OUP user menu

The labelling and reporting of euthanasia by Belgian physicians: a study of hypothetical cases

Tinne Smets , Joachim Cohen , Johan Bilsen , Yanna Van Wesemael , Mette L. Rurup , Luc Deliens
DOI: http://dx.doi.org/10.1093/eurpub/ckq180 19-26 First published online: 3 December 2010

Abstract

Background: Belgium legalized euthanasia in 2002. Physicians must report each euthanasia case to the Federal Control and Evaluation Committee. This study examines which end-of-life decisions (ELDs) Belgian physicians label ‘euthanasia’, which ELDs they think should be reported and the physician characteristics associated with correct labelling of euthanasia cases, the awareness that they should be reported and the reporting of them. Methods: Five hypothetical cases of ELDs: intensified pain alleviation, palliative/terminal sedation, euthanasia with neuromuscular relaxants, euthanasia with morphine and life-ending without patient request were presented in a cross-sectional survey of 914 physicians in Belgium in 2009. Results: About 19% of physicians did not label a euthanasia case with neuromuscular relaxants ‘euthanasia’, 27% did not know that it should be reported. Most physicians labelled a euthanasia case with morphine ‘intensification of pain and symptom treatment’ (39%) or ‘palliative/terminal sedation’ (37%); 21% of physicians labelled this case ‘euthanasia’. Cases describing other ELDs were sometimes also labelled ‘euthanasia’. Factors associated with a higher likelihood of labelling a euthanasia case correctly were: living in Flanders, being informed about the euthanasia law and having a positive attitude towards societal control over euthanasia. Whether a physician correctly labelled the euthanasia cases strongly determined their reporting knowledge and intentions. Conclusion: There is no consensus among physicians about the labelling of euthanasia and other ELDs, and about which cases must be reported. Mislabelling of ELDs could impede societal control over euthanasia. The provision of better information to physicians appears to be necessary.

Introduction

Medical end-of-life decisions (ELDs) that possibly influence the remaining life-span frequently precede dying in many countries.1–3 They include decisions to withhold or withdraw potentially life-prolonging treatments, to alleviate pain or other symptoms with a possible life-shortening side effect, to administer life-ending drugs without explicit patient request, and to perform physician-assisted suicide or euthanasia. The decision to perform palliative or terminal sedation can also be made, which is the administration of drugs to keep a patient continuously in deep sedation or coma until death.4,5 Euthanasia is legal only in the Netherlands, Luxembourg and Belgium.6–8 Since 2002, physicians in Belgium may under strict legally defined circumstances administer life-ending drugs at the explicit request of a patient.8 The practice of euthanasia embraced 1.9% of all deaths in Flanders, Belgium, in 2007.9 To enable societal control over euthanasia, physicians must report each euthanasia case to the Federal Control and Evaluation Committee (Review Committee) which determines whether the physician has complied with the requirements of the law.8

A post-mortem survey conducted in Flanders (the Dutch-speaking part of Belgium) suggested that about half of all euthanasia cases are not being reported.10 The main reason euthanasia cases were not being reported was that many cases were not labelled as euthanasia by the physicians involved in the decision.10 A similar reason for not reporting euthanasia cases was found in the Netherlands.11–13 In actual practice the dividing lines between different ELDs, such as between the alleviation of pain and symptoms and euthanasia may not always be easy to define.11,14–16 Euthanasia is considered to be ‘the administration of life-ending drugs by a physician with the explicit intention of ending a patient’s life, at the latter’s explicit request.’8,17 The death of the patient must also be the result of the administration of the drugs.8 Not all physicians may be aware of this definition, and even if they are, its interpretation can be complex for physicians who are for instance uncertain about the actual effect of giving certain drugs on the ending of life.18 For euthanasia, the administration of barbiturates followed by neuromuscular relaxants is advised19,20 because they ensure a peaceful and certain death. In practice physicians also perform euthanasia with opioids,13,21 howbeit that opioids are advised against for euthanasia because they have an uncertain life-shortening effect and can have unpleasant side effects.19,20 When opioids are used with the intention to end life, the actual life-ending effect may thus not always be very clear for the physician.18 Non-reporting may not only be caused by confused definitions, but also by other factors, such as a physician’s unwillingness to report euthanasia cases, for example, out of fear of criminal prosecution or because the reporting procedure is perceived as too burdensome.

If euthanasia is not labelled as euthanasia, this would be problematic as it could hamper effective societal control over euthanasia practice.

Mislabelling other ELDs as euthanasia would also be problematic because it could make physicians hesitant to apply certain end-of-life interventions as they might be afraid that they are performing euthanasia.

How physicians label different ELDs, which of these decisions they think must be reported as euthanasia, which decisions they would report themselves, and the multitude of personal, attitudinal and knowledge determinants of correct labelling of euthanasia, and of reporting and non-reporting of euthanasia, have not been investigated in Belgium. Insight into those issues can contribute to the further understanding of euthanasia practice in Belgium and to the debate about the prospect of efficient societal control over euthanasia in countries, which are debating the legalization of the practice.

The aim of this article is to answer the following research questions:

  1. Which ELDs do physicians in Belgium think are euthanasia and should be reported, and which ELDs would they report themselves?

  2. Which personal, attitudinal and knowledge variables predict whether physicians label a euthanasia case as euthanasia, know euthanasia must be reported and would report it themselves?

Methods

Study design

In 2009 we sent a questionnaire to a representative sample of 3006 physicians registered as working in Belgium, who had graduated in their specialty at least 12 months before the sample was drawn and were likely to be involved in the care of dying patients on the basis of their specialty: general practitioners, anaesthesiologists, gynaecologists, internists (including oncologists), neurologists, pulmonologists, neuro-psychiatrists, psychiatrists, cardiologists, radiotherapists and surgeons were included. The sample was stratified for province and speciality; for each of the 10 provinces a random proportional sample was drawn within each speciality.

A lawyer was involved as intermediary between responding physicians and researchers in the mailing procedure to guarantee the anonymity of the physicians. According to the Total Design Method an intensive follow-up mailing in case of non-response was performed with up to three reminders.22

Finally, non-responders were sent a one-page questionnaire to assess non-response bias. Alongside reasons for not participating to the survey, questions were asked about two key items in the survey, the physician’s acceptance of euthanasia and whether or not he/she had ever received a request for euthanasia.

Measurement instrument

The pre-structured questionnaire drew partly on questionnaires previously used.23,24 The questionnaire was developed in Dutch and forward-backward translated into French to avoid differences due to language. Questions were asked about the physician’s socio-demographics, work-related characteristics and attitudes, and practices concerning euthanasia. Physicians were presented with five hypothetical cases of a patient in the final stage of a terminal disease (Supplementary Box 1). In each case (except in case 5 where the patient is unconscious), the patient explicitly asks the physician to end his/her life. This factor was kept constant, as we especially wanted to focus on the act that the physician performs in response to such a request for life ending. The scenario varied between the drugs administered, the mode of administration, and the effect of the administration of the act, thus covering the different types of ELDs. As we were interested in how physicians label a medical decision based on objective facts, we did not explicitly mention the intention of the physician because intentions are known to be multilayered and ambiguous. Using the intention of the physician would also have been too influential as to the correct labelling of the cases.

For each case we asked the physician which label best describes the act (euthanasia, palliative/terminal sedation, life-ending without explicit request, intensification of pain and symptom treatment, other), whether they thought it conceivable that they would perform a similar act themselves, whether the act should be reported to the Review Committee and whether they would report the act themselves. In order to select relevant and realistic cases, a variety of cases was presented to several experts in the field of palliative care. We selected cases based on suggestions from the experts and on the literature.

Statistical analysis

Significant differences between response population and total sample were found for region but not for specialism. A weighting factor was used to correct for this response bias by region, making the data representative for all physicians in the sample.

Differences between physicians’ answers on the different hypothetical cases were tested by Fisher’s Exact test. P ≤ 0.05 were considered to indicate statistical significance. Multivariate logistic regression was performed to estimate associations between a physician’s characteristics and their labelling or reporting of euthanasia cases. Odds ratios (ORs) and 95% confidence intervals are presented. The analyses were performed using SPSS 16.0 (SPSS Inc., Chicago, IL, USA) and StatXact 6.

Results

Response rate and response bias

Of the 3006 questionnaires sent, the non-response study found out that response was impossible for 223 respondents: 149 of those physicians did not receive the questionnaire, 1 physician was ill, 1 was deceased and 72 were no longer active as a physician or worked in a specialism that was not included in our study.

Of the remaining 2783 questionnaires, 914 were returned. Of the non-responders, 583 replied to the non-response questionnaire. Not being involved in the care of dying patients and never responding to questionnaires were the main reasons for non-response. Those indicating they no longer worked as physicians (N = 32) or had not received the questionnaire (N = 25) were subtracted from the denominator. Thus, the response rate of the study was 34%.

Non-responders were somewhat less likely to agree that euthanasia is acceptable (87.4% vs. 93.0%, P = 0.001) and were more neutral towards the statement than responders (8.8 vs. 4.0%, P = 0.001). No significant difference between responders and non-responders was found for the question whether or not the physician had ever received a request for euthanasia (48.3% of responders vs. 46.0% of non-responders ever received a request; P = 0.405).

Characteristics of responding physicians

Sixty-four percent of responding physicians were men, 49% were Roman Catholic, 62% were general practitioners, 61% had >20 years experience as a physician, 48% had received training in palliative care and 19% had cared for more than ten terminal patients in the last year (table 1).

View this table:
Table 1

Characteristics of the studied physicians

CharacteristicN (%)
Number of studied questionnaires914
Socio-demographics
    Sex
        Men576 (63.5)
        Women323 (35.6)
    Age (years)
        <35110 (12.4)
        36–50323 (36.7)
        51–65398 (45.1)
        >6651 (5.8)
    Religious affiliation/philosophy of life
        Roman Catholic428 (49.1)
        Other denomination21 (2.4)
        Religious, but no specific denomination104 (12.0)
        Humanist163 (18.7)
        Not religious155 (17.8)
    Region
        Flanders480 (52.8)
        Wallonia305 (33.6)
        Brussels123 (13.6)
Work-related characteristics and experiences
    Speciality
        General practitioner561 (61.8)
        Clinical specialist347 (38.2)
    Years experience as physician
        ≤10148 (16.6)
        11–20202 (22.6)
        21–30287 (32.1)
        31–40216 (24.1)
        >4040 (4.5)
    Training in palliative care
        Yesa433 (48.1)
        At medical school133 (30.4)
        In post-graduate education375 (85.8)
        Other training46 (10.5)
    Member of palliative team/service
        Yes47 (5.3)
    Number of terminal patients cared for in the last 12 months
        0202 (24.5)
        1–9463 (56.2)
        ≥10160 (19.4)
  • a: More than 1 answer possible

Labelling and reporting of ELDs

Eighty-one percent of physicians labelled the case in which the physician administers a sleep-inducing drug and a neuromuscular relaxant at the explicit request of the patient as ‘euthanasia’; 9% labelled this case as ‘palliative/terminal sedation’ (table 2). The case in which the physician ends the patient’s life using morphine was labelled as ‘euthanasia’ by 21% of physicians. This case was more often labelled as ‘intensification of pain and symptom treatment’ (39%) or as ‘palliative/terminal sedation’ (37%). The acts described in the other cases were also sometimes labelled as ‘euthanasia’, but less frequently so (between 6 and 11%). The case in which the physician gradually and proportionally raised the dose of morphine was most often labelled as ‘intensified pain alleviation’; the case in which the physician administers midazolam until death was labelled as ‘palliative/terminal sedation’ by 63% of the physicians. The case in which the physician ends the life of a comatose patient by disproportionally raising the dose of morphine and adding valium to the infusion was labelled as ‘intensification of pain and symptom treatment’ by 43% of the physicians and less often as ‘life-ending without patient request’ (17%).

View this table:
Table 2

Differences between cases according to physicians’ willingness to report the case and their labelling of the act

Case 1: intensified pain alleviationCase 2: palliative/ terminal sedationCase 3: euthanasia 2 (with morphine)Case 4: euthanasia 1 (with neuromuscular relaxants)Case 5: life-ending without patient requestP-value
Which label describes best the act that the physician performs?<0.001
    Euthanasia10.79.720.580.96.2
    Palliative/terminal sedation24.073.237.29.132.4
    Life-ending without explicit request1.12.31.83.316.6
    Intensification of pain and symptom treatment63.413.138.54.943.3
    Other0.91.72.01.81.5
Do you think the death of this patient must be reported to the Federal Control and Evaluation Committee according to the euthanasia law? If you had performed this act, would you report?<0.001
    This death must be reported, and I would report10.99.118.168.411.0
    This death must be reported, but I would not report0.82.21.74.52.0
    This death must not be reported69.862.856.114.160.3
    Do not know18.425.124.113.026.6
Willingness to report the death of the patient according to physicians’ labelling of the act<0.001
    Label ‘euthanasia’, would report7.86.815.769.73.3
    Label ‘euthanasia’, would not report2.92.74.511.62.7
    Label ‘other than euthanasia’, would report8.08.99.54.813.3
    Label ‘other than euthanasia’, would not report81.381.670.313.980.6

For all cases which were labelled as ‘euthanasia’ or ‘life-ending without explicit request’, the physicians were less likely to find it conceivable that they would perform a similar act themselves than for cases which they labelled differently (P < 0.001) (not in tables).

Seventy-three percent of physicians were aware that the case in which the physician administers a sleep-inducing drug and a neuromuscular relaxant at the explicit request of the patient had to be reported to the Review Committee (table 2). Twenty-two percent of physicians who did not label this case as ‘euthanasia’ were aware that the case had to be reported, while 14% of those who did label the case as ‘euthanasia’ said it did not have to be reported or that they did not know whether the case had to be reported or not. Sixty-eight percent of physicians indicated that they would report the case themselves. Fifty-eight percent of those who indicated that they would not report the case themselves did find it conceivable that they would perform a similar act themselves (not in tables). Physicians who labelled the other cases as ‘euthanasia’ indicated most of the time that they would also report the case.

Factors associated with correct labelling of euthanasia cases

In a multivariate logistic regression analysis using the case in which the patient explicitly requests that their life be ended and the physician administers a sleep-inducing drug and subsequently a neuromuscular relaxant, factors associated with a higher likelihood of labelling this case correctly as ‘euthanasia’ were living in Flanders (OR 2.69), being sufficiently informed about the euthanasia law (OR 1.69), and having a positive attitude towards societal control over euthanasia (OR 1.74) (table 3).

View this table:
Table 3

Factors associated with correct labelling of euthanasia casesa

Euthanasia 1 (case 4)Euthanasia 2 (case 3)
N (%)OR (95% CI)N (%)OR (95% CI)
Region
    Wallonia202 (70.4)160 (21.5)NS
    Flanders413 (89.0)2.69 (1.52–4.75)82 (18.2)NS
    Brussels87 (74.4)0.92 (0.53–1.58)31 (27.4)NS
I am sufficiently informed about the euthanasia law
    Yes333 (86.0)1.69 (1.11–2.57)71 (19.0)NS
    No366 (77.2)1100 (21.7)NS
Attitude towards control over euthanasiab
    Pro507 (84.4)1.74 (1.16–2.62)124 (21.5)NS
    Against187 (73.9)149 (19.9)NS
Attitude towards euthanasia
    Pro648 (82.1)NS147 (19.2)1
    Against57 (70.4)NS26 (32.5)1.87 (1.07–3.30)
Number of terminal patients cared for in the last 12 months
    0152 (83.1)NS55 (30.6)1
    1–9365 (80.8)NS77 (17.7)0.49 (0.32–0.75)
    ≥10127 (80.9)NS24 (15.6)0.40 (0.22–0.71)
  • a: Multivariate logistic regression. Presented figures are numbers and percentages correctly labelled, odds ratios and 95% confidence intervals. Independent variables, which have no significant relationships are not presented in the table. Sex, years experience as physician, training in palliative care, religious affiliation, specialty and whether a physician had ever performed euthanasia in practice were entered in the regressions but were not significant and were therefore eliminated from the table

  • b: K-means cluster analysis. Physicians are divided in two groups (pro or against control over euthanasia) according to their attitudes on three statements, assessed on a 5-point Likert Scale. Statement 1: ‘Euthanasia is a private matter between patient and physician that does not need to be controlled by the Control and Evaluation Committee.’ Statement 2: ‘Societal control over the euthanasia practice is necessary.’ Statement 3: ‘Reporting euthanasia cases contributes to the carefulness of physicians’ medical behaviour at the end of life’

  • NS = not significant

Using the case in which the physician ended the patient’s life at his/her request using morphine, factors associated with a higher likelihood of labelling this case as ‘euthanasia’ were being against euthanasia (OR 1.87), and not having cared for terminal patients in the last year (OR 1.00).

Factors associated with reporting of euthanasia cases

Physicians who labelled the euthanasia case in which the physician administers a sleep-inducing drug and a neuromuscular relaxant at the explicit request of the patient correctly were substantially more likely to know that this case had to be reported to the Review Committee than those who labelled it incorrectly (table 4). Controlling for labelling, other factors associated with higher likelihood of knowing that this case had to be reported were being female (OR 1.76), living in Flanders (OR 2.76), being sufficiently informed about the euthanasia law (OR 2.36), having a positive attitude towards euthanasia (OR 2.01), and having a positive attitude towards societal control over euthanasia (OR 2.53). Factors associated with willingness to report this euthanasia case were the same, except for attitude towards euthanasia, which had no influence.

View this table:
Table 4

Factors associated with reporting of euthanasia casesa

Euthanasia 1 (case 4)Euthanasia 2 (case 3)
This case must be reportedI would report this caseThis case must be reportedI would report this case
N (%)OR (95% CI)N (%)OR (95% CI)N (%)OR (95% CI)N (%)OR (95% CI)
Sex
    Men399 (70.9)1392 (71.0)1119 (21.2)NS134 (24.3)NS
    Women240 (77.2)1.76 (1.11–2.80)232 (80.0)2.07 (1.30–3.30)56 (18.2)NS72 (26.3)NS
Religious affiliation/philosophy of life
    Roman Catholic308 (74.4)NS302 (75.3)NS82 (19.8)1.23 (0.64–2.35)96 (24.6)1.66 (0.84–3.31)
    Other denomination13 (61.9)NS13 (68.4)NS10 (50.0)1.16 (0.54–2.50)10 (52.6)2.54 (1.14–5.66)
    Religious, but no specific denomination76 (75.2)NS69 (75.0)NS15 (15.2)5.72 (1.59–20.51)18 (19.6)9.10 (2.36–35.14)
    Humanist104 (67.1)NS105 (69.5)NS34 (21.8)0.96 (0.38–2.41)47 (31.1)1.63 (0.65–4.06)
    Not religious122 (79.7)NS118 (79.2)NS25 (16.4)128 (19.3)1
Region
    Wallonia170 (59.9)1172 (62.1)161 (21.2)NS72 (26.4)NS
    Flanders397 (85.6)2.76 (1.74–4.37)381 (83.7)1.73 (1.11–2.68)89 (19.3)NS104 (23.5)NS
    Brussels69 (57.0)0.80 (0.45–1.42)69 (64.5)1.02 (0.60–1.90)25 (21.4)NS30 (28.6)NS
I am sufficiently informed about the euthanasia law
    Yes320 (82.5)2.36 (1.52–3.68)316 (82.7)2.14 (1.40–3.26)80 (20.9)NS95 (25.8)NS
    No313 (65.8)1302 (67.0)188 (18.6)NS105 (23.6)NS
Attitude towards euthanasiab
    Pro590 (74.6)2.01 (1.04–3.89)576 (75.3)NS136 (22.8)NS162 (28.6)NS
    Against48 (60.0)147 (62.7)NS33 (13.3)NS36 (15.2)NS
Attitude towards control over euthanasiac
    Pro480 (79.2)2.53 (1.65–3.89)477 (81.4)3.32 (2.20–5.02)136 (22.8)2.32 (1.33–4.03)162 (28.6)2.39 (1.33–4.29)
    Against148 (59.0)1135 (56.5)133 (13.2)136 (15.2)1
Labelling of the act
    Intensification of pain and symptom treatment2 (4.9)15 (12.2)117 (5.3)128 (8.9)1
    Euthanasia604 (85.9)92.57 (21.59–396.89)584 (85.8)43.15 (14.58–127.71)120 (69.4)45.92 (24.68–85.44)126 (77.8)44.07 (23.30–83.34)
    Life-ending without explicit request11 (40.7)7.84 (1.43–42.87)12 (44.4)5.18 (1.30–20.69)4 (26.7)3.09 (0.59–16.16)5 (38.5)4.40 (1.02–19.07)
    Palliative/terminal sedation16 (20.0)3.13 (0.65–15.09)19 (25.3)2.7 (0.82–8.96)30 (9.6)1.71 (0.89–3.29)41 (13.7)1.36 (0.76–2.44)
    Other6 (37.5)16.31 (2.57–103.54)4 (28.6)4.09 (0.79–21.15)1 (5.9)1.24 (0.14–11.04)3 (20.0)1.93 (0.39–9.51)
Number of terminal patients cared for in the last 12 months
    0125 (70.2)NS131 (77.5)NS54 (29.7)NS69 (42.1)1
    1–9337 (73.4)NS329 (73.6)NS76 (16.8)NS88 (20.2)0.35 (0.20–0.61)
    ≥10120 (76.9)NS109 (73.2)NS27 (17.3)NS26 (17.3)0.26 (0.13–0.55)
  • a: Multivariate logistic regression. Presented figures are numbers and percentages, odds ratios and 95% confidence intervals. Independent variables, which have no significant relationships, are not presented in the table. Years of experience as physician, training in palliative care, specialty and whether a physician had ever performed euthanasia in practice were entered in the regressions but were not significant and were therefore eliminated from the table

  • b: Attitude towards euthanasia was measured on a 5-point Likert scale using the statement: ‘The administration of life-ending drugs at the explicit request of a patient is acceptable for patients with a terminal disease with extreme, uncontrollable pain or other uncontrollable suffering.’ Physicians who answered that they ‘agreed’ or ‘totally agreed’ with the statement were classified as pro euthanasia; physicians who answered they ‘disagreed’, ‘totally disagreed’, or were neutral towards the statement were classified as being against euthanasia

  • c: K-means cluster analysis. Physicians are divided in two groups (pro or against control over euthanasia) according to their attitudes on three statements, assessed on a 5-point Likert Scale. Statement 1: ‘Euthanasia is a private matter between patient and physician that does not need to be controlled by the Control and Evaluation Committee.’ Statement 2: ‘Societal control over the euthanasia practice is necessary.’ Statement 3: ‘Reporting euthanasia cases contributes to the carefulness of physicians’ medical behaviour at the end of life’

  • NS= not significant

Physicians who labelled the case in which the physician ended the patient’s life at his/her request using morphine as ‘euthanasia’ was more likely to know that this case had to be reported than those who labelled this case differently. Controlling for labelling, other factors associated with higher likelihood of knowing that the case had to be reported were being religious without having a specific denomination (OR 5.72) and having a positive attitude towards societal control over euthanasia (OR 2.32). Factors associated with willingness to report this euthanasia case were the same. In addition, not having cared for terminal patients in the last year also increased a physician’s willingness to report this case.

Discussion

Our study shows that there is a lack of agreement among physicians in Belgium about the classification of euthanasia and other ELDs, and about which cases must be reported as euthanasia to the Federal Review Committee. Seven years after implementation of the euthanasia law in Belgium, 2 out of 10 physicians, likely to be involved in the care of dying patients, did not label a hypothetical case in which a physician ends the life of a patient at that patient’s explicit request using neuromuscular relaxants (case 4) as ‘euthanasia.’ Three out of 10 physicians did not know the case had to be reported to the Federal Review Committee. Most physicians labelled the euthanasia case in which the physician ends the life of a patient at that patient’s explicit request using morphine (case 3) as ‘intensification of pain and symptom treatment’ (39%) or as ‘palliative/terminal sedation’(37%); only 21% of physicians labelled this case as ‘euthanasia’. Most physicians who knew the euthanasia case with neuromuscular relaxants (case 4) had to be reported indicated that they would report the case themselves. In particular, the correct labelling of the euthanasia case was strongly associated with whether a physician knew the case had to be reported and whether they would report the case themselves.

While we used a large representative sample of physicians and included only those specialties which are likely to be involved in the care of dying patients, a limitation of our study is that the response rate was only 34%, limiting the generalizability of the results. However, comparison of the responders and non-responders through our non-response survey suggests that the sample of responders was similar to the group that did not respond in terms of region, and in terms of whether or not they had ever received a request for euthanasia. Furthermore, we used hypothetical cases that are reductions of the complex situations that may occur in clinical reality.11,25,26 However, all cases were tested with several specialists in the field of palliative care, who found the cases realistic and could answer the questions adequately. Furthermore, we cannot exclude the possibility of a social desirability bias, especially for the question of whether or not the physician would report the case. Finally, intended behaviour and real behaviour may not be identical, as real behaviour is known to be influenced by situational factors.27–29

Identical hypothetical cases were not uniformly labelled by physicians. Some cases were inaccurately labelled as ‘euthanasia’. A consequence of this may be that these cases may also be unnecessarily reported to the Review Committee. Far more problematic is that this incorrect labelling of normal medical practice as euthanasia could prevent physicians from applying these adequate and often necessary end-of-life interventions. As shown in our study, a physician’s willingness to perform ELDs such as palliative sedation or intensified pain alleviation was much lower if they labelled the case as ‘euthanasia’ or ‘life-ending without explicit request.’ This finding has implications that stretch well beyond the countries with a law on euthanasia; better knowledge about euthanasia and about the use and effects of opioids can contribute to better treatment of pain and other suffering.30–32

In light of the Belgian law on euthanasia, an important finding is that 2 out of 10 physicians labelled the euthanasia case with neuromuscular relaxants (case 4) incorrectly and 3 out of 10 were unaware of the legal reporting obligation.

The euthanasia case in which the physician ended the patient’s life at his/her request using morphine (case 3) was labelled as euthanasia by only one in five of the physicians. When a patient requests that their life be ended, and the physician in response disproportionally increases the morphine dose instead of administering neuromuscular relaxants the distinction between euthanasia and normal intensification of symptom treatment may become blurred. Cases in which the physician performs euthanasia with opioids are often not perceived as euthanasia by the physician.10 Some physicians see a ‘grey area’ or continuum between palliation and euthanasia and find that the distinctions between the two are not always very clear cut.33

A post-mortem survey on ELDs in Flanders found that most unreported euthanasia cases were not regarded as euthanasia by the physicians themselves.10 Our findings also show that physicians who regarded the euthanasia cases as euthanasia were substantially more likely to know that these cases had to be reported and were substantially more willing to report the cases themselves than those who labelled the cases differently. Correct labelling is thus important to enable adequate societal control over the practice of euthanasia. The aforementioned study also found that euthanasia cases that were not labelled as euthanasia were dealt with less carefully than the cases that were.10 If physicians have another definition of euthanasia than the definition determined by the euthanasia law then they will not be inclined to comply with the requirements of the law such as the mandatory consultation of a second independent physician. Correct labelling of euthanasia cases is thus also pivotal in guaranteeing the carefulness of the euthanasia practice.

A considerable number of physicians who labelled the cases correctly as ‘euthanasia’ did not know that they had to be reported, indicating a lack of knowledge of the law, a conclusion also suggested by the fact that not being sufficiently informed about the law was associated with lower awareness of the reporting obligations. Both correct labelling of ELDs and knowledge with regard to legal requirements thus seem important factors in explaining reporting behaviour.

We found considerable regional differences in labelling and reporting knowledge and intentions with regard to the euthanasia case with neuromuscular relaxants (case 4): compared with physicians from Wallonia, Flemish physicians were 26% points more aware of the legal requirement to report the case (86 vs. 60%) and 22% points more willing to report it (84 vs. 62%). This is also reflected in the Review Committee’s biennial reports: only ∼15% of all reported euthanasia cases had been reported by French-speaking physicians. 34–37 The fact that Flemish physicians were more likely to label the euthanasia case correctly partly explains the difference in reporting. However, even after controlling for labelling, geographic region was still strongly associated with reporting knowledge and willingness. The geographic differences in labelling and reporting could be influenced by a difference in information dissemination.34 Flanders shares the same language with the Netherlands, so Flemish physicians may have had better access to information from the Netherlands, which has seen a decades-long history of public debate about euthanasia, than have French-speaking physicians from Wallonia.34 Furthermore, the establishment of the LifeEndInformationForum (LEIF) in Flanders, a network of physicians trained to give expert advice and consultation on euthanasia and other ELDs, may also have played a role in informing Flemish physicians.34,38–40

Physicians were less likely to know that the euthanasia case with neuromuscular relaxants (case 4) had to be reported when they were against euthanasia and were less likely to label the euthanasia case correctly, to know that the case had to be reported and to be willing to report it themselves when they were against control over euthanasia. Those physicians who are against euthanasia or control over euthanasia may be less open to information about the euthanasia law and the legal reporting obligation than those with a more positive attitude, and will hence be less inclined to report their euthanasia cases in actual practice. If physicians who are unaware that euthanasia cases must be reported or who would not report euthanasia case themselves were not willing to perform euthanasia in actual practice, our findings would be less problematic. However, a majority of these physicians could conceive of performing the euthanasia case themselves.

In conclusion, the reporting procedure for euthanasia is based on the premise that ELDs can be uniformly labelled and that physicians are able to classify those decisions according to the legally defined categories. Our hypothetical case study shows that identical cases are not uniformly labelled and that there is no complete agreement about which ELDs are considered to be euthanasia and which ELDs should not be labelled as euthanasia. Physicians sometimes label intensified pain alleviation or palliative/terminal sedation as euthanasia. Those physicians are less willing to perform these acts in practice. Incorrect labelling of normal medical practice as euthanasia could thus pose a barrier to effective pain treatment. Physicians who did not perceive the euthanasia cases in the study as euthanasia were less willing to report these cases themselves than those who did. This finding has profound repercussions for the working of the current system for societal control over euthanasia. Agreement about the labelling of ELDs is thus pivotal in countries where euthanasia is legal. Furthermore, our results show that there are large regional differences in labelling and reporting of euthanasia cases, which might be remedied by information campaigns specifically targeted at physicians from Wallonia. Further research should focus on investigating how exactly physicians come to label ELDs and which factors are decisive in their labelling.

Supplementary data

Supplementary data are available at EURPUB online.

Funding

Institute for the Promotion of Innovation by Science and Technology in Flanders (Instituut voor de aanmoediging van Innovatie door Wetenschap en Technologie in Vlaanderen) (SBO IWT nr. 050158); It had no role in study design; in the collection, analysis and interpretation of the data; in the writing of the article; and in the decision to submit for publication. This study is part of the ‘Monitoring Quality of End-of-Life Care (MELC) Study’, a collaboration between the Vrije Universiteit Brussel, Ghent University, Antwerp University, the Scientific Institute of Public Health, Belgium, and VU University Medical Center Amsterdam, the Netherlands. The researchers are independent from the funders.

Conflicts of interest: None declared.

Key points

  • Half of all euthanasia cases performed in Belgium is not being reported to the Federal Control and Evaluation Committee.

  • Whether non-reporting of euthanasia cases is mainly caused by mislabelling of ELDs or is rather a problem of unwillingness to report euthanasia cases, is hitherto unknown.

  • Our study about hypothetical cases shows that there is disagreement among physicians in Belgium about the classification of euthanasia and other ELDs, and about which cases must be reported as euthanasia to the Federal Control and Evaluation Committee.

  • Among other variables, in particular the correct labelling of a euthanasia case is strongly associated with whether a physician knows a euthanasia case must be reported and whether they would report such a case themselves.

  • Mislabelling of ELDs not only impedes societal control over the euthanasia practice, but also hampers adequate end-of-life treatment.

Acknowledgements

We thank lawyer Wim De Brock for his cooperation in the data collection, and Jane Ruthven for her linguistic help. We especially thank the physicians who provided the study data.

References

View Abstract