Ethical issues in restraining patients for dialysis

By Jordan A. Parsons – 21st October 2020

It needed four people to restrain him.  I explained to him that he didn’t have capacity to make the decision to refuse dialysis so we were going to give it in his best interests and we would sedate him to do that.  He curled up in a ball in the top corner of the bed and pulled the sheet over his head.  Two psychiatric nurses and two security guards held a limb each so that he couldn’t kick out.  I got an injection into his upper arm and a cannular into his hand.  He was shouting through most of it, ‘Leave me alone – I want to die’.

This is a doctor’s description of using restraint to provide dialysis to a patient with kidney failure (whom we call “Paul”), provided in an earlier blog on this website here.

I read this blog post with particular interest as I am currently researching best interests decisions in relation to dialysis. In carrying out a scoping review as part of this research, it struck me how frequently the question of dialysing uncooperative patients arose. Indeed, there have been many cases like Paul’s. One in the United States is discussed by O’Dowd and colleagues, with the situation described as follows:

‘We ended up deciding that he would be treated against his will which at times involved dragging a kicking, screaming, hitting person, who may have been HIV+, down to dialysis, strapping him down for 4 hours, putting needles in his arms, and dialyzing him. The dialysis staff was not happy about this and it wasn’t clear that it was the best thing or the right thing to do’.

Perhaps, then, the avoidance of restraint is more highly valued in some cases than the life extension that dialysis would provide (be that from the perspective of the patient, the patient’s loved ones, or the clinical team). Dialyzing a severely uncooperative patient would therefore not always be in that patient’s best interests.

Among patients without capacity to consent, both verbal and physical resistance to dialysis are not uncommon among those with kidney failure. Neither is lack of capacity. The average age of a patient with kidney failure in the UK is over 60, so there is a high prevalence of dementia, along with other neurodegenerative diseases and various vascular comorbidities. However, there are other causes of cognitive impairment that might lead a patient to not have the necessary mental capacity to make their own decision about dialysis, e.g. learning difficulties and psychiatric conditions. In the case quoted above, the patient (“Paul”) has been diagnosed as having a borderline personality disorder. Another blog on this website discusses a case in the Court of Protection concerning a patient with ‘a complex psychotic illness’.

Two aspects of Paul’s treatment particularly interest me. First, that the restraint required is rather extreme. Second, that the patient had expressed a clear desire to be restrained as he wants to be dialyzed.

Restraint ‘at the boundaries of acceptable practice’

To dialyze Paul requires, on occasion, both physical and chemical restraint. In doing this, Paul’s clinical team felt ‘at the boundaries of what they consider to be acceptable practice’. The use of general anaesthetic to enable dialysis carries with it a risk of brain injury and death. For these reasons the clinical team sought a declaration that it was not in Paul’s interests to have dialysis forced upon him.

However, even though the restraint in this situation is extreme and the risks high, the alternative is certain death. It is very much a matter of determining the best of two options both that would ordinarily be avoided where possible.  So how best to respond to such cases?

For now, I will consider this question in the abstract, thereby setting aside the fact that Paul has, when capacitous, expressed a desire for dialysis even if restraint is necessary. Cases like Paul’s are not a simple matter of a patient with dementia getting restless and causing a minor disruption. Further, the impact of resistance on this scale extends beyond the patient in question. One might argue that in such a scenario it would be appropriate to consider the wider impact of a patient’s noncompliance. In particular, (1) dangers posed to others and (2) moral distress among the clinical team.

Danger to others

What has been made clear in Paul’s case is the danger of restraint to himself – brain injury and/or death. A best interests decision is about what is best for the patient in question – but from a broader ethical viewpoint it’s important also to consider that a particularly resistant patient might prove a danger not only to themselves, but those around them.

For example, if a patient attempts to disconnect themselves from a dialysis machine, there is the potential for a needlestick injury to staff members. This, in turn, might present a risk of HIV transmission if that patient is HIV+ (the patient’s HIV status may be unknown). Anyone present at the time of such a patient’s dialysis sessions could be put at risk of this, including not just staff, but also other patients, and any family or friends accompanying the patient.

At present, and potentially for an extended period, there is an added risk of staff becoming infected with COVID-19 if they are required to be more physically closely involved in a patient’s care. I have explored the dialysis-specific ethical issues in an article published elsewhere with colleagues, so will not dwell on them here.

The occurrence of behaviour such as Paul’s may also cause significant distress to other patients present at the time, potentially causing them to become agitated and so impacting on their own care (i.e. abruptly disconnecting themselves from a dialysis machine).

Given these risks to persons beyond the patient in question, the clinical team may face a difficult decision. Even if dialysis is considered best for the patient, there surely must be a point at which danger to others supersedes this.

Sedation is a route to enabling dialysis without significant physical struggle, but this option carries its own risks which must surely be accounted for. It has on several occasions been argued that cooperation ought to be a prerequisite for dialysis, and that the need for restraint is evidence that dialysis is no longer appropriate (e.g. Spike 2007; MacPhail et al, 2015).  I am not sure that I agree to quite that extent, but certainly a line must be drawn when a patient is hugely disruptive and poses a danger to those around them.

An argument might also be made from a resource allocation perspective, as a patient requiring restraint will necessitate the involvement of several members of the clinical team. However, equally, it is argued that a nephrologist should, when making decisions for individual patients, not consider wider issues of cost and resource allocation (McDougall, 2005).  I agree that such a decision would be most appropriately made at a policy level, and that the use of resources is not pertinent to the best interests of an individual patient.

Moral distress

Moral distress arises when the course of action a clinician is being asked/instructed to take does not align with what they consider to be the ethically appropriate course of action (Ducharlet et al, 2020).  There is an important distinction between distress and moral distress: a clinician may well find it distressing to provide care that they agree is in the best interests of the patient, but that is not moral distress.

In a case such as Paul’s, moral distress might be experienced by the clinical team if they are of the opinion that restraint for dialysis to such an extent is not in the patient’s best interests. This would be a very reasonable opinion to have when physical and chemical restraint are required for a patient who is hugely resistant to care, and it seems to have been the opinion of Paul’s treating clinician.

A nurse who is providing care that has been authorised by the patient’s doctor may experience moral distress if she does not believe it is the right thing for the patient. A doctor who authorises care that she does not consider best for the patient as a result of pressure from the patient’s family may also feel moral distress. In the case of Paul, the Court of Protection authorised both physical and chemical restraint of Paul, so any member of the clinical team who disagrees with the Court’s decision may experience moral distress in carrying out the authorised care. Of course, an individual member of the clinical team could express an objection to being involved in the restraint of Paul and hope that she may not have to participate if there are enough other members of the team. However, there is no absolute legal right to conscientious objection when it comes to restraint for treatment.

This potential for moral distress raises the question of how (or if) it ought to factor into decision-making. After all, the nature of moral distress is such that the individual experiencing it may have a useful perspective on what care is and is not appropriate for a particular patient. At the very least, moral distress ought to be acknowledged, and an environment should be fostered in which all members of a clinical team feel comfortable voicing their perspectives in the knowledge that they will be listened to. Whilst moral distress is an occasional inevitability, clear and open communication might at least partly mitigate it.

Restraint at the patient’s request

Matters become more complex when restraint is at the patient’s request. Paul presents his clinical team with two conflicting choices – one when he is ill (ill-Paul) and another when he is well (well-Paul). Which Paul should prevail if the aim is to respect Paul’s autonomy?

In this case, the Court sided with well-Paul’s choice, permitting the conflicting choice of ill-Paul at the time of treatment to be overruled. In doing so, an apparent distinction was made between what can be considered Paul’s first- and second-order preferences. Paul’s first-order preference (meaning that which is somewhat base and instinctive) was to not be dialysed, whereas his second-order preference (meaning that which is rather more reasoned and, to use a contentious term, authentic) was to be dialysed so that he could spend more time with his son. Given the account of well-Paul as ‘articulate, thoughtful and intelligent’ during the hearing, the Court’s decision is understandable. Few would deny that his deeper preference is for treatment and that his occasional objections are an unfortunate result of his mental state. Sometimes, however, it may not be as clear.

When the line between the (cognitively) “well” and “ill” patient is more blurred, it may not be as apparent which choice is truly (most?) autonomous. The clinical team – and the Court – may be less comfortable authorising restraint based on one of two conflicting choices expressed by the patient if there is less conviction and demonstrable reasoning behind it. Well-Paul was quite clear as to why he wanted dialysis, but a rather more diffident expression of the want for dialysis may not be perceived as sufficient to outweigh the conflicting choice against dialysis expressed at the time of care. Whilst a best interests decision is not a simple matter of substituted judgement, conflicting choices by a patient will inevitably complicate the decision where one is not clearly more representative of the patient’s true preferences.

Further, there is a question of whether well-Paul is actually entitled to demand that ill-Paul suffers. When ill-Paul clearly demonstrates distress at receiving dialysis, is he to be afforded no autonomy? It has previously been argued that where dialysis is considered to cause unnecessary suffering by clinical staff and the patient’s family it would be appropriate to cease dialysis even if the patient had previously made it clear that she wanted to continue (Kaye and Lelia 1986).  Therefore, even if the more authentic preference of a patient is clearly for dialysis, there may still be a point at which providing dialysis to that patient in their later extremely distressed state is inappropriate. This is an interesting point to consider as it brings into question the relative importance of the patient’s own preferences.

Can Paul’s “authentic” choice change?

A further challenge might arise if the patient’s authentic preferences change. Now that restraint has been authorised to dialyse Paul against his objections, what would happen if well-Paul decided that he no longer wanted dialysis?

Given how things have played out thus far, the clinical team may not recognise any future objections to dialysis as authentic. This may especially be the case if Paul’s mental state deteriorates so that there is no longer a clear distinction between well-Paul and ill-Paul.

Moreover, as time goes on it may be that Paul genuinely does change his mind and wish to discontinue dialysis. It is one thing to say that you want to be restrained for dialysis, but another to experience it repeatedly. A reasonable person might well change their mind. In essence, it is verging on torture.  Interestingly, a study in the US and Germany found the prevalence of post-traumatic stress disorder as a result of haemodialysis to be 10.4%. (Tagay et al,  2007).  An important component of patient autonomy is the right to change one’s mind. In situations such as Paul’s, it is very possible that this right could be (unintentionally) overlooked.

Deciding whether to dialyze a cognitively impaired patient when restraint is necessary is not straightforward. There are myriad factors at play, and decisions need to be made on an individual basis. It is also important that the harm of restraint itself is a key consideration in such decisions due to the potential for lasting trauma. Further, whilst I am as yet unsure of where exactly I think the line should be drawn, I am convinced that there must be a point at which harm to others is reason enough to discontinue dialysis for physically resistant patients.

Jordan A. Parsons is a PhD student based at the University of Bristol’s Centre for Ethics in Medicine. His doctoral research is an empirical bioethics project questioning if and when it is in the best interests of a patient with end-stage kidney disease who lacks decision-making capacity to forego dialysis in favour of conservative management.  He tweets @Jordan_Parsons

---