Euthanasia in adults with psychiatric conditions: A descriptive study of the experiences of Belgian psychiatrists

Abstract

To investigate the experience of psychiatrists who completed assessment procedures of euthanasia requests from adults with psychiatric conditions (APC) over the last 12 months. Between November 2018 and April 2019 a cross-sectional survey was sent to a sample of 753 psychiatrists affiliated with Belgian organisations of psychiatrists to gather detailed information on their latest experience with a completed euthanasia assessment procedure, irrespective of its outcome (i.e. euthanasia being performed or not). Information on 46 unique cases revealed that most APC suffered from comorbid psychiatric and/or somatic disorders, and had received different kinds of treatment for many years prior to their euthanasia request. Existential suffering was the main reason for the request. The entire procedure spanned an average of 14 months, and an average of 13.5 months in the 23 cases that culminated in the performance of euthanasia. In all cases, the entire procedure entailed multidisciplinary consultations, including with family and friends. Psychiatrists reported fewer difficulties in assessing due care criteria related to the APC’s self-contemplation – for example, unbearable suffering on top of the due care criteria related to their medical condition; incurability due to lack of reasonable treatment perspectives. In a few cases in which euthanasia was the outcome, not all legal criteria were fulfilled in the reporting physicians’ opinions. Both positive and negative experiences of the assessment procedure were reported: for example, reduced suicide risk for the APC; an emotional burden and a feeling of being pressured for the psychiatrist. This study confirms that euthanasia assessment in APC entails a lengthy process with diverse complexities, and psychiatrists require support in more than one respect if the assessments are to be handled adequately. Thorough evaluation of current guidelines is recommended: that is, to what extent the guidelines sufficiently address the complexities around (e.g.) assessing legal criteria or involving relatives. We formulate various avenues for further research to build on this study’s insights and to fill remaining knowledge gaps.

Introduction

Adults with psychiatric conditions (APC) can be found legally eligible for euthanasia in Belgium if all the legal criteria as listed in Box 1 are fulfilled. ¹ As some contest whether and when an APC can meet all legal requirements, the practice remains subject to controversy, fiercely dividing clinical and ethical opinions,^2–7 and sometimes resulting in legal examination.^8,9 Meanwhile, the proportion of euthanasia cases in APC within all reported performed euthanasias remains small but has increased from 0.2% during the period 2002–2007 to 2.1% in 2015^10,11 before declining to 1.2% in 2017. ¹²

Empirical evidence regarding Belgian euthanasia practice with APC is limited. To our knowledge, only two retrospective studies exist, which were limited in scope because one only reported about performed euthanasia cases, and the other only reported about requests from a single practice.^10,13 One study revealed that the consultation process takes an average of 9 months, involving an average of four consultation sessions with multiple actors (e.g. patient, clinicians, family and friends). The study also showed that, whilst 48 of 100 euthanasia requests were accepted, 73% had been carried out, 21% had been withdrawn voluntarily, 2% had to be withdrawn due to imprisonment, and 4% of the requestors died by suicide.^13.

The performed euthanasia cases concerned adults of different ages, mostly women, suffering from multiple chronic psychiatric disorders, mainly major depressive and personality disorders.^10,13

A recent Belgian survey study gauging psychiatrists’ attitudes and experiences on this topic pointed out that psychiatrists struggle with these practices, due to the difficulties of reconciling euthanasia assessment with the patient-psychiatrist relationship.^14,15 In addition, almost three out of four Belgian psychiatrists question the adequacy of euthanasia assessment in current practice, which is in line with previous Dutch studies that indicate dissension among physicians regarding whether the legal criteria were/can be met.^16–18

So far, few Belgian studies have investigated the reasons that APC request euthanasia, and none have focused in detail on the challenge of the assessment for the psychiatrists involved. As psychiatric consultation is imperative and legally mandatory for determining the APC’s eligibility for euthanasia, we in this study use psychiatrists’ experiences to gain additional insights into current practices.

This study aimed to come to a description of completed euthanasia assessment procedures by asking a large representative sample of psychiatrists about their most recent experience during the last 12 months regarding: (1) the APC’s background in terms of diagnoses and treatment history; (2) the APC’s reasons for requesting euthanasia; (3) the main characteristics of euthanasia assessment procedures and finally, (4) the psychiatrists’ perceived difficulties and/or other experiences regarding the assessment.

Methods

Study design and participants

Case-based data on individual APC’s completed euthanasia assessment procedures were obtained through a cross-sectional survey of Belgian psychiatrists, consisting of a paper-and-pencil and web survey. The survey was sent to 753 potential respondents: 499 Flemish-speaking psychiatrists affiliated with the Flemish Psychiatry Association (FPA), and 254 French-speaking psychiatrists of the Royal Society of Mental Health of Belgium (SRMMB). The FPA’s members comprise an estimated 80%–90% of all psychiatrists active in the Flemish-speaking part of Belgium. No estimated percentages could be given with regard to the SRMMB’s members, due to a lack of current trustworthy registration of practitioners in the French-speaking part of Belgium.

Only reports from psychiatrists working in Belgium and having been involved in at least one completed euthanasia assessment procedure for an APC in the previous 12 months were included in the study.

Survey instrument

We based our questionnaire partly on an existing Dutch questionnaire. ¹⁹ We validated the final instrument with a selected group of 15 psychiatrists and their trainees for cognitive validation purposes (i.e. for participants to identify potential problems regarding item interpretation, item redundancy, completeness of the survey, feasibility of generating correct answers and time estimation) and adjusted it accordingly.

The survey was divided into two parts: one general part to be completed by every psychiatrist, whether or not they had been involved in concrete evaluation of euthanasia requests (see Appendix A in OSF); and one part focussing on their last concrete involvement with a completed euthanasia assessment procedure during the past 12 months, if applicable (see Appendix B in OSF). This study reports on their last concrete involvement (see Appendix C in OSF for the English version of the questionnaire).

The survey questions were preceded by the following sentence: ‘The questions below relate to your most recent experience with a completed euthanasia procedure (regardless of its final outcome) of an adult patient, predominantly suffering from a psychiatric condition, other than dementia, in the past 12 months’. Capitals were used to make clear that APC encompass the following two adult patient groups: (1) patients whose euthanasia request is predominantly based on suffering caused solely by their psychiatric conditions, other than dementia; and (2) patients whose euthanasia request is predominantly based on suffering caused primarily by their psychiatric conditions, and secondarily by somatic comorbid conditions.

Procedure

Data collection

Data were collected between November 2018 and April 2019. The FPA members were first sent a link to LimeSurvey’s online platform. ²⁰ Non-responders received a first reminder via e-mail 2 weeks after the initial invitation and a second, including a paper-and-pencil version by post, 3 weeks after. The SRMMB members were only sent the paper-and-pencil version, by post, as the SRMMB database only contained postal addresses, and non-responders received a reminder 2 weeks afterwards (See OSF, Appendix D, for a more detailed research protocol).

Data management

Data were imported from LimeSurvey into SPSS, cleaned according to the principles of a predetermined data analysis plan (See OSF, Appendix E), and completed with the cleaned data gathered from the returned paper surveys.

Statistical analysis

No sample size calculation/power analysis was done, as we intended to survey the entire eligible professional group. As duplicate cases could occur (i.e. the same individual euthanasia case being reported by at least two psychiatrists), we performed a manual check to identify euthanasia cases with identical or near-identical data by crossing the values of the following variables: (1) specific characteristics of the responding psychiatrists (e.g. specific role in the euthanasia procedure), (2) specific characteristics of the euthanasia procedure (e.g. the duration of the procedure, the number and nature of formal (and additional) advices obtained, the final outcome) and (3) specific characteristics of the APC (eg. psychiatric and somatic diagnoses, duration of the treatment trajectory). As the manual check revealed no duplicates, all reported cases were included in this study. All gathered data were analysed by means of standard descriptive statistics, including data that describe the sample of responding psychiatrists. The answers to the open question were used to elaborate further on the given responses by means of thematic analysis.

Ethics

This research project was performed in accordance with the Declaration of Helsinki and received ethical approval from the Medical Ethics Committee of the Brussels University Hospital with reference BUN 143201837302 and the Medical Ethics Committee of the Ghent University Hospital with reference 2018-1165.

Results

The appended flowchart in OSF illustrates the response sample procedure of Belgian psychiatrists who filled out the optional part of the survey. The Supplemental Material in OSF shows the characteristics of our sample of 46 psychiatrists. Most were men (65%), mainly working in a private or group clinical practice (63%) and/or psychiatric hospital care (63%) for more than 10 years (91%).

Clinical characteristics for APC requesting euthanasia

As shown in Table 1, in 89% of the completed case questionnaires the APC’s psychiatric disorders were specified with depressive disorders (N = 23) and personality disorders (N = 18) being the most common. Nearly half of the APC (48%) suffered from somatic co-morbidities, from chronic fatigue syndrome to Parkinson’s disease to overall multi-morbidity. At their first consultation for euthanasia, 91% were in treatment, most often including psychotropics (80%) and/or other medical drugs (28%) and/or psychotherapy (67%). The mean and median length of treatment history were 11 and 7 years, respectively.

Table 1.

Clinical characteristics of adults with psychiatric conditions with assessed euthanasia requests.

	All requests(N = 46)	Euthanasiacases (n = 23)
	N (%)
Patient’s pathology
Specified psychiatric conditions	41 (89.1)	21 (91.3)
Depressive disorders	23 (50.0)	9 (42.8)
Personality disorders	18 (39.1)	7 (30.4)
Schizophrenia spectrum and other psychotic disorders	6 (13.0)	4 (17.4)
Trauma- and stressor-related disorders	6 (13.0)	2 (8.7)
Anxiety disorders	4 (8.7)	2 (8.7)
Bipolar and related disorders	3 (6.5)	2 (8.7)
Feeding and eating disorders	3 (6.5)	0 (0.0)
Neurodevelopment disorders	2 (4.3)	0 (0.0)
Substance-related and addictive disorders	1 (2.2)	1 (4.3)
Unspecified psychiatric conditions	5 (10.9)	2 (8.7)
Somatic co-diagnoses	22 (47.8)	11 (47.8)
Severe brain injury	5 (10.9)	2 (8.7)
Physical deterioration	3 (6.5)	1 (4.3)
Pain, incl. consequences of failed suicide attempts	3 (6.5)	3 (13.0)
Palsy	2 (4.3)	1 (4.3)
Parkinson	2 (4.3)	1 (4.3)
Hearing problem	2 (4.3)	1 (4.3)
Chronic fatigue syndrome/fibromyalgia	2 (4.3)	1 (4.3)
Diabetes/morbid obesitas	2 (4.3)	1 (4.3)
Cancer	1 (2.2)	1 (4.3)
Overall multimorbidity	1 (2.2)	1 (4.3)
Patient’s treatment history at first consultation
No active treatment a	4 (8.7)	1 (4.3)
Active treatment	42 (91.3)	22 (95.6)
Psychotropics	37 (80.4)	21 (91.3)
Other drugs	13 (28.3)	7 (30.4)
Psychotherapy	31 (67.4)	18 (78.3)
Other interventions b	14 (30.4)	8 (34.8)
Length of the patient’s treatment history
Mean (SD)	10.6 years (9.8)	8 years (6.9)
Median (min-max)	7 years (1 month–32 years)	5 years (1 month–25 years)
<1 year	5 (11.0)	2 (8.7)
1–2 years	6 (13.0)	3 (13.0)
2–5 years	8 (17.3)	6 (26.1)
>5–10 years	5 (11.0)	3 (13.0)
10+ years	16 (34.7)	6 (26.1)
Missing	6 (13.0)	3 (13.0)

^aIn two cases explained as follows: the patient did receive psychiatric treatment in the past.

^bOther interventions were specified as follows: neurosurgical treatment and/or electroconvulsive therapy, ambulant and/or residential admittance in a psychiatric unit, nursing and/or other care in a psychiatric home care setting, alternative psychotherapy, mobile team.

A similar picture emerges with regard to the euthanasia requests that culminated in the performance of euthanasia (n = 23). These cases mainly concerned APC with comorbid disorders (70%), and close to half (48%) suffered from severe physical co-morbidities: for example, cancer and chronic pain-related problems, some of which were related to injuries incurred by a previous suicide attempt. At the time of their first consultation for euthanasia, all but one APC were in treatment. The mean and median length of treatment history were 8 and 5 years, respectively, with a minimum of 1 month and a maximum of 25 years.

Main reasons for requesting euthanasia

Most psychiatrists (87%) indicated more than 3 reasons for the request, with a minimum of 1, a maximum of 12, and an average of 6–7. Table 2 lists the indicated categories, in descending order of prevalence. No perspective for improvement (87%), a very low level of quality of life, just being in ‘survival mode’ (72%), and existential suffering (63%) were most often reported, and even to a greater extent if the APC died by means of euthanasia (96%, 83% and 74%, respectively). In the 23 performed euthanasia cases, ‘No purpose left in life’ (78%) was also more often indicated. When asked to report the two main reasons for euthanasia requests, the most frequent were: existential suffering, and no perspective for improvement. Whereas loneliness was ranked third in all reported cases, pain-related problems closed the top three ranking with regard to the 23 performed euthanasia cases.

Table 2.

Reasons for requesting euthanasia in adults with psychiatric conditions. ^a

	All requests(N = 46)	Euthanasiacases(n = 23)
	N (%)
Indicated reasons for requesting euthanasia b
No perspective for improvement	40 (87.0)	22 (95.6)
No quality of life, only in ‘survival mode’	33 (71.7)	19 (82.6)
Existential suffering (suffering from life itself, meaninglessness)	29 (63.0)	17 (73.9)
Stalled on many life domains (work/relationships/...)	27 (58.7)	13 (56.5)
No purpose (left) in life	26 (56.5)	18 (78.3)
Feelings of depression	22 (47.8)	12 (52.2)
Loss of dignity	22 (47.8)	15 (65.2)
Loss of autonomy, control over own life	21 (45.7)	10 (43.5)
Loneliness	18 (39.1)	9 (39.1)
No (longer) wanting to be a burden	16 (34.8)	10 (43.5)
Gradual deterioration	16 (34.8)	10 (43.5)
Total exhaustion	10 (21.7)	6 (26.1)
Fear of suicide	9 (19.6)	4 (17.4)
Disability/immobility	9 (19.6)	5 (21.7)
Other (e.g. pain)	8 (17.4)	2 (8.7)
Indicated main reasons of the euthanasia request
Existential suffering	16 (34.8)	7 (30.4)
No perspective for improvement	11 (23.9)	7 (30.4)
Loneliness	7 (15.2)	2 (8.7)
No quality of life, only ‘surviving’	6 (13.0)	4 (17.4)
Pain related problems c	5 (10.9)	5 (21.7)
Fears	5 (10.9)	2 (8.7)
Feelings of depression	5 (10.9)	4 (17.4)
Gradual deterioration	5 (10.9)	3 (13.0)
Lack of purposes left in life	5 (10.9)	2 (8.7)

^aPsychiatrists could indicate as many predesignated categories as applicable.

^bThe answers on the open question, no. 9: ‘In your opinion, what were the two main reasons for the patient to request euthanasia?’ were tallied. If the answers did not fit one of the categories of question no. 8, it was also coded and counted (missings: n = 3). This yielded additional motives for the patient’s euthanasia request, namely: (1) all kinds of fears, other than the fear of suicide, for example: fear of repetitive traumatic events; (2) being through with treatment due to for example, treatment resistance, even if the patient is improving on the physical level; (3) complicated grief; (4) self-hatred; and (5) financial difficulties.

^cSome pain related problems were ascribed to the consequences of failed suicide attempts.

In addition, the open question yielded additional motives for the request: namely, all types of fears other than suicide (e.g. potential repetitive traumatic events), being finished with treatment (due to, for example, treatment resistance on the level of the APC’s psychopathology, even if the APC is improving on the physical level), complex grief, self-hatred and financial difficulties.

Main characteristics of the APC’s euthanasia assessment procedure

Based on the answers of the responding psychiatrists, the mean and median length of the euthanasia assessment procedure were 14 and 7 months, respectively, and if the patient died by euthanasia, 13.5 and 6 months, with a minimum of 2 weeks and a maximum of 5 years (see Table 3). The psychiatrist was usually (61%) the patient’s treating physician. In all cases, other professionals were consulted, most often the general practitioner (63%) and the psychiatrist’s colleagues (46%), and to a greater extent when death by euthanasia was the outcome. Note that, even in the case of performed euthanasia, the palliative care team was involved during the euthanasia assessment procedure (21.7%). In addition, family and/or friends were also often consulted (74%), and a third of these family and/or friends also during a concluding session after a final decision (54%). When euthanasia was the outcome, the APC’s social inner circle was consulted in almost all cases (91%), and in 43.5% also after the final decision had been reached.

Table 3.

Characteristics of the euthanasia procedure in adults with psychiatric conditions.

	All requests(N = 46)	Euthanasia cases(n = 23)
	N (%)
Duration of the euthanasia procedure
Mean (SD)	13.9 months(16.2)	13.5 months(15.9)
Median (min-max)	7 months (2 weeks–5 years)	6 months(2 weeks–5 years)
<1 month	2 (4.3)	2 (8.7)
1–2 months	2 (4.3)	0 (0.0)
2–6 months	11 (23.9)	5 (21.7)
6–12 months	11 (23.9)	6 (26.1)
1–2 years	7 (15.2)	4 (17.4)
>2 years	9 (19.5)	5 (21.7)
Missings	4 (8.7)	1 (4.3)
Involvement of professionals and carers in the euthanasia procedure
Specific role of the psychiatrist in the euthanasia procedure a
Treating physician (of the patient’s psychopathology, not regarding the euthanasia procedure)	28 (60.9)	15 (65.2)
Attending physician of the psychiatrist’s own patient (actively assessing the euthanasia request)	10 (21.7)	5 (21.7)
Attending physician of a patient from a colleague-physician (idem)	11 (23.9)	4 (17.4)
Preliminary advising physician	5 (10.9)	3 (13.0)
Procedural advising physician	13 (28.3)	6 (26.1)
Performing physician	4 (8.7)	4 (17.4)
Involvement of other professionals
None	0 (0.0)	0 (0.0)
The patient’s general practitioner	29 (63.0)	18 (78.2)
Independent colleague-psychiatrist(s)	21 (45.7)	13 (56.5)
Independent LEIF-physician(s), trained and experienced in end-of-life care issues	15 (32.6)	10 (43.5)
Psychologist(s)	15 (32.6)	8 (34.8)
Nurses	11 (23.9)	9 (39.1)
Other physicians of the patient	9 (19.6)	4 (17.4)
Independent physicians of specialised end-of-life centres	9 (19.6)	4 (17.4)
Ethics committee	8 (17.4)	6 (26.1)
Palliative care team	7 (15.2)	5 (21.7)
(Psycho-)Social service(s)	4 (8.7)	3 (13.0)
Another internal advisory committee	3 (6.5)	3 (13.0)
Others b	5 (10.9)	4 (17.4)
Involvement of family and/or friends
No, although the patient did have family or friends	9 (19.6)	2 (8.7)
No, patient did not have family or friends	3 (6.5)	0 (0.0)
Yes, during the euthanasia procedure	22 (47.8)	11 (47.8)
Yes, during and after the euthanasia procedure	12 (26.1)	10 (43.5)
Psychiatrists’ opinion on the substantive due care criteria being fulfilled
Voluntary, sustained and repeated request	41 (89.2)	21 (91.3)
Unbearable suffering	40 (87.0) c	22 (95.6)
Mental competency	38 (82.6) c	20 (87.0) c
Incurability of the disorder	30 (65.2) e	16 (69.6) d
No reasonable therapeutic options left	29 (63.0) c	19 (82.6)
Medical futility	28 (60.9) f	18 (78.3) d
Outcomes of the procedure
Formal advices on the euthanasia requests
Yes, without additional advices	23 (53.5) h	8 (34.8) g
Yes, with additional advices obtained	15 (34.9) h	11 (47.8) g
No	3 (6.9)	0 (0.0)
Don’t know	2 (4.7)	0 (0.0)
Nature of advices given or obtained c
Only positive advices	30 (65.2)	21 (91.3)
Only negative advices	8 (17.4)	0 (0.0)
Mixed positive and negative advices	2 (4.7)	1 (4.3)
Patient still alive? c
No, the patient died by means of euthanasia	23 (50.0)	23 (100)
No, the patient died otherwise i	5 (10.9)
Yes, the procedure is still on going j	8 (17.4)
Yes, the patient had withdrawn the request	3 (6.5)
No idea (not informed)	6 (13.0)

^aPsychiatrists could indicate as many predesignated categories as applicable. For example, at the start or during the course of a euthanasia assessment procedure, the treating physician can decide to also engage as performing physician.

^bOthers: patient’s treating physician (of the patient’s psychopathology), colleague-psychiatrists for informal advice, the religious/spiritual caregiver’ at the affiliated psychiatric centre, members of the ambulant or residential psychiatric care facility, or the case was distributed at the responsibility of the hospital in question.

^cMissing n = 1.

^dMissing n = 2.

^eMissing n = 3.

^fMissing n = 4.

^gThe number of advices was not specified in n = 3.

^hThe number of advices was not specified in n = 4.

ⁱIn some cases specified as death by suicide. In one case the patient died after having the euthanasia request withdrawn.

^jIn these cases, the assessment procedure is concluded but the final decision is not yet made or the practical modalities are to be discussed, for example, the decision when or where to die.

According to the respondents, the substantive due care criteria, as prescribed by the law on euthanasia were fulfilled in 61%–89% of all cases and in 70%–96% of performed euthanasia cases. The criteria ‘medical futility’, ‘incurability of the disorder’ and the ‘absence of reasonable therapeutic options’, were met to the lowest degree (in 61%–65% of all cases, or in 67%–70% cases if corrected for missings). Note that, whereas the legal criterion ‘incurability of the disorder’ was considered sufficiently met in 70% of all performed euthanasia cases (76% if corrected for missings), its operationalised criterion (as suggested in the guidelines on how to adequately assess euthanasia requests from APC) was considered sufficiently met in 83%.

In four out of five cases, at least two legal advices were given or obtained, mostly positive ones (70%). In all performed euthanasia cases, at least two positive advices from other physicians were obtained, except in one case in which both positive and negative advices were obtained. In five cases, in the responding psychiatrists’ opinion, not all of the substantive due care criteria were sufficiently met. The APC’s young age, remaining treatment options according to the state-of-the-art protocol, as well as certain clinical conditions (i.e. personality or bipolar disorder) were reported as contra-indications.

In cases in which the APC died otherwise – for example, suicide (data not shown for reasons of privacy, as n = 5, and the cause of death is not reported in all cases), negative advices were obtained more often, or the absence of hopelessness or remaining reasonable treatment options were reported. In three of the latter cases, psychiatrists reported an improvement in the medical condition due to a new treatment programme.

As for outcomes, 61% of the APC died by means of euthanasia (50%) or otherwise (e.g. suicide). In 26% of the cases, the APC were still alive. In 13%, the reporting psychiatrist was out of the loop regarding the final decision. One psychiatrist reported two final outcomes, as the APC had withdrawn the euthanasia request a few weeks prior to suicide.

Perceived difficulties and/or other experiences

As revealed in Table 4, difficulties in the adequate assessment of the substantive due care criteria were in most cases related to the characteristics of the medical condition. One-quarter of the psychiatrists that were involved in the 23 cases that culminated in euthanasia reported having difficulties in the assessment of the legal criteria ‘medical futility’ (26%) and ‘incurability of the disorder’ (22%), and with its operationalised criterion ‘lack of reasonable therapeutic perspectives’ (26%).

Table 4.

Psychiatrists’ experiences or difficulties perceived during the assessment procedures in adults with psychiatric conditions.

	All requests(N = 46)	Euthanasia cases(n = 23)
	N (%)
Experienced difficulties inassessing criteria a
Lack of a reasonable therapeutic perspective	16 (34.8)	6 (26.1)
Medical futility	15 (32.6) b	6 (26.1)
Incurability of the disorder	14 (30.4)	5 (21.7)
Unbearable suffering	9 (19.5)	1 (4.3)
Voluntary, sustained and well-considered request	7 (15.2)	1 (4.3)
Mental competence	4 (8.6)	1 (4.3)
Experienced forms of pressure
Patient requesting euthanasia under pressure from others	4 (8.7) c	1 (4.3) b
Pressure from the patient to approve euthanasia	24 (52.2)	12 (52.2)
Pressure from patient’s family or friends to approve euthanasia	7 (15.2)	4 (17.4)
Pressure from patient’s family or friends to reject the euthanasia request	4 (8.7) b	3 (13.0)
Pressure from colleagues to reject the euthanasia request	4 (8.7) c	1 (4.3)
Pressure from colleagues to approve euthanasia	3 (6.5) c	3 (13.0) b
Pressure from the care institute to reject the euthanasia request	2 (4.3) c	1 (4.3) b
Pressure from the care institute to approve euthanasia	0 (0.0) c	0 (0.0) b
Other experiences
High emotional burden for yourself	33 (71.7)	15 (65.2)
A lowered risk of suicide with the patient	26 (56.5)	14 (60.9)
New therapeutic opportunities with the patient	12 (26.1)	2 (8.7)
Re-establishment of relationships between patient and significant others	12 (26.1)	9 (39.1)
Fellow patients also requesting euthanasia b	4 (8.7)	2 (8.7)
Emotional support sought?
No	21 (45.7)	10 (43.5)
Yes, inner personal circle	14 (30.4)	9 (39.1)
Yes, colleagues	17 (37.0)	8 (34.8)
Yes, external professional help	1 (2.2)	1 (4.3)
Yes, others	1 (2.2)	1 (4.3)
Attitude towards psychiatric euthanasia changed after this specific case?
No	36 (78.3)	17 (73.9)
Yes d	10 (21.7)	6 (26.1)

^aThis variable was measured by means of a Likert-Scale using scores from 1 to 5, with minimum score = 1 (None) and maximum score = 5 (A great deal). In this table, only the N and % of scores ≥4 are presented.

^bMissing: n = 1.

^cMissing: n = 2.

^dIn some cases the change in attitudes was specified as follows: in 6/10 cases (or 4/6 cases when n = 23) the attitude towards euthanasia in APC changed in a (n) even more risk-aversive way. In 2/10 cases (or 1/6 cases when n = 23) the attitudes changed in a (n even) more favourable way. Finally, in 2/10 cases (or 1/6 cases when n = 23) mixed attitudes due to both favourable and unfavourable experiences were reported.

Half of the psychiatrists (52%) reported feeling pressured by the APC to approve euthanasia. When they felt pressured by the APC’s family or friends, this concerned pressure to decide in favour (15%) or against (9%) approving the APC’s request.

The whole assessment procedure posed a heavy emotional burden on the majority of the psychiatrists (72% and 65% for those confronted with performed euthanasia cases) and more than half of the psychiatrists (irrespective of the outcome) sought emotional support to cope with it. Positive effects were also reported, such as a lower suicide risk (57% and 60.9% for the ones that reported on performed euthanasia cases).

Whereas the re-establishment of relationships between patient and significant others was reported to a greater extent by the psychiatrists who reported on performed euthanasia cases (39% vs 26%), new therapeutic opportunities were reported to a lesser extent (9% vs 26%).

After conclusion of the procedure, the attitudes of the majority of the psychiatrists (78%) towards euthanasia had not changed. If it had changed, most psychiatrists reported that they were willing to engage in future euthanasia procedures, albeit more carefully (e.g. by taking more time to reflect thoroughly on the request, adopting more inter- and supervisions, being less quick to refer to end-of-life consultation centres). Others looked back upon the experience more favourably and described it as beautiful and enriching for all actors involved, including for themselves.

In addition, qualitative analysis of the answers to the open question ‘Would you like to add any clarification or comments about this particular case?’ revealed that some psychiatrists, irrespective of their change of mind, expressed the need for a change in law, for example, implementation of more strict criteria for APC, per the recommendations of the guidelines that were published in the year prior to the survey in order to make these recommendations legally enforceable.

Discussion

Summary of main results

Of all 46 completed euthanasia assessment procedures in APC, most concerned patients who suffered from comorbid psychiatric and/or somatic disorders and who had received different forms of treatment for many years prior to their request. ‘Existential suffering’ and ‘no prospect of improvement’ were reported as the main reasons for the request. In all cases, the entire procedure entailed multidisciplinary consultations, including family and friends.

Psychiatrists reported fewer difficulties in assessing due care criteria related directly to the APC themselves than in assessing the criteria related to their medical condition (e.g. incurability). Both positive and negative experiences during the assessment procedure were reported: for example, a reduced suicide risk for the APC versus emotional burden and feeling pressured by the APC and/or their relatives for the psychiatrist.

As for the final outcomes, half of the completed euthanasia assessment procedures culminated in the performance of euthanasia after at least two legally required advices were obtained, all positive bar one.

Interpretation of findings

Our study has shown the complexity of euthanasia assessment procedures in different regards. One noteworthy illustration is that euthanasia assessment procedures may span multiple months or even years. This can be related to the APC not being expected to die in the foreseeable future, and that some mental disorders tend to fluctuate in severity or even resolve over time, which warrants extreme caution. The majority of the APC, irrespective of the outcome, have been treated for their conditions for many years, giving psychiatrists involved in the assessment a lot of ground to cover. In line with Dutch results,^19,21 our study confirms that, when euthanasia was performed, the assessment procedure took an average of more than 1 year, with a few conspicuous exceptions. In two cases, assessment was reported as concluded in <2 weeks. This would be a violation of the Law, which requires a minimum waiting period of 1 month between the formal request for, and the performance of, euthanasia. However, this is highly unlikely to occur in practice; it is more plausible that the question was accidentally answered from the sole perspective of the individual psychiatrist and their task-specific involvement, instead of for the entire assessment procedure.

Another marked result is that, in 5 of 23 performed euthanasia cases, not all of the legal criteria had been sufficiently met in the responding psychiatrist’s perception. This may raise questions about the legality of some euthanasia cases in APC. However, we have not gauged the opinion of the other clinicians involved in those cases, and we do know that the necessary formal advices were obtained in all cases. These cases again illustrate the complexity of the procedures and therefore the likely lack of consensus between the physicians involved,^14,15 which, according to our study, primarily concerns the incurability of the condition and the lack of reasonable perspectives for improvement.

The psychiatrists also reported specific challenges regarding euthanasia assessment, in terms of the difficulties encountered in determining the extent to which the legal criteria are met in APC cases.

In line with former studies, the APC present with various psychiatric and somatic comorbidities. ¹³ As comorbidity is perceived as an important challenge in medicine in general,^24,25 it also seems to pose a challenge in euthanasia assessment. However, this study confirms former research,^22,23 which maintains that the reasons for the APC’s euthanasia request are not entirely dependent on clinical symptoms alone (e.g. loneliness) and that the APC’s problems are deeply rooted and branched into various aspects of the patient’s past and current life. These findings point to the responsibility of our societies (and thus not only of the field of psychiatry) to address the problems that confront APC such as loneliness. This multidimensional picture undoubtedly compounds the difficulties for psychiatrists in determining (e.g.) the incurability of the APC’s condition and to what extent there are reasonable treatment alternatives, which are reported in about 1 in 3 cases (and which may lead to dissensions, as discussed above). Symptoms of psychiatric disorders tend to change over time – even leading, in some cases, to remission and clinical and/or social rehabilitation – and this underscores the challenge to operationalising this legal criterion in the field of psychiatry, as stated in previous studies.^14,15 The question is whether or not the present guidelines are sufficient to support psychiatrists in these assessments.

Relatively few psychiatrists (9%) reported difficulties in assessing another central legal criterion – mental capacity – which is noteworthy given the predominant focus on competence in clinical and societal debate. A marked finding is that some respondents referred to specific diagnoses as contraindications for APC to be competent, and therefore eligible for euthanasia, a much-debated issue of which the last word has not yet been said.^7,16,26,27 Ruling out APC for euthanasia on the basis of a diagnostic label can be problematic, as diagnostic classification is often contested due to low reliability and validity.^28–30 Though the nature of (some) psychiatric diagnoses may indeed affect mental capacity, it has been stressed in all Belgian guidelines on euthanasia ³¹ that this cannot be grounds to rule out all APC for euthanasia by definition. In any case, utmost caution is needed; and the perceived absence of mental competence in a few cases might suggest the need for a standardised capacity evaluation. To our knowledge, only one Dutch and one Belgian study on this topic have shown that the assessment of this criterion differs among individual physicians (i.e. to some extent due to their personal values and belief system), ³² and, in some cases, seems even flawed, which has led to dissensions among physicians on the evaluation outcome. ¹⁷

Our study brought an underexposed issue to light: namely, the high emotional strain on almost three quarters of the participating psychiatrists. Our findings suggest that one source of such strain is that the whole euthanasia procedure can be seen as a ‘balancing act’ in terms of suicide prevention on the one hand and taking sufficient time for rigorous euthanasia assessment on the other. For example, both reduced suicidality and opportunities for rehabilitation during euthanasia assessment were reported, which is also in line with former research findings.^13,19,22 Anecdotal accounts reveal that suicide risk may be one of the reasons responding psychiatrists feel pressured by the APC into granting the request. Previous research shows that some patients die by means of suicide, even when the euthanasia request has been granted, which suggests that these APC perceived the euthanasia procedure to be too long and/or too arduous.^13,25 The relatively high number of negative advices in this group would corroborate this interpretation. However, it is important to note that we gauged neither for past suicide attempts nor for actual suicide risk in this survey. As for the latter, other potential explanations need to be taken into account: that is, for some, the euthanasia procedure itself might reduce the risk of suicide for that period, but for others it could actually increase the risk of suicide. Another likely source of strain is pressure coming from relatives, either to approve or to deny the APC’s request.

That said, it remains unclear whether the support available to psychiatrists is sufficient and which aspects of the assessment cause the most emotional strain. Current research and guidelines predominantly emphasise the implementation of the legal and due care criteria, thereby largely ignoring the moral and personal challenges for psychiatrists themselves.

Implications for practice, policy and research

As for policy and practice, the finding that some legal criteria were not (sufficiently) met in the perception of the psychiatrists involved seems to corroborate concerns about whether requests are always assessed and monitored adequately and rigorously. ¹⁴

It is deemed quintessential to gauge whether, and to what extent, the recently published guidelines ³³ have sufficiently addressed and effectively tackled the many challenges regarding decision-making and the abovementioned moral dilemmas. For example, the ‘incurability of the disorder’ criterion has been operationalised in these guidelines, but there may be a need for further refinement, or maybe even for an alternative legal term that better suits the field of psychiatry. The same could be said about ‘reasonable state-of-the-art treatment options’– with the inherent relative proximity of ‘therapeutic tenacity’ and ‘therapeutic negligence’– in the context of psychiatry’s lack of objective knowledge regarding prognoses and treatment outcomes. Moreover, the guidelines provide very little on involving and dealing with relatives, while our study revealed that they are often involved and can add pressure on psychiatrists during the assessment. However, as most of the guidelines were published just a few months prior to this survey, it is yet unclear whether the psychiatrists were familiar with them.

In order to expand upon this study’s generated insights, the need for further research is considerable. Qualitative in-depth research into the factors that might further support and enable psychiatrists and other professionals in adequately assessing such requests is needed. This will also allow us to gain deeper insight into the emotional impact these procedures can have on psychiatrists, on the APC and those close to them, and on the therapeutic relationship. ³⁴ Given that the psychiatrists reported successful rehabilitation in some APC, future research should also focus on protective factors – such as engagement in a supportive social network or acquiring resilience and coping skills – that can lead to increased quality of life which may decrease the wish to die. With regard to the ambiguity of law and the difficulties of its implementation in psychiatric practice, especially in the most complex cases, the research method of casuistry may help to address the unclear legal and ethical challenges. Also, large-scale studies should provide more reliable estimates of requests and granting rates and enable the factors influencing the outcomes of the euthanasia requests to be identified. Of the 46 APC applying for euthanasia in this study, two-thirds obtained at least two positive advices and could be considered formally approved for euthanasia. This result may suggest a high approval rate – but that is misleading, as prior evidence indicates that the vast majority of requests are denied, rejected, or withdrawn before a formal outcome is reached.^{13,19,21,35,36} Moreover, obtaining two positive advices does not automatically mean that the APC have been approved for euthanasia, as the physician entrusted with the clarification of the APC’s euthanasia requests may seek to obtain additional (i.e. more than the two legally required) advices. Lastly, future research might also focus on examining the impact and consequences of ungranted requests – as, for example, the APC might be left to their fate with their death ideation, while their physicians refuse to engage in discussion about it.

Strengths and limitations

This is the first study to provide an in-depth analysis of the experiences of Belgian psychiatrists regarding the complexity of euthanasia practice in adults with psychiatric conditions. It reveals new insights into many aspects of the assessment procedure and the impact it has on the psychiatrists involved. We gathered information on 46 assessed cases and 23 performed euthanasia cases predominantly based on psychiatric disorders that were checked for uniqueness by crossing essential variables. As for representativeness, according to the latest official Euthanasia Review Committee Report, 27 APC died in 2016 and 26 in 2017 by means of euthanasia ¹² (p. 46). Assuming that the numbers remained similar in the period of our survey, this may suggest that our study comprises close to all euthanasia cases based on psychiatric disorders. However, given the potential response and selection bias in our study, we cannot make assertions about the representativeness of the captured cases in relation to the entire euthanasia practice in psychiatry.

Though this analysis provides rich insights into psychiatrists’ practice and challenges in dealing with euthanasia requests from APC, the authors wish to stress that the data do not readily allow for evaluation of: (a) the legality of performed euthanasia cases, or (b) the factors predictive of requests leading to euthanasia.

Some results should be interpreted with caution due to the potential sources of bias: response bias given low response rates, but also selection bias as we suspect respondents often refrained from reporting about concluded cases where the APC are still alive. Assuming that the psychiatrists were much more inclined to report on euthanasia requests that have been carried out than on those put on hold, we have thus not captured a large proportion of completed evaluations that have not culminated in euthanasia.

This is supported by anecdotal evidence, and annual reports from clinical practice reveal that a large proportion of these patients put their procedure on hold after 1 to 2 consultations. ³⁷ This combination of potential biases renders the interpretation of half of the requests leading to euthanasia untrustworthy and overestimated, all the more so considering existing sources reporting lower rates.^19,21,38 Finally, and although the survey was pre-tested for cognitive validity, we cannot exclude the possibility of misunderstandings remaining with regard to the interpretation of individual items.

Conclusions

This study has revealed the complexity of euthanasia assessment in APC, due to the variety of (comorbid) diagnoses and often severe somatic co-diagnoses, the variety of reasons for requesting euthanasia (also appealing to the responsibility of our society), the difficulties in assessing the legal and due care criteria, and the emotional impact of euthanasia assessment on psychiatrists. Not only does it involve people with long histories of medical diagnoses and treatment, but assessment also requires a large amount of time. When the euthanasia request culminated in the performance of euthanasia, the entire procedure spanned an average of 13 months (which is much longer than the legally required 1 month) and entailed multidisciplinary consultations (e.g. psychologists, palliative care team), including with family and friends (which is not required by law). Our findings indicate that psychiatrists require support in more than one respect if euthanasia requests by APC are to be handled adequately: To what extent can or do the guidelines provide answers to assessment complexities? Is there a need for specific education in assessment? Are legal clarifications in order? Future (qualitative) research can aid by focussing on the psychiatrists’ and the APC’s experiences and needs in this regard.

Due to the considerable risk of bias, this analysis should be read as an account of the types of cases and issues encountered in psychiatric euthanasia practice, and not necessarily as a reflection of the entire psychiatric euthanasia assessment practice. A more robust mapping of euthanasia assessment procedures in APC would be better achieved through studies with large reliable denominators generating estimates of (e.g.) granting rates and insight into factors influencing the granting of requests.

Author biographies

Monica Verhofstadt, MA, MSc, holds a Master in Clinical Psychology. Since October 2017, she joined the Belgian End-of-Life Care Research Group as a doctoral researcher. In the preceding years she did volunteer work at Vonkel, a Belgian organisation dedicated to supporting patients and their relatives with end-of-life issues. There, she conducted intake interviews with patients applying for euthanasia and joined the first Belgian research projects on euthanasia requests from and the euthanasia practice regarding patients with psychiatric conditions.

Kurt Audenaert is Senior full professor in Psychiatry and Forensic Psychiatry, and the Head of Clinic Adult Psychiatry in the GhentUniversity Hospital (Belgium). He holds e.g., amaster’s in Medical Sciences, in Psychiatry andin Criminology. He also holds a PhD in medical sciences (functional brain imaging in psychiatry: a functional-psychopathological approach) and is a trained psychotherapist (family therapy).

Kris Van den Broeck, PhD, is psychologist and behavioural therapist, visiting professor at the University of Antwerp, Antwerp, Belgium and managing director of the Flemish Psychiatric Association. He is involved in the training of (future) (general) practitioners and psychiatrists. His topics of interest are the organisation of (mental) health care, interprofessional collaboration amongst health professionals, and appropriate communication towards patients in care. Ethical issues often take an important place in these themes.

Luc Deliens holds an MA in Sociology, MSc in Human Ecology and PhD in Health Sciences. He is trained in medical sociology and Professor of Palliative Care Research. Since 2000, he is the founding Director of the End-of-Life Care Research Group of the Vrije Universiteit Brussel (VUB) and Ghent University, in Belgium (“http://www.endoflifecare. be“www.endoflifecare.be).

Freddy Mortier is Full professor of ethics at Ghent University, Belgium. He studied both at the Ghent University and Paris-Sorbonne and holds a PhD in philosophy at Ghent University. He is a former member of the Belgian Advisory Committee for Bio-ethics, member of Ghent University Hospital Ethics Committee. He is a member of the End-of-Life Care Research Group (Free University Brussels and Ghent University, Belgium).

Koen Titeca holds a master’s in medical sciences and in Psychiatry and is a trained psychotherapist. He is the head of the Emergency Psychiatry Department in the General Hospital of Groeninge, Kortrijk (Belgium). He is alsotrained LEIF-physician (Life End Information Forum) and gives courses regarding ‘euthanasia and psychiatry’ to LEIF-physicians and nurses. He co-authored the Flemish Psychiatric Association’s Guideline on how to adequately manage euthanasia requests and procedures from patients with psychiatric conditions.

Dirk De Bacquer, PhD, is a Senior Full Professor in Epidemiology and Biostatistics at the Department of Public Health and Primary Care, Faculty of Medicine and Health Sciences, Ghent University, Belgium. He is also theHead of the Epidemiology and Prevention section and of the Biostatistics Unit, Ghent University, Belgium. He was theFormer Chair of the Department of Public Health, Ghent University, Belgium, and Fellow of the European Society of Cardiology.

Kenneth Chambaere is Interdisciplinary Professor of Public Health, Sociology & Ethics of the End of Life at the End-of-Life Care Research Group of Ghent University & Vrije Universiteit Brussel (VUB). He has an MSc in Sociology, a Postgraduate in Logic, History and Philosophy of Science, and a PhD in Medical-Social Sciences. He was a Postdoctoral Fellow of the Research Foundation Flanders (FWO). Prof. Chambaere’s current research focusses on three main themes: (1) end-of-life practices, (2) palliative care in and by the community and (3) end-of-life care for people in vulnerable positions.