Article Text

Download PDFPDF

Choosing death in depression: a commentary on ‘Treatment-resistant major depressive disorder and assisted dying’
  1. Matthew R Broome1,
  2. Angharad de Cates2
  1. 1 Department of Psychiatry, University of Oxford, Oxford, UK
  2. 2 Warwick Medical School, University of Warwick, Coventry, UK
  1. Correspondence to Dr Matthew R Broome, Department of Psychiatry, University of Oxford, Oxford OX3 4JX, UK; matthew.broome{at}

Statistics from

Schuklenk and van de Vathorst's paper is a very welcome addition to the literature on the assisted dying debate and will be of great interest to clinicians working in the field of mental health.1 Many psychiatrists will have had patients who have asked them to allow them to die, to desist in their efforts to prevent their suicide, and one of us has had personal experience, outside of professional life, of being asked to aid in someone's attempt to end their life in the context of an episode of mood disorder. The person with depression asking professionals, friends and family members to aid them in ending their life is a very real phenomenon. In their discussion, Schuklenk and van de Vathorst's paper uses two principles that we endorse and add weight to their argument: namely, the reality of suffering in depression and the parity of mental and physical illness. Indeed, the case Schuklenk and van de Vathorst make is a strong one and our commentary will focus on the premise of the argument around choosing death and some empirical clarification about competence in depression and the nature of treatment-resistant depression (TRD) and its prognosis.

The conceptual concern we wish to discuss in relation to assisted dying, endorsed by the state and legislature, is not linked to any moral or religious concern, or even on grounds related to ‘slippery slope’ or ‘abuse’ arguments. Rather the worry is how one could operationalise decisions around assisted dying as based on competence and capacity and view such decisions as rational. I think this concern is not limited to depressive illness, but to all decisions regarding choosing non-existence. Schuklenk and van de Vathorst offer the following useful criteria to regulate assisted suicide:1

  • The patients are competent to evaluate their current situation.

  • The patients are competent to evaluate their future prospects based on the scientific evidence available at the point in time when they request assistance in dying.

  • The patients’ decision is voluntary and informed.

  • The patients’ quality of life is such that they do not consider it worth living, and the likelihood of improvement is exceedingly small or non-existent.

  • The patients repeat their requests over a reasonable period of time.

Our worry lies with the first three points and is perhaps encapsulated in a line at the end of Wittgenstein's Tractatus Logico-Philosophicus,Death is not an event in life: we do not live to experience death (6.4311, p.87).2

The reason this idea of Wittgenstein's seems key is that for issues of competence and capacity, the decision maker has to weigh information about the outcome of their choices. Death is non-existence; epistemologically there is no information about that prospect to be weighed alongside a continued, suffering-filled, existence. There is no ‘view from nowhere’ from which objectively these outcomes can be weighed and fed into a judgement about these two choices. Hence, it is not a question as to whether state-assisted dying should be allowed, or whether indeed we can understand why people may want it, but rather that it seems misleading to view those decisions as being able to be rational and to be operationalised in a non-problematic manner. This concern is not limited to assisted dying in depressive illness (although temporal decision making may be particularly impaired in severe depression3), but to all cases of assisted dying and the very possibility of thinking of futural outcomes which include one's own non-existence.

The second points to make are empirical. If one accepts Schuklenk and van de Vathorst's argument, an issue is whether a person with TRD who meets the criteria above is likely to exist? This question relates to two themes apparent in the paper: the issue of competence in depression, and viewing TRD and the absence of remission as equivalent to failure to ameliorate suffering. Taking these in turn, first we'll discuss competence in depression. Appelbaum's important paper on competence is cited4 and Schuklenk and van de Vathorst rightly say that empirical data are lacking to suggest that ‘depression per se renders patients incompetent’.1 Capacity decisions are local, that is, topic-specific, and Appelbaum's study looks at decisions made to take part in research. Further, Appelbaum's sample is likely to not be representative of a group of patients with treatment-resistant illness. Hence, we can't generalise findings from this study to a group with TRD making decisions regarding dying. Yet, there are some data to guide us in this. A recent systematic review found evidence from clinical ethics and empirical studies indicating that decision-making capacity, or the ability to appreciate the personal significance and context of information received, is often impaired in those with severe depressive illness.5 Further, and relating to weighing decisions about future events, problems in temporal abilities and decision making have been demonstrated in those with severe depression, but not in those with mild or moderate forms of the illness.3 Second, we discuss the important issue of the construct of TRD and whether such a diagnosis carries with it the idea of little hope of recovery. There is difficulty in finding a universal definition of TRD.6 ,7 Despite this, a key element to a likely definition is failure to achieve remission from symptoms. Yet, remission is not equivalent to response and some degree of amelioration of suffering. For the STAR*D studies, remission is defined as a score of lower than 7 on the Hamilton Rating Scale of Depression,8 that is, scoring outside the pathological range, and other authors conceptualise remission as full recovery.6 Thus, it cannot be assumed that those with TRD, and hence the absence of remission, haven't experienced marked symptomatic benefits and alleviation of their suffering from treatments: they may, for example, move from a severe depression to a mild episode in terms of severity and still be classed as TRD. Hence, as a construct, TRD doesn't do a terribly good job of marking out a clinical state that is wholly ‘irreversible’ or ‘debilitating, incurable’.1 Further, depression is a disorder in which feasibly new treatments may be discovered, as the recent interest in ketamine and glutamatergic targets attests.9 Although it may be hard to achieve remission for all patients, as the STAR*D studies demonstrate and Schuklenk and van de Vathorst correctly report, this is not to say that the vast majority of patients with depression, and even those with TRD, do not benefit from interventions.7 ,10 ,11

To conclude, putting our conceptual concerns to one side about the possibility of rational decision making about non-existence, there is good evidence to suggest that TRD may not meet the criteria of a disorder characterised by lack of improvement. Further, competence in decision making is a crucial element when considering assisted dying, and it can't be assumed that those with depression are able to make such decisions. Indeed, hopelessness, closure of the future12 and suicidal ideation are key features of the illness and it could be argued, in contrast to other areas where capacity is assumed as a default, that in these cases it should be assumed to be absent unless assessed thoroughly, and that this assessment should include a thorough psychiatric review. Hopelessness and suicidal ideation are core symptoms of the illness such that if the wish to die, and belief in lack of improvement, were rebadged as non-pathological, in the assessment of competence, then that same individual would necessarily have a less severe illness, as based on symptom severity. Conversely, if they were deemed to remain psychopathology, then the individual would lack competence. Hence, there is a bind for decision making in depression—those who may be competent to make decisions would necessarily have a less severe illness that then wouldn't meet the other criteria of Schuklenk and van de Vathorst. For these reasons and others, some suggest that psychiatric assessment should be generalised to all cases of assisted dying, and this perspective is seen in the evidence given to the UK House of Lords as well as in wider international views.13–15 Hence, we are in agreement with Schuklenk and van de Vathorst's conclusion, and the validity of their argument, but where the issue lies for us is that it is very unlikely that there is such a patient with TRD who is both competent to make decisions about ending their own life, and that the same individual has no prospect for relief of their suffering.


We would like to thank Dr Rebecca Roache and Dr Dominic Wilkinson for their help and guidance during the preparation of this manuscript.


View Abstract


  • Contributors The paper is jointly authored with MRB writing the first draft and preparing final draft for submission.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles