Article Text

Download PDFPDF

Doctors’ perceptions of how resource limitations relate to futility in end-of-life decision making: a qualitative analysis
  1. Eliana Close1,
  2. Ben P White1,
  3. Lindy Willmott1,
  4. Cindy Gallois2,
  5. Malcolm Parker3,
  6. Nicholas Graves4,
  7. Sarah Winch3
  1. 1 Australian Centre for Health Law Research, Queensland University of Technology, Brisbane, Queensland, Australia
  2. 2 Faculty of Social and Behavioural Sciences, University of Queensland, Brisbane, Queensland, Australia
  3. 3 Faculty of Medicine, University of Queensland, Brisbane, Queensland, Australia
  4. 4 Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Queensland, Australia
  1. Correspondence to Eliana Close, Queensland University of Technology, Australian Centre for Health Law Research, Brisbane, QLD 4001, Australia; eliana.close{at}


Objective To increase knowledge of how doctors perceive futile treatments and scarcity of resources at the end of life. In particular, their perceptions about whether and how resource limitations influence end-of-life decision making. This study builds on previous work that found some doctors include resource limitations in their understanding of the concept of futility.

Setting Three tertiary hospitals in metropolitan Brisbane, Australia.

Design Qualitative study using in-depth, semistructured, face-to-face interviews. Ninety-six doctors were interviewed in 11 medical specialties. Transcripts of the interviews were analysed using thematic analysis.

Results Doctors’ perceptions of whether resource limitations were relevant to their practice varied, and doctors were more comfortable with explicit rather than implicit rationing. Several doctors incorporated resource limitations into their definition of futility. For some, availability of resources was one factor of many in assessing futility, secondary to patient considerations, but a few doctors indicated that the concept of futility concealed rationing. Doctors experienced moral distress due to the resource implications of providing futile treatment and the lack of administrative supports for bedside rationing.

Conclusions Doctors’ ability to distinguish between futility and rationing would be enhanced through regulatory support for explicit rationing and strategies to support doctors’ role in rationing at the bedside. Medical policies should address the distinction between resource limitations and futility to promote legitimacy in end-of-life decision making.

  • medical futility
  • resource allocation
  • health care rationing
  • clinical decision-making
  • withholding treatment

Statistics from

Request Permissions

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


A tension in end-of-life care is reconciling what is best for the individual patient with what is best for society as a whole. Advances in medicine allow life to be prolonged, frequently in poor health, raising concerns over whether scarce health resources are being used wisely.1 2 Yet, as illustrated by the recent case of Charlie Gard,3 at the individual level, a decision to withhold or withdraw life-sustaining treatment can be difficult, and patients (or their substitute decision makers) may want treatment doctors believe is inappropriate.4 There are two distinct, but related, ethical rationales that a doctor could rely on to refuse such treatment.3 5–7 The first is the doctor believes it is not in the patient’s interests, because it would cause harm or is unlikely to provide sufficient benefit (sometimes called ‘futile’ or ‘potentially inappropriate’8 treatment). The second, and more controversial, rationale is the doctor believes the expected efficacy of the treatment is too small to merit the expenditure of public resources (distributive justice).3 5–7 Although largely overshadowed by the decades of debates about the concept of futility, interest in the relationship between these two non-treatment rationales has been present in the literature for some time.6 7 9–11 One concern is that futility could act as a mask (consciously or unconsciously) for rationing,12 denying patients beneficial or sought-after treatments on the basis of cost to the system.

The idea that resources are taken into account at the bedside challenges the traditional view that doctors should solely be patient advocates, who leave allocation decisions to others.13 However, societal interests are an undeniable backdrop to doctors’ decisions, with many professional guidelines mandating doctors to act in the patient’s best interests while maintaining the efficient use of scarce health resources.14 Additionally, empirical evidence demonstrates that doctors ration at the bedside and are more likely to do so when treatments have a small expected benefit or a low chance of success, or when the patient has a poor quality of life or is over 85 years of age (circumstances that can underscore disputes).15 Doctors also have highly variable preferences for how to distribute scarce resources.16 Furthermore, in systems with privatised healthcare, doctors might also be influenced by personal profit. Doctors are therefore ‘dual agents’17 (or in privatised systems, ‘triple agents’) which, in the end-of-life context, underlies concerns that futility and rationing are insufficiently distinguished in practice.6 7 10 11

The distinction between a decision based on interests and one based on distributive justice is particularly important when navigating end-of-life discussions and disputes. The basis for the decision is important because it implies a different decision-making process; a decision based on patient interest requires patient input, while one based on resources requires a fair administrative process.5 6 ,18 However, for several reasons, futility and rationing are difficult to separate in practice. The provision of futile treatment has resource implications (both in cost to the health system1 and opportunity cost to other patients19), both concepts require value judgments about what constitutes patient benefit,6 9 and both grounds for non-treatment can apply in the same circumstances.3 6 For example, an expensive experimental treatment with a low probability of success could be both against a patient’s interests (because it is unlikely to work and may cause harm) and unjustified from a societal point of view (because it is not cost effective). While recent policy developments have been aimed at supporting discussions about futility and potentially inappropriate treatments,8 resources are inadequately addressed.6 The lack of explicit mechanisms to address resources in practice adds credence to the concern that, in some cases, rationing may be subsumed in the concept of futility.

Despite the importance of distinguishing futility and rationing, there is little empirical information exploring how doctors conceptualise the relationship between these two rationales. Three previous studies found some doctors include resources when defining futility, but what this meant remains unexplored.20–22 In particular, to what extent do futility judgments mask or incorporate resource concerns in clinical practice? This paper addresses this gap. Drawing on a larger project looking at how doctors conceptualise futile treatment,20 and the reasons they provide it,4 it reports on how doctors view the relationship between futility and resources, both conceptually and in their descriptions of their practice.



Doctors were recruited from three tertiary public hospitals in Brisbane, Australia, using purposive sampling in departments that treat patients at end of life or, in the case of medical administrators,i encounter end-of-life disputes.

Table 1

Additional examples of the relevance of resources to doctors’ current practice (theme 1)

The sample was developed in consultation with an experienced reference group of interested clinicians. Minimum recruitment targets in the target specialties were set to obtain diverse views within specialties and across the sample. Doctors volunteered to be interviewed following information from department heads and colleagues asking those interested to contact the research team directly.

Data collection

Participants provided signed informed consent and were individually interviewed in a private setting (typically the participant’s office) by EC, who has expertise in psychology and law. EC had no personal or professional relationship to the interviewees. Interviews were audio-recorded and lasted about an hour (range=0.5–2 hours). Using the convergent interviewing technique,23 areas of convergence and divergence in participants’ responses were probed until saturation of themes was achieved. The interviews followed a semistructured interview guide developed by all authors. EC piloted the interview guide with two participants, with feedback from CG. Consistent with the convergent interviewing technique, the interview guide was refined based on issues raised in the pilot and early interviews.

The interviews began with an open-ended question.23 Doctors were asked to recount a situation when ‘a person got treatment at the end of life you didn’t think they should have had’. As this study is part of a wider investigation into futile treatment at the end of life, the interviews asked doctors how they defined futile treatment, why they provided it, strategies they used to avoid providing it and the perceived impact of law and policy on their decisions. Doctors were given two resource-related prompts: ‘Some say resources are a factor in assessing whether or not to offer treatment that may be futile, what do you think…?’ and ‘Some say that by providing treatment that is futile (even when there is some justification) others miss out on beneficial treatment, what do you think…?’

Data transcription and analysis

Interviews were professionally transcribed verbatim, and participants were invited to amend their transcripts.

This study involves a secondary analysis of the interview data, looking specifically at perceptions of the relationship between futility and resource constraints. Using QSR International’s NVivo 11 qualitative data analysis software and a thematic analysis,24 EC analysed all text that mentioned resources. This included whether it came from resource-specific prompts, doctors’ definitions of futility or other references to resources elsewhere in the transcripts. From this first analysis, a series of themes and subthemes concerning resources was developed. EC then discussed and refined the initial codes with LW and BPW, examining both high-level coding structure and individual extracts. EC, LW and BPW then conducted a second round of analysis, looking at selected individual quotes and broader transcript passages to ensure consistency in approach. Inconsistencies and disagreements were resolved iteratively.


Between May and July 2013, 96 interviews were conducted at three participating hospitals, from a range of specialties: emergency (15); intensive care (12); oncology (10); palliative care (10); internal medicine (9); renal medicine (9); respiratory medicine (9); surgery (8); cardiology (5); geriatric medicine (5); and medical administration (4). Since we targeted doctors who had experience providing end-of-life care, most participants were established in their career (87 consultants, 9 registrars), with an overall average of 22 years’ experience working as a doctor. Their age ranged from 30 years to 72 years, with an average of 49 years.

The analysis produced the following themes: (1) perceptions of the relevance of resources to doctors’ current practice; (2) perceptions of the relationship between resources and the concept of ‘futility’; and (3) resource-related distress and proposed solutions.

Theme 1: perceptions of the relevance of resources to doctors’ current practice (table 1)

Whether resources are relevant

Doctors’ views on taking resources into account when deciding to limit life-sustaining treatment differed. Some (n=15) said they never took resources into account in end-of-life decisions; they were patient advocates, and resources were not relevant to their specialty or practice environment. Other doctors (n=25) mentioned general awareness of health system resourcing, but believed questions about distributive justice did not affect practice and required resolution by higher authorities. An intensive care unit (ICU) consultant remarked:

… [T]here’s an enormous amount of money and resources and time … not available to other people who would get more yield … but it is … harder at an individual …level to apply that. … I don’t feel myself that that’s a major thing that plays on my mind, or on any …individual patient… (#25, ICU consultant)

Many doctors (n=54) indicated they did take resources into account, viewing themselves as both a patient advocate and as a steward of scarce health resources:

… [I]f you’re responsible, you have to … take account of the fact that we don’t have indefinite resources … the old adage of I’m just focused on my patient and I’ll do whatever it takes for my patient … just doesn’t take account of the whole aspect. (#74, emergency consultant)

Within this group, 28 participants clarified resources were relevant to their decisions but not the prime consideration. One doctor noted:

… [C]osts are worth considering but they’re by no means the most important. So patient well being, patient comfort, are the prime criteria for deciding but you can’t ignore costs… (#77, respiratory medicine consultant)

Situations when doctors perceive resources as the main driver of decisions

Despite divergence of opinion about the role of resources in their practice generally, doctors provided similar accounts of when resources were a main driver of decisions, which some referred to as ‘rationing’ treatment. Limited availability and high cost were relevant factors, with perceived rationing occurring in decisions about admission to ICU, organ transplantation, dialysis and high-cost procedures such as implantable cardioverter defibrillators and ventricular assist devices.

Theme 2: perceptions of the relationship between resources and the concept of ‘futility’ (table 2)

Of the 54 doctors identified in theme 1 who thought resources were relevant to end-of-life decisions, nearly half (n=23) addressed explicitly the relationship of resources to the concept of futility. Two doctors emphasised that resource allocation must be distinguished from futility, with one stating:

Table 2

Additional examples of the relationship between resources and the concept of ‘futility’ (theme 2)

You’ve got to divorce withholding therapy from resource allocation. … although they’ve got to be separated, they are coming together because of limited resources. (#3, ICU consultant)

By contrast, some doctors (n=21) included resources in assessments of futility. Several described doing this consciously, for example, in how they defined futility (n=11):

[To assess futility] … one factor would be the premorbid function. Number two would be their likely level of function … after the intervention. [Three is] … the extent of the resources you’re actually going to use to get a certain outcome. (#27, emergency consultant)

Five doctors who incorporated resources into their definition of futility emphasised that patient-related considerations still usually took priority. However, one emergency consultant commented that when the treatment had a low chance of success, resources were sometimes the deciding factor that would lead to treatment being labelled futile (#96, emergency consultant, table 2).

Five doctors spoke of how futility was sometimes consciously used to ‘dress up’ or conceal rationing decisions to patients and families, particularly when denying ICU admission. This occurred when doctors believed the degree of potential benefit to the patient was too small to warrant the resources required:

I think sometimes we use the argument of futility to covertly say that this is inappropriate… it’s a difficult term to use and one that probably should be avoided. (#44, ICU consultant)

You see there’s a lot of situations where you can be … pretty confident the treatment will work but you need to dress it up as a ‘will not work’ futility… (#82, cardiology consultant)

Distinct from this conscious behaviour, a few doctors thought that sometimes their colleagues unconsciously conflated resources and futility. For example, one ICU doctor suggested that the language of ‘burden of treatment’ and ‘futility’ concealed unacknowledged rationing:

… [I]t’s hard to divorce the whole [futility] argument from resources … you’ll never hear that … articulated … but it’s what people actually mean … people will talk about the burden of the treatment and the chance of it working. The ultimate futility in someone who’s 98… who is going to die soon anyway. … [But] if the world had an infinite supply of money and for that matter, dialysis machines then … of course [you would provide the treatment]. (#19, ICU consultant)

Theme 3: resource-related distress and recommendations to address it (table 3)

Some doctors (n=22) described the tension between futility and resources as a cause of distress, which came from two sources. The first was the perceived waste and opportunity cost of providing unnecessary treatment to dying patients:

Table 3

Additional examples of resource-related distress (theme 3)

[What troubles me most about futile treatment is] giving them a bad death and … the ethics of futility and wasting resources. It takes a lot of nursing time. It occupies hospital beds. It takes a lot of doctors’ time away from those who really need it, not to mention the economics of it. (#76, internal medicine consultant)

… [T]he longer that I’m in it the more the resources bug me. … Because I understand how much I have to fight for any resources we get in palliative care. So when you see somebody getting a single chemotherapy for $30 grand which could give me a third of a nurse in the community… [it’s hard to reconcile]. (#90, palliative medicine consultant)

Participants were frustrated particularly about avoidable waste due to lack of advance discussions about goals of care, unrealistic patient (or family member) expectations or overly aggressive treatment by colleagues. Some highlighted pressing questions about fairly prioritising resources, especially given the ageing population, citing examples of perceived injustice in existing resource allocation. They described feeling troubled that societal norms now favour more intervention at the end of life. Yet doctors described lacking institutional supports to address these problems and were highly sceptical that governments would act in such a complex and taboo area.

The second source of resource-related distress was frustration about doctors’ role as gatekeepers or rationers of treatment. Doctors perceived pressure to ration life-sustaining treatment to counterbalance the increasing tendency to ‘do everything’ for dying patients. Doctors worried that they were not able to fairly prioritise multiple patients and that individual medical decisions could be unduly subjective and potentially biased:

… [I]mplicitly it’s expected that doctors will ration, which I don’t think is right, because doctors are subjective and just as prone to biases as anyone else… I really don’t like the idea that whether to continue or withdraw life-sustaining measures, differs according to the subjective biases of an individual doctor. So whoever you see determines what happens… (#30, renal consultant)

Doctors were reluctant gatekeepers and suggested administrative and government supports to make these decisions more transparent and less dependent on individual doctors. They provided several recommendations to address resource-related distress (table 4).

Table 4

Doctors’ recommendations to address resource-related distress


These findings provide an empirical basis for the assertion that doctors do not always clearly distinguish the concepts of futility and rationing in end-of-life decisions.6 7 10 While earlier studies observed that some clinicians include resources in their definition of futility,20–22 these did not address how doctors balanced resources against patient considerations (including when resources are the primary reason to limit treatment) or whether futility operated as a mask for rationing. This multispecialty study is significant as it provides a deeper understanding of both these issues.

The specialists in the study differed on the role of resources in decisions about withholding and withdrawing treatment, with many (n=40, 42%) maintaining that they were solely patient advocates, but the majority (n=54, 56%) regarding themselves as ‘dual agents’.17 ,ii Even doctors who took resources into account emphasised they prioritised patient considerations over resources, which reflects the traditional view that the patient should come first.13 Doctors rejected the notion they engaged in bedside rationing, putting resources ahead of patient needs. There was greater consensus on the existence of explicit rationing, however, such as organ allocation and ICU triage. One explanation for these findings is that doctors believe bedside rationing at end of life is morally questionable. This hypothesis is supported by the finding that doctors were uncomfortable as gatekeepers in the absence of procedural guidelines and administrative supports.

Our findings on the relationship between futility and rationing nevertheless suggest some of these doctors consciously and unconsciously ration at the bedside, under the guise of futility. Five participants discussed situations of bedside rationing, when resources were the primary factor supporting their recommendation to limit ‘futile’ life-sustaining treatment. Doctors also thought their colleagues rationed unconsciously. Reasons for this were the lack of precision in the term ‘futility’ and the lack of societal and medical acceptance of reasonable limits of medicine. This finding reinforces that some doctors internalise the social taboo about discussing resources in end-of-life decisions. While using the futility rationale to encompass resources may be a less confronting discussion (for both doctor and patient/family) than invoking rationing,11 failing to distinguish between these two concepts is problematic.

Significantly, doctors also reported distress from the tension between futility and distributive justice. This contributes to the moral distress literature, which has previously focused on nurses’ experiences and on reactions to patient suffering.25 Moral distress can be defined as ‘anger, frustration, guilt, and powerlessness that health care professionals experience when they are unable to practice according to their ethical standards’26 due to structural or institutional barriers.25 Our study adds to a line of emerging research suggesting doctors may experience moral distress in response to conflict between obligations to society and patient.26 27 Participants described distress arising both from providing (or observing their colleagues provide) too much treatment, resulting in wasted resources, and from denying treatment, absent appropriate institutional supports to ensure consistent decision making. Doctors’ worries about inconsistent and overly subjective decision making show that concerns about the ‘roster lottery’28 in end-of-life decision making are another source of moral distress.

Implications for future practice

A number of strategies address our findings about the tensions between resourcing and futility. Our participants made several recommendations to shift responsibility for rationing decisions to higher authorities (table 4). However, strategies that address doctors’ own inevitable role as gatekeeper are equally important. First, doctors might be encouraged to critically reflect on their decisions to limit life-sustaining treatment. Would the suggestion that treatment is ‘futile’ be raised if resources were unlimited? Where resources are the main driver of decisions doctors should be supported and encouraged to make this explicit. Critical reflection could be encouraged through discussions with colleagues, better understanding of patients’ values and procedural decision-making guidelines. Another important step is to ensure that policies from medical bodies about withholding and withdrawing life-sustaining treatment address and clarify what is meant by ‘futility’ and the role of resources in non-treatment decisions. An international multisociety critical care statement8 partly addresses this concern by distinguishing between treatments that are physiologically ineffective (and are objectively ‘futile’) and those that are ‘potentially inappropriate’ (involving a value judgement). While this is a more precise rendering of futility, which we endorse, we agree with Rubin and Truog6 that the multisociety statement does not sufficiently address resources, and a separate procedure for rationing decisions is needed. More research is needed on the extent to which end-of-life policies provide guidance concerning patient interests and distributive justice as two distinct non-treatment rationales.

Second, there should be more support for doctors experiencing moral distress caused by tension between futility and resources. Doctors in this study called for regulatory change yet were sceptical that the government would take the necessary steps in such a controversial area. In the interim, policies from medical bodies could be one way to provide more support and guidance. For example, the British Medical Association statement ‘Withholding and Withdrawing Life-prolonging Medical Treatment: Guidance for decision making’ provides a page-long section on resource management, which acknowledges the resource implications of futile treatment and provides procedural guidance.29 Addressing this moral distress also requires broader solutions than regulatory reform, including the provision of ongoing education, an open culture of discussion with opportunities for debriefing and strong ethical and clinical leadership. Even if more precise and comprehensive laws and policies are in place, doctors still carry the burden of engaging with patients and families about these issues, so it is important that strategies address both individual and institutional factors.

A final strategy, endorsed by our participants, is to promote more societal discussion about acceptable interventions at the end of life and where funding priorities should lie. One way to do this, which is gaining traction, is to convene citizen juries.30 A more controversial proposal is to enable doctors to discuss end-of-life costs with their patients.31 Beyond the practical problem of how to provide transparency to doctors about costs of treatment, this approach would also require a shift in how some doctors regard their role. Indeed, as Baily argues, to effect any change in this area such a shift may be necessary, as ‘the long-standing reluctance of physicians as a profession to give up the concept of physicians as pure patient advocates has been a major obstacle to the development of an equitable framework for controlling cost’.32 A better strategy might be to promote procedures for end-of-life decision making that guard against arbitrary or potentially discriminatory decision making. This would involve doctors explicitly identifying, critically evaluating and reflecting on the many competing factors in their decisions, including patient welfare, family views, resource implications, pressure from their peers and their own moral beliefs. While some variability in decisions is expected, identifying and challenging decisions that are based on rationing into a separate process cannot occur without first transparently interrogating one’s own assumptions and decision-making process. Regardless of the strategy, discussions about appropriate end-of-life decisions require input from both medical professionals and society; to accomplish this, there must be increased visibility and awareness about the resource implications of decisions.


Our sample consisted of volunteers who may have an interest in futile treatment. Additionally, resources were embedded in a larger study on futility. Nevertheless, all except two doctors addressed resource constraints, many in detail. Furthermore, we interviewed doctors in a tertiary public hospital setting. Even so, the sample included broad specialty representation (and went beyond the common focus on ICU perspectives). Future research could address the perspectives of doctors who work in private settings who may have different views given the potential added influence of personal financial incentives.


Doctors sometimes take resources into account when making decisions about the futility of life-sustaining treatment, and the word ‘futility’ can conceal rationing. Doctors perceive that there are inadequate supports to promote transparency when they seek to limit life-sustaining treatment, which can cause moral distress. Medical policies should help doctors to more explicitly distinguish between patient interests and distributive justice as the underlying rationale for limiting treatment and should promote transparent discussion and more explicit policy development about both rationales for non-treatment.


The authors acknowledge the support of the Futility Reference Group at the Royal Brisbane and Women’s Hospital, in particular, Leonie Callaway, Jane Turner and Jeffrey Lipman. We would also like to thank Michael Daly (Princess Alexandra Hospital) and John Fraser (The Prince Charles Hospital) for their support at their respective institutions, which helped promote awareness of the research. The authors also acknowledge Nicole Shepherd for her contributions to the wider research project and her NVivo expertise. The authors would also like to thank the participants for their time and insights.



  • i Medical administration is a distinct medical specialty in Australia, comprised of practitioners who have specialist training in the leadership and management of health services, including hospitals. Medical administrators are members of the Royal College of Medical Administrators, a nationally-recognised specialist medical college:

  • ii Two participants did not express a clear view about this issue.

  • Contributors All authors created the study design and qualitative data collection methods for the broader project. EC conceived of this study and oversaw all aspects, including writing. BPW and LW contributed to the data analysis and manuscript writing. CG contributed to key elements of the data analysis approach. All authors discussed the content of the article, critically revised multiple drafts of the manuscript and approved the final version.

  • Funding This research was funded by the Australian Research Council Linkage Projects scheme (project no. LP121000096) and the Royal Brisbane and Women’s Hospital. EC was supported by an Australian Government Research Training Program Scholarship and a National Health and Medical Research Council Centre of Research Excellence in End-of-Life Care Top-Up Scholarship.

  • Competing interests None declared.

  • Ethics approval Multicentre approval from Royal Brisbane and Women’s Hospital (approval no. HREC/12/QRBW/429) and from participating universities.

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

  • Patient consent for publication Not required.