Introduction
Cognitive impairment is recognised as a challenge in the care of adult patients with kidney disease.Footnote 1 This is in part due to the average age of someone commencing kidney replacement therapy for kidney failure being 63.4,Footnote 2 meaning an increased likelihood of neurodegenerative diseases entailing dementia.Footnote 3 Chronic kidney disease itself is also a recognised risk factor for cognitive decline.Footnote 4 Further, whilst conclusive evidence is lacking, it is also thought that dialysis as a treatment for kidney failure can itself contribute to cognitive impairment.Footnote 5 As such, it is not uncommon that decisions have to be made on a basis other than the patient’s own valid consent – something acknowledged by a wealth of literature.Footnote 6 In England,Footnote 7 the process of decision making where an adult patient lacks decision-making capacity is set out by the Mental Capacity Act 2005 (MCA 2005).
Per the MCA 2005, any ‘act done, or decision made … for or on behalf of a person who lacks capacity must be done, or made, in [their] best interests’.Footnote 8 That is, the decision made by the treating clinician is to centre on ‘P’ (the person) and what is best for them. The interests of other parties are not to be guiding. Those close to P, such as family and friends, are to be consulted, and thus fulfil what we will refer to as a ‘consultee’ role.Footnote 9 There will be times when the decision of the treating clinician and the views of consultees differ, and in that case a compromise might sometimes be attempted. Compromise may be a useful, pragmatic response to intractably opposed views, but it may also risk ‘compromising’ the patient.
This paper reports findings from a qualitative study of best interests decision making in the context of adult kidney care in England. The Best Interests in Renal Dialysis (BIRD) Study explored the views and experiences of healthcare professionals and family members on the process of kidney care decision making under the MCA 2005 as it relates to decisions around kidney failure. Findings relating to family members are published elsewhere.Footnote 10 In this paper, we focus on the perspectives of healthcare professional participants in relation to the idea of compromise – a particularly prominent explanatory theme in our findings. Drawing on this theme, we illustrate the notion of ‘rotten medical compromise’ and argue that it is problematic because it instrumentalises P in best interests decisions.
1. Best interests context
Where it is thought that an adult may lack decision-making capacity in a healthcare setting, the MCA 2005 becomes applicable.Footnote 11 By adults, we mean those aged 18 and above. However, it is worth noting that the MCA 2005 applies from the age of 16.Footnote 12 The MCA 2005 provides a means of proceeding where valid consent to care cannot be provided. It outlines how a determination of incapacity must be reached to confirm that valid consent is not possibleFootnote 13 and stipulates that where this is the case any decision must be made on the basis of the individual’s best interests.Footnote 14 As such, it is of central importance in the context of our empirical work.
To first consider the question of whether an individual has the requisite decision-making capacity, there are certain background principles to account for before and during any capacity assessment. Of primary importance is the starting point that P is considered to have the requisite decision-making capacity unless determined otherwise;Footnote 15 there is no burden of proof on P regarding their capacity.Footnote 16 Further, the question of P’s capacity is both time and decision specific.Footnote 17 That is, a capacity assessment must relate to a specific decision at a given time, and is not a broad determination of P’s mental faculties.Footnote 18
Where a capacity assessment is to be carried out, the MCA 2005 provides two tests that must be applied. First is the functional test. This requires the assessor to explore whether P can: (1) understand the relevant information; (2) retain that information; (3) weigh that information in some manner of decision-making process; and (4) communicate the resulting decision.Footnote 19 The functional test is essentially ‘all or nothing’ – where P falls short on any of the four elements, they are considered to have failed the functional test. In such instances, the second test – the diagnostic test – is to be applied. This test requires the reason for P failing the functional test to be ‘an impairment of, or a disturbance in the functioning of, the mind or brain’.Footnote 20 It is essential that there is a causative nexus.Footnote 21 Only then can P be considered to lack the requisite decision-making capacity. If P cannot understand the relevant information because of the way it was explained, they do not necessarily lack decision-making capacity.Footnote 22 If P cannot communicate their decision because the professionals concerned do not understand their method of communication (such as the use of sign language), they do not necessarily lack decision-making capacity. Only where the failure of the functional test satisfies the diagnostic test can P be said to lack decision-making capacity – other eventualities may be indicators of a need for further supports to enable P to make their own decisions.
The ordering of the tests is pertinent, with the functional test preceding the diagnostic one. This was confirmed by the Supreme Court in 2021.Footnote 23 Incidentally, this is a departure from the ordering as explained in the (current and original) Code of Practice.Footnote 24 However, an updated Code is under discussion, the draft of which reflects the Supreme Court’s ruling.Footnote 25 Ahead of the updated Code’s publication, the Court of Appeal has made clear that the clarified ordering should be followed.Footnote 26 Starting with the functional test helps to uphold another of the MCA 2005’s principles, that a lack of capacity cannot be established solely based on a condition of P’s ‘which might lead others to make unjustified assumptions about his capacity’.Footnote 27 For example, the fact that P has a diagnosis of dementia is not necessarily indicative of a lack of decision-making capacity. Requiring the functional test to be applied first helps, at least to some extent, to keep inappropriate assumptions at bay.
The approach to decision-making capacity within the MCA 2005, as well as its best interests decision-making framework that we will come to shortly, are by no means unanimously endorsed. Of note, the inclusion of a diagnostic element of capacity assessment remains controversial,Footnote 28 even though the MCA 2005’s approach remains distinguishable from a status-based alternative. Commentators have also highlighted the apparent discordance with UK commitments to the United Nations Convention on the Rights of Persons with Disabilities,Footnote 29 which has been suggested to prohibit the use of mental capacity as a requirement of legal capacity under Article 12.Footnote 30 However, a detailed exploration of these critiques would be overly tangential given our focus.
Once there has been a determination of incapacity following the abovementioned tests, the MCA 2005 provides for a process of decision making.Footnote 31 Where a named individual (or individuals) has been empowered to make decisions on behalf of P, it will fall to them in line with the nature of the legal authority vested in them. This might be an attorney where P has put in place a lasting power of attorney for health and wellbeing,Footnote 32 or a court-appointed deputy.Footnote 33 It is where there is no such named individual that things can become more complicated, and such instances were the focus of our empirical work. Per the MCA 2005, these circumstances call for a best interests decision under section 4. This is a decision made by ‘D’, who takes responsibility for the decision. Whilst not explicit in the MCA 2005, in the healthcare setting D has been taken to be the treating professional. This is stated in the MCA 2005 Code of Practice:
Where the decision involves the provision of medical treatment, the doctor or other member of healthcare staff responsible for carrying out the particular treatment or procedure is the decision-maker.Footnote 34
To avoid liability when providing care or treatment, D must be satisfied as to their reasonable belief that such care or treatment is in P’s best interests.Footnote 35 That is, on a strict reading of the requirements of the MCA 2005, any healthcare professional providing care or treatment must individually be satisfied as to their reasonable belief.Footnote 36 This is where section 4 assists in detailing a range of factors for consideration. The relevant section of the MCA 2005 first clarifies that P is to participate in the decision as far as is reasonably practicable.Footnote 37 Further, in questions of life-sustaining treatment – as dialysis is – D should not be ‘motivated by a desire to bring about [P’s] death’.Footnote 38 Beyond this, factors are divided into, broadly: (1) P’s perspective,Footnote 39 and (2) consultees’ views.Footnote 40
On the matter of P’s perspective, this includes consideration of any past or present expressions of preference, such as a written statement made when P had capacity.Footnote 41 In addition, D should consider any beliefs or values held by P that would likely influence decision making should P have capacity.Footnote 42 Finally, the MCA 2005 incorporates a catch-all of ‘other factors’ P would likely take into account in making a decision, were they able to.Footnote 43 All of these are subject to the caveat of what is ‘reasonably ascertainable’,Footnote 44 recognising that in certain care contexts such information may be limited.
This exploration of P’s perspective is to be supplemented with the views of so-called consultees. This is a broad grouping of people with an interest in P’s welfare. Whilst family and close friends are the obvious candidates to be consulted, the MCA 2005 allows for a broad interpretation. Depending on P’s situation, this might include a colleague, long-term carer, or even a faith or community leader.Footnote 45 D should also consult any attorney or court-appointed deputy that is identifiable but does not have authority over the decision in question.Footnote 46 The purpose of consultation is for D to build a better understanding of P’s perspective – preferences, beliefs, values, etc. – and, ultimately, ‘what would be in the person’s best interests’.Footnote 47 Engagement with consultees is similarly caveated by what is ‘practicable and appropriate’Footnote 48 to allow for the sheer range of contexts in which best interests decisions might be made.
What is apparent from these factors is the so-called ‘P-centricity’ of best interests decision making.Footnote 49 That is, the overriding goal is to make the decision that is best for P. When best interests decision making under the MCA 2005 was first before the Supreme Court, this P-centricity was made clear in there being ‘emphasis on the need to see the patient as an individual, with his own values, likes and dislikes, and to consider his best interests in a holistic way’.Footnote 50 A best interests decision is not, for example, to be based on the impact a particular course of action would have on P’s loved ones. Certainly, there are ways in which the impact on P’s loved ones would become a factor. As clarified in Aintree, by reference to the Law Commission’s 1995 Report on Mental Incapacity, a best interests decision is to consider any relevant ‘altruistic sentiments and concern for others’ held by P.Footnote 51 If it became apparent, then, that P would account for impact on loved ones in making the decision himself, it would be appropriate for D to factor this into a best interests decision. This is a nod to a relational understanding of autonomy, whereby we often make decisions about our own lives with regard for those close to us.
Alternatively, the position taken by loved ones may affect the options available as part of the best interests decision. Consider a question of discharge from hospital, whereby P’s family coming forward saying they would be willing to provide care at home as required would introduce an additional option to explore; by contrast, the family might refuse to take this on, in which case that option is taken off the table and options involving local authority provision would have to be explored.Footnote 52 The distinction is that the final decision is not based on what works best for P’s loved ones. No matter how willing a family is to take on caring responsibilities upon discharge, D may decide this is not in P’s best interests if, for example, the level of care required is more than they believe the family can manage. Similarly, it might be decided that life-sustaining treatment is not in the best interests of P even if the family objects and expresses a willingness to provide P with care. The ultimate responsibility of D in a best interests decision is to prioritise the best interests of P, and P alone.
Whilst the views of and impact on P’s loved ones should not be determinative of a best interests decision, that does not mean that they should be entirely set aside. As noted above, a key factor stipulated by the MCA 2005 is the perspective of consultees.Footnote 53 Whilst the authority and responsibility for the decision remains with D, discussion of options for P’s care with consultees is important to better understand P’s wishes and feelings.Footnote 54 So important is such consultation, that where there are no reasonably identifiable consultees, a similar role should be filled by an Independent Mental Capacity Advocate (IMCA) in prescribed situations.Footnote 55 The IMCA is then required to consult with both P and the care team, examine any appropriate records, evaluate all information and perspectives in relation to the treatment options in question, and prepare a report outlining findings and, if appropriate, a recommendation.Footnote 56 Though with both consultees and IMCAs, it remains that their role is consultative and they are not providing proxy consent. The decision, whilst informed by consultees and IMCAs, must be made by D on the basis of P’s best interests.
The law is clear in vesting decision-making authority in D. However, the involvement of consultees does create the possibility of disagreement and, indeed, conflict. Where P’s family and care team differ in their opinions on P’s best interests, there may arise significant tension. One possible response by D may be to go with what the family wants even if it is not what they, as the legal authority, consider to be in P’s best interests. Such instances are the subject of this paper and the empirical findings we will soon explore. Nonetheless, for now it is worth acknowledging the Court’s view on such compromise. This has been addressed by Hayden J recently in the case of GUP v EUP & Another:
Whilst these two perspectives [the family and the medical team] of EUP’s medical needs have diverged, I am concerned that the treatment she has received reflects a convergence between the two. In other words, the treatment plan has an air of compromise about it, a negotiation between the family and the medical team. There may, sometimes, be a place for that, but not if the person at the centre of it becomes marginalised. P (the protected party) must always be afforded care, which is identifiably in her own best interests.Footnote 57
Hayden J thus reiterates the position that the family are to have a consultative but not decisional role in best interests decisions, and that P’s own best interests must be determinative. As we come to explore in this paper, the reality may not always align with this legal ideal.
Finally, to briefly outline the role of the Court of Protection. Established by the MCA 2005 to hear cases concerning the Act,Footnote 58 the Court serves several purposes. Broadly, it maintains the same powers as the High Court.Footnote 59 The Court can make declarations as to P’s (in)capacity and the lawfulness of any care plan.Footnote 60 Also of relevance for our purposes the Court can, by order, make decisions on behalf of P,Footnote 61 having due regard for the same expectations the MCA 2005 would place on any other decision maker around the principles and P’s best interests.Footnote 62 Importantly, the Court does not infringe on medical expertise in making orders. The care options available are determined by professionals, meaning the Court cannot compel healthcare professionals to provide care that they have deemed clinically inappropriate.Footnote 63 Applications to the Court ordinarily require permission,Footnote 64 with exceptions being those from P,Footnote 65 those legally empowered to act for P,Footnote 66 and any person named in an order of the Court (where the application concerns this order).Footnote 67
2. MethodsFootnote 68
(a) Design
This study employed qualitative methods to explore the views and experiences of people who have been involved in best interests decision making in adult kidney care. By adult kidney care, we mean the care of patients aged 18 and above, and we do not include young people who may have some engagement with adult services as part of their transition.Footnote 69 Semi-structured interviews were conducted utilising an evolving topic guide, with participants prompted to explore areas of interest whilst affording them the freedom to recount their personal views and experiences as they felt comfortable.
(b) Participants and recruitment
Participants were recruited from two renal units in England, chosen on the basis of their serving different patient populations. For inclusion, participants had to have been involved in at least one best interests decision relating to dialysis and/or CKM as a healthcare professional. Family members were also recruited, but they are not considered in this paper, which focuses on healthcare professionals.
Resource constraints prevented the use of an interpreter, so participants had to be able to consent to undertake the interview in English. We do not believe this impacted on the work as participants would be required to communicate in English as part of their job roles anyway. Participants also had to be able and willing to consent to participation.
Purposive sampling was used to maximise diversity, resulting in the participation of a combination of renal consultants, renal trainees, ward nurses, and specialist nurses, amongst which were individuals with particular responsibility within their unit for, for example, CKM.
Local collaborators first identified and approached potential participants based on personal knowledge of teams and reviews of patient records. Anyone interested in participation was then put in touch with JAP and provided with participant information materials. Interviews were scheduled after confirmation of eligibility, with participants offered the choice between an in-person or online interview. Research suggests online interviews do not suffer significantly in terms of quality or quantity of data.Footnote 70
All participants provided consent immediately prior to their interview, either in writing or orally depending on the interview location. Oral consent was audio recorded and stored separately to the interview recording. Following their interview, all participants were given a £20 shopping voucher as a token of thanks.
(c) Data generation
The data generation period ran from June 2021 to May 2022. Interviews lasted for between 0h41m and 1h23m, with a mean length of 0h59m. JAP conducted all interviews. Participants were unknown to JAP, who introduced himself to participants as a non-clinical PhD researcher. Interviews were structured around a topic guide which was iteratively developed throughout the data generation period in response to interviews.
(d) Data analysis
Interview transcripts were subject to reflexive thematic analysis, utilising Braun and Clarke’s stages of: data familiarisation, inductive coding, constructing themes, reviewing themes, and naming themes.Footnote 71 This is a commonly used approach that allows for sufficient flexibility in meeting the needs of a given study whilst providing necessary methodological rigour.
Audio recordings were transcribed verbatim by a transcription serviceFootnote 72 and checked for accuracy by JAP. This process of checking, having conducted the interviews some time before, enabled satisfaction of the data familiarisation stage. NVivo software (version 12) was then used to facilitate inductive, data-driven coding by JAP. The approach used was, in practice, a combined inductive-deductive analysis, similar to that articulated by Fereday and Muir-Cochrane.Footnote 73 That is, whilst being led by the data, initial coding was informed by an earlier scoping reviewFootnote 74 – the thematic mapping of this scoping review was not rigidly applied, but acted as a starting point. Initial codes were then discussed between the authors before themes were constructed and agreed. The organisation and naming of themes focused on balancing the descriptive and analytical; themes needed to accurately portray what participants were saying, but also recognise that their construction is intertwined with the researchers’ analytical input.
(e) Ethical approval
This study was reviewed by the Health Research Authority’s London – Camberwell St Giles Research Ethics Committee, with approvals granted on 22 December 2020 (REC reference: 20/LO/1233).
3. Findings: the instrumentalisation of P
For the purposes of this paper, we are focusing on reporting two themes generated through our analysis of the healthcare professional data, which are linked by compromise as an explanatory theme: ‘the path of least resistance’ and ‘dialysis trials as conflict resolution’.
Participants included kidney doctors at various stages in their careers, as well as nurses in a range of specific roles in kidney care (see Table 1). In the reporting that follows, ‘D’ and ‘N’ are used to denote quotes from doctor and nurse participants, respectively.
Study participants
* Some doctors who participated also held secondary roles, such as lead for supportive care. Participants self-identified their roles, and those represented in the table are how they described them when asked.
** These figures represent the details of five of the six participants. The audio was unclear at the relevant point in the recording of one interview, and it was not possible to follow up with the participant afterwards.
(a) The path of least resistance
As much as professionals want to do what they perceive to be the ‘right’ thing for their patients, the reality of the decision-making landscape was often characterised as being under pressure from several directions.
Notably, where disagreement between the care team and the patient’s familyFootnote 75 arises it is generally thought of as not worth allowing conflict to escalate to the point of damaged relationships and potential court proceedings. The result, then, is a situation whereby professionals can often find themselves pursuing what they described as the ‘path of least resistance’ (D02) or ‘easiest way forward’ (N01) even if they do not consider it to be in the patient’s best interests. Professionals mostly commented on the realisation of such a scenario when the family are insistent on the patient undergoing dialysis – whether that be an initiation or continuation – when the care team feels that the patient will struggle to tolerate the burdens of dialysis and/or has an overall low quality of life, such that the burdens of dialysis outweigh any benefit.
[I]n the textbook it is always easy, say, ‘family will never be able to make the decision for patients’. But actually, we all know as a medical practitioner if there is a strong objection from the family things are always quite difficult to move on. (D05)
Professionals characterised this approach as defensive medicine, commenting on what they consider an unfortunate state of affairs whereby they must always have an eye on the potential for formal complaints from family members.Footnote 76 This was reflected on, by some, as something of a shift that had occurred during their careers; noting an increased culture of complaints when patients and their families are not satisfied with the care received. Whilst it was acknowledged that a right to complain is important, some professionals expressed concern that, in the context of patients who lack decision-making capacity whose family members are making complaints, a general movement towards defensive medicine can prove detrimental for the patient.
If there is a default [of providing dialysis to avoid conflict], someone who hasn’t got a voice is particularly vulnerable to that. (D02)
It should be about protecting the patient, but I think … in the current world, it’s about protecting yourself. I’m afraid it really is. (D04)
Indeed, this practice of defensive medicine was, for some professionals, about the avoidance of legal action as much as maintaining positive relationships with the family members of patients. Some reflected on situations where this approach resulted in a best interests decision being made that they did not actually feel was in the best interests of the patient.
Sometimes we end up outside of what we would think is necessarily the right thing to do, but actually the families are so insistent that our other option is to end in the Court of Protection and have some extremely horrific situation for everyone involved. (D08)
This is also something that was observed by nurses who were involved in discussions but did not have responsibility for the final decision. One described a situation where they felt the doctor who had to make the decision was put in a difficult position by the patient’s family.
I think more that the consultant was being led by the family. I really do. Because knowing that consultant, I don’t think he probably would have done it. I think he was being pushed into a corner. Whether that was right or wrong, I don’t know. (N09)
Whilst taking this path and proceeding in line with the views of family members, even if not in agreement with them, professionals noted that pressure to do so can come from various directions. As much as family members may seek to dominate the decision-making process, there was a feeling that the hospital itself – meaning management and the hospital lawyers – may sometimes pressure healthcare professionals to do what the family want in the event of conflict. Professionals suggested that it can sometimes feel as though there is an ever-looming threat of legal proceedings which the hospital wants to avoid, even at the expense of patients who may find themselves undergoing treatment that is hugely burdensome.
The Trust solicitors are very scared of litigation. So, I think there are some times when we do them [the family’s wishes] because we feel backed into a corner, and it is very difficult. Then you’re very defensive and reticent because your views are very different to the family’s views. (N04)
There appears to be a suggestion, then, that members of the care team do not always feel able to do what they think is in the best interests of a patient. Some professionals spoke of the distress that can come from not following the wishes of P’s family members. They recounted the experiences of either themselves or their colleagues having received complaints, in some cases resulting in Court of Protection hearings. Those who had not themselves been involved in cases that reached this point still had an impression of the process – and the impact – through word of mouth.
I’ve never been [to the Court of Protection], but I’m told it’s something that you never forget. It’s an experience. I had a girlfriend – it broke her. She actually jacked in being a nurse because it was so traumatic. So touch wood, I’ve never been through that situation. (N03)
However, the feeling of wanting to do the ‘right’ thing for the patient was recognised as necessitating the Court of Protection in some cases.
If it’s the only way that I have to ensure that the person is looked after properly then we’ll do it, we’ll go there [Court of Protection] because we have to. The ultimate person I’m looking after is that patient. (D08)
However, this doctor did reflect on the confrontational nature of pursuing legal avenues of conflict resolution and how this has a negative effect even if the outcome is what the care team consider to be in the patient’s best interests. D08 highlighted how a decision not to dialyse a patient does not mean abandonment because ‘you still need to look after them’ with ‘all the other good stuff’. Indeed, the provision of such non-dialytic care still requires a continuing relationship with the patient’s family, which legal proceedings can hugely damage, and can be detrimental to the patient’s interests.
One of the things that going to the Court of Protection fundamentally means is that you and the family have lost that therapeutic relationship. (D08)
This professional stressed that this potential for a severed relationship makes the Court of Protection a last resort for them – a feeling echoed by most professionals. Instead, participants felt it better to ‘try and reason with them [the family]’ (D11) so as to avoid things reaching this stage.
Building on the idea of protecting oneself from possible complaints and legal action, professionals stressed the importance of extensively documenting discussions to the point that records are ‘fantastically accurate and legible’ (D04). One doctor highlighted the possible benefits down the line, recounting his experience of a subsequent complaint.
The family wrote to the GMC about anyone that had been involved in the care, and we were, as a unit, eight, nine of us were investigated by the GMC for our role in his care and subsequent death. Having everything in writing, even though it was time-consuming, was fantastically helpful. (D04)
Accurate documentation was also highlighted as important in demonstrating that a best interests decision taking place at all was appropriate. One nurse described situations in which she was not clear, on reading the patient’s notes, how decisions had been made. This left her feeling strongly that all stages of the process should be well documented – not just the final best interests decision itself.
My opinion is, it should be documented at that stage, that the patient hasn’t got capacity. I think they should make a capacity assessment. You know, ‘we have done a capacity assessment and that patient has not got the capacity’. … It should be clearly documented that that process took place and what steps we did to involve the patient and what is important to the patient. (N01)
Professionals overall expressed a feeling that they are often in the difficult situation of having to provide treatment that they do not think is appropriate. The need to maintain relationships with family members and avoid complaints can cause them to follow the path of least resistance and agree to demands of family members that go against their own view of the patient’s best interests. Where participants spoke of appropriateness, there was not always a clear sense as to the line between thinking something is clinically inappropriate (and thus should not be provided) and thinking something is clearly not in P’s best interests. Nonetheless, there was at least the implication that participants could realistically see how care that is deemed clinically inappropriate could be provided in these situations (including trials of dialysis, as will be discussed in the next section). Though, interestingly, concerns about legal implications centred on family complaints where care was not provided, with no real discussion of the potential for liability where care that is provided may give rise to civil and/or criminal liability.
(b) Dialysis trials as conflict resolution
As something of a middle ground between complete acquiescence to the will of family members and risking the Court of Protection in conflict situations, many professionals raised the idea of dialysis trials. It was suggested that offering to provide dialysis on a trial basis, with the intention of revisiting the decision soon after, could defuse a situation and help maintain the relationship between the care team and the family. In effect, dialysis trials were discussed as a compromise.
So, occasionally, it’s a trial where it’s not very clear where there might still be some areas of conflict, both within that individual or around that individual, and they’re not sure and not committed, then sometimes a trial of the therapy is quite helpful. I don’t do it as a routine. (D03)
Trials of dialysis are a polite way of avoiding conflict and proving, one way or another, to that mediated group, that you’ve done everything you possibly can and it wasn’t a success. Because it’s a trial, it has only been for a couple of months, so we probably haven’t caused harm, yes? But it’s a fudge. It’s a fudge. (D04)
[E]ven if I’m convinced that this is not the right thing for this patient, I do accommodate. I’ve been known to accommodate the opinion of a family to the point that, yes, we try dialysis. Obviously, one tends to try to circumscribe it in a sense that, ‘yes, we will try this. If this happens, this happens, then we know we’ve tried and we’ll stop’. … I always find that a less combative, less adversarial, approach and that’s what I tend to use in my practice. (D09)
These professionals admitted that trials may be started in situations where they do not feel it is good for the patient. Indeed, one said they would ‘not dig my heel in’ (D07) and decide not to dialyse the patient when the family objects even if ‘complication and the risk from dialysis might still be high’ (D07). That this course of action was often discussed in terms of something that had to be done in these instances of conflict suggests that those around the patient – their consultees – can strongly influence decisions, to the point that professionals may set aside what they feel is best for the patient.
Whilst there was widespread recognition that dialysis trials do happen, and for various reasons, some professionals expressed discomfort at their increasing prevalence.
[I]f there’s doubt, sometimes people then go for a trial, which I don’t particularly agree, the overuse of trial dialysis. … If you are in doubt, people will just say, ‘dialyse if you’re in doubt’, and people say, ‘give them a trial’, which means that you let them dialyse. … So, your limit of dialysing people becomes very, very low. (D07)
This discomfort, at least in part, stems from a feeling that ‘[o]nce it starts, it’s very hard to stop’ (D07). Professionals explained the difficulties of ending a dialysis trial in situations where the reason for stopping is not immediately clear to family members who may be pushing for continued dialysis. For example, one doctor suggested that the ‘trajectory becomes clear’ (D10) in most cases, such as regular hospital admissions during the trial making apparent the ‘natural end of the trial’ (D10).
However, recognising the difficulties in discontinuing a dialysis trial, one doctor outlined the very particular approach he takes to avoid further conflict when the decision is revisited. When starting a trial, he sets very clear parameters as to what ‘improvement’ is going to be measured by – ensuring there is not a general (and very subjective) focus on whether the patient feels better.
The definition of how they feel, the fact that they got accustomed to having dialysis is often taken as they’re feeling better or feeling well. You need to be very specific. That’s my personal view, is that if you’re very specific about when you would stop, what are the things that you are looking for after a month and to either continue or discontinue, that’s very helpful. (D11)
This feeling that it is difficult to stop dialysis once it starts fed into discussion by many professionals about differences and similarities between withholding care and withdrawing care. There appeared to be a preference for the language of discontinuing or not carrying on, rather than withdrawing, due to the negative connotations of the latter.
There were very mixed views on the withholding/discontinuation distinction, as to whether they ought to be viewed as in any way different, assuming all other aspects are similar. One doctor was explicit that neither is ‘more existentially challenging, or ethically questionable, or anything like that’ (D02). For some, then, they are comparable. However, others were less sure.
I think I would argue perhaps more strongly to not start the treatment, rather than them to not continue it. Because this is the difference between starting something and stopping something. (D01)
For some, this distinction comes from feelings conveyed by family members, which can make it more difficult to pursue the route of discontinuing dialysis.
Almost, for the family, it [stopping dialysis] feels like active killing for them, that you’re actively stopping something so that they can die from not having it, rather than you not offering it before. (D07)
Many spoke of instances where dialysis was being withdrawn out of recognition that the care is no longer improving the patient’s condition, but that family members sometimes struggled to accept this and instead viewed it as akin to some form of euthanasia.
Despite the strong element of compromise in starting dialysis trials, some professionals did explain that trials can provide useful information to make a definitive best interests decision. As such, in some instances they may be indicated and thereby not constitute compromise.
[B]y that point [after a period of dialysis trial], people have lived with dialysis. They know what it means to have the hospital transport turn up at 6:45 in the morning 15 years after you’ve retired and having to be ready for that. Get driven in, do the dialysis. The machine goes down and the shouting patient over there. (D02)
Relatedly, some spoke of trials in terms of allowing time to make a particularly difficult decision, especially if the patient’s condition is quickly deteriorating.
[W]hat might happen is there might be a decision made to do some dialysis to stabilise the situation and make it less likely that the patient would, in the short term, deteriorate, to allow that conversation to happen. So, I suppose you might do that. I cannot recall a situation of having done that, but I have seen that done. (D01)
Dialysis trials, then, were characterised as fulfilling two purposes. First, they prevent the relationship between the care team and the patient’s family breaking down in the event of disagreement by meeting somewhere in the middle. Secondly, they can provide more – and very useful – information about how the patient tolerates dialysis and whether it improves the patient’s condition, and time to reach a decision. The former represents a compromise position wherein the professionals proceeds with what they believe is contrary to the best interests of the patient. The latter, where it is the motivation for a trial, represents an agreement that a dialysis trial is in the patient’s best interests. Certainly, there was a recognition that in some circumstances a trial may be useful in confirming the clinical appropriateness of dialysis for a particular patient. With many kidney patients having complex care needs, often because of multimorbidity, the suitability of dialysis for a patient may not always be clear. Trialling dialysis on the basis of such uncertainty may reasonably be considered in a patient’s best interests and is distinct from compromising in situations where the care team are confident that a patient is not a suitable candidate for dialysis.
4. Discussion: rotten medical compromise
Many professionals spoke of instigating dialysis trials as a means of conflict management where the family were insistent on dialysis against the view of the care team. In situations where the deciding professional felt dialysis was not in P’s best interests but the family disagreed, some spoke of going against what they considered P’s best interests to appease the family. In considering the ethical status of such decisions, we will initially afford the view of the professional a certain epistemic privilege. Whilst professionals are fallible – much to their own admission in interviews – professional opinion must count for something. Medicine is not a perfect science, but we accept that research and clinical experience provide informed predictions of how things will pan out. This is, importantly, in relation to the professional’s clinical view – a prediction as to how dialysis will affect P’s condition is not the same as a view of P’s best interests.
There are two senses of compromise at work here: (1) being compromised; and (2) reaching a compromise. First, the pejorative sense, where the professionals acquiesce to the views of the family, initiating dialysis (on a trial basis) to avoid conflict. The professionals can be said to have allowed themselves to be compromised through being voluntarily complicit in what they consider to be a wrong – that is, not in the best interests of the patient. Secondly, the dialysis trial can be viewed as a mutual compromise position, where both parties move from their ideal position to agree on some kind of middle ground. Some may suggest that this is not a compromise position at all in the context with which we are concerned, as one party – the family – is ‘winning’. Compromise is usually characterised as each side giving some ground, meeting somewhere in between, though not necessarily the middle. According to Benjamin, compromises ‘require concessions by both sides and be able to be seen by the opposing parties as somehow splitting the difference between them’.Footnote 77 Indeed, Golding suggests that in a compromise neither party can have the outcome they would ideally want.Footnote 78 If this is correct, dialysis trials started in order to appease P’s family may not be examples of compromise, because the family essentially get what they want. It is, however, a compromise position if genuinely carried out as a trial with a fixed end point, where both parties are moving from their stated preference – to either not dialyse at all or to dialyse indefinitely.
Given the presence of compromise in best interests decisions about dialysis, there is a need to consider whether such compromise can be considered ‘good’ – or, as a minimum standard, ethically and legally defensible. Golding highlights how the literature presents ‘a fundamental disagreement on the rationality and morality of compromise’. For some, compromise may be viewed as weakness, whereas others regard it as a means of rational agreement in the face of conflict. Whatever side one takes, compromise is an unavoidable feature of everyday life for most of us. Nonetheless, there remains a question over whether it is appropriate when it comes to P’s treatment in the context of a best interests decision.
The view of the professionals interviewed was very much that compromise, whether in the form of acquiescing to the family or commencing a dialysis trial, is far from ideal. Several acknowledged that, in doing so, they have felt they were going against the best interests of P. Nonetheless, it happens, and the ‘path of least resistance’ is chosen. The justification offered by our participants is the need to maintain a relationship with the family and, consequently, avoid an escalation to the point of the Court of Protection, which is perceived as benefiting no one. Viewing court proceedings as unpleasant for all parties, professionals were willing to be flexible and move in the direction of the family – either in part or completely – to avoid a legal process. This aligns with other studies, such as one with intensive care unit doctors, which found that pressure from family and threat of legal action were reasons for making admission decisions that they considered inappropriate.Footnote 79
Perspectives found in the literature on this matter are somewhat sparse, though there is some discussion of the issue. Whilst much of it is from the US context, which differs markedly from the approach of the MCA 2005, it still provides useful insight of a more ethical slant. Kaye and Lella, for example, suggest that it may be mandatory to set aside the known preferences of P for the benefit of P’s family unless there is certain suffering for P.Footnote 80 Others feel keeping P alive for the benefit of the family to be entirely inappropriate.Footnote 81 Whilst not framed as a discussion of compromise, these perspectives are at least indicative of diverse positions on these sorts of questions. Whilst many argue that the decision should not be made by the family as a matter of principle, there is less discussion in the literature of the extent to which professionals can be expected to resist insistent families. Recognition of defensive medicine is notably more common in the literature than proposed responses to it, with a seeming acceptance of its inevitability. Conscious of the importance of an is/ought distinction here, there is a need to find a more satisfactory account. For this, we turn to the work of Margalit.
Margalit writes of so-called ‘rotten compromises’ as those which are to be avoided.Footnote 82 Whilst Margalit discusses the idea of compromise in the political context, referencing major historical events such as the Munich Agreement, there are points of comparison in how he frames his discussion that prove useful for our purposes in the clinical context.
A rotten compromise, per Margalit, is ‘an agreement to establish or maintain an inhuman regime, a regime of cruelty and humiliation, that is, a regime that does not treat humans as humans’.Footnote 83 He argues that there is often a tension between peace and justice, with compromise as the ‘go-between’.Footnote 84 Whilst justice is an ideal, it may not be achievable, and for the sake of peace (which, in the political context Margalit is concerned with, will generally be a matter of stability) something less than just may have to be accepted. In many cases, such compromise can be deemed acceptable, in part because compromise is an essential part of life. Rotten compromises, however, are never permissible for Margalit – even if for the sake of peace.
We are in no way suggesting that the issues Margalit is concerned with represent direct comparisons with those we are discussing. Whilst a poorly-made best interests decision may cause P harm, this is not remotely on the scale of things, such as genocide, which constitute rotten compromise in the international relations sphere and with which Margalit is concerned. However, Margalit’s thinking can be useful in considering dialysis trials as compromise where the care team and P’s family are in staunch disagreement.
Before drawing this comparison, it is worth reiterating exactly the sort of decisions we are referring to. Our focus is situations where the deciding clinician moves towards the family’s view, initiating (a trial of) dialysis that they believe – very strongly – to be against P’s best interests. What we are not talking about is situations where discussion with P’s family has led the deciding clinician to question their position. Indeed, research has found that nephrologists can struggle more generally in disagreeing with patients and their families in dialysis decisions because they can feel ‘trapped in their uncertainty and afraid of using their authority’.Footnote 85 If this is the case and a trial of dialysis is started because the deciding clinician is genuinely unsure what is in P’s best interests and has been convinced by the family that it is worth trying active treatment, this is not a compromise – it is open and honest practice that is seeking to do the best for P in the situation, even if the outcome may not always be desirable. It is only when appeasement goes against what the deciding clinician is confident is in P’s best interests, and they believe that what is being done is inappropriate,Footnote 86 that a compromise arises that we take issue with. These situations are our focus.
Back to the application of Margalit’s work to the kidney care context: let us consider justice as the ‘best’ decision for the patient – what is more objectively in the best interests of P. The elusive nature of this concept makes it idealistic in much the same way as justice in Margalit’s framing. That is, the deciding clinician’s view of what is in P’s best interests cannot truly be considered objective, even if the MCA 2005 speaks of the best interests test as objective.Footnote 87 Peace, on the other hand, constitutes a maintenance of a working relationship between the patient’s care team and family – it represents stability in decision making. The question is, then, whether a nephrologist pursuing a dialysis trial purely for the avoidance of conflict, when they are sure the best interests of the patient lie in not dialysing, represents too much of a move from of justice in favour of peace, or too much of a dismissal of P’s best interests in favour of a working relationship P’s family. If it does, then it constitutes a best interests version of a rotten compromise – this we will hereafter refer to as a rotten medical compromise.Footnote 88
Of course, a similar situation can arise outside of the best interests context. A patient deemed to have decision-making capacity may still find themselves as a pawn in disagreements between their clinician and family, particularly where that patient displays a decidedly agreeable temperament. There is reason to be concerned by this reality too, but perhaps less so than where it is a best interests decision. At least in the case of the patient with decision-making capacity, it is more possible for them to interject and speak up on their own behalf and this be determinative. Such a patient’s diffidence and willingness to agree with, for example, their family, can still largely be considered a valid decision – particularly where a capacity assessment has been carried out and no determination of incapacity followed. When we are concerned with a best interests decision, the patient lacks the ability to break the disagreement by making the decision themselves. As such, the patient who lacks decision-making capacity is far more vulnerable to this process of compromise which may prove harmful to them. We recognise that agreeable patients with capacity may fall foul of similar compromises, but a full exploration of the complexities of such a situation are beyond the scope of this work.
Similarly, we at this stage set aside examples which we have alluded to previously in which clinically inappropriate care may be provided. Instances in which a professional could be found guilty of battery for having acted in this way fall into the realm of criminal law. They do not occupy the grey area with which we are concerned.
So, with our focus clarified, what ought we to consider too much of a move from justice in favour of peace in this context? Margalit’s definition of a rotten compromise cannot be directly applied here as it, arguably, is a standard that would never be met in the clinical context; it would be a stretch to suggest that occasional poor best interests decisions constitute an ‘inhuman regime’.Footnote 89 Nonetheless, the framing of a compromise that ‘does not treat humans as humans’Footnote 90 can be drawn on. Based on this, we might consider a rotten medical compromise to be a best interests decision that: (1) constitutes an attempt to maintain a working relationship between P’s care team and family; and (2) entails any manner of harm to P. In such a scenario, P can be said to have been instrumentalised (used as a means) to achieving that peace. Such a situation is more treating the family than P and prioritising the avoidance of complaints and legal action over P – an apparent application of the words of the poet George Herbert: ‘a lean compromise is better than a fat lawsuit’.Footnote 91
It might be asked whether what we are proposing allows too little room for manoeuvre. After all, it is reasonable to recognise that part of the job of a healthcare professional is to ‘treat’ the family as well;Footnote 92 patients, for the most part, do not exist independently and family often feature prominently in medical decisions, particularly where a possible outcome is death. The professionals interviewed accepted this view. As such, if the provision of dialysis may only cause P mild discomfort, might not the appeasement of P’s family be a legitimate justification for compromise? We contend that the answer is no.
Medicine is inherently harmful, in that almost every treatment results in some level of iatrogenic harm, however trivial. Such harm is generally considered justified on the basis of the intended results of that treatment – eg curing an individual of a disease. Central to such routine and informal harm-benefit calculations is consideration of proportionality.Footnote 93 We seek to avoid undue and excessive harm in pursuit of treatment or cure. It is this underlying focus on relieving suffering with minimal harm that forms the foundations of the doctor-patient relationship. Sacrificing this solely to avoid conflict with a patient’s family is to undermine this foundation of medical practice, as it fails to centre P in decisions about their own careFootnote 94 and, rather, instrumentalises P to serve the interests of others.
This conceptualisation of rotten medical compromise mirrors what Fiske and Tetlock refer to as a ‘taboo trade-off’, defined as that which ‘violates deeply-held normative intuitions about the integrity, even sanctity, of certain forms of relationship and the moral-political values that derive from those relationships’.Footnote 95 The doctor-patient relationship is undermined by rotten medical compromise because nonmaleficence – however one comes to define it – is recognised as central to the integrity of that relationship.
Effectively, the concern is in both the instrumentalisation of the patient in such rotten medical compromises and the possible long-term implications of undermining the doctor-patient relationship in this way (for example, if a particularly high-profile case sees the care team depicted as the villains by the media).
We recognise that in opposing such compromise we are effectively opposing the solution put forward by professionals to navigate this complex problem. For many of those interviewed, it is essential to find a workable way forward, even if it is not considered in P’s best interests. Whilst professionals certainly recognised the problems with using dialysis trials as a compromise, they apparently did not feel empowered to pursue an alternative course. The exception to this is more extreme cases, in which some professionals explained they would insist that dialysis is inappropriate and would not provide it. But this was not a consistently expressed position, with many professionals finding it too difficult to take such a stance. Indeed, P’s best interests should not be contingent on being such an extreme case – rotten medical compromise should be avoided across the board.
This ultimately comes back to respect for P’s preferences, as discussed above. In making a decision on the basis of what makes P’s family happy, P’s preferences may fade into the background. It is, of course, possible they will align, but if the motivation is appeasing the family rather than doing what is best of P, then that important P-centricity is lost. Our point is that once the interests of P are set aside in the interests of avoiding conflict with the family, the care team is engaging in rotten medical compromise, which contravenes its duties to P.
Despite our case against rotten medical compromise, there remains the question of institutional pressure. Many healthcare professionals spoke of complying with the demands of family members at the instruction of their hospital’s legal team (or based on the anticipated instruction), which ultimately takes it out of their control. One might suggest, therefore, that the healthcare professional is not the compromising party as they are, in effect, acting out of necessity.Footnote 96 Margalit usefully explores this, suggesting that acting out of necessity in this manner constitutes coercion: ‘coercion as a condition in which, being left with no reasonable alternatives, we do, against our better judgment, what others want us to do’.Footnote 97 Coercion cannot be considered compromise due to the conditions in which it plays out; the professional may want to resist the demands of the family, but is rendered unable to reasonably do so by the instructions of their employer.
Where there is such institutional pressure in the direction of rotten medical compromise, there is a clear need for change at a level higher than the individual professional (or even the multidisciplinary team). To ensure decisions are made in a way that can, if scrutinised, be said to be in P’s best interests, professionals must be supported by their employer to do so.Footnote 98 Where P’s doctor assumes the role of D in a best interests decision, that doctor must be able to say that they endorse any decision they make – if they do not reasonably believe the care decision made is in P’s best interests, any treatment being provided could constitute battery. That is not to say that a hospital ought to unquestioningly back its doctors when they oppose the view of P’s family. There will be situations when the family’s perspective is valid, and the professional will be in the wrong. However, this simply suggests a need to consider both sides, rather than immediately aligning with family to avoid legal proceedings.
As such, the support we speak of is not the hospital’s legal team simply doing what they can to enact the doctor’s decision, nor is it allowing the doctor to abdicate responsibility and have the decision made for them. Rather, it is about a strong working relationship that enables all parties to effectively discharge their own (legal) duties. Professionals must not feel scared to approach the legal team, nor should they consider it pointless because they anticipate being told to do what P’s family ask. Equally, the legal team should not take issue with having to provide support, nor should they immediately side with either party for purely financial-risk-averse reasons. Professionals should be able to approach their legal team and explain what they consider the appropriate course of action to be and expect to be advised on the most appropriate legal avenue to achieve this or, where relevant, have explained to them the reasons why it is not possible. Whilst we appreciate this may sound optimistic, it should not be considered so. An open, collaborative approach between clinical and legal expertise such as this can help to prevent instances of rotten medical compromise – and also identify where the family’s perspective is in need of greater exploration.
Conclusion
Some manner of compromise appears to be a noteworthy feature of best interests decision making in adult kidney care. As Hayden J’s comments in GUP v EUP & Another Footnote 99 suggest, compromise is not in itself something to shun. Indeed, it can be a practical means of proceeding where disagreement is strong. Our concern, however, is with what we have termed rotten medical compromise – which is, we argue, always something to be avoided.
We stress the important distinction between flexibility in best interests decisions that arises from the deciding healthcare professional being genuinely convinced that dialysis is worth trying, and those where the compromise is simply the path of least resistance and pursued solely to placate P’s family. The former may be justified, but the latter exemplifies rotten medical compromise. Out of recognition of P’s continued preferences as distinct from capacity, a decision made on the basis of avoiding conflict with the family cannot be defensible in any but the most trivial of decisions – certainly not that of dialysis. To do so is to instrumentalise P rather than centre them in questions about their own care. There is certainly something to be said for a professional having a duty of care to the family of P, but this should never extend to undermining P’s own best interests.
On a practical level, addressing rotten medical compromise may be feasible through a combination of interventions, perhaps most obviously through improved education of healthcare professionals around the MCA 2005 – ensuring their understanding of the requirements of the legislation and the roles of different parties in best interests decisions. But also, we consider there to be a need for structural change such that healthcare professionals feel able to resist the demands of families where they do not consider them to be in P’s best interests. That is, avoiding rotten medical compromise requires an environment in which healthcare professionals feel able to engage their hospital’s legal team for constructive discussion and appropriate support to enable what they, in their clinical judgement, consider to be in P’s best interests. We do not explore these possible solutions in detail but acknowledge them as a subject for future work.
Whilst our focus here is on best interests decisions made in the context of kidney care, it is feasible that a similar phenomenon arises in other healthcare contexts. Disagreements between families and healthcare professionals are by no means unique to kidney care, so we would imagine the path of least resistance would be taken in other areas too – albeit manifesting in a different form depending on the nature of decisions being made. Certainly, there are indications of similar in intensive care.Footnote 100 This, we suggest, is an area in need of further examination.
Open access
