constant comparison technique and emerging themes regarding the meaning and usage of the words deny and denial were identified. This paper examines the theme of denial as an obstacle to palliative care. In the articles, denial was described as an impediment to open discussion of dying, dying at home, stopping ‘futile’ treatments, advance care planning and control of symptoms. I suggest that these components of care together constitute what has come to be perceived as a correct ‘way to die’. Indeed, the very conceptualisation of denial as an obstacle to these components of care has been integral to building and sustaining the ‘way to die’ itself. The personal struggle with mortality has become an important instrument in the public problem of managing the dying process.


The contemplation of death is for most of us inherently disconcerting. It could therefore be construed as natural or at least functional to repress such an unpleasant reality both on an individual and on a societal level (Freud 1948, Blauner 1966). However, while discussion of the terror of death can be traced to antiquity (Aristotle 1115a6-29), it is only in the past four decades that Western societies have come to regard themselves as inherently ‘death-denying’ (Aries 1974). Although this characterisation has generated considerable debate in the sociological literature surrounding whether we remain a death-denying society and indeed whether we ever were (Kellehear 1984, Walter 1991, Seale 1998), this debate has largely been ignored in the clinical medical literature (Callahan 2000), which continues to perpetuate the ‘denial of death thesis’ (Zimmermann and Rodin 2004). Moreover, with few exceptions (Armstrong 1987), there has been little sociological discussion of why death denial has become such a prominent topic and how the term denial is used in the clinical setting. Discourse analysis provides a method with which these questions can be systematically examined.

Potter and Wetherell (1987) emphasise that ‘there is no method to discourse analysis in the way that we traditionally think of an experimental method or content analysis method. What we have is a broad theoretical framework concerning the nature of discourse and its role in social life, along with a set of suggestions about how others can be convinced findings are genuine’ (1987: 175, emphasis in the original). Thus, discourse analysis is a method that is tightly interwoven with theory, and the theoretical stance behind the methodology will have a direct impact on the resultant analysis. It is partly for this reason that there have developed, in the last few decades, a number of different forms of discourse analysis that are diverse in theory and method. These have been in various disciplines including psychology (Potter and Wetherell 1987, Parker 1992), sociology (Lupton 1993, 1994), linguistics (Fairclough 1990) and media studies (van Dijk 1991). Common to all of these forms of discourse analysis is a recognition that ‘an integral and intertwined relationship exists between discourses – the way we speak or visually represent phenomena – and practices – the actions and activities surrounding these phenomena’ (Lupton 2003: 20).

My own theoretical perspective relies on Foucault (1973, 1979, 1990) for whom one goal was to determine ‘the conditions of possibility of medical experience in modern times’ and ‘disentangle the conditions of its history from the density of discourse’ (1973: xix). In particular I draw on Foucault's later work which poses what he termed ‘genealogical’ questions: ‘How are these discourses used? What role do they play in society?’ (Dreyfus and Rabinow 1983: xxv). These questions implicitly involve an analysis of power, which, from a Foucauldian perspective, is ‘not embodied in individuals, social groups or institutions, even though it is clearly played out in institutional contexts, . . . but operates through norms and technologies which shape the body and the mind’ (Annandale 1998: 37–8). In my research, one form of normalising power manifests itself in dominant discourses on attitudes toward death and dying. My intent is to question the hidden value systems in the common practice of calling patients ‘in denial’ of death and relate this discourse to the historical possibilities of its existence.

According to Parker (1992) ‘the advantage of discourse analysis is that it reframes the object . . . and allows us to treat it not as the truth, but as one ‘truth’ held in place by language and power’ (1992: 22). Whether or not denial is ‘really’ standing in the way of proper palliative care will not be the focus of this paper. Rather, I ask why this discourse should exist in the first place: how death denial came to be seen as a problem and how this relates to the situation of palliative care in the contemporary healthcare environment.

The study

This study is part of a larger project examining death attitudes in the palliative care setting. The study methods have been previously described (Zimmermann 2004). A Medline search (1971–2001) was performed combining the text words ‘deny’ and ‘denial’ with the subject headings ‘terminal care’, ‘palliative care’ and ‘hospice care’, and was restricted to articles in clinical Western journals discussing death denial in adults. The 57 articles initially retrieved were narrowed down to 30 articles, for which the references and countries of authorship and publication are listed chronologically in the Appendix. The articles appeared in palliative care journals (13/30), general medical journals (6/30), nursing journals (5/30) and psychiatry journals (6/30). The majority (20/30) were either written by US authors or published in US journals. The countries of publication included the US (19/30), the UK (7/30) and Canada (4/30). Seventeen of 30 authors were from the US and the remaining authors were from Canada (5/30), UK (4/30), France (1/30 – published in a Canadian journal), Australia, (1/30 – published in a US journal), Japan (1/30 – published in a US journal) and Switzerland (1/30 – published in a UK journal) (see Appendix).

The articles retrieved were closely read and coded according to the procedure described by Potter and Wetherell (1987). ‘Bodies of instances’ (1987: 167) relating to the meaning and usage of the term denial were collected together and categorised by photocopying all pages containing relevant instances and placing them in a separate file. This file then served as the basis for more detailed analysis. The analysis itself followed Parker's (1992) criteria for distinguishing discourses. Journal articles were closely read and the ‘connotations, allusions and implications which the texts evoke’ were explored. Intertextual linkages of meaning were analysed in terms of how discourses ‘embed, entail and presuppose other discourses’, both in the interplay of texts with each other and in relation to broader systems of knowledge. In the course of the analysis, a discourse that was manifest as a ‘coherent system of meanings’ (Parker 1992: 10) on denial of dying was elaborated.

Two dominant themes were elucidated: denial as a psychological coping mechanism and denial as an obstacle to palliative care. The first theme has already been presented in detail (Zimmermann 2004). There were three dominant subthemes: denial as an unconscious defence mechanism, denial as healthy and denial as temporary. Denial was seen to be an unconscious psychological state that was normal and healthy when terminal illness was first diagnosed, but became ‘maladaptive’ and ‘pathological’ when it persisted beyond a certain time frame. I suggested that the definition of death denial as an individual psychological coping mechanism may allow for the labelling of patients who resist participation in the planning of their death as being in denial. In this paper I will demonstrate how the label of denial is put to use by presenting the theme of denial as an obstacle to palliative care. Denial was perceived to be standing in the way of several components of palliative care: (1) open discussion of dying, (2) dying at home, (3) advance care planning, (4) symptom management and (5) stopping ‘futile’ treatments. I present each sub-theme below with specific references to the texts from which it was derived. In the Discussion, I suggest that these components of care form what may be considered a proper ‘way to die’, which in turn represents a managed death. Rather than being an obstacle to palliative care, death denial – and indeed its very conceptualisation as an obstacle – may be seen as a construction that is instrumental in upholding the ‘way to die’ and efficiently manage the dying process.

Denial as an obstacle to open discussion of dying

Denial was described as an obstacle to the open discussion of death and dying, which in turn was perceived as being central to the practice of palliative care. The following passages attest to the perceived importance of open communication for palliative care:

Helping a patient towards a fuller awareness of and adjustment to a fatal prognosis is the beginning of a communication process which is in itself an integral part of palliative care (Claxton 1993: 206).

Patients and families have to engage in decision-making over a wide range of issues in palliative care from what treatments they will accept (or refuse), pain management, whether to participate in clinical trials, to where they chose to die. At all points along this decision-making chain the provision of useful information and the facilitation of effective communication are essential to creating satisfying outcomes (McGrath et al. 1999: 29).

It was not clear from the latter paper what represented a ‘satisfying outcome’; rather, open discussion and the engagement of patients in ‘decision-making’ was in and of itself described by nurses as ‘positive and satisfying’ with patients who were willing openly to discuss their death being described as ‘at peace’ or even ‘quite happy about the situation’ (McGrath et al. 1999: 28). While the ability to convey information to the dying was described as ‘useful’, denial was explicitly portrayed as an impediment to the communication of that information. Denial was seen to contribute to ‘interference in generating a useful discussion on dying’ between patients and nurses and was referred to as one of several ‘blocks to open communication’ that nurses could be trained to overcome (McGrath et al. 1999: 26).

The same theme is seen in the report of a qualitative study examining ‘how caregivers perceive their communication with dying family members’ (Beach 1995: 13). Instances where death was not discussed among patients and family caregivers were referred to as ‘dysfunctional communication patterns’, which deprived the patient and family of a chance to openly ‘talk out’ anxieties and ‘realistically plan for the future’ (Beach 1995: 14). In this article, the very existence of lack of communication appears to have been equated with denial. For example, under the heading ‘Caregiver denial via communicative avoidance strategies’, the following passage was cited:

Spousal Caregiver:  You know, what is there to say? I mean, when it's a fatal disease, there isn't much you can say. You'd like to, but, you can't get mopey, you know, because, I cried, but I decided that doesn't help.

Interviewer:     Do you wish he'd talk about it more?

Spousal Caregiver:  No. I don't, because I go to the doctor with him and I know what the doctor says. It would just be a rehash of what's been said, you know . . . As I said, the doctor's very open. There's no secrets, only the truth. Tells you what it's like, you know, and we go on from there (Beach 1995: 16–17).

Here the avoidance of communication was taken to stand on its own as evidence of denial on the part of the caregiver; it could be interpreted, however, as acceptance. The caregiver has accepted the physician's opinion regarding the illness; believes that any further talk would just be a ‘rehash of what's been said’ and does not see much point in further discussion. The fact that avoidance of communication was interpreted as evidence of caregiver denial may reflect a bias towards open communication in general, with denial serving as a label for any non-communicator.

Denial as dishonest

Not only was communication perceived as useful while denial was perceived as an obstacle, but there was as well a moral argument invoked against denial, equating it with ‘lying’. For example:

From the Middle Ages to the mid-19th century, the attitude toward death changed from something that was commonplace, ordinary, expected and accepted, to something that was shameful and forbidden. Lies began to surround the process of dying (Friel 1982: 767).

Resistance to talk was also associated with denial and optimism. Participants articulated the tension that exists between the desire to honestly inform patients and the patient's psychological mechanisms for coping with death (McGrath et al. 1999: 26).

One caregiver explains that she was brought up with strict moral values, including the: belief that lying by commission or omission is unacceptable. Although this woman can intellectually accept that denial is useful to her palliative clients, she feels morally that denial is a form of lying. It makes her uncomfortable and she feels that she should encourage the client to face the truth (Murray and Neilson 1994: 34).

Thus denial was seen to be not only a barrier to generating a useful discussion about dying but was also seen as encouraging a model of care that was not open and honest, which is contrasted with the desire for honest communication on the part of nurses. The link of societal and individual denial in the article by Friel (1982) was also present in other articles. For example: ‘Our society is death-denying, so it is not unexpected that patients and families are reluctant to acknowledge the finality of a terminal diagnosis’ (Burgess 1994: 48). The argument that the reluctance to acknowledge or tell a prognosis of a terminal nature constitutes a lie implicitly draws on the socio-historical claims of Aries (1974), Gorer (1980) and others that we are a death-denying society and that this is to our detriment. As I will discuss below, however, there have been several sociologists who have asserted that contemporary Western society either no longer is or never was death-denying (Parsons and Lidz 1967, Walter 1994, Seale 1998). Moreover, as Armstrong (1987: 651) has pointed out, the denial of death argument rests upon the assumption that ‘truth and silence are in opposition and that it is only since 1960 that it has been possible to speak the truth about death’. Instead of a replacement of silence by truth, there may have been a ‘reconfiguration of what could or could not be said’ (1987: 652), which, as I will argue, both stemmed from and contributed to the construction of a new way to die.

Denial as an obstacle to home death

Denial was also perceived to be an obstacle to the achievement of death at home rather than in the hospital (Hinton 1994a,b, Reese 2000). In one study patients were followed by serial interviews in the last months of life and the interviewer then rated variables including denial on a linear analogue scale from 1–9. It was concluded that:

Patients rated at the first interview as showing more realistic coping attitudes likewise more often stayed at home . . . Only two of the 22 who stayed at home initially denied or partly suppressed the prognosis . . . compared with 30 of the 55 subsequently admitted (χ2= 11.56; p < 0.001) (Hinton 1994b: 204).

In Reese's study, entitled ‘The role of primary caregiver denial in inpatient placement during home hospice care’, the investigators set out to test the hypotheses that both primary caregiver denial and patient denial would be positively related to admission of the patient to an inpatient setting and to death in an institution rather than at home. Denial was measured by a standardised scale, which was completed twice by the hospice nurse, first regarding the patient and then regarding the primary caregiver, to arrive at denial scores for each of them:

This study ( . . . ) revealed that primary caregivers in denial of the patient's terminality were more likely to place hospice patients in inpatient treatment. Moreover, patients placed in inpatient settings were more likely to die there, rather than at home as planned. These findings suggest an impact of primary caregiver denial upon patient self-determination, and indicate the importance of addressing denial in counselling with primary caregivers of terminally ill patients (Reese 2000: 15).

While there was no relationship found between patient denial and inpatient placement, family denial was found to be an obstacle to care at home and eventual home death and this was framed as an intrusion upon patient autonomy. However, patients were not interviewed to determine whether or not they preferred to remain at home. Rather it was stated a priori in the introduction:

Hospice patients have expressed the desire for palliative care in the home and to die in the home. Their primary caregivers have made a commitment to uphold these wishes. Primary caregivers in denial, however, may violate the treatment plan and seek life-prolonging institutional care rather than palliative care in the home (Reese 2000: 20).

There is in this article a preferred treatment plan: home hospice care and death at home rather than in the hospital. Denial on the part of the family member was seen to result in the ‘placement’ of patients in inpatient treatment for ‘life-prolonging care’, ‘violating’ this plan of home-based care. The assumptions that patients always desire home death and that deaths in hospital violate their rights run counter to results found in other studies. For example, patients’ preferences for the place of death may change as they approach death (Townsend et al. 1990) and hospice patients may have many reasons, both practical and existential, for desiring the opportunity to die in an inpatient setting (Lawton 1998, Fried et al. 1999, Thomas, Morris and Clark 2004). However, whether or not they represent the truth, these assumptions, in and of themselves, contribute to the construction of a certain way to die, which denial is perceived to be obstructing.

Denial as an obstacle to advance-care planning

Denial was seen to present a barrier to advance-care planning, both in the sense of generally impeding planning for death and in the more structured sense of filling out advance directive papers. While it was acknowledged that denial can have ‘adaptive benefits’ and can be used ‘temporarily to cope until the reality of one's situation can be acknowledged and accepted’ (Reese 2000: 17) it was also seen to be ‘dysfunctional’ (Reese 2000: 18, Beach 1995: 14, Davidhizar 1998: 45), ‘maladaptive’ (Burgess 1994: 46), to ‘have disadvantages’ (Murray and Neilson 1994: 34) or even to ‘threaten harm’ (Goetzke 1995: 20) when it went ‘beyond the function of a temporary haven to making inappropriate treatment decisions, avoiding preparations for death, or avoiding responsibility’. (Reese 2000: 18). The use of the term responsibility is interesting. The patient and family are expected at a certain point in the course of disease to accept and prepare for death in a responsible fashion. This is also stressed in the following articles:

While denial can be an healthy adaptive response and beneficial to the patient, caregivers also know that denial has disadvantages. An unwillingness to accept that one may die, for example, can prevent the palliative patient from making a will or putting personal business in order (Murray and Neilson 1994: 34).

People using denial may become dysfunctional when a crisis occurs that recalls buried feelings they cannot deny. To cope with the crisis, there may be other reactions and inappropriate behaviours. While denial can preserve hope, it becomes dysfunctional when the person needs to make decisions about real situations (Davidhizar 1998: 45).

There is intertextuality here with the theme of denial as ‘temporary’ (Zimmermann 2004). Denial early in the course of disease is ‘functional’ but becomes ‘pathological’ if sustained to the point where ‘end-of-life decision-making’ (meaning preparation and planning for death) is affected. Otherwise stated, it has become the responsibility of patient and family to engage in formal planning for death, and out of this imperative for death planning a discourse of denial has emerged, labelling those who do not participate.

Denial as an obstacle to making advance directives

There were two articles in which advance directives were explicitly discussed. One, entitled ‘Death, denial and defeat: older patients and advance directives’, was a phenomenological study of 17 people over 65 years old who shared their ‘reasons for not having formalized their end-of-life decisions’ (Winland-Brown 1998: 36). Denial was identified as a prominent theme, and elaborated on as follows:

To deny thinking about unpleasant topics leads to inaction, which is in fact an action. Ignorance of an issue does not make it disappear (Winland-Brown 1998: 38).

Thus, the action of preparing for death and the role of denial in obstructing death preparation is again emphasised. Another article entitled ‘Denying the dying. Advance directives and dialysis discontinuation’ states:

Despite the enthusiastic impression in the literature that most patients wish to declare their preferences and complete advance directives, it has been our impression, to the contrary, that important and underappreciated psychosocial barriers, such as denial and avoidance, commonly prevent people from formulating and expressing their wishes (Cohen et al. 1997: 27).

The article goes on to document the results of a structured interview that was administered to a sample of patients on maintenance dialysis and their attending physicians to obtain documentation of their end-of-life preferences. It was found that:

A majority of the patients reported that they had neither thought of stopping dialysis nor had they discussed stopping dialysis with their nephrologist or family members . . . Of the 121 patients, only 7 (6%) had completed advance directives (health care proxies) (Cohen et al. 1997: 29).

Once again, denial was blamed, though this time it was on the part of both patients and treatment staff:

In most cases, the dialysis patients and their treatment team staff were preoccupied with the struggles of daily life and had avoided or denied considerations of terminal illness and death (Cohen et al. 1997: 27).

The fact that so few patients had completed advance directives was described as illustrating a ‘denial-like coping style’ (1997: 30) and the article closes with a review of the psychiatric literature on death and denial, subtitled ‘Psychiatry, denial and death’.

Once more, denial is construed as a pre-existing psychological condition, which in this paper acts as a block to filling out advance directive papers. There may, however, be other reasons why advance directives are not completed. In a recent qualitative study, for example, it was found that patients may not find advance directives helpful or desirable, with one concern being that care preferences may change (Seymour et al. 2004). Nevertheless, the existence of the imperative to fill out these documents may contribute to the construction of denial in the healthcare setting, with those not participating labelled ‘in denial’.

Denial as an obstacle to symptom management

In addition to acting as an obstacle for communication and advance planning, denial was also held responsible for impeding the management of physical and psychological symptoms. Intertextuality with denial as temporary and persistent deniers as ‘noncompliant’ was again apparent. Consider the following abstracts:

Several studies have noted inappropriate treatment or care, noncompliance with treatment plans, and/or poor health care outcomes for patients resulting from family denial . . . (Reese 2000: 18).

Of course, there are situations where denial can cause harm. Patients or families may incur astronomical costs in search of a cure or a better prognosis. They’ll ignore pain because it may mean the disease is getting worse. They’ll lose valuable time that could be spent on unfinished business – saying good-byes or resolving rifts (Goetzke 1995: 20).

Another article discussed two case reports of patients in whom denial was held partially responsible for the choice of alternative therapy over conventional therapy at a time when the latter may have been curative (Jenkins et al. 1998). When the disease advanced and palliative care was involved, denial on the part of patient or family stood in the way of treatment of pain in one patient and stent placement for acute obstructive renal failure in another.

One article describes in depth a situation where denial interfered with both communication and arriving at a correct diagnosis (Vigano et al. 1995). The patient presented was a 60-year-old man with terminal prostate cancer who had been complaining of increasing physical pain ‘without any evidence of significant disease progression, opioid toxicity or previous history of drug addiction’ (1995: 168). Attempts to increase his opioid dose resulted in drowsiness, with no relief in pain:

The lack of activity relative to functional assessment, reluctance to leave the room or bed, and a decreased level of communication with the family and team suggested that psychological disturbances may be the major contributors to his pain. The team (physician, nurse, social worker, pastoral carer and occupational therapist) was unsuccessful in identifying any potential emotional factors affecting his physical suffering because of the patient's persistent denial, as reported by a self-report visual analogue scale (VAS) for depression. Furthermore, the patient had difficulty communicating because of a language barrier (Vigano et al. 1995: 168).

Although the passage in italics is somewhat confusing (is it denial or depression that is being diagnosed by the VAS?), the sequence of events becomes clear further on in the article. The patient was in pain and increases in the dose of opiates resulted in drowsiness with no relief in pain. The treating team suspected that the patient was depressed and that this was a case of somatisation; a VAS scale for depression was, however, negative and no other test for depression was performed. The negative result of the VAS scale was assumed to be due to ‘the patient's persistent denial’ and furthermore, due to difficulty communicating because of a language barrier. The possibility that the patient might not have been depressed was apparently not considered. The negative VAS for depression, rather than being interpreted at face value as a negative VAS for depression, became a ‘report’ of denial.

After treating the patient with methylphenidate (a stimulant), the success in treating the presumed depression is no more evident from the VAS than the depression was in the first place. However, denial is again invoked:

MP (methylphenidate) was used with the intent of both counteracting opioid-induced sedation and ameliorating psychological distress. The subsequent improvement in pain can be related to a positive action of MP on mood, even if this was not reflected in the VAS for depression because of the patient's persistent denial (Vigano et al. 1995: 169).

Once again, the reversal of the patient's depression is assumed to be present, but masked by the patient's presumed denial. There can be no proving or disproving of this assertion of the presence of denial. Somewhat paradoxically, however, the very conception of denial as a diagnostic obstacle may have allowed it to serve an important function, allowing for the successful diagnosis and treatment of an otherwise elusive depression.

Denial as an obstacle to stopping ‘futile’ treatments

I have argued, treating denial as a construction, that denial may be used as a label for those patients and families who resist the dominant discourses on the right way to die. In the themes presented thus far, it has generally been patients and their family caregivers who were in denial. In the theme of denial as an obstacle to stopping life-sustaining treatment, the finger of denial is pointed at physicians, the healthcare system and Western society in general. Consider the following excerpts:

This fear and denial within our society has led us to ignore or minimize the huge and rapidly growing burden of suffering imposed by terminal illness. Our fear of death has dictated our society's massive expenditure on high-technology resources in the last months of life in frantic, futile attempts to prolong life (Scott 1994: 35).

Dying patients and their families repeatedly express their need for supports based on compassion and caring, yet healthcare efforts focus on often ineffective technological interventions and procedures. . . . Physician denial of death and dying has a significant impact on clinical decision making and misleads healthcare administrators about priorities (Super et al. 1996: 50).

Death is too often denied in American culture, and futile efforts to rescue the dying from its grasp are costly and can sap resources from basic healthcare that should be made available to all people. Many times not everything technologically possible should be done and dying should be allowed to occur (Post 1991: 36).

Physicians who for personal and neurotic reasons need to deny their own mortality will get caught in a counterphobic battle against death, always insisting on one more therapeutic effort. Such measures do not serve to prolong life but prolong the act of dying (Friel 1982: 770).

All of these articles are based on the stark opposition of curative care and technology against compassion and caring. It is stressed that the technological interventions are ‘often ineffective’ resulting in ‘frantic’, ‘futile’ and ‘massively expensive’ attempts to prolong life. As with the sub-theme of denial as an obstacle to open discussion of dying, this sub-theme of denial as an obstacle to stopping futile treatments has obvious intertextuality with the socio-historical claims that we are a death-denying society (Gorer 1980, Kubler-Ross 1969, Aries 1974). Moreover, the fact that the accusation of death-denial is directed explicitly at the healthcare system and its workers may reflect a frustration with the lack of integration of the principles of palliative care into general medical practice, a point on which I will elaborate in the discussion.

Denial as an obstacle to stopping intravenous hydration

In one paper, the issue of stopping intravenous hydration was specifically addressed (Morita et al. 1999). In this study, contributing factors to decision-making regarding rehydration for terminally ill cancer patients were examined using a multi-step method. First, a discussion took place between hospice inpatients who ‘could not maintain satisfactory oral intake’ (1999: 510), family members and physicians. The latter ‘informed patients and their family members of the patient's current condition and the risk versus benefits of rehydration therapy’ (1999: 510). Second, physicians ‘either did or did not recommend forced rehydration from a medical standpoint alone’, and made recommendations against rehydration in 78 per cent of patients. Third, physicians performed a structured interview where they explored concerns of patients and their family members with regard to rehydration, and recorded ‘possible contributing factors to decision-making’ (1999: 510), of which denial was one. After multiple regression analysis, only ‘patient denial’ emerged as an independent determinant for receiving hydration. Denial was defined as:

the individual asserting that one did not have the physical disorder or behaviour in a way that indicated that one minimized the significance of the disorder in spite of obvious physical manifestations or the patient's having been adequately informed (Morita et al. 1999: 510).

This definition was taken directly from a publication entitled ‘Maladaptive denial of physical illness: A proposal for DSM-IV’ (Strauss, Spitzer and Muskin 1990). There may, however, be an assumption on the part of the physician that a patient who insists on rehydration despite recommendations to the contrary is minimising the significance of his or her illness. Again, rather than being a ‘contributing factor to decision making’, denial could be an outcome or construction of the decision-making process itself. After having discussions with patients and giving recommendations, those patients who resisted these recommendations and desired rehydration could be labelled ‘in denial’. Indeed, the act of involving family members and patients in decision-making while retaining a bias about the outcome of this discussion may lead to the construction of denial in the healthcare setting.


In this paper I have examined the usage of the term denial of death in articles in the palliative care literature dating from 1971–2001, a period during which the palliative care movement was developing increasing momentum. Denial, in these articles, was portrayed as presenting an obstacle or impediment to several elements of palliative care as it is practised in contemporary society. The specific elements to which denial was seen to present an obstacle – open discussion of dying, death at home, stopping ‘aggressive’ treatments, advance care planning, symptom management – together constitute a certain way to die that is typically considered to represent quality palliative or end-of-life care (SUPPORT Principal Investigators 1995, Emanuel and Emanuel 1998). One can then ask, as I have alluded to in the presentation of the results: Is denial a psychological obstacle that contributes to decision-making and impedes palliative care (as it has been portrayed in these articles) or can it be regarded as a construction of the decision-making process itself and of the manner of care delivery? In order to answer this question it is necessary to examine where the label of death denial came from, at what time in history it began to be perceived as an individual and social problem and whether and how this problematisation was related to the development of the field of palliative care. While a detailed discussion of the history and evolution of the use of the term denial is beyond the context of this paper, I will give a brief overview of the subject as it pertains to this discussion.

Sigmund Freud first described ‘defence mechanisms’ for the ego's struggle against unendurable realities (Freud 1953) and Anna Freud later delineated these defences in her book The Ego and the Mechanism of Defence (Freud 1948). For Freud, however, it was not death but sexuality that constituted the basic repression and it was not until the 1960s and 70s that a ‘great deal of collective bustle’ (Lofland 1978) arose over death and its denial in the modern industrialised world. Its psychological beginnings can be traced to the first scientific symposium on ‘Death and behaviour’ organised in 1956 by Herman Feifel at the American Psychological Association (Feifel 1959). During the subsequent 40 years, there followed an explosion of literature on death and dying in the psychological (Kastenbaum and Aisenberg 1972, Becker 1973), sociological (Glaser and Strauss 1965) and psychiatric (Kubler-Ross 1969, Weisman 1972) literature, centred on the problem of death denial. While these authors focused on the individual, others took an interest in denial of death at the societal level and a ‘shift in prudery’ (Gorer 1980) was said to have taken place: ‘the more society was liberated from the Victorian constraints concerning sex, the more it rejected things having to do with death’ (Aries 1974: 93).

In the clinical literature, in particular, the discourse on death denial took the form of a moral imperative. We were told that while we used to ‘face death with equanimity’ (Kübler-Ross 1969: 16), our society now suffered from ‘such a fear and denial of death, it has to use defences which can only be destructive. Wars, riots, and increasing numbers of murders and other crimes may be indicators of our decreasing ability to face death with acceptance and dignity’ (Kübler Ross 1969: 14). It was an injustice to the dying that their death was not spoken about or planned for and the ‘conspiracy of silence’ was portrayed as a conspiracy of the living against the dying (Feifel 1959). It was the fault of a death-denying medical system that dying had become a ‘technological nightmare with no awakening’ (American Hospital Association 1996). Against this was contrasted palliative care, which was described as ‘a powerful antidote to this pattern, representing both a return to older traditions of care and a fresh, less troubled response to death’ (Callahan 2000). Following from this moral discourse came a ‘right’ and a ‘wrong’ way to die, the right way being exemplified in the creation of the ideal of the ‘good death’ (Saclier 1976, Emanuel and Emanuel 1998, Smith 2000) or of ‘dying well’ (Byock 1997) and the wrong way represented by the demon of death denial. The implicit and at times explicit aim of the propagation of the right way to die was to ‘educate the community to a healthier view of death, even a planned death, as a part of life and a responsible conclusion to it’ (Saclier 1976: 3).

This discourse, which originated in the United States and quickly permeated the rest of Western society, was not without sociological critique. Already in the 1960s, Parsons and Lidz (1967) presented the alternative view that American society had an ‘orientation to death that is fundamentally not a “denial” but a mode of acceptance appropriate to our primary cultural patterns of activism’ (1967: 135). More recently, Seale maintained in a similar vein that institutional sequestration of the dying, often cited as an example of the denial of death, is in fact an ‘active facing and management of mortality based on realistic acceptance of death’ (Seale 1998: 101). And Kellehear (1984), in his systematised rebuttal of the denial of death thesis went so far as to say that ‘the sociological content in the death denying society view is minimal’ (1984: 713).

Walter (1994), on the other hand, argued that while death did ‘die’ or become taboo in the modern era, we have now entered a new ‘neomodern’ age where there has been a ‘revival’ of death due to discourses on palliative care and euthanasia. Seale (1998) expands on the theme of revivalism in his text on the sociology of dying and bereavement:

Dying people . . . are offered the opportunity, by writing themselves into a revivalist cultural script, to participate in the construction of the caring team as joint adventurers . . . These hoped-for journeys, through stages of denial to eventual acceptance for example, represent for committed onlookers a ritual transformation of death into life, hope or ‘fertility’ . . . (1998: 121).

Such dying patients who embrace revivalism may be exemplified in Giddens's conception of the individual in late modernity, where ‘self-identity becomes a reflexively organised endeavour’ and ‘reflexively organised life-planning, which normally presumes consideration of risks as filtered through contact with expert knowledge, becomes a central feature of the structuring of self-identity’ (Giddens 1991: 5). Seale also draws on the work of Rose (1989) commenting that psychological discourse, including revivalist discourse on death awareness and acceptance, both acts as a pervasive script in late-modern Anglo-American society and involves a ‘deeply rooted cultural obligation to be free’ (Seale 1998: 120). However, Seale provides evidence for a ‘rival script on unaware dying, which is given less legitimacy by official or professional authority’ (1998: 173). It is patients who adhere to this rival script, resisting dominant revivalist pressures to discuss their dying and participate in end-of-life planning, who may be subject to negative labelling with the label ‘denial’ (Zimmermann 2004).

In this paper I am approaching the analysis of the theme of denial as an obstacle to palliative care from an angle similar to Armstrong's (1987). Rather than arguing for or against the death-denying nature of our society, Armstrong critiques the opposition of truth and silence in the discourse on death denial from a Foucauldian standpoint, concluding that ‘discourse, in its inimical link with power, can never lie outside itself’ (1987: 656). Similarly my concern in this paper is not with whether we are in reality death-denying or death-accepting on a social or individual level. Rather, I am interested in the usage of the term denial: whether and how the term is used in the palliative care literature and what possible connections this may have to broader discourses in healthcare and the management of dying people. Despite the revivalist scripts described above, it remains a truism to say that contemporary Western societies and individuals are death-denying, a view which continues to be propagated in the hospice (Connor 1998) and medical literature (McHue 1995, Emanuel and Emanuel 1998, Callahan 2000). Moreover, the fear and denial of death are staged as specifically modern and problematic, despite the fact that these are sentiments that have probably always existed (Elias 1982). This discourse is illustrated in the articles in this study: the theme of denial as an obstacle to palliative care reflects the overall discourse on death denial as a problem for society in general.

In order to examine why this problematisation of death denial occurred at this particular time in history and continues to this day, it is helpful to refer to Foucault's analysis of sexuality (Foucault 1990). There are many parallels between the manner in which death and sexuality have been treated in modern society. Both sexuality and death were problematised in psychoanalytic terms as representing basic repressions; both of these repressions were considered harmful for modern society and there was a concern in both cases with liberation from a perceived constraint. In the case of sexuality it was sexual pleasure that was being liberated; in the case of dying, it was liberation from a prolonged, technological death. In both cases coming face to face with our sexuality or mortality was seen to be the cure, which was accomplished by bringing these taboo topics into open discourse.

Foucault describes the incitement to speak about sex as ‘a public interest’ which arose out of the phenomenon of ‘population’ as an economic and political problem in the 18th century: ‘population as wealth, population as manpower of labour capacity, population balanced between its own growth and the resources it commanded’. And at the ‘heart of this economic and political problem was sex’. It was necessary to analyse and take control of birth rates, legitimate and illegitimate births, marriage, methods of contraception, the ‘precocity and frequency of sexual relations’:

One had to speak of sex; one had to speak publicly and in a manner that was not determined by the division between licit and illicit [ . . . ]: one had to speak of it as a thing to be not simply condemned or tolerated but managed, inserted into systems of utility, regulated for the greater good of all, made to function according to an optimum (1990: 24, emphasis in the original).

Two centuries later, with the ‘character’ of death (Marshall 1980) having changed from sudden and unexpected to prolonged and preventable, it has become important to assert control not only over life's beginning but also over its end. The dying process has been prolonged, not only by the ability to delay death by life-sustaining measures but also by earlier diagnosis of terminal disease due to advances in diagnostic technology. And with the prolongation of the dying process, there is a public interest in managing this large number of patients now classified as ‘dying’. In ancient times when patients died in their beds at home with little intervention, there was no reason for a public interest in whether the dying were in denial or not – though on the individual psychological level this denial might have been equally present. It is only with the transfer of the management of death from the private to the public realm that the interest in whether or not death was spoken about became prominent (Armstrong 1987). The necessity in modern medical care of the transfer of patients from acute care wards to palliative units and hospices (which necessitate the acknowledgement of a short prognosis), the discontinuation of nutrition and hydration, the signing of ‘do-not resuscitate’ orders and completion of advance directives all result in a concomitant imperative to acknowledge and speak about dying.

As I have briefly outlined above, denial is a term that began to be promulgated in the United States in the 1950s and 1960s, concomitant with the development of the hospice and palliative care movements (Zimmermann and Rodin 2004). Although the purpose of this article is not to compare the usage of the term denial in journals of different countries of authorship (which would require a more extensive search also including languages other than English), it is noteworthy that the majority of articles retrieved by the denial text word search and analysed in this study were written by US authors and/or published in US journals. One could therefore argue that the discourse on death denial that I have described predominantly reflects an American perspective. Unlike the situation in the UK, where palliative care is integrated into mainstream medicine, end-of-life care in the United States is dichotomised into a hospice and a non-hospice alternative, hospice here referring not to a place but to a philosophy of care. In order to receive palliative care under the Medical Hospice Benefit, which came into place in the US in 1982, patients must have a prognosis (determined by their physician) of six months or less and forego usual Medicare coverage for non-symptom-oriented services related to treatment of their terminal illness, including palliative chemotherapy and participation in clinical trials (Browner and Carducci 2005). If they would like to go on receiving cancer-directed treatment, they may not receive the benefits of hospice care, including expertise in pain and symptom management, social services and counselling (Casarett et al. 2002). This dichotomised method of care, together with the high cost of non-hospice care compared to hospice care, may provide a strong incentive for the transfer of patients to hospice services and the labelling of resistant patients as in denial.

The articles retrieved in this search were not only from palliative care journals, but also from nursing journals, general medical journals and psychiatry journals, illustrating the incorporation of palliative care discourse into the broader clinical literature. Nevertheless, the incorporation of hospice and palliative care ideals into clinical practice is far from complete. Most patients still die in hospitals rather than at home, and do not receive palliative care by in-hospital palliative care consultation teams. In the US, while hospice is the formalised way of end-of-life care, only approximately 25 per cent of dying Americans use it (Lynn 2001) and, although physicians state that patients ideally should receive hospice care for three months before death, most patients are enrolled in hospice care less than one month before they die (Lamont and Christakis 2002). The apparent difficulty of incorporating palliative care ideals into mainstream practice serves as fodder for the death-denial argument. This is illustrated in the sub-theme of denial as an obstacle to stopping futile treatments, where the continuation of a senseless battle against death in hospitals is blamed on death denial on the part of healthcare workers and hospital administrators.

Thus, denial is conceived as an obstacle to a proper way to die that has been endorsed in the palliative care and general medical literature. However, the very conceptualisation of denial as an obstacle may be integral to building and sustaining the way to die itself. There is in contemporary Western society a public interest in the control of the dying process, which is linked to a change in our way of dying, and this control may be achieved by bringing death into discourse, by increased talk about death and dying. In keeping with Foucault's conception of power as disciplining rather than coercive (Foucault 1979) patients, health professionals and ‘responsible’ members of society in general are all invited to participate in this discourse. Out of the necessity for managing death in modern society, a new narrative on death denial has emerged, which continues to be professed in the clinical literature despite sociological arguments to the contrary. The fear and denial of death – probably no more novel or distinctively modern than the pleasure and repression of sex – have been thrust into the centre of public and professional discussions on death and dying. Death denial has been erected as an obstacle by the very discourse that continues to propose to tear it down and remains in this discourse because it is an integral component of it. The personal struggle with mortality has become an important instrument in the public problem of managing the dying process.


I thank Ann Robertson and two anonymous referees for helpful comments on earlier versions of this paper.

References for articles recovered in search in chronological order*

*Marked according to US (1) UK (2) or Canadian (3) authorship (1st number) and journal publication (2nd number). For authorship from another country, the name is written in full.

  • 1Abrams, R. Denial and depression in the terminal cancer patient. A clue for management. Psychiatric Quarterly, 45, 3, 394–404, 1971. (author – 1) (journal – 1)
  • 2Vispo, R. Critique of ‘denial and depression’Psychiatric Quarterly, 45, 3, 405–9, 1971. (1) (1)
  • 3Silberfarb, P. and Levine, P. Psychosocial aspects of neoplastic disease III. Group support for the oncology nurse. General Hospital Psychiatry, 2, 3, 192–7, 1980. (1) (1)
  • 4Silberfarb, P. and Greer, S. Psychological concomitants of cancer: clinical aspects. American Journal of Psychotherapy, 36, 4, 470–8, 1982. (1) (1)
  • 5Friel, P. Death and dying. Annals of Internal Medicine, 97, 5, 767–71, 1982. (1) (1)
  • 6Marshall, J. The dying elderly patient. American Family Physician, 28, 5, 161–5, 1983. (1) (1)
  • 7Connor, S. Measurement of denial in the terminally ill: a critical review. Hospice Journal, 2, 4, 51–68, 1986. (1) (1)
  • 8MacKillop, W., Stewart, W., Ginsburg, A. and Stewart, S. Cancer patients’ perceptions of their disease and its treatment. British Journal of Cancer, 58, 3, 355–8, 1988. (2) (2)
  • 9De Hennezel, M. Denial and imminent death. Journal of Palliative Care, 5, 3, 27–31, 1989. (France) (3)
  • 10Post, S. American culture and euthanasia. The changing definition of a ‘good death’. Health Progress, 72, 10, 32–8, 1991. (1) (1)
  • 11Connor, S. Denial in terminal illness: to intervene or not to intervene. Hospice Journal, 8, 4, 1–15, 1992. (1) (1)
  • 12Claxton, J.W. Paving the way to acceptance. Psychological adaptation to death and dying in cancer. Professional Nurse, 8, 4, 206–11, 1993. (2) (2)
  • 13Heim, E., Augustiny, K., Schaffner, L. and Valach, L. Coping with breast cancer over time and situation. Journal of Psychosomatic Research, 37, 5, 523–42, 1993. (Switzerland) (2)
  • 14Stoll, B. Can unorthodox cancer therapy improve quality of life? Annals of Oncology, 4, 2, 121–3, 1993. (2) (2)
  • 15Hinton, J. Can home care maintain an acceptable quality of life for patients with terminal cancer and their relatives? Palliative Medicine, 8, 3, 183–96, 1994. (2) (2)
  • 16Hinton, J. Which patients with terminal cancer are admitted from home care? Palliative Medicine, 8, 3, 197–210, 1994. (2) (2)
  • 17Burgess, D. Denial and terminal illness. American Journal of Hospice and Palliative Care, 11, 2, 46–8, 1994. (1) (1)
  • 18Scott, J.F. More money for palliative care? The economics of denial. Journal of Palliative Care, 10, 3, 35–8, 1994. (3) (3)
  • 19Murray, M. and Neilson, L. Denial: coping or cop out? Canadian Nurse, 90, 2, 33–5, 1994. (3) (3)
  • 20Vigano, A., Watanabe, S. and Bruera, E. Methylphenidate for the management of somatization in terminal cancer patients. Journal of Pain and Symptom Management, 10, 2, 167–70, 1995. (3) (1)
  • 21Goetzke, E. When your patient is in denial. American Journal of Nursing, 95, 9, 18–21, 1995. (1) (1)
  • 22Beach, D. Caregiver discourse: perceptions of illness-related dialogue. Hospice Journal, 10, 3, 13–25, 1995. (1) (1)
  • 23Super, A. and Plutko, L.A. Danger Signs. Coalition points to causes and consequences of inadequate care of the dying. Health Progress, 77, 2, 50–54, 1996. (1) (1)
  • 24Cohen, L., McCue, J., Germain, M. and Woods, A. Denying the dying. Advance directives and dialysis discontinuation. Psychosomatics, 38, 1, 27–34, 1997. (1) (1)
  • 25Jenkins, C.A., Scrafe, A. and Bruera, E. Integration of palliative care with alternative medicine in patients who have refused curative cancer therapy: a report of two cases. Journal of Palliative Care, 14, 4, 55–9, 1998. (3) (3)
  • 26Davihizar, R. and Giger, J.N. Patients’ use of denial: coping with the unacceptable. Nursing Standard, 12, 43, 44–6, 1998. (1) (2)
  • 27Winland-Brown, J.E. Death, denial and defeat: Older patients and advance directives. Advanced Practice Nursing Quarterly, 4, 2, 36–40, 1998. (1) (1)
  • 28McGrath, P., Yates, P., Clinton, M. and Hart, G. ‘What should I say?’: Qualitative findings in palliative care nursing. Hospice Journal, 14, 2, 17–33, 1999. (Australia) (1)
  • 29Morita, T., Tsunoda, J., Inoue, S. and Chihara, S. Perceptions and decision-making on rehydration of terminally ill cancer patients and family member. American Journal of Hospice and Palliative Care, 16, 3, 509–16, 1999. (Japan) (1)
  • 30Reese, D.J. The role of primary caregiver denial in inpatient placement during home hospice care. Hospice Journal, 15, 1, 15–33, 2000. (1) (1)