Shaw et al argue that ‘it is wrong to assume that patients suffering from mental health issues (including depression) cannot consent to assisted suicide’.Reference Shaw, Trachsel and Elger¹ But being depressed is the strongest correlate of decision instability, of changing from acceptance of euthanasia to rejecting it at follow-up.Reference Blank, Robison, Prigerson and Schwartz² As a matter of fact, the rate of psychiatric patients who, after seeking euthanasia or assisted suicide (EAS), no longer wished to die and/or withdrew their requests is quite high.Reference Thienpont, Verhofstadt, Van Loon, Distelmans, Audenaert and De Deyn³ In general, caregivers should be aware of the risks of EAS for patients with a mental health issue.

Beyond the paradox of use of EAS criteria corresponding to clinically targets of therapeutic intervention, available data on psychiatric EAS from Belgium, the Netherlands and Luxembourg highlight real issues of such practice. Even EAS defenders criticise the procedure,Reference Vandenberghe⁴ agreeing that: (a) a rigorous standardised evaluation involving a biopsychosocial perspective is lacking; and (b) all available treatments are not always tried and access to care not systematically assured. Decision-making capacity evaluation in patients requesting assisted suicide is even more complex in the presence of psychiatric disorder.

Medicine's ongoing assumption that clinicians and patients are rational decision-makers is questionable. All humans (including patients and clinicians) are influenced by seemingly irrational preferences in making choices about risk, time and trade-offs. By extension, the existence of rational suicide is uncertain. Decisions are considered to be rational when they rely on two core dimensions: being realistic and having minimal ambivalence.Reference Clarke⁵ But how can we rationally consider the options ‘to be or not to be’? Suicide is known to be an ambivalent choice. In addition, considering that ‘I would be better off dead’ is not sensible because there is no knowledge of ‘being’ after death. The term ‘understandability’ could thus be rather used than ‘rationality’ for suicide. However, the ability to understand someone's wish to die does not mean that suicide is for the best.

Moreover, the irremediable dimension of suffering justifying EAS is unclear because suffering may be improved for some patients when they are heard and taken seriously in their death request. Altogether, it suggests that EAS defenders may be misled by personal beliefs, feelings and values. Are EAS advocates reignited caregivers having forgotten the Hippocratic oath ‘primum non nocere’? It is important to note that mental illnesses are now recognised to be chronic and disabling, belonging to a group of serious medical illnesses such as cancer, but do not benefit from the same research approach. Whereas the goals of biomedical research for severe somatic illnesses are generally cure and prevention, very little research for the mental illnesses has set the bar this high. Thus, to propose an irremediable and definitive solution (death) to a complex and poorly understood phenomenon (suffering) is going in the wrong direction.