The Feature article by Bastidas-Bilbao and colleagues^{Reference Bastidas-Bilbao, Castle, Gupta, Stergiopoulos and Hawke1} proposes a sophisticated framework for evaluating the impact of the Canadian policy for medical assistance in dying (MAiD) for people with mental illness. The article is timely, as concerns about mentally ill patients having access to MAiD have led to a hiatus in its provision, and attention has shifted to devising a suitable programme that can be broadly agreed upon and implemented uniformly. To achieve this, an expert panel was convened by the Canadian government to provide advice regarding ‘protocols, guidance and safeguards to apply to requests made […] by persons who have a mental illness’² (p. 19). In 2022, this panel published a report in which they laid out ‘a broad set of principles to structure the practice of MAiD’² (p. 12) for those with mental illness and although they supported its provision, they set limits to its application and made 19 recommendations for how MAiD should be enacted. The report is detailed but the recommendations may not provide sufficient safeguards. To illustrate this, we briefly comment on the inadequacy of the key eligibility criterion for MAiD, namely, the definition of a ‘grievous and irremediable medical condition’ and argue for further research.

In its report the panel defined irremediability in line with the Canadian criminal code, as having three components: incurability of the illness, disease or disability; irreversibility of the decline in capability; and enduring and intolerable suffering. And although the panel acknowledged the ‘interdependence of the three elements’² (p. 53), it nevertheless made three separate recommendations.

Recommendations 2 and 3, which relate to establishing incurability and irreversibility respectively, both require the person requesting MAiD (requestor) and the person assessing eligibility for MAiD (assessor) to arrive at a ‘shared understanding’ of the definition of incurability and irreversibility. Further, in both recommendations, the respective definitions depend on the treatments tried. At the same time, the panel argues that ‘it is not possible to provide fixed rules’² (p. 55) for the number and type of treatment attempts or over what time these should have been trialled, concluding that these ‘must be assessed on a case-by-case basis’. In addition, recommendation 4 describes suffering as ‘a personal experience’ and notes that ‘it is subjective’² (p. 57), which means that if a person states that their suffering is unbearable then this must be taken at face value.^{Reference Dembo, Schuklenk and Reggler3}

Therefore, in effect, the determination of eligibility for MAiD is left solely to clinical judgement and consensual agreement between the requestor and the assessor, and no objective standard is set for either incurability or irreversibility. This is problematic because even though recommendations 10, 11 and 12 propose safeguards, such as having independent assessors, involving other healthcare professionals in decision-making, and seeking collateral information, the process of determining eligibility for MAiD remains vulnerable to abuse as there is no specified objective standard of care that a mentally ill person must satisfy. This approach is also not in keeping with assessment procedures employed to assess suitability of psychiatric patients for other irreversible interventions. For instance, to be eligible for deep brain stimulation or psychosurgery, which are sometimes used to treat intractable psychiatric conditions, the individual is expected to have trialled all other available evidence-based treatment options.^{Reference Figee, Riva-Posse, Choi, Bederson, Mayberg and Kopell4,Reference Müller, van Oosterhout, Bervoets, Christen, Martínez-Álvarez and Bittlinger5} In severe depression, for example, the person should ideally have trialled several courses of psychotherapy, pharmacotherapy and electroconvulsive therapy. In contrast, the Canadian expert panel seems to have placed greater emphasis on the requestor's preferences and the ‘conditions they consider acceptable’ because ‘capable persons are usually entitled to refuse interventions they do not wish to receive’² (p. 58). But ending one's life is not a usual situation and given that the ontology of mental illnesses is unknown and psychotropic treatments are largely non-specific, setting a treatment standard for each kind of mental illness seems prudent. Moreover, medical procedures should be provided in the best interests of patients. Clearly patients can desire things, including refusal of medical procedures, that are not in their best interests. The Canadian expert panel conflates promotion of best interests or well-being with respect for autonomy.

The argument against having to meet certain standards and undergo specified therapies is that it constrains patient autonomy. However, given that mental illnesses are generally not terminal, meaning there is ample time to trial available therapies, it seems a reasonable expectation and an appropriate safeguard to have to undergo specified treatments that are known to be effective in some patients. In practice, the treatments a person should have to undergo to be eligible for MAiD could be determined for each psychiatric disorder by drawing on local treatment guidelines and by asking experts to advise what they consider to be a reasonable minimum requirement. This could then serve as a starting point for discussion among the professionals involved in a particular case.

Having to navigate a variety of medical opinions and striking a balance between different moral positions highlights the challenges of offering MAiD to those with a mental illness. At the same time, it underscores the need for research, which is further reinforced by the observation that MAiD can have unexpected and counterintuitive effects. For example, in some cases, merely offering the possibility of MAiD to a person with a mental illness may provide them with sufficient hope to re-evaluate their desire to die. In jurisdictions such as The Netherlands and Belgium, where MAiD is available for those with a mental illness, preliminary findings suggest that being granted access makes people feel their suffering is acknowledged and they have been heard.^{Reference Verhofstadt, Pardon, Audenaert, Deliens, Mortier and Liégeois6} As a consequence, they sometimes decide that they can cope with their suffering for a while longer. Conversely, for others, granting eligibility and offering access to MAiD may accentuate their sense of hopelessness as they feel their carers have given up and that doctors have run out of alternatives.^{Reference van Veen, Ruissen and Widdershoven7}

As far as we know, presently, many mental illnesses cannot be cured, and most treatments only offer symptomatic relief. Further, as mental illnesses are not terminal, in the same sense as, for example, advanced cancer, if irremediability is to be a meaningful criterion in determining eligibility for MAiD, it will need to be clearly defined.^{Reference Grassi, Folesani, Marella, Tiberto, Riba and Bortolotti8} In addition, many important issues will need to be clarified not only for mental illness as a whole but for specific psychiatric disorders.^{Reference Grassi, Folesani, Marella, Tiberto, Riba and Bortolotti8} For example, in severe clinical depression, suicidal ideation is a common symptom. Indeed, it is a diagnostic symptom of depression that is thought to be at least caused by the illness, if not intrinsic to it. Therefore, when a depressed person expresses a well-considered and competent desire to die there needs to be some means by which this can be reliably differentiated from suicidal thinking. Similarly, in the course of dementia, it is necessary to determine the point at which a person's suffering becomes unbearable, but they still retain sufficient capacity to express an authentic desire to die.

Again, these questions are best answered if psychiatric patients facing such issues are studied in circumstances where MAiD is available to them, as this will provide authentic and clinically meaningful insights. To this end, the evaluation of any such programme, along the lines suggested in the Feature,^{Reference Bastidas-Bilbao, Castle, Gupta, Stergiopoulos and Hawke1} must include research. A key argument against MAiD is that it does not align with the moral obligations of the doctor's role, namely the ‘duty of physicians […] to preserve life [and] therefore to prevent suicide, not to cause it even if requested’.^{Reference Grassi, Folesani, Marella, Tiberto, Riba and Bortolotti8} This is usually countered by the argument that their duty is ‘to reduce intolerable suffering [and that MAiD should be regarded] as a form of care’.^{Reference Grassi, Folesani, Marella, Tiberto, Riba and Bortolotti8} Nevertheless, many physicians opt out of participating in MAiD because of ethical concerns. Therefore, to better align MAiD with the professional role of doctors, the process of determining eligibility could be framed as a clinical assessment, the purpose of which is to find a treatment that aids recovery. Adopting an approach that attempts to alleviate suffering and improve care by ensuring that all therapeutic strategies that can be reasonably actioned have been properly considered is likely to instil a sense of hope and facilitate the engagement of both patients and clinicians in research endeavours.

Therefore, although the Feature^{Reference Bastidas-Bilbao, Castle, Gupta, Stergiopoulos and Hawke1} proposes an important framework for the evaluation of a MAiD programme for mental illness, it needs to be applied to programmes that allow for a much deeper understanding of the nature of psychiatric disorders, the suffering they confer and the desire to die that they can sometimes generate. And thus, alongside developing evaluative frameworks, further research needs to be conducted into the core aspects of MAiD both in jurisdictions where it is already being offered to those suffering from mental illness and where there are plans to introduce it.^{Reference Sinyor and Schaffer9}