Article Text

Download PDFPDF

Depression and competence to refuse psychiatric treatment
  1. A Rudnick
  1. Correspondence to:
 Dr A Rudnick, Department of Behavioral Sciences, Tel Aviv University School of Medicine, Tel Aviv 69978, Israel;


Individuals with major depression may benefit from psychiatric treatment, yet they may refuse such treatment, sometimes because of their depression. Hence the question is raised whether such individuals are competent to refuse psychiatric treatment. The standard notion of competence to consent to treatment, which refers to expression of choice, understanding of medical information, appreciation of the personal relevance of this information, and logical reasoning, may be insufficient to address this question. This is so because major depression may not impair these four abilities while it may disrupt coherence of personal preferences by changing them. Such change may be evaluated by comparing the treatment preferences of the individual during the depression to his or her treatment preferences during normal periods. If these preferences are consistent, they should be respected. If they are not consistent, or past treatment preferences that were arrived at competently cannot be established, treatment refusal may have to be overridden or ignored so as to alleviate the depression and then determine the competent treatment decision of the individual. Further study of the relation between depression and competence to refuse or consent to psychiatric treatment is required.

  • Coherence
  • competence
  • consent
  • depression
  • preferences
  • refusal

Statistics from

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.

Depression is a common and serious mental disorder. The more severe forms of depression, termed collectively major depression, have a life time prevalence of roughly 10% in the general population.1 Such depression causes considerable morbidity and mortality to the individuals afflicted, mostly from suicide attempts and physical ill health.2 It also imposes considerable emotional and financial burdens on families and on society in general.3 Major depression can be successfully treated by psychiatric interventions, such as antidepressant medications, electroconvulsive therapy, and some psychotherapies.4 Considering all this, it seems obvious that major depression should be treated. Yet individuals afflicted with major depression may refuse psychiatric treatment, sometimes because of their illness. Hence, we are faced with the problem whether to respect or override refusal of psychiatric treatment by individuals afflicted with major depression.

There is widespread agreement that treatment refusals should normally be respected,5 that is providing they are arrived at competently6; this agreement is based on the widely accepted principle of respect for autonomy.7 Thus, the problem whether to respect or override (or, rather, ignore) refusal of psychiatric treatment by individuals afflicted with major depression can be formulated—assuming the principle of autonomy is of first priority—as the question whether such refusal is arrived at competently. For if it is, it should be respected (if no other party, such as a depressed mother's child, is seriously harmed by the refusal), and if it is not, it should be overridden/ignored, according to the prevalent autonomy-oriented bioethics. In order to address the problem whether refusal of psychiatric treatment by individuals afflicted with major depression should be respected or overridden/ignored, this paper will illustrate and discuss the question whether such refusal is arrived at competently.


The patient is a 53 year old north American white woman, divorced, with an adult son and daughter, who has been an inpatient in a tertiary-care mental health centre for the last year because of a prolonged major depressive episode without psychotic features, consisting of diminished pleasure and interest, insomnia, reduced weight, fatigue, lack of energy, poor concentration, and suicidal ideation, all of which considerably impair her ability to live in the community—to the point of starting a fire at home with a cigarette. It is unclear whether this was intentional due to suicidality or neglectful due to poor concentration. She is pessimistic and indifferent regarding the improvement of her condition. She has a history of recurrent major depression, with a couple of suicide attempts since the age of 19, as well as abuse of hypnotic medications. Her previous depressive episodes responded best to electroconvulsive therapy. She has no other notable clinical history and there is no identified recent trigger for her current depression. She has an unremarkable personal and family history, aside from her father having abused alcohol, and she has never worked. During the current hospitalisation, she was given various antidepressant medications that did not improve her condition, after which she and her children consented to her being given electroconvulsive therapy. With electroconvulsive therapy, her depression improved to the extent that she slept better and went back to her old habit of reading books (reflecting improvement in anhedonia and concentration), but her other symptoms persisted. She remained pessimistic and indifferent as to the outcome of treatment, resulting in her wish to be left alone and in her eventual refusal to continue electroconvulsive therapy after eight sessions in spite of attempts to inform her of, and demonstrate to her, the benefits of electroconvulsive therapy, which she knew.8


The question of the competence of depressed individuals to refuse psychiatric treatment has not been explored much.9 This may be due to the fact that the notion of competence to refuse (or consent to) treatment was originally required to address mainly cognitively impaired or psychotic individuals, some of whom are more in the public eye because of an increased risk of danger to others when their mental impairment is not treated, such as in schizophrenia.10 It may also be due to the related fact that such mental impairments, and particularly psychosis, impair competence to consent to treatment most conspicuously, in that they are frequently associated with lack of awareness of the mental illness, that is, poor insight11; less conspicuous impairments of competence to consent to treatment may have been overlooked. Be that as it may, such poverty of discussion is disturbing and requires rectification, considering the repercussions for wellbeing and freedom of choice related to either respecting or overriding/ignoring the psychiatric treatment refusals of depressed individuals.

The notion of competence to consent to treatment refers also to competence to refuse treatment, although the threshold for the latter is considered by some to be higher than the threshold for the former, as the risks of refusing treatment may be greater than the risks of consenting to treatment, at least for serious illness12; admittedly, others have disputed the validity of this difference in threshold for competence.13 This threshold controversy can be ignored for the purposes of our discussion. What has become practically the standard notion of competence to consent to treatment includes four components: the ability to express a choice; the ability to understand the information involved; the ability to appreciate the personal relevance of this information, and the ability to reason logically in decision making.6 As noted above, psychosis—in the narrow sense of delusional belief14 is associated with poor insight and hence with inability to appreciate the personal relevance of the information about the mental illness and its treatment. If so, major depression with psychotic features (such as nihilistic delusions about the depressed person or the world being at an end), which may be the most severe form of depression, impairs competence to consent to treatment. The refusal of psychiatric treatment by individuals afflicted with major depression with psychotic features is then not arrived at competently and should be overridden/ignored.

But what about major depression without psychotic features? That is, how does such depression fare according to the standard notion of competence to consent to treatment? And is this standard notion of competence sufficient to address the question whether refusal of psychiatric treatment by individuals afflicted with major depression is arrived at competently and hence should be respected?

Major depression is characterised by affective symptoms of sadness and diminished interest or pleasure (anhedonia); by physical symptoms of changes in sleep, in appetite/weight, in psychomotor activity, and fatigue or loss of energy; and by cognitive symptoms of feelings of worthlessness or excessive/inappropriate guilt, diminished thinking/concentration or indecisiveness, and suicidal ideation/behaviour.1 It is further characterised by the cognitive triad of pessimism regarding the self, the world, and the future.15 These symptoms may not impair competence to consent to treatment according to the standard notion of competence, as they do not necessarily disrupt the four abilities mentioned above—those of expression, of understanding, of appreciation, and of reasoning—so long as they are not accompanied by psychotic features such as delusions of guilt or of nihilism. At the most, extreme indecisiveness that may be related to catatonia may disrupt expression of choice, but this is very uncommon in major depression without psychotic features. Even suicidality as such does not impair standard competence, since there may be circumstances where it follows from sound understanding, appreciation and reasoning, such as in some cases of terminally ill patients (who may be concomitantly depressed).16 Finally, major depression may sometimes disrupt appreciation of the benefits of treatment due to undervaluing positive outcomes while focusing on negative outcomes of treatment and thus skewing the weights given to treatment outcomes in favour of treatment risks.17 Yet such a cognitive distortion does not necessarily occur, because depressive symptoms such as diminished interest, fatigue or loss of energy, if they are dominant, may only result in indifference and lack of drive to act on the appropriately appreciated benefits and risks of treatment, leading to a default choice of no treatment without any associated significant cognitive distortion.18

Thus, in the case illustrated above, the patient consistently expressed her refusal of continued electroconvulsive therapy, she clearly understood the information given to her about the benefits of the treatment, she appreciated the personal relevance of this information as she accepted the diagnosis that she was depressed and that treatment may help her, and she reasoned logically in her decision making in the sense that her conclusion to refuse treatment followed logically from her premises (which included her not caring about the treatment outcome). Accordingly, this patient may be deemed competent to refuse psychiatric treatment according to the standard notion of competence to consent to treatment. Yet, such a judgment misses the importance of the point that she did not care about her outcome, which may be central, because giving value to outcomes that significantly affect one's life (and death) may be considered part of normal decision making (and hence of competence). So the standard notion of competence to consent to treatment may be insufficient to address this issue.

The standard notion of competence to consent to treatment has been argued, even by its leading proponents, to be cognitively skewed.19 Consideration of emotion may be required to supplement it. That is, the four abilities noted above are largely cognitive, yet emotional or motivational factors may enhance or disrupt decision making capacity or competence. For instance, hope may enhance competence by driving us to extraordinary decisions and accomplishments, even though this may sometimes be against the odds. And fear may disrupt competence by deterring us from simple decisions and activities even though these may be commonplace. Still, affective factors are known to be related to cognitive abilities and may interact with the four abilities noted above, for example, when emotional avoidance such as denial results in impoverished ability to appreciate the personal relevance of information.20 Hence, consideration of emotion as such may not be sufficient as a supplement to the standard notion of competence.

Perhaps the standard notion of competence to consent to treatment can be criticised as lacking in other respects. It seems that the four abilities noted above refer to the output (expression) and process (understanding, appreciation, and reasoning), but not to the input (information and preferences), of decision making. Input information is addressed within the broader doctrine of informed consent,6 but input preferences, which may be characterised as ends assumed by the individual, are largely ignored in this framework. This may be due to the fact that, since Hume, many have considered such preferences or ends as not open to critical discussion.21 Indeed, it is difficult to argue otherwise regarding the (isolated) content of such preferences if pluralism—and hence respect for individuals—is to be upheld. But such content does not exhaust the features of preferences. A pertinent feature of preferences which may be open to critical discussion is the extent of their coherence with the set of related preferences held by the person. Input preferences that do not cohere considerably with most other preferences held by the person in the present or in the past are suspect, and following that, decision making resulting from them is suspect too, suggesting incompetence on the grounds of not being consistent with what that person would usually decide. Hence, coherence of preferences may be a neglected component of competence to consent to treatment, and if major depression without psychotic features considerably disrupts such coherence, then the refusal of psychiatric treatment by individuals afflicted with such major depression may not be arrived at competently and then should not be respected as such (and the standard notion of competence to consent to treatment may be insufficient to address the question at hand regarding the impact of depression on such competence). Admittedly, the trouble with positing the notion of coherence of preferences as a component of the notion of competence to consent to treatment is that it suggests that some such coherence is normally expected from an individual. Yet there is no agreement that this has to be so, for example, as observed in discussions of personal identity.22 Indeed, some conflict of preferences is the rule rather than the exception in any individual. Furthermore, this suggestion may be taken too strictly to imply that individuals cannot competently change their preferences, which is absurd, as preferences do and should change, depending upon circumstances, critical reflection, and personal development. Also, there is to date no clear measure of such coherence of preferences. Still, some coherence of preferences is required for an individual to make decisions reliably, which is necessary (although not sufficient) for accountable—and hence competent23—decision making. The question then remains whether and to what extent major depression without psychotic features disrupts the coherence of preferences of the afflicted individual, so as to evaluate the impact of such depression on the competence of that individual to refuse psychiatric treatment for the depression.


Pervasive emotional states or moods impact on preferences by regulating their relative weights and perhaps less commonly by generating new preferences, thus modifying the set—and hence the coherence—of preferences held by the individuals experiencing these moods.24 This is well known for the awe and elation associated with experiences of enlightenment, which commonly transform the way individuals view the world and themselves and, as part of that, their preferences. As major depression is associated with a similarly pervasive mood, it may modify the preferences held by the depressed individual, for example, towards preferring death, and hence change their coherence perhaps to the point of disruption (considering that such depression is maladaptive). Also, the core symptoms of major depression, that is, sadness and diminished interest or pleasure, which foster passivity, may diminish preferences in general, including those that refer positively to productive activities, such as work, recreation, and life itself.25

A useful way to evaluate such an impact of major depression on the coherence of preferences and hence on the competence of an individual to refuse psychiatric treatment is to compare pertinent preferences of the individual during depression to preferences regarding the same subject matter held by that individual when not depressed. This strategy deals particularly with preferences regarding death and treatment preferences (the content of which may serve both as input and output of the decision making process, yet in the former it serves as preference whereas in the latter it serves as decision), and evaluates consistency, which is a clear—although narrow—measure of coherence; admittedly, sometimes the comparison to past preferences may not be helpful, particularly if they are from the remote past, as some preferences may normally develop and change as individuals age. This strategy has been used in the study of mildly to moderately depressed individuals who are seriously physically ill, where it was demonstrated that their pertinent preferences, particularly those related to death and treatment, did not change after successful treatment of their depression.26 There are no similar studies concerning more severe (but not psychotic) depression, so it is unclear whether these findings hold for severe major depression without psychotic features; still, these findings at least support the claim that the milder forms of depression associated with serious physical illness do not change the pertinent preferences of the afflicted individuals and, following that, do not impact on the coherence of their preferences and hence may not disrupt their competence to refuse treatment. Surprisingly, there does not seem to be similar research on depressed individuals who are not physically ill, although such research is most desirable if one is to be able to assess to what extent major depression on its own changes the preferences of the afflicted individuals and, following that, evaluate the impact of major depression on the coherence of preferences and hence on the competence of these individuals to refuse psychiatric treatment. Still, this strategy may be helpful in the individual evaluation of the competence of patients with major depression without psychotic features to refuse (or consent to) psychiatric treatment.

Such an evaluation of competence referring to the coherence of preferences may consist of the following steps (related to the depressed patient refusing or consenting to psychiatric treatment).27 First, attempt to establish the treatment preferences of the individual from the (preferably recent) past when he or she was clearly not depressed. This can be done most easily if an advance directive is available. Second, if these past treatment preferences have been established, determine whether they are consistent with the current treatment consent or refusal (then, if they are consistent, it is suggested that the current treatment consent or refusal be respected, and if they are not consistent, other things being equal, it is suggested that this is due to the current depression and that the current treatment consent or refusal be overridden/ignored). Third, if these past treatment preferences cannot be established, attempt a therapeutic trial (if this serves the best interests of the patient, that is, the risk-benefit ratio seems favourable), after which, if the trial is successful, the patient's competent treatment preferences can be established and compared to his or her preferences during the depression; only then can the competence of the patient during the depression be determined in these circumstances.28 Note that a special legal procedure may be required for this step, as a substitute decision maker cannot be appointed then due to competence not being determined before the therapeutic trial succeeds, and as such a trial may have to be attempted in the face of treatment refusal on the part of the depressed patient (who may turn out to have been competent to refuse treatment during the depression if he or she maintains the preference for treatment refusal after the depression is alleviated).

How does the patient in the case illustrated above fare according to this type of competence evaluation? She has no advance directive in place, and although she consented to the same treatment in the past, it is not clear that she was then competent to consent to treatment as she was then depressed, and there is conflicting evidence as to her treatment preferences, particularly regarding electroconvulsive therapy, when she was not depressed. Therefore, no past treatment preference can be established with sufficient certitude to determine whether her current treatment refusal is consistent with past competent preferences, and following that, whether she is competent to refuse psychiatric treatment during her current depression. Hence, a therapeutic trial of electroconvulsive therapy may be justified in spite of her refusal of it. What happened in fact was that her son agreed and succeeded in persuading her to agree to continue electroconvulsive therapy, which was then discontinued after 12 sessions as it did not result in further improvement in her depression and as she reverted to refusing it. So her competence to refuse psychiatric treatment remained largely undetermined.


The question of the competence of depressed patients to consent to or refuse treatment for their depression has not been sufficiently addressed to date. It raises difficulties for the standard notion of competence to consent to treatment, as the latter does not address preferences that serve as premises of decision making, whereas depression—particularly major depression without psychotic features—may considerably impact on decision making through its impact on such preferences. Thus, major depression may sometimes change current treatment preferences of patients so that they are not consistent with their past treatment preferences, and then a determination has to be made as to which preferences of the patient to follow (so long as it can be shown that the inconsistency is not due to other changes, such as changes in circumstances which may provide reasonable grounds for change of mind). In such cases, when past treatment preferences of the patient from periods without depression can be established, it is suggested that they should override current treatment preferences of the patient. If past treatment preferences cannot be established with sufficient certitude, a therapeutic trial of treatment for the depression may be justified, even if the patient currently refuses the treatment, as in such circumstances the determination of competence during the depression may be possible only after successful treatment of the depression, when it can be evaluated whether the treatment preferences during and after the depression are consistent.

Further study of this suggested addition to the notion of competence is required. Empirical research of changes in treatment preferences of severely depressed (but not psychotic) patients, following successful treatment of their depression, may shed light on the question whether severe depression without psychotic features usually disrupts competence to refuse or consent to psychiatric treatment. Legal research could explore the possibility of formalising such a suggestion, in particular the potentially controversial idea of a therapeutic trial that may override/ignore current patient treatment preferences in order to assess competence. And social research could clarify the attitudes of the various parties involved, such as depressed patients, their families, clinicians, policy makers, and the general public, towards such a suggestion, so as to determine some of the practical obstacles such a suggestion may confront. Finally, conceptual investigation of the application or modification of this suggestion so as to address more chronic forms of depression, such as dysthymia, may test its limits in that depression-related preferences in such forms of depression may be so ingrained and longstanding that successful treatment may be viewed as disrupting coherence of preferences rather than reinstating it, which might make such patients on this account competent when depressed and incompetent if and when they are not depressed! Be that as it may, some such suggestion may be helpful in evaluating the competence of depressed patients to refuse or consent to psychiatric treatment.


Thanks are due to Lawrie Reznek, MD, DPhil, for his helpful suggestions and comments on this paper.


Other content recommended for you