Experiences, Personal Attitudes, and Professional Stances of Swiss Health Care Professionals Toward Voluntary Stopping of Eating and Drinking to Hasten Death: ACross-Sectional Study

Context. Voluntary stopping of eating and drinking (VSED) is the self-determined decision of an individual with the decision-making capacity to cause premature death. During the course of VSED, the person is dependent on the support of relatives and health professionals. To date, little is known of the attitudes of Swiss health professionals on this topic. Objectives. The objective of this study was to assess the experiences, personal attitudes, and professional stances of Swiss health care professionals toward VSED. Methods. We conducted a nationwide cross-sectional survey by questioning family physicians and the heads of outpatient care and long-term care (e.g., nursing directors, institute directors, or head nurses) about VSED (n 1⁄4 1681; response rate 40.1%). Descriptive data analysis and hypothesis testing (occupational group, age, sex, professional years, VSED experience, and regions) were subsequently conducted. Results. Individuals who are willing to die are granted the right to professional accompaniment during VSED (agreement 97.8%), and their death is usually classified as a natural form of dying (63.5%) and only rarely (5.4%) as suicide. Family physicians have significantly more moral concerns during accompaniment compared with the heads of outpatient and longterm care (P < 0.001). Conclusion. Swiss health care professionals support the autonomy and self-determination of patients, which is also reflected in their positive attitude toward VSED, even if they have moral reservations when accompanying patients. J Pain Symptom Manage 2021;61:270e278. 2020 The Authors. Published by Elsevier Inc. on behalf of American Academy of Hospice and Palliative Medicine. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).


Introduction
In addition to satisfying basic physiological needs, eating and drinking have social significance. 1,2 Eating together is described as an important part of the day that is conducted in familiar communities and is described as valuable. 2 In this respect, there is a direct effect on the social environment and health care professionals when an individual refuses to eat and drink. 3e5 For various reasons, patients forego eating and drinking. This is often observed in the dying process. 6,7 Furthermore, physical illness, 8e13 mental diseases, 14e16 or disorders involving organ structures (e.g., oral cavity, gastrointestinal tract) can be responsible for reduced or complete cessation of eating and drinking. 17 In addition to these reasons, there is another form of food and fluid refusal: 18,19 voluntary stopping of eating and drinking (VSED). VSED is characterized by the choice to refuse food and liquid; although the person can still eat and drink, he or she refrains from doing so with the intention of dying. 20,21 Therefore, this decision is a self-determined and autonomous one by an individual capable of judgment who decides to end her or his life prematurely regardless of an existing basic illness. 22e25 Most people who choose this path have underlying diseases that favor the decision of VSED. 26e29 Other than possible ailments of old age, approximately onequarter of these people have no underlying diseases or disability. 26e28,30 All these people are united in their decision to opt for VSED to end unbearable suffering. The underlying symptoms or emotional conditions are pain, suffering without any prospect of improvement, fatigue, fear of dependence or a feeling of fullness of life, and many others. 26,28,30e32 The decision to engage in the VSED is made solely by the person who wishes to die. During VSED, the individual is dependent on the support of relatives and health professionals. 33e35 Health professionals in particular have a great responsibility in this regard; they take care of symptom management and accompany the relatives during the dying process. 26,35e38 Regardless of whether the person wishing to die suffers from physical symptoms at the beginning of the VSED, these can develop anew, progress, or be reduced if possible. During VSED, individuals willing to die mainly suffer from pain, thirst, weakness, delirium, and somnolence. 39 According to the literature, most people realize their wish to die through VSED at home or in long-term care institutions. 26e28 In these settings, people are mainly cared for by family physicians and outpatient or long-term care nurses. However, it is highly controversial how VSED should be classified, which has led legal and ethical disputes, in which VSED is discussed as (physician-assisted) suicide, 40e42 passive euthanasia, 43,44 or natural form of dying. 33,45e47 Regardless of the pending classification, physician-assisted suicide and passive euthanasia are legal in Switzerland, as they are in five states in the U.S. (Oregon, Washington, Montana, Vermont, and California), Canada,Columbia,Luxembourg,Belgium,and The Netherlands. 48 In contrast to the U.S. states, the law in Switzerland neither sets an age limit for the person wishing to die nor requires a life-limiting illness. 48 In this respect, the legal situation in Switzerland does not argue against accompanying an individual during VSED and emphasizes that no health professional is obliged to accompany an individual during VSED. 49 According to a review of literature, there are no recommendations for health professionals in Switzerland on how to manage people who choose this route, and little is known of the attitudes of this target group toward this topic. Therefore, the aim of this study was to explore the experiences, personal attitudes, and professional stances of Swiss health professionals toward VSED.

Study Design
This was a cross-sectional study on VSED among Swiss health care professionals.

Sample and Setting
The main health care professionals involved in outpatient and long-term care were included, including the heads of outpatient and long-term care (e.g., nursing directors, institute directors, or head nurses) and family physicians (with titles in general internal medicine or further training to become a practical doctor). All three professions are well organized through professional organizations. We sent a staggered invitation between January 2017 and July 2018 to the following professionals: About 1616 members of outpatient care services, including 426 members of public outpatient care organized by Spitex (https://www.spitex. ch), 175 members of private outpatient care organized by Associations Spitex priv ee Suisse (https://www.spitexprivee.swiss/fr/), and 1015 members of freelance outpatient care organized by CURACASA (https://www.curacasa.ch). About 1562 members of long-term care organized by CURAVIVA (https://www.curaviva.ch/). About 1411 family physicians organized by M edecins de famille et de l'efance Suisse (https://www.medecinsdefamille.ch/qui-sommesnous/lassociation).
We excluded all participants who did not fulfill the inclusion criteria, including participants with other functions, such as nurses, care assistants, or physicians from other disciplines, as well as deceased or retired people. The participants were invited to participate in the survey by professional organizations by electronic mail or post, and each potential participant received two reminders.
A response rate of 20% was targeted throughout Switzerland. Because of cultural and linguistic differences within Switzerland, where the national languages of German (mostly spoken in Northwestern Switzerland, Zurich, Eastern Switzerland, and Central Switzerland), French (mostly spoken in the Lake Geneva region and Espace Mittelland), and Italian (mostly spoken in Ticino) are spoken, a response rate of 20% per region was also targeted. There are seven regions in Switzerland that comprise one or more cantons with an average population of 1,041,144 (Appendix Fig. 1). 50

Data Collection
Question and answer scales of the standardized questionnaire are presented in Table 1. The development of the questionnaire has been described elsewhere 51 and includes questions on experiences, personal attitudes, and professional attitudes regarding VSED. The questionnaire was created online and based on the survey software Questback (EFS 10.9; Questback GmbH, K€ oln, Germany). On the front page, the project and its objectives were described. Participation was voluntary and anonymous, and the participants had to actively provide their consent to the study. Because of the low response rate among family physicians, which is described in more detail elsewhere, 52 we created a paper-andpencil questionnaire. The answered questionnaires were scanned and then read in, edited, and exported as SPSS files by using the EVASYS software (Electric Paper [Schweiz], Lachen, Switzerland).

Data Analysis
Descriptive data analysis was conducted by using the statistical software IBM SPSS statistics (Version 26, Statistical Package for Social Science; IBM Deutschland GmbH, Ehingen, Germany). To describe the participantsdthe personal attitudes and professional stances of health professionalsdappropriate statistical methods were used, such as the means, SDs, percentages, and frequencies.
Subsequently, tests were conducted to determine whether there was a connection between the participants' response behavior and their characteristics. The Kruskal-Wallis test and then the Dunn-Bonferroni test were used for the professional group (1 ¼ family physician, 2 ¼ long-term care, and 3 ¼ outpatient care) and the seven regions of Switzerland (1 ¼ Lake Geneva region, 2 ¼ Espace Mittelland, 3 ¼ Northwestern Switzerland, 4 ¼ Zurich, 5 ¼ Eastern Switzerland, 6 ¼ Central Switzerland, and 7 ¼ Ticino; Appendix Fig. 1). The group comparisons between the seven regions are made because in addition to the aforementioned linguistic differences, 53 there are cultural differences, which are also reflected in the number of physician-assisted suicides: Ticino has the lowest number of assisted suicides (0.26% of all deaths), and the Zurich region has the highest number of assisted suicides (1.42% of all deaths; Table 2). 54 Therefore, it is of interest whether there is a connection between the number of physician-assisted suicides and the VSED in the region. The number of physician-assisted suicides from the five U.S. states can be compared with the Italianspeaking region of Ticino. 48 The Mann-Whitney U test was applied to sex (1 ¼ female; 2 ¼ male), age (1 ¼ younger than 50 years; 2 ¼ 50 years and older), work experience (1 ¼ less than 20 years; 2 ¼ 20 years and more), and VSED experience (1 ¼ no; 2 ¼ yes). For the nominal item How would you classify VSED? (1 ¼ suicide, 2 ¼ passive euthanasia, 3 ¼ natural dying, and 4 ¼ others), a Chi-squared test was conducted based on all the aforementioned participants' characteristics. The analyses were performed with a 95% CI. The correlation coefficient (CC) r was calculated to evaluate the effect strength. Significance was set at a ¼ 0.05. Missing values were coded as such and hence automatically excluded from analysis. The number of missing values in the analysis was indicated by the number of values included.

Validity, Reliability, and Rigor
The questionnaire was checked by the standard pretest 55 and content validity index. 56,57 A definition of VSED was inserted at the beginning of the questionnaire to avoid misunderstandings.

Description of the Participants
Of the 4589 participants invited to the study, 4191 were eligible to participate (Fig. 1), of which 1681 (40.1%) participated. The response rate exceeded 20% in all seven regions (Appendix Fig. 1). The average processing time for answering the online questionnaire was 23 minutes; we assumed that the paper survey used the same amount of time. The average age of the participants was 54.6 years (SD 9.5; range 25e87). Just more than half (52.9%) were females, 47% were males, and two participants (0.1%) stated a diverse gender. Most participants had either already accompanied an individual during VSED (n ¼ 728; 43.3%) or were at least familiar with the topic (n ¼ 694; 41.3%), and VSED was unknown by 15.4% (n ¼ 259). Fig. 2 shows the classification of VSED by the participants. The category something else included the right to self-determination/autonomy (4.2%), an expression of life fatigue (0.6%), it depends on the case (0.4%), and a dying individual is abandoned (0.1%). There was a significant correlation between the classification  Table 1).

Personal Attitudes
Participants with VSED experience (mean ¼ 1.85) were significantly more likely to choose VSED as an option for themselves than those without (mean ¼ 1.67; U ¼ 276561.000; P < 0.001; r ¼ 0.20), and participants from the Lake Geneva region were less likely to do so than those in Espace Mittelland and the other German-speaking regions (c 2 [6] ¼ 15.964; P ¼ 0.014) ( Table 1; Appendix Table 1). In addition, VSED was significantly more compatible with the worldview of participants with VSED experience (mean ¼ 4.64) than for those without (mean ¼ 4.39; U ¼ 296060.000; P < 0.001; r ¼ 0.10). In the Lake Geneva region and Ticino, VSED was significantly less compatible with the participants' worldview (c 2 [6] ¼ 64.164; P < 0.001) and involved more moral doubts during the accompaniment of an individual undertaking VSED (c 2 [6] ¼ 77.003; P < 0.001) than in the other regions. In addition, family physicians show significantly more moral concerns during accompaniment, and this contradicted their professional ethics more than those of professionals in outpatient care (z ¼ 5.501; P < 0.001; r ¼ 0.2 and z ¼ 6.956; P < 0.001; r ¼ 0.2) and long-term care (z ¼ 6.438; P < 0.001; r ¼ 0.2 and z ¼ 9.994; P < 0.001; r ¼ 0.3). VSED was more likely to be interpreted by participants with VSED experience (mean ¼ 4.26) as a dignified death (mean [without VSED experience] ¼ 3.87; U ¼ 266425.000; P < 0.001; r ¼ 0.20) and is less likely to be interpreted as such by participants from the Lake Geneva region (c 2 [6] ¼ 45.058; P < 0.001).

Professional Stances
Most (90.5%) of the participants granted individuals the right to medical and nursing care during the VSED process, and the willingness to do so in the Lake Geneva region and Ticino was significantly lower than in the other regions (c 2 [6] ¼ 27.589; P < 0.001; Table 1; Appendix Table 1). Determining the judgment of people with VSED desires was significantly more important for participants without VSED experience (mean ¼ 4.30) than for those with experience (mean ¼ 4.14; U ¼ 307953.000; P ¼ 0.001; r ¼ 0.10) and was more important to Italian-speaking and French-speaking regions than to German-speaking regions (c 2 [6] ¼ 25.950; P < 0.001). The VSED decision of an individual was more likely to be accepted (mean ¼ 4.84) and respected (mean ¼ 4.88) by participants with experience than by those without experience (mean ¼ 4.64; U ¼ 297149.000; P < 0.001; r ¼ 0.20 and mean ¼ 4.79; U ¼ 316786.500; P < 0.001; r ¼ 0.10). Participants with VSED experience (mean ¼ 1.57) were more likely to recommend VSED as an option than those without experience were (mean ¼ 1.41; U ¼ 273852.000; P < 0.001; r ¼ 0.20), but that was less so in the Lake Geneva region than in other regions (c 2 [6] ¼ 59.084; P < 0.001). The willingness to care for an individual during VSED was very high overall (92.9%), with a significantly higher approval rate from women

Discussion
In this large national cross-sectional study of 1681 participating health care professionals, we examined their experiences, personal attitudes, and professional stances regarding the VSED. The main findings of this study can be summarized as follows. The views of health care professionals were very much in line with the VSED option. The classification of the VSED was comparable to the current international discussion, in which there were also supporters of calling VSED a natural death, because no active intervention by health professionals is required. 33,45e47 Another option is to regard VSED as passive euthanasia 43,44,58 because the accompanying persons do not take an active part but accept death by refraining from lifesustaining measures. Others posit that VSED could be compared with physician-assisted suicide 41 because there are similarities regarding the intention to die, and others called it suicide, the health care professionals accompanying the individual do not have an active part in supporting the person in dying, but in accompanying them. 40,42,59 If an individual chooses VSED, most Swiss health professionals can accept this decision and posit that it justifies an individual's right to medical and nursing care during this period, 60 although just more than one-quarter of the participants have moral concerns about accompanying an individual during VSED. An explanation for this finding may be that health care professionals feel morally obliged to care for the individual during VSED so that the individual does not suffer unnecessarily. 16 We want to emphasize that health care professionals should not be forced to accompany an individual during VSED but should do because of their convictions, comparable to abortions without medical indication. 61 Moral concerns may also depend on the health care professional's acceptance of the desire to die, which is usually easier when unbearable suffering is at the forefront and more difficult when people who want to die without physical suffering are more likely to be satisfied with life. 43,62 This study is comparable to international studies in terms of the proportional distribution of participants who have already accompanied an individual during VSED. 26,30,43,63 Dying, which was described by the participants as dignified, was described in other studies as peaceful and with little suffering. 26,28,30,64 A notable finding from the analyses was that the factors of professional group, age, sex, and work experience showed few significant differences, and those differences had a very low effect strength. Furthermore, it was striking that the response behavior of participants with VSED experience differed significantly in almost all aspects from that of participants without VSED experience, resulting in an even more positive attitude toward VSED. This finding suggests that the participants only accompanied those individuals willing to die during VSED in which they agreed with the decision. It is not clear how many patients were not accompanied. This aspect, and the reasons for the rejection of patients, should be included in further research. Based on this, our conclusion is that no major resistance to this option of accelerated dying by Swiss health professionals is to be expected. What should also be considered is that linguistically and culturally different attitudes have been identified within Switzerland. In the French-speaking region of Lake Geneva and the Italian region of Ticino, there were more moral concerns about VSED, which contradicted the attitude of the institutions. No correlation could be drawn between the number of physicianassisted suicides in the regions and VSED. There were only parallels in Ticino and Central Switzerland, which showed few physician-assisted suicides and a lower acceptance of VSED. The number of physician-assisted suicides in the Lake Geneva region was high, and the acceptance of the VSED was lower than that. 54 Because we did not ask the denomination of the participants, we compared the data with the religious affiliation from the census. 65 No similarities were identified between the participants' response behavior and the two main denominations in Switzerland: Protestant and Roman Catholic. In this respect, it is possible that neighboring countries have linguistic and cultural similarities. Although there have been no studies on VSED in France or Italy, there is one study from Germany with physicians, 43 which suggested that German physicians agree that individuals have a right to medical and nursing care during VSED. Except for these, no other comparable results were identified. Therefore, what remains unclear is whether there are culturally similar attitudes toward VSED in addition to language. The results suggest that qualitative interviews with health care professionals in French-speaking and Italian-speaking Switzerland are necessary to better assess the background of the different views.
The strength of this study is the high response rate of 40% in a national survey 66 conducted in three languages and with three professional groups. This rate enabled us to directly compare physicians and nurses for the first time. Surprisingly, this study found more similarities than differences. The limitation of the study was that the response rate between the professions differed greatly. The required response rate was achieved by all three groups, with family physicians achieving the highest response rate of 74%. This finding may be due to the changed recruitment strategy of family physicians, which changed to a paper survey after the failure of the online survey. Perhaps we could have increased the response rate in outpatient care and long-term care by switching to a paper survey as well.
Our results have important implications because knowledge of the positive personal attitude of the professional groups mainly confronted with VSED cases can contribute to political measures. In 2018, the Swiss Academy of Medical Sciences included the topic of VSED in its guideline 49 ''Management of dying and death'' as a controversially discussed action, but there was a lack of practical recommendations. Such recommendations could help clarify uncertainties about whether an individual should be accompanied while they are engaging in VSED.

Disclosures and Acknowledgments
The authors express their thanks to the participants of the study, without whose involvement the study would not have been possible. They are also very grateful to the professional organizations that helped to provide access to the participants. They express their final thanks to Prof. Dr. Wilfried Schnepp (deceasedy), who, from the very beginning of this study, with great support, critical comments, and a great deal of commitment, helped the study to become what it has become. This study was sponsored by Research in Palliative Care of the Swiss Academy of Medical Sciences, supported by Stanley Thomas Johnson Foundation and the Gottfried and Julia Bangerter-Rhyner-Foundation.
The sponsor played no role in the study design; data collection, analysis, or interpretation; or writing or publication of this report.
Deidentified study data may be available on reasonable request and after relevant ethical review board approval. For further information, contact one of the study's primary investigators: sabrina.staengle@ zhaw.ch.
The authors declare no conflicts of interest. Ethical approval: All procedures performed in this study involving human participants were in accordance with the ethical standards and reviewed and approved by the responsible institutional review board of the Greater Region of Eastern Switzerland (EKOS 17/083).