In a society where EAS is legalised and a clearly viable option in the eyes of a large proportion of the population, as is the case in the Netherlands (where currently around 1 in 25 deaths are through EAS ), there should be clear and compelling reasons if a person with an intellectual disability or autism spectrum disorder is to be denied this option.
Our findings are in line with those of Kim et al.  who reviewed 66 Dutch case summaries of patients receiving EAS for psychiatric conditions. In a discussion of capacity evaluations in these 66 cases, two of the cases we examined (2013–22 and 2014–83) are described in detail . They conclude that
“the practice of EAS for psychiatric disorders involves complicated, suffering patients whose requests for EAS often require considerable physician judgement. The retrospective oversight system in the Netherlands generally defers to the judgements of the physicians who perform the EAS.”  (p.367)
Dutch physicians thus carry significant responsibility in deciding whether or not a patient’s EAS request can be met. Here we discuss the difficulties they face in applying the due care criteria to the situations of vulnerable patients.
Voluntary and well-considered EAS request
The RTE Code of Practice 2015 states that “decisional competence means that the patient is able to understand relevant information about his situation and prognosis, consider any alternatives and assess the implications of his decision”  (p12). The RTE highlights the importance of careful capacity assessments if the patient has an intellectual disability . The case summaries have shown how difficult such assessments can be, and it is therefore particularly unhelpful that so little detail is given about how the conclusion was reached that these patients had capacity.
Under Dutch legislation , as in many other jurisdictions including the UK , capacity is decision-specific. There is a presumption that capacity should be deemed to exist unless there is evidence to the contrary; in other words, the test for capacity involves disproving someone’s competence for a particular decision. Dutch government guidance states that capacity assessments should be based primarily on a patient’s ability to comprehend the decision, rather than on the outcome of the patient’s decision, as focusing on the latter risks the assessor’s norms and values being decisive. However, it goes on to explain that:
“This does not mean that the assessor must put aside his own norms and values completely. In dialogue between the assessor and the patient, their respective norms and values can influence each other; but in assessing decisional capacity, the assessor’s norms and values must not be unilaterally decisive. However, the nature of a patient’s intended decision can be a reason for assessing his decisional capacity.”  (p.7, translated by ITW)
This guidance seems to recognise the perhaps difficult-to-avoid fact that a physician’s ideas about what is best for the patient may influence the patient’s choices.
The most commonly used and influential model for assessing whether a patient has the capacity to exercise autonomy in making healthcare choices is the MacArthur model, which consists of four abilities (also known as the “Appelbaum criteria”): to understand the illness and treatment-related information; to appreciate the significance of that information; to weigh up options using reasoning and logic; and to communicate a choice  (see Additional file 2). The Appelbaum criteria were referred to by Dutch physicians and the RTE as their model for capacity assessments. The use of this model for EAS decisions is not without its difficulties. Whilst impairment of decision-making capacity lies on a continuum, the judgment of decision-making capacity is an all-or-nothing concept (either the patient has capacity to take a particular decision, or he doesn’t). Where on the continuum the cut-off point for competence lies, is therefore a matter of physician judgement. Appelbaum  has argued that the stringency of capacity tests should vary directly with the seriousness of the likely consequences of the patient’s decisions. In the reported case summaries, it appears that the bar is not set high. In one case, the physician stated explicitly that a patient did not have all four Appelbaum abilities, but she was still deemed to be competent. This “lower threshold view” was also noted by Doernsberg et al.  in their review of capacity evaluations of psychiatric patients requesting EAS in the Netherlands. However, for decisions involving EAS, applying capacity laws and guidance is challenging. Legislation and guidance governing capacity assessments have not been written specifically for assessment of EAS requests; indeed, the Mental Capacity Act 2005 (England and Wales)  specifically states at Section 62 that nothing in the Act affects the law relating to assisting suicide. Wide variations and inconsistencies have been reported in the way mental capacity in patients requesting hastened death is conceptualised, but many consider that EAS should only be open to those with a high degree of mental capacity to make such a request . In other words, for a life-or-death decision such as EAS, the bar should be set high, with patients meeting all aspects of stringent capacity assessments. Whether the MacArthur model is in fact appropriate for such assessment is open to debate and requires a comparative prospective research study. The model has been criticised for its focus on cognition, understanding and reasoning, ignoring the importance of emotions, values, biographical and context-specific factors [26,27,28,29]. A study by Tan et al.  showed that competence to refuse treatment may be compromised in patients with anorexia nervosa because of their difficulties with thought processing and changes in values, which was not captured by the standardised tests for competence. Breden and Vollmann  argue that in reality, patients often base a preferred choice on emotions, values or intuitive factors rather than on a rational analysis of all the options. This debate is important and may well lead to future changes in the way decision-making capacity is conceptualised and assessed, but at present, the MacArthur model is reflected in legal frameworks as well as guidance to Dutch physicians. Because of the complexities of capacity assessments and the judgement required of physicians, even within the seemingly clear Appelbaum criteria, it is especially important that the specifics of such assessments are described with greater transparency than we have found within the case summaries. Stringent capacity testing and a high bar for decision-making capacity, under the current model, would mean that the patient has all four Appelbaum abilities. We could not find evidence of such stringent testing in the Dutch case summaries. Most contained no detail of how the Appelbaum criteria were operationalised.
The RTE reports put a heavy emphasis on the consistent or repeated nature of an EAS request, but for people with an intellectual disability, the difficulties with decision-making are more commonly in the area of ‘appreciating the significance of the information’ or ‘reasoning with the information and weighing up treatment options’ (see Additional file 2). An example of how a consistent treatment refusal can be wrongly seen as a competent and autonomous decision is given in a UK study of patient safety issues, where a senior hospital consultant initially thought a patient with intellectual disabilities was being clear and consistent in his treatment refusal:
“[The patient] had cancer and needed surgery. I didn’t realise that he didn’t have the capacity to say ‘no’ to the operation. He didn’t want the operation, and I just thought that was that.”  (p.98)
It wasn’t until this patient was seen by the hospital’s intellectual disability nurse specialist, who found that the patient did not fully appreciate the consequences of his decision, that he was found to lack capacity; the patient, who would have died without surgery, had the operation after all, following a ‘best interest’ decision.
A persistent request does not necessarily imply capacity; it might even be indicative of a lack of capacity, if the patient’s intellectual disability leads to difficulties in considering or weighing up alternatives. In case 2015–24 (the woman with tinnitus), the consistency of the patient’s EAS request took precedence over her ability to consider alternatives, which the physicians agreed was impaired due to her low level of intelligence. This was not helped by the fact that “often the wrong treatments had been instigated” and “practitioners had not encouraged her to try and persevere with these treatment(s) a bit longer”. If stringent capacity assessment criteria were to be applied in this case, it may well be that the patient’s inability to appreciate the significance of the information in relation to her own situation, and to weigh up treatment options, would render her incapable to make an EAS decision, however persistent her EAS request. In case 2013–22, the patient interpreted light flashes related to her eye disease as “people shining lights into her house” and was disturbed by this; she “could not be convinced” by the physician’s efforts to help her understand the cause of her symptoms. Under stringent capacity testing, such inability to understand the nature of her condition could have led to the conclusion that the patient lacked decisional capacity. As the Dutch government guidance states:
“In assessing whether the patient understands the nature of his health condition, one aspect that must be checked is whether he can see the relevant causal links, or conversely, whether he ascribes his health condition to a cause other than the ‘objectively’ determined one. A lack of insight into his condition is a factor that can be important in assessing decisional capacity.”  (p.12, translated by ITW)
It is evident from the RTE verdicts that it does not require patients to meet all four Appelbaum criteria. The difficulties many people with intellectual disabilities have in rationally manipulating information with regard to their situation, the consequences of their decision and the possible alternatives make them particularly vulnerable when the bar for capacity assessment is not set high. Of all four Appelbaum criteria, appreciation ability is undoubtedly the most difficult to understand and measure. This conclusion was also reached in a systematic review of decision-making capacity of patients with depression, which found that appreciation of information was the ability that was most impaired, but difficult to measure within standard capacity tests .
Making a “voluntary and well-considered request” is closely related to the concept of autonomy. Autonomy in decision-making is complex for people with intellectual disabilities and not easily achieved. Whether a person is able to have a high level of self-determination is not necessarily dependent on their intellectual ability. A study of 301 adults with intellectual disabilities found that experience and opportunities to make everyday choices contributed significantly to greater self-determination and autonomy; intellectual capacity, on the other hand, was not a significant factor . People with intellectual disabilities generally have less experience of and control over major decisions that affect their lives than those without intellectual disabilities , and therefore may lack the necessary skills in making life-changing decisions.
A recognition that people with cognitive conditions may need not only highly skilled capacity assessments but also robust support in the decision-making process is not clearly addressed within the Dutch case summaries. The presence of a ‘mentor’ at patient-physician meetings was mentioned in two case summaries, but family was notably absent. It may be that family and carers were involved but that this was not reported in the case summaries; or that the reporting physicians wanted to be sure that the patient was not under undue influence of family or carers (‘voluntary request’). It seems, however, that for vulnerable patient groups, the perspectives of those in the person’s social circle can be invaluable. In the UK, inquiries into health and social care provision for people with an intellectual disability have found consistently that including and listening to their family and carers is important in ensuring patient safety [33, 34]. The guidelines for testing impairment under the Oregon Act  include the recommendation to consult, where possible and with the patient’s consent, the patient’s significant others about their views on the patient’s competence and the patient’s decision. Similarly, the Dutch government guidance on capacity assessments includes the statement that “the decision about capacity is often difficult and should be discussed with the various professionals and carers involved… this is particularly important for serious decisions” . We found no evidence in the case reports that physicians included the perspectives of family and carers who knew the patient well.
Unbearable suffering without prospect of improvement
Under the Dutch EAS system, it is up to the physician to decide whether the patient’s suffering is indeed unbearable; this is then verified by the independent consultant. In the 2016 annual report, it is explained that the review committees
“do not re-examine the same issues as the physician who made the original decision. The RTE cannot do this, as the patient is no longer alive.”  (p.24)
The RTE therefore scrutinises whether or not the physician could have reasonably come to a conclusion about the degree of suffering, but does not decree whether the physician’s verdict is correct. This places the responsibility for assessing suffering solely on the shoulders of physicians. It is important to remember the gravity of the physicians’ decision: in agreeing that suffering is unbearable and without prospect of improvement, they ultimately agree that it is better for the patient to be dead than to live.
There are obvious difficulties with implementing this criterion. Suffering is a deeply subjective experience, and it could be argued that no person can ever truly understand the experience of someone else’s suffering, nor assess whether such suffering is severe enough to make the patient’s life not worth living . Cormack and Fléchais  point out that suffering needs to be interpreted in the context of the patient’s personality structure, coping mechanisms and psychosocial environment, and argue that the Dutch requirement of suffering being ‘palpable’ to doctors raises concerns about counter-transference. The word ‘palpable’ is a perhaps inadequate translation from the Dutch ‘invoelbaar’, which implies that the physician needs to ‘feel’ or ‘imagine’ their patient’s suffering, from the patient’s perspective. The difficulties with this requirement are clear from those case reports where physicians struggled to understand the patient’s perspective. Case 2016–48 (a woman with autism spectrum disorder) is a striking example. The physician visited the patient several times but was unable to understand the patient’s view of her own suffering. He trusted a psychiatrist who told him that the experience of suffering is different for people with autism spectrum disorder. A physician’s ability to empathise and ‘feel’ his patient’s suffering is a guideline specified in the RTE Code of Practice , but it seems that because of the patient’s specific disabilities, it was suggested that this was not possible. Physicians’ acceptance that the suffering of patients with particular conditions may not be obvious to them could have the worrying implication that for people with some conditions or disabilities, a different (lower) standard for judging the severity of the suffering is accepted.
It has been made legally clear that a somatic or psychiatric medical condition must be the cause of the suffering that leads to the EAS request, but there is a lack of clarity around the question whether lifelong disability would qualify as a reason for EAS. If the answer is ‘yes’, the implications for people with disabilities could be very serious. In this context, it is of concern that for several of the patients’, suffering was described in terms of their lifelong condition, rather than their psychiatric illness (see, for example, cases 2013–21, 2013–22 and 2014–77). Autism spectrum disorders and intellectual disability, with concomitant difficulties with social communication and relationships, could make it more difficult for patients to cope with the changes that come with ill health and ageing, or to weigh information, or to understand and accept possible alternatives. In the case summaries, the term ‘suffering’ was used to describe the normal variations in behaviour and perceptions seen in people with autism spectrum disorders which are an inherent part of the person. There were no explanations of how a lifelong condition such as autism spectrum disorder could be seen as “no longer” curable, nor how it might be treated palliatively. Statements about the lack of prospect of improvement, such as “intractable symptoms”, “refractory to treatment” and “palliative treatment”, are meaningless in the context of lifelong disability. This raises the prospect of diagnostic overshadowing, where there is a negative bias impacting on a clinician’s judgement regarding co-occurring conditions in people with an intellectual disability or mental health condition, and symptoms due to a specific condition are attributed to another .
It seems clear from the Dutch case summaries that the patients were indeed suffering, and that they viewed their suffering as unbearable. When suffering is the result of a progressive medical condition, its assessment may be relatively straightforward. It is not surprising, perhaps, that case 2016–03 (a patient with an advanced tumour) is highlighted in the RTE annual report as an example of someone with intellectual disabilities who could choose EAS. Difficulties arise when the suffering, or the fact that the suffering cannot be relieved, is related to the nature of autism spectrum disorder or intellectual disability itself – as was the case for those who were highly dependent, had difficulties with social functioning, difficulties in coping with social circumstances, or a tendency not to cooperate with treatments.
Assessing the nature of suffering is a key question and has significance for disabled people, whether intellectual or otherwise. As we have seen, EAS requests are often based not on pain but on loss of dignity and autonomy. An unintended consequence of this rationale for EAS is to imply that lives lived with disability must be undignified, and that disabled people must experience suffering, despite assertions to the contrary by people living with long term disability . Numerous reports in recent years have suggested that the lives of people with an intellectual disability are valued less across society, and that their short life expectancy results from inappropriate value-laden decision-making by healthcare professionals [30, 41, 42].
Informing the patient about his/her situation and prospects
All patients need adequate information, but informing patients with intellectual disabilities may require considerable effort. It is crucial that the information is given in a format that the patient can understand, if patients with intellectual disabilities are to be properly supported to make decisions of such a serious nature. Even people with mild intellectual disabilities, who are recognised as being capable of making independent and autonomous decisions in some circumstances, may require information presented clearly in pictures or other reasonable adjustments in order to be able to fully understand and appreciate their situation. The lack of detail about how patients were informed about their situation and prospects makes it very difficult to understand how physicians have done this, or what expert assistance they need in the future.
No reasonable alternative
The key challenge in relation to this due care criterion is that the conclusion must be reached by the physician together with the patient. This implies a level of partnership working of which many people with intellectual disabilities have little experience during their lifetime, making this a difficult criterion to implement. Partnership working and ‘shared decision-making’ are important in achieving patient autonomy and patient-centred care. It requires an excellent relationship between the physician and the patient. For people with intellectual disabilities and/or autism spectrum disorder, especially those who also have long-standing psychiatric problems, building such relationships is likely to require time, and therefore, applying the usual standards and timeframes for EAS trajectories may put vulnerable patient groups at a disadvantage, or even at risk. In several cases, especially those where the patient’s own physician did not want to enter the EAS trajectory, the relationship between the patient and the physician carrying out the euthanasia was fairly recent. Furthermore, several case reports seem to imply that there were frictions between the physicians and patients, especially around the question whether further treatments were possible and reasonable. The patient who called her psychiatrist “an oaf who doesn’t understand a thing” is an example of this (case 2013–22). It is difficult to see how an assessment of the patient’s prospects can be made, together with the patient, in such complex circumstances.
Consulting at least one other independent physician
As we have discussed above, assessing capacity and supporting autonomy outside an ongoing relationship with the patient can be problematic for people with intellectual disabilities, and these problems will be magnified if there are also mental health issues. It appears, therefore, that the requirement of a second opinion from an independent consultant (responsible for assessing whether the due care criteria had been met) may not be appropriate for people with cognitive conditions. It will often be too difficult to gain sufficient understanding of the patient and come to a fully informed view in just one meeting, no matter how carefully the consultant has studied all the case notes. It could be argued that it is difficult to assess any patient’s request in one meeting, whether they have a disability or not, but it seems that people who have difficulty with communicating and relating to others are at a particular disadvantage.
The most obvious limitation is that the case summaries are unlikely to tell the whole story. They are rather brief, with a frustrating amount of ‘standard text’ that sometimes fails to convey the individual nature of each situation. It may well be, for example, that family and carers were involved or that capacity assessments were far more extensive than was reported, but we could only base our assessment on what was written in the case summaries. It was striking that much illuminating detail was included in those case summaries where physicians had been asked by the RTE to give further explanations. It would be highly valuable to conduct in-depth interview-based research with physicians, families and carers who have been involved with EAS for patients with autism spectrum disorder or intellectual disabilities; this would provide crucial further insight into the patients’ character, life story, social circumstances, important relationships, and crucially, the assessment of the due care criteria. It is possible that the concerns we have highlighted could be dismissed if the full facts of the cases were known. However, as Doernberg et al.  point out, since these summaries are meant to act as ‘case law’, not including extensive capacity discussions leads to the conclusion that the RTEs do not seem to expect high levels of scrutiny from physicians, nor a high threshold for capacity.
It was not the aim of this paper to establish how prevalent EAS is for people with autism spectrum disorder or intellectual disabilities, and we are unable to conclude how often such cases arise. We found nine relevant case summaries out of 416 (2.2%), which is roughly in line with the prevalence of intellectual disabilities among the population. However, those 416 published summaries represent only 1.6% of the total number of EAS cases over the 5 year period. It may be that the other 98.4% are more straightforward and less likely to involve such complex situations; on the other hand, it is also possible that intellectual disability is under-reported. In several of the case reports, the presence of “mild intellectual disability” was only mentioned in passing, in a description of the patient’s medical history. It is also possible that many physicians do exercise extreme caution when a patient with an intellectual disability or autism spectrum disorder requests EAS, as indicated by the fact that two thirds of the reviewed cases had their initial EAS request refused. A study of outcomes of the 645 EAS requests to the End of Life Clinic during its first year (2012) found that about half such requests were refused. Factors associated with a request being rejected included being single, a psychological condition, and having loneliness or loss of mental capacity . The lack of information on prevalence does not detract from the aim of this paper, however, which was to examine how the EAS due care criteria are (and can be) applied in these cases.