When Terri Schiavo died on March 31^st 2005, she had been at the centre of a protracted legal struggle and a heated public debate for 8 years. At the onset was a conflict between her husband and parents about removing her feeding tube after more than a decade in a persistent vegetative state. In the course of the controversy, however, the case engaged several levels of jurisdiction, the US-Congress, President George W. Bush, the US-Supreme Court, and even the Vatican. A great variety of political and ideological groups accompanied the whole process and voiced heterogeneous, at times completely incompatible views and agendas. Civil liberties movement representatives underlined the right to an autonomous decision and a dignified death [1]; Christian advocacy groups stressed the sanctity of human life and the absolute prohibition of killing [2]; Jewish commentators referred to Halakha to emphasize the moral difference between withholding and withdrawing treatment [3].

The fierce, at times seemingly irreconcilable culture wars sparked by the Schiavo case exemplify why end-of-life (EOL) decision making constitutes a major challenge for bioethical deliberation and political governance in modern liberal democracies. On the one hand, these decisions touch upon some of our most fundamental convictions regarding life and death, the meaning of being a human being and a person, and the principles of leading a good, decent and dignified life: How can we maintain the autonomy and dignity of dying patients in an environment dominated by professional paternalism and medical bureaucracy? Is it morally permissible to stop or withhold life sustaining treatment, assist terminally ill patients in committing suicide, or even conduct euthanasia? And to what extent can and should professional principles or political regulations intervene in these deeply personal and familial matters? On the other hand, however, social research indicates that public attitudes towards these issues are also deeply rooted in cultural, especially religious and spiritual traditions, as well as in specific historical experiences, and therefore often show great variety between different socio-cultural contexts [4–7].

The increasing sensitivity to cultural differences in bioethical debates on EOL decision making appears highly desirable [8]. Although there may be isolated attempts to exclude religious viewpoints, e.g., from relevant advisory committees, they frequently play a prominent role in the contemporary expert discourse [9]. At the same time, however, it is essential to develop adequate conceptions of culture and religion in order to avoid oversimplifying notions and cultural stereotyping. Indeed, many contributions to the debate still seem to express an insufficient, overly simplified picture by drawing on essentialist conceptions of culture and religion as monolithic, homogeneous entities with unchangeable defining features [10]. It is still common to identify “culture” with national characteristics or a religious denomination, or to equate religion with an official theological doctrine and expert discourse [11]. There are numerous popular publications claiming to present “the” Jewish, Christian, or Buddhist view on EOL [12, 13]. This simplified perspective may promote a kind of socio-cultural determinism. The assumption that being “American,” “German,” or “Israeli,” as well as being “Jewish,” “Christian,” or “secular,” automatically suggests particular positions towards EOL was already critically tested in comparative studies [14]. When used as a basis for guidelines for clinical practice or ethics consultation [15], or even as an input for algorithms of EOL decision making for incapacitated patients [16], such stereotypical understandings can yield highly problematic effects. The same holds true for political players who claim to speak on behalf of certain cultural or religious communities and to advocate their alleged positions [17].

In light of the longstanding controversial political and academic debate on liberalism and cultural pluralism in bioethics, there has been comparatively little empirical research on actual cultural specificities and differences in public attitudes towards medical EOL decision making [18]. Against the backdrop of the political and legal situation and the academic discourse, this study aims to explore similarities and differences in the moral perspectives of lay persons. For this purpose, we confront the bioethical and biopolitical expert discourse with concrete public views of and attitudes towards EOL decision making in different cultural settings. This consideration of “public understandings of ethics” allows a critical assessment of the plausibility and practicability of academic ethical arguments and political rationales [19]. The aim is to develop a more fine-grained understanding of cultural identity at the intersection of different religious and cultural commitments. Moreover, including the perspectives of affected persons who have already been in situations of EOL decision making can help to understand the perceptions and priorities of the patients themselves – a fundamental requirement for all ethical approaches that call for strengthening patient autonomy [19].

Our considerations are based on a qualitative socio-ethical study in the United States (California), Germany, and Israel. We begin with an overview of the bioethical expert discourses on EOL decision making and the pertinent practical, socio-political, and legal frameworks in the three countries, pointing out similarities as well as differences. Against this background, we analyse similarities and differences in public attitudes towards EOL decision making, with a focus on advance directives (AD), termination of life-sustaining treatment and euthanasia, as well as the role of the state. This research design allows us to examine the complex interplay between the diverse influences and interpretations of culture and religion. We will discuss our findings in light of the bioethical expert discourse on EOL and draw conclusions for professional decision making.

EOL decision making has been at the centre of bioethical discussions for several decades [20]. Especially in the USA, the call for more consideration of culture and religion in bioethics has received attention in the broader context of multiculturalism. Different cultural backgrounds and religious denominations have presented their stances [21–23], revealing a great diversity of positions [8, 24, 25]. Against this background, it has been suggested that in the spirit of culturally sensitive medical care, each group should have the right to be treated according to their respective worldviews and value systems [15]. However, once we step beyond our particular cultural communities and enter the public and political sphere, we seem to lack a consensual basis for formulating generally binding legal regulations [22]. A similar situation occurs on an international level: Even modern liberal democracies offering high standards of medical technology and care for EOL show at the same time important differences with regard to the public debate and the legal regulation of EOL decision making. In the following, we give a brief overview of the most important pertinent discussions and legal regulations in the USA (California), Germany, and Israel in order to provide a background and frame of reference for the empirical analysis of lay perspectives in all three countries (see [26]).

In the US, questions of EOL form part of the culture wars between liberal pro-choice groups and primarily Christian pro-life groups. The case of Karen Ann Quinlan in 1975 was at the onset of the recent political and public controversy and proved precedential for subsequent legislation and jurisdiction. Several federal states implemented legal regulations regarding the right to die and the conditions for the termination of life-prolonging treatment and physician-assisted suicide. The state of California played a particular role in this development. The California Natural Death Act (1977) was the first of its kind in the US. It allows directives instructing physicians not to use artificial methods to extend the natural dying process. A series of court rulings increased the range of physician obligations to obtain patient consent to treatment, expanded informed consent to include the risks of rejecting recommended procedures, and determined that life-sustaining treatment may not be stopped without evidence that the patient is in a persistent vegetative state and wants termination. Since 1991, the Federal Patient Self-Determination Act has required medical and care providers to inform patients in writing about their rights to participate in decisions concerning their medical care and about the provider’s policy on ADs, to include information in patients’ records about whether they have executed ADs, and to educate their staff and community about ADs. California Probate Code Sections 4600–4678 state that withholding or withdrawing healthcare is permitted “pursuant to an advance health care directive, by a surrogate, or as otherwise provided, so as to permit the natural process of dying.” Mercy killing or euthanasia are prohibited. California Health and Safety Code Sections 442–442.7 state that healthcare providers should offer comprehensive information and counselling regarding EOL care options. Finally, California Probate Code Section 4700–4701 gives provisions for an Advanced Health Care Directive.

In Germany, questions of EOL attracted major public attention with the case of physician Julius Hackethal promoting and practicing euthanasia in the mid-1980s. The issue was frequently discussed with reference to the crimes of physicians during the Nazi era. Religious groups also played an important part in the debate, especially the Roman Catholic Church and the Protestant Churches [27]. Both hold that euthanasia should be legally banned but support pain relief and accept withholding or withdrawing medical treatment in accordance with a patient’s wishes. In 2005, the German National Ethics Council issued recommendations for the legal regulation of EOL decisions, thus preparing the ground for the legislation on ADs in Germany. The report advocated the provision of basic care for all dying patients and emphasised the right to refuse treatment [18, 28]. The law on ADs (2009) [29] allows a variety of ADs as expressions of individual autonomy [30]. Any treatment decision has to respect an existing AD, irrespective of the stage of illness. Institutions like the Federal Ministry of Health, the medical associations, civil rights organizations, patient advocacy groups, and the Churches provide different forms of ADs which are all legally recognized. Physician-assisted suicide is not criminalized by German penal law, but there is an ongoing controversy among medical associations about its evaluation in light of professional ethics and law [31, 32]. Recent legislation has banned organized and commercial forms of assisted suicide but paved the way for single-case decisions.

In Israel, there is also a separate law on ADs, the Dying Patient Act (2006). It was the result of deliberations at the Public Committee on the Care of the Dying Patient, where different religious, ethical and professional perspectives were presented. According to the Orthodox Jewish (Halakhic) tradition, the sanctity of life has more ethical weight than individual autonomy. Active euthanasia is strictly forbidden. Withdrawing treatment is categorised as an active, life-shortening intervention and therefore deemed unacceptable. Withholding treatment, however, is considered as passive and thus permissible under certain conditions [28, 33, 34]. The Dying Patient Act regulates EOL care within the Jewish Halakhic restrictions on autonomy. ADs only apply within rather restrictive bureaucratic constraints [33–36]. In general, the Israeli law has two characteristics: First, the application of ADs is limited to terminal patients in the final 6 months of life; second, only withholding treatment is permissible. The interruption of an ongoing treatment, e.g., mechanical ventilation or artificial feeding, is not accepted, even if it is the patient’s will. As a compromise, a timer was proposed – a technical device turning a continuous treatment into a discontinuous one. Israeli ADs are completed on a complicated bureaucratic form issued by the Ministry of Health. Since the commencement of the act, the Ministry has received more than 7.200 ADs. However, many of them do not meet the complex formal requirements and are returned to the senders for correction (Eti Biton, registrar, Center for Advance Directives, Ministry of Health, personal communication January 1, 2017). A new potential development in Israel concerns the legalization of physician-assisted suicide. In June 2014, the Knesset’s Ministerial Committee for Legislative Affairs passed a bill legalizing physician-assisted suicide which was rejected in the Knesset in 2016. The law would have allowed doctors to administer a lethal injection to terminally ill patients in the last 6 months of their lives.

A common point of reference in all three countries’ debates is the Holocaust and the involvement of doctors in the Nazis’ programme of involuntary “euthanasia” [37]. In the US-American debate, drawing analogies to Nazism plays an important role in politically charged EOL controversies [38]. In the German discourse, the historical experience of atrocities committed by medical professionals is used as evidence supporting slippery-slope arguments against physician-assisted suicide or euthanasia. Remarkably, the views of Israeli Holocaust survivors regarding euthanasia (which only played a minor role in the legislation) are opposed to those of German professionals [39]. They argue that there are significant differences between the Nazis’ involuntary “euthanasia” programmes and physician-assisted suicide in accordance with a patient’s will. Cultural factors such as religion and the experience of the Holocaust also influence the attitudes of healthcare professionals [40] and the role of patient-support groups [41].

In order to explore lay perceptions on EOL decision making, we conducted twelve focus groups (FGs) with religious (Jewish/Christian) and non-religious persons (personally affected and non-affected) in the United States (California), Germany, and Israel between 2010 and 2012. FGs are moderated group discussions with 6–10 participants. They are an established tool in qualitative research to explore common arguments and public topoi on a general level [42, 43]. The participants in our FGs were recruited by posters, flyers, and information sheets outlining the central aims and methods of the study and distributed in different public places. All interested persons responded voluntarily by submitting a flyer with their basic socio-demographic data and contact information. IRB approval had been obtained in advance from the Jewish Home San Francisco as well as San Francisco State University for the US groups, the University Medical Centre Göttingen for the German groups, and Ben-Gurion University for the Israeli groups.

Participants in the FGs were selected to reflect different contexts with regard to being affected and socio-economic background. “Being affected” was defined as having personal experience with EOL decision making, either due to own decisions or caring for a terminally ill friend or relative (compared to “non-affected” persons who had no such personal experience) [44]. In the US, two groups consisted of residents of the Jewish Home San Francisco and two of volunteers at the same facility. Participants in all US-groups identified as Jewish, but only roughly half of them as religious. The volunteers had primarily professional experiences with terminal care. In Germany and Israel, one non-affected group was composed of participants identifying as religious and the other of participants identifying as secular. The participants in the religious groups identified as Christian in Germany and as Jewish in Israel. Ultra-orthodox Jews were not represented in the Israeli sample. The participants in our FGs were by no means regarded as representative of the overall populations of the US, Germany, or Israel. Rather, we used the sample to explore how being religious or not interacts with other factors such as legal frameworks and historical backgrounds. For example, we were interested whether participants identifying as religious would refer to national policies or arguments of religious authorities in the respective public debates.

All FGs lasted between one and two hours. A semi-structured questionnaire with similar scenarios and questions was used. The discussions were audio recorded and transcribed. The German and Israeli FGs were translated into English to facilitate analysis. All material was pseudonymised using a number/letter code and later anonymized for publication (speakers are identified as M = male; F = female, as rel = religious, sec = secular, and as living in US = USA; GE = Germany, IL = Israel). Transcripts were coded using Atlas.ti software and analysed thematically in order to identify discursive themes and main lines of argument recurring within and across the groups [45, 46]. The focus was on moral perspectives on ADs, physician-assisted suicide, euthanasia, withholding and withdrawing life-sustaining treatment, and political regulation of EOL decision making [26]. To juxtapose lay moralities and the expert discourses, topics emerging through inductive coding were compared with the general categories of the bioethical discourse collected from analyses of public policies and expert committee statements [18, 28]. This helped to examine whether and how common ethical principles were embedded in public opinions and how they differed with regard to cultural factors in the context of EOL decision making.

The FGs had between 5 and 9 participants, with 82 participants (29 male and 53 female) in total, 23 in the US, 29 in Germany, and 30 in Israel. Participants were between 20 and over 90 years old. Their educational backgrounds differed, but the self-recruitment led to a slightly higher proportion of persons with an academic education.

A central topic of concern throughout all group discussions in all three countries was the meaning, scope, and limits of individual autonomy in EOL decision making. At the same time, however, autonomy was interpreted in different ways and different arguments were brought forward in favour of or against autonomous decision making in the different FGs, depending on varying situational aspects and socio-cultural contexts. The subsequent sections explore these similarities and differences with regard to ADs, termination of life-sustaining treatment and euthanasia, as well as the role of the state in EOL.

Our findings draw attention to the variety and complexity of cultural and religious aspects of EOL decision making. In contrast to stereotypical prejudices, e.g., about rational secularists and unthinking “religionists,” moral reasoning among those identifying as religious often proved to be rather varied and nuanced. Narrow interpretations of individual autonomy in the sense of personal self-determination were complemented with and sometimes relativised by other aspects, such as responsibility for others, natural law, or divine will. Furthermore, purely consequentialist views levelling the moral relevance between withholding and withdrawing treatment were put in perspective.

Of course, the specific qualitative design of our study and the composition of the samples have limitations that need to be taken into account when interpreting the findings. Thus, we only conducted group discussions with secular, Jewish, and Christian persons in Germany, Israel, and the US (only in California). Moreover, there was a slight bias towards higher education which might have influenced the way EOL-issues were perceived and discussed. Finally, our sample is not symmetrical as it does not include Jewish persons in Germany or Christian persons in the USA or Israel. Further, more systematic empirical research will be needed in order to draw a richer and more differentiated picture of the respective influences of and interactions between the religious commitments and the national situatedness of participants. Moreover, similar studies including additional countries and denominations could help to explore whether laypeople from other national or religious backgrounds and in different sociocultural settings would express divergent perspectives on EOL, e.g., Muslims in the US, Israel, and Germany (for an Islamic theological perspective on EOL, see [23]).

On a practical ethical level, our findings suggest that there are good reasons for being suspicious of academic positions, professional stances, and practical guidelines that seek to derive persons’ moral attitudes towards EOL mainly from their cultural or religious backgrounds. Cultural and religious perspectives seem to be far too complicated to predict individual preferences regarding limitation of care or euthanasia. As our findings show, this may not just be a pragmatic problem of dealing with complexity in the sense that we need more statistical information about additional factors potentially modifying cultural or religious convictions. Indeed, no statistical model, however complex, seems to do justice to the fundamental reflexivity of modern cultural and religious attitudes: the fact that people not just belong to certain cultural or religious communities, but frequently position themselves in relation to their respective commitments by interpreting, reviewing, modifying, or even abandoning them. In this sense, our speakers often explicitly articulate their own stances as “Christian” or “Jewish” while at the same time qualifying them with reference to particular authoritative traditions or cultural circumstances, e.g., as a Jewish minority in a Christian majority culture. Thus, rather than going through great pains to develop elaborate decision making algorithms, our findings suggest that the conditions and standards of actual decision making need to be improved. One important way of doing so would be by strengthening the reflexive and communicative practice of completing ADs [47], as newer efforts of so called Advance Healthcare Planning conceptualize this rather as process than as a fixed document [48].

On a theoretical level, these findings can provide an important clue for a contemporary bioethics that is willing to engage seriously with cultural and religious diversity. It suggests that the discipline has to be informed by adequate theoretical conceptions and empirical insights. In this sense, culturally informed bioethics should not rely on a do-it-yourself approach to culture, but systematically take into account the approaches and results of contemporary cultural theory. This can help to question essentialist and monolithic conceptions in discussions about EOL decision making. Furthermore, it facilitates the differentiation between various factors that play into “culture,” as well as the analysis of their dynamic interaction and development, e.g., effects of migration, acculturation, collective history, family story, and individual biography [49, 50]. The same holds true in view of the religious dimension: In order to develop an adequate understanding of religion and avoid stereotyping, bioethical theory needs to include the perspectives and insights of religious studies. They can help to differentiate between diverging interpretations of religious sources like the Bible or the Halakha, and to distinguish authoritative theological positions from lived religious experience and practice influencing personal attitudes, e.g., in the context of cultural Judaism [51]. Hence, cooperation between bioethics and religious scholars can be fruitful when it comes to working out a more fine grained understanding of religious factors in bioethical decision making processes [52].

The theoretical implications of this study thus encourage bioethics to develop in a direction that can be called “comparative empirical bioethics” [53]. In contrast to the classical division between deductive and inductive approaches in bioethics, this approach uses both, narrative, empirically induced findings about moral experience as well as theoretically informed categories to understand and critically reflect moral attitudes. A major attempt is to understand, analyse, and reflect on “implicit” assumptions made in everyday life morality as well as in expert statements. This methodology helps to flesh out some parts of the “whole” picture, while others still remain vague and need to be examined. Here, we focused mainly on religion, but other studies might evidently focus on the empirical implications of ethnicity, gender, patient identity, or their intersection. This is exactly the kind of inductive uncertainty which makes empirical bioethics so difficult to embrace for a deductive approach, but also a necessary challenge to test theoretical pre-assumptions.