Terminally Ill Adults (End of Life) Bill

Baroness Hollins Excerpts

Committee stage
Friday 6th February 2026 Lords Chamber

[Baroness Hollins (CB)]

My Lords, I added my name to Amendment 87, which details the exclusion of those with diagnosed eating disorders from accessing assisted dying, because I consider this a tricky area deserving of very careful debate. There are serious and specific risks that the Bill poses to people with anorexia nervosa and other eating disorders—conditions with the highest mortality rate of any psychiatric illness, but which are not terminal illnesses. They are complex fluctuating disorders, characterised by impaired insight and a distorted relationship with food, weight and how people view themselves.

Written and oral briefings that we have received make it clear that this is not just a theoretical concern. The Complex Life and Death Decisions group briefing on eating disorders documented how, in jurisdictions with assisted dying laws, eating disorders have been redefined as terminal once patients stop eating and drinking. This is despite the availability of effective evidence-based treatments and well-documented cases of recovery, even after decades of illness. A systematic review published last year in Frontiers in Psychiatry identified at least 60 individuals with eating disorders who died by assisted death in Oregon, Colorado and California—jurisdictions where, as in this Bill, assisted dying is restricted to terminal conditions. A third of those who died were in their teens or 20s, and all were women.

Cases such as this demonstrate the danger of conflating treatment refusal with terminal illness, and of mistaking the voice of a life-threatening psychiatric disorder for a wish to die. Starvation affects the brain in profound ways, including causing rigidity of thinking and allowing the illness itself to dominate choices. This can happen even to the point where perceived thinness is valued over life.

Yet people with anorexia can appear highly articulate and lucid, which can mask these deeper impairments and make capacity assessments unreliable. Many are judged to have capacity in other areas of life even while their decision-making around food, weight and choosing self-preservation is seriously distorted by the disorder. Someone may seem capable of making an end-of-life decision while simultaneously lacking capacity for the very decisions that would keep them alive. We need to acknowledge this complexity and proceed with the utmost care, recognising that these individuals are among those most at risk under the Bill.

Those who refuse to eat or drink because of mental illness must also be excluded from the Bill’s scope so that we can properly safeguard them and ensure that they receive the treatment and support that they need to recover. Amendment 87A would offer a valuable clarification, ensuring that this would apply regardless of whether the refusal of food or drink is voluntary or arises from a mental disorder. This would prevent any ambiguity that could otherwise allow conditions such as catatonia or other serious mental illnesses to be misclassified as terminal under the Bill. It would provide further safeguards for people with mental disorders.

There are mental-health conditions where life-threatening physical complications can occur without the restriction of food and drink. Amendments 90, 92 and 103 also cover this key area and would ensure that assisted dying is limited to people who are genuinely terminally ill due to disease progression and not because they choose to refuse care, food, fluids or treatment. A person cannot be considered terminally ill if their life-threatening condition arises from the secondary effects of mental illness.

[Baroness O’Loan (CB)]

My Lords, the issues raised in this group are extensive and a matter for great concern. Refusing food and hydration is a sure way to die, as, in many cases, is refusing treatment such as dialysis and insulin. Refusing food or treatment is a choice that a patient is entitled to make, if they have the capacity to make such a decision. However, that cannot translate into an obligation on the state to help the person to kill themselves. That was recognised, as the noble and learned Lord, Lord Falconer, said, by the Bill’s sponsor in the other place, hence the wording of Clause 2(2). However, the noble and learned Lord’s Amendment 87 would remove the safeguard offered by Clause 2(2), as it would prevent the Bill applying only where the person has a mental disorder and the act of refusing food or hydration causes them to have an illness or disease.

Many states start by offering assistance to die in very limited circumstances, but, only a short time afterwards, the grounds for the state helping people to kill themselves expand—in some cases, almost inconceivably. Do you have no wheelchair? We will help you die. Do you have no home? We will help you die. Are you suffering from PTSD after military service? We will help you die. In many jurisdictions, anorexia, which is a treatable illness, is accepted as justification for assisted suicide and so on. All the examples that I have just given—no wheelchair, no home and suffering PTSD—were grounds for assisted dying in Canada recently.

There are many suggestions for how to address the terrible vacuum through which all sorts of conditions that are not terminal could, as a consequence of refusing food, treatment or hydration, become terminal. I put my name to one such amendment, Amendment 101, in the name of the noble Baroness, Lady Parminter, and all the amendments in this group, other than Amendment 87, seek to address the issue. However, it is clear that the list of conditions and situations currently provided in Clause 2 as exceptions to the definition of terminal illness is not adequate. The noble and learned Lord, Lord Falconer, has even removed from the Bill the more general safeguard inserted by the MP for Batley and Spen. The noble and learned Lord will have to move far beyond his Amendments 87 and 87A if he is to satisfy the Committee that the definition of the situations that do not qualify as terminal illness is in any way satisfactory.

The prevalence of eating disorders goes way beyond what one might perhaps reasonably expect. NHS statistics for 2023 tell us that the prevalence of any eating disorder in the age group of 11 to 16 year-olds was 2.6%, but that in the case of girls it was 4.3%, compared with only 1% in the case of boys. If we move to the group of 17 to 19 year-olds, the figure is 12.5% in the whole group but, among girls, the number suffering from an eating disorder rises to 20.8%, compared with 5.1% among boys. Those are NHS figures. A review of current literature up to 2022 investigating suicide in eating disorders found that the prevalence of suicidal ideation was 51% and the prevalence of suicide attempts was 22%. I am very grateful to the young intern doing a PhD who provided me with these figures from the NHS. They demonstrate the magnitude of the problem of eating disorders today and, even more terribly, the 51% prevalence of suicidal ideation. Half of all people suffering from anorexia or other eating disorders will contemplate suicide, and more than one in five of those suffering from an eating disorder will attempt to take their own lives, although the disorder that they suffer from can be treated if they get proper treatment.

The Bill effectively creates a class of people—particularly disabled people, diabetics, those on dialysis, and others who need continuous life-sustaining treatment—who can at will cause themselves to have a terminal illness as defined by the Bill. Experts and the courts have agreed that refusing such treatment and creating a terminal illness out of a desire to die does not vitiate capacity. Can the noble and learned Lord, Lord Falconer, explain how creating a class of disabled people for whom voluntary suicide is legal and state-supported is consistent with the UK’s obligations under Articles 2 and 14 of the European convention?

--- Later in debate ---

[Baroness Hollins (CB)]

My Lords, it is a pleasure to follow my noble friend, who has presented a different approach to assessing capacity. I was a member of the post-legislative scrutiny committee on the Mental Capacity Act. Both in my professional capacity as a learning disability psychiatrist and as a family carer, I have observed the misapplication of the Act on rather too many occasions. It is quite a complicated Act and it is often misunderstood. I have serious concerns about its applicability and appropriateness for this purpose. I recently co-authored an editorial on this subject in the journal BMJ Supportive and Palliative Care. My Amendments 117 and 892 propose a bespoke decision-making framework, designed specifically for assisted dying, which would of course need to be evidenced in practice.

As the mother of a neurodiverse son, I share the concerns expressed by the honourable Member Daniel Francis in the other place. He is the parent of a daughter with a learning disability, and said:

“The concern that haunts me every single day is, ‘Who will make, and how will they make, those decisions to support my daughter when my wife and I are gone?’ That concern, shared by thousands of others, led me to the conclusion that the Mental Capacity Act 2005 was not written for this scenario”.—[Official Report, Commons, 16/5/25; col. 691.]


As a parent myself, I completely identify with what he said. I really worry about the suggestibility of so many people with learning disabilities and the possibility that they might be offered an assisted death. I think they should probably not be eligible. For example, they might mask their lack of understanding. If you ask my son whether he wants to do something, and if he understands it, he will be delighted to say that he does, when I know he has not got a clue. He wants to be consulted and invited and to participate and to have the same opportunities as everybody else, when he does not have the first understanding of what actually he is being offered. That is the reality of living day by day with somebody who does not understand. It is really difficult to know.

From my research looking at the Netherlands data, we found quite a significant number of people who have been offered euthanasia, and have been euthanised, for whom a capacity assessment had not been done, certainly not by anybody who had any actual experience of working with neurodiverse people. This is just not acceptable.

The Mental Capacity Act was designed to support and protect people when they lack decision-making capacity. Conceptually, it is a deficit-based framework intended to identify the absence of capacity so that decisions can lawfully be made on a person’s behalf and in their best interests. It begins from a presumption of capacity, and only if there is a justifiable reason for doubting capacity for a specific decision at a specific moment is there any requirement to assess the person’s understanding. There is no such thing as being “capacitous” or “incapacitous”. The Bill, as written, requires capacity to be demonstrated, but this contradicts the presumption of capacity in the Act and creates a legal inconsistency. The Bill does not resolve these ambiguities, leaving it unclear whether professionals are expected to apply the Mental Capacity Act as a deficit-based framework or to repurpose it as a positive authorisation tool.

The Royal College of Psychiatrists has serious concerns regarding the use of the Mental Capacity Act and recommends a formal review of its suitability as the decision-making framework for assisted dying, which is why I have proposed an alternative. Assessing capacity is also inherently subjective and vulnerable to bias. Judgments are influenced by the assessor’s values, the context and the relationship with the person being assessed. These biases are likely to be amplified where assessments are undertaken by a limited group of professionals working in a context in which requests for assisted dying are framed as expressions of autonomy.

Such a setting risks subtly predisposing assessors towards affirmation rather than critical scrutiny, particularly where capacity is treated as something to be confirmed. An article published in the British Journal of Psychiatry this week by two prominent psychiatrists and psychotherapists highlights exactly this: the role that unconscious psychological processes play in decisions related to assisted dying. These influences are not currently being considered when assessing decision-making capacity but are extremely relevant.

Near the end of life, cognitive impairment is common and undermines genuine decision-making capacity. A patient may articulate clear preferences despite a compromised ability to understand, retain or weigh information. A study of hospice patients with no clinically obvious cognitive disorder of impairment, published in the American Journal of Geriatric Psychiatry, found that over half had significant previously undetected cognitive impairment on formal neuropsychological testing. In the same journal, another study of terminally ill cancer patients receiving palliative care found that cognitive impairment was common, with just under half showing impairment in understanding and appreciation, and around four in five showing impairment in reasoning. Importantly, there was poor agreement between the physician’s judgment of capacity compared with more structure neurocognitive testing.

Those and other studies show that evidence-based structured assessments are essential in end-of-life context for decisions with irreversible consequences. It is inappropriate for the Bill to require psychiatrists to assess capacity. We are trained to diagnose and treat mental illness as well as prevent suicide. There is no requirement in the general psychiatry specialty curriculum for expertise in assessing decision-making capacity or detecting cognitive impairment in terminal illness. So, requiring psychiatrists to confirm capacity in assisted dying means treating a wish to end one’s own life as an autonomous and valid choice in one context while continuing to regard it as a symptom of mental illness in another. What a confusing position to find oneself in.

Other professionals, such as practitioner psychologists, may be better positioned to evaluate decision-making in end-of-life contexts due to their expertise—a position supported by the British Psychological Society.

The Mental Capacity Act and the current approach to capacity in the Bill are not fit for purpose, which is why I have proposed a different approach. Obviously, this would require expert development and testing.

--- Later in debate ---

[Lord Falconer of Thoroton (Lab)]

The framework that is to be applied by the co-ordinating doctor and the independent doctor is the framework for assessing capacity under the Mental Capacity Act. That is what the Bill says, and I am resisting the changes in Amendment 115, in the name of the noble Baroness, Lady Finlay.

The noble Baroness, Lady Hollins, is proposing something rather different from that which the noble Baroness, Lady Finlay, is proposing. She proposes that the Secretary of State must, by regulations, establish a framework for the assessing and determination of capacity; that having capacity is to be read in accordance with that framework and not the framework under the Mental Capacity Act; that regulations must involve setting out an evidence-based methodology; and that any assessment must be conducted by a practitioner psychologist who holds particular qualifications and requires the use of validated, standardised instruments capable of detecting subtle or partial impairments in cognition, reasoning or decision-making capacity.

I understand from the speech that the noble Baroness made, and from her splendidly detailed amendment, that particular specialists applying a different regime from that which already applies, with specialised tests, should be the only people who can give a capacity assessment here. Again, in my respectful submission, that falls foul of the same arguments that I advanced before. The first of those is that the Mental Capacity Act is tried and tested in dealing with the question of whether someone is able to make the decision. Secondly—and I am glad the noble Baroness, Lady Finlay, gave me the opportunity to answer this—the co-ordinating doctor and the independent doctor will require training, and the panel will have experience in making these decisions. Thirdly, they will be making decisions having received proper advice.

[Baroness Hollins (CB)]

A bespoke decision-making framework would initially be legally untested, but the same is true of the Mental Capacity Act, which, as my noble friend Lady Cass said, has never been operationalised or judicially tested in the context of assisted dying. The idea that the two doctors involved would be trained ignores my point that this kind of complexity is not something that can be learned simply through an extra little module. It requires years of experience to help people develop the understanding and ability to make these kinds of assessments. Further, the existence of bespoke decision-making frameworks is not novel in UK law. The Mental Health Act already operates in power over the Mental Capacity Act, recognising that certain contexts require a different approach to decision-making.

[Lord Falconer of Thoroton (Lab)]

I do not want to be disrespectful, but that was just a repeat of the speech that has already been made and it did not contain a question. I completely respect the arguments of the noble Baroness, Lady Hollins, but she should really only ask me questions, rather than make a speech in the middle of my speech. I apologise for having to say that, but she has not asked me a question.

[Baroness Hollins (CB)]

Let me frame it as a question then. Does the noble and learned Lord agree that there are already bespoke decision-making frameworks and that this is not novel in UK law?

[Lord Falconer of Thoroton (Lab)]

I agree that there are bespoke decision-making processes in existence. It is not the law; it is how it works in practice. Through codes of practice, it will be possible to develop appropriate methods of dealing with it.