Introduction

Over the past several decades, medical education at undergraduate and postgraduate stages has broadened its scope to include learning and teaching in the areas of professionalism, communication skills, the humanities, ethics, and law. While the inclusion of ethics and law is declared by consensus statements (Stirrat et al. 2010; A Working Group—ATEAM 2001; Royal College of Physicians and Surgeons of Canada 2005) and required by most accreditation bodies (Australian Medical Council 2012; General Medical Council 2015), the importance accorded them (Preston-Shoot and McKimm 2011) and models of teaching remain uneven across educational institutions. What is common to this education but has so far gone unremarked and unexplored is its status as an application of external disciplines to core factual/scientific/clinical knowledge and practice. The earliest developed and still the foremost field within the broad category of applied ethics is medical ethics. Medical law is sometimes integrated with medical ethics and related teaching and sometimes taught in a stand-alone fashion. But either way, where it is considered, law too has followed the model of applying legal principles to scientifically formulated clinical action plans.

While this applied model is not historically surprising, we argue that a shift to what we term the “constitutive model” is overdue. By this we mean that, in particular, established legal principles and rules should help constitute the clinical environment, in a no less normative and authoritative way than the medical/scientific aspects of the case at hand. Within medical education and practice, this shift would reflect broader challenges to the starkness of the traditional philosophical distinction between facts and values.

We support this shift in approach in three sections. (1) We briefly characterize the introduction of education in ethics and law to medical programmes and clinical practice over recent decades as following an applied model, which derives at least in part from the post-Cartesian split between facts and values, with values coming to be regarded as epistemologically and practically inferior to empirically verifiable facts. (2) We provide selected results from an Australian empirical research study concerning doctors’ knowledge of and attitudes to the law in the particular context of withholding or withdrawing life-sustaining treatment from adults who lack decision-making capacity. (3) We align these results with the applied model of ethics and law, critique the applied model, and offer a theoretical basis for the preferred, constitutive model. This new model is the normative conclusion from a process of empirical ethics in action (Parker 2009; Goldenberg 2005), with the broad historical story of the development of applied models of ethics and law together with selected results of a rigorous empirical research project informing the argument for preferring it to the established model.

It is important at the outset to be clear about the historical, conceptual, and practical aspects of the complex relation between law and ethics. We consider that both ethics and law as normative institutions have been historically sidelined in relation to medical/scientific “facts” in the hierarchy of evidence in clinical medicine and medical education. Historically, of course, greater emphasis has been laid on medical ethics as the normative element to be “applied,” as medical ethics and bioethics were the first recognizable disciplines to challenge various aspects of medical practice from the 1960s. However, we have focused on law as the primary normative element that needs to be integrated/reinstated in the evidentiary hierarchy for two related reasons. Firstly, law is the distillation at any one time of widespread community ethical considerations. We argue that it can thus function in the same kind of way as medical and scientific facts, indeed as a requirement for or prohibition of action, while ethics necessarily involves consideration of various factors prior to codification into specific rules of behaviour. Second, subcategories of ethics such as “professional ethics” or “good medical practice” can nevertheless contradict current law, and we draw attention to this phenomenon as it emerges from the empirical data. Law and ethics are certainly distinct, since laws are regularly challenged on ethical grounds and are modified accordingly. But we argue that in medical education and practice this does not justify the discretionary attitude to law often taken by clinicians revealed in our empirical work. We are suggesting then that law, having followed ethics into the applied field, should now be accorded a primary, constitutive position in the evidentiary hierarchy.

Ethics and Law in Medical Programmes and Clinical Practice

Miles and colleagues described medical ethics education in the United States as “coming of age” by the end of the 1980s (Miles et al. 1989); in 1995 Fox and colleagues claimed that medical ethics education there had a twenty-five-year history (Fox et al. 1995). While other countries initially lagged the United States, by the middle of the 1990s medical ethics had become a recognized component of medical education across Western jurisdictions, albeit with uneven prominence in the curriculum. Consensus groups formalized this recognition in published statements, and medical educators are now required by accreditation bodies to provide instruction in the principles of medical ethics, the fundamental legal responsibilities of doctors, and professional aspects of practice. The professional aspects include an emphasis on attitudes and behaviour, which is additional to the initial focus of medical ethics curricula on the knowledge and cognitive skills necessary for ethical decision-making. To the extent that these attitudinal and behavioural aspects are taken to be the responsibility of whole programmes rather than that of ethics/law courses alone, describing models of ethics teaching as “add-ons” to traditional curricula might now be reasonably challenged. But the knowledge and skill components of ethics and law continue to follow an applied ethics formula, with discrete courses usually developed and implemented beside existing core scientific elements of education programmes even where the ethical and legal aspects themselves are integrated (Preston-Shoot and McKimm 2011) and despite continuing calls for greater integration (Dowie 2011; Campbell 2012). This is mirrored in clinical practice by consultation arrangements that provide external expert advice to clinicians (Doran et al. 2015).

Part of a plausible explanation of this development lies with the story of twentieth century Anglo-American moral philosophy. The emergence of bioethics and modern medical ethics early in the second half of the century is generally agreed to have been immediately triggered by, inter alia, the exposure of unconscionable wartime research practices, technological progress that could extend life beyond acceptable quality, and increasingly negative perceptions of medical paternalism. These developments occurred in the context of the civil rights movements, as well as expanding access to education and public reaction to U.S. involvement in south-east Asian conflicts (Jonsen 1998, 34–89; Jonsen 2000, 99–109). Just as medical science and clinical practice were rapidly expanding therapeutic possibilities, strong critiques developed, for example, of the unearned social status and power of doctors and the medical profession (Illich 1976; Illich et al. 1977; Kennedy 1981) and of the aggressive but often harmful prolongation of life (Callahan 1993). But these responses occurred in an intellectual environment influenced by the strong philosophical distinction between facts and values, the historical legacy of Cartesian dualism’s separation of body and mind, and the consequent escalation of the focus on the body as amenable to scientific exploration. The explosive success of the biological and organic medical sciences over the past three centuries effectively helped quarantine many aspects of our mental life from intellectual exploration, including the experiences, preferences, and values of patients (Shelton 2013). This has meant that within medicine itself, psychiatry has generally been attributed lower status than “organic” medicine, as a result of critiques of the objective, scientific reality of psychiatric conditions (Fulford 1989, 3–24; Nickens 1984; Porter 1997, 520–524). But it has also meant that patient preferences and values have, until relatively recently, been accorded much less normative authority than that of medical expertise. Early in the twentieth century, Moore philosophically distilled this trajectory in his contention that no natural facts (or any other sort) could be the basis for ethical conclusions (Moore 1959). It reached its apogee with some of the logical positivists, who not only distinguished facts from values but equated value judgements with personal choices or approval, categories seen as distinctly inferior to those of logic and science (Mormann 2007) due to their being subjective and therefore not amenable to reasoned argument (Putnam 2002) or to empirical verification.

The philosophical dominance of this demotion of the status of values by practitioners of meta-ethics during the first half of the twentieth century meant that the resulting “impotence of normative theories discouraged forays into real problems” (Jonsen 1998, 79–110). But with real problems becoming more prominent and acute in the context of the broad social movements mentioned above, particularly in healthcare, normative ethics resurfaced to tackle them, alongside other contributions from theology, law, sociology, and kindred disciplines. However, while fundamental principles and putative solutions were advanced through new versions of broad theories such as utilitarianism and human rights, the background distinction between facts and values has proven resilient, reflecting the indelible influence of the Cartesian legacy. This has helped to maintain the continuing dominance in medical education and practice of its foundation in science, with the emerging normative considerations being applied to this factual base. Anecdotally, current teachers of medical ethics and/or law are familiar with continuing (albeit uneven) resistance to new courses and remain concerned about the vulnerability of those and existing courses within programmes.

Implementing a constitutive model for ethics and law in education and practice would reduce this insecurity by engendering the acceptance of ethical and particularly legal knowledge as necessary and normative clinical components of the delivery of patient care, no less so than the scientific facts about diseases and their management. As examples, it should be straightforward—in response to a patient’s request for a termination of pregnancy, or for the removal of life sustaining treatment, or for off-label drug treatments—that the relevant legal rules should play a combined evidentiary-dispositive role, so that the responsible clinician is in no doubt about what clinical action should or should not be taken. This requires a nuanced appreciation of the nature of ethical and legal rules and of the similarities, differences, and interaction between these two categories of rules in the clinic. This would help clinicians to understand the law, in particular, as a routine component of their clinical resources, but also have the effect of disrupting the discretion that the law has historically vouchsafed to the medical profession through deference to “good medical practice.” In the next section, this discretion is illustrated by the responses of doctors in an empirical research project that aimed to determine their attitudes to and knowledge of the law governing withholding or withdrawing life-sustaining treatment from patients who have lost decision-making capacity.

Doctors’ Knowledge of and Attitudes to Law

Recently the authors have undertaken a cross-sectional postal survey of specialist doctors from seven specialty groups active in end-of-life care in the three eastern Australian states: Queensland, New South Wales, and Victoria. Specialists in emergency, geriatric, palliative, renal and respiratory medicine, medical oncology, and intensive care constituted the final sample (n = 2702). These specialists were those listed for their categories on the AMPCo Direct databaseFootnote 1 at the time of survey distribution (July 2012). The overall response rate was only 32 per cent, although participant demographic characteristics were strongly congruent with the overall AMPCo sample, with the exception that there were fewer young respondents than in the sample. Detailed description of the methodology and the results of the survey have been reported in previous publications (White et al. 2014; Parker et al. 2015; Willmott et al. 2016a, b). Here we provide selected results from the survey that support the present argument in the following contexts: (1) law in Western jurisdictions has an increasingly important official role in decisions to withhold or withdraw such treatment; (2) nevertheless, our research confirms that clinicians are known to sometimes act in ways that they consider ethically and professionally appropriate but which conflict with that official role—including the legal right of patients to refuse treatment, either contemporaneously or in an advance health directive—and which diminish the role of law, as we have previously reported (White et al. 2016).

The survey demonstrated significant knowledge gaps about this area of law across all specialty groups but with variation across groups. Specialists in palliative and geriatric medicine had significantly higher knowledge than the all-group average, while emergency and respiratory medicine specialists had significantly lower scores. Across the whole cohort, there were significant linear associations between doctors’ perception of their legal knowledge, the number of end-of-life decisions they had made, the recency of post-graduate training on the law in this area, and their actual level of knowledge.

Specific questions concerning attitudes to the law revealed strong support for the very general proposition that law has a place in the practice of medicine (88 per cent strongly agreed or agreed) and strong disagreement for the general proposition that law is not relevant to making these end-of-life decisions (77 per cent strongly disagreed or disagreed). However, there was less support for more specific aspects of the role of law. For example, 82 per cent of respondents strongly agreed or agreed that resolving disputes through legal processes takes too long, and 50 per cent strongly agreed or agreed that following the law can lead to inappropriate treatment decisions. Slightly less than half (47 per cent) of respondents strongly agreed or agreed that the law provides a useful framework for decision-making, and only 42 per cent strongly agreed or agreed that the law supports good medical practice. Factors other than law were more important for many respondents in making decisions in this clinical context. For example, 60 per cent of respondents strongly agreed or agreed that medical and family consensus matters more than the law. As well as higher knowledge scores, palliative care physicians and geriatricians also had the most positive attitudes to the law, while intensive care specialists had the least positive attitudes. Irrespective of specialty, older doctors (>60 years) and those with continuing professional development (CPD) training in this area of law were also more positive about the role of law in end-of-life practice.

The survey allowed for open-ended comments by participants in relation to the law in the clinical context in question, how it could be improved, and its role in medical practice. Such comments were provided by 252 of the 867 respondents. A coding process to determine whether a respondent’s comments revealed a positive, negative, or neutral attitude to the law in this area indicated that 113 respondents had a positive attitude towards the law, while 85 respondents had a negative attitude (Willmott et al. 2016a, b). (Fifty-four responses either did not relate to attitudes about the law or combined positive and negative comments, preventing the determination of overall attitude.) A number of participants urged better education for doctors about the law, in more accessible formats, and reiterated the need for doctors to know the law. These comments are consistent with the general support for a role for law indicated above but do not distinguish between patient-centred motivations for acting lawfully and more prudential motivations to avoid legal sanctions from acting unlawfully. Some of the negative open-ended comments are particularly pertinent to the argument of this paper. They demonstrate a desire for clear boundaries between the practice of medicine and legal considerations and a belief that law is either irrelevant in medical decision-making or, at best, a distant second to clinical considerations (see Box 1).

Box 1
figure 1

Participant comments indicating negative attitudes to law in clinical decision-making

Full size image

Discussion: From Applied to Constitutive Law and Ethics

We distil the following broad observations from the selected quantitative results and qualitative comments:

Despite general support for a place for law in medicine, a significant number of doctors regard law as being of limited utility in supporting good medical practice and at times as contributing to poor clinical decisions. This negativity is reduced to some extent by familiarity with the law, greater clinical involvement, and recent continuing medical education in this area of law.

The legal perspective is seen as one amongst many and a proportion of clinicians think that the law should be downgraded or ignored if the pursuit of the primary goal of patient welfare (as clinicians perceive it) requires this. Accordingly, they hold that following the law should be seen by doctors as discretionary; for some it is simply an unnecessary aspect of clinical decision-making.

In summary, a proportion of clinicians perceive “good medical practice” as central and constant, in contrast to law, which is discretionary, over-emphasized, and/or irrelevant.

These observations reflect the applied model within which ethics and law have been included in medical education and clinical practice; that is, they are secondary considerations that are external to the core practice of medicine. Not surprisingly, engagement with something that is applied to a traditional, revered core of scientific knowledge and expertise, particularly when the application process is perceived as instigated by external agencies, may well be seen as a matter of choice or discretion. Legal and ethical experts remain predominantly non-medical experts, who may or may not be called on for advice, and law and ethics courses may be amongst those most at risk when resource pressures squeeze education programmes. Even where they seem to be accepted and relatively safe, they are usually separate courses, and they are often taught by those from non-healthcare backgrounds, reflecting the applied model. These features instantiate not just a persisting separation of core scientific and medical facts from legal and ethical values but also a clearly adjunctive status for the latter.

One way in which it may appear that this separation is being broken down is through the development of medical–legal partnerships (MLPs), through which lawyers are incorporated as part of healthcare teams. However, the U.S. National Center for Medical Legal Partnership characterizes the need to embed “lawyers and paralegals alongside health care teams in clinics” as based in:

… how law functions as an important social determinant of health, and how lawyers can effectively collaborate with clinicians, case workers, patient navigators, and other members of the health care team to both prevent and remedy the many health-harming factors that have their roots in legal problems. (National Center for Medical Legal Partnership 2017)

Apart from the emphasis here on health-harming social conditions, including the law itself, which sometimes call for distinct legal support for patients, this collaborative arrangement currently remains closer to the applied model than the constitutive model we have in mind. However, by demonstrating how civil legal needs sometimes profoundly affect health and encouraging practices like prescribing legal solutions to some health problems it does point to a closer integration of law with medicine and would encourage a greater awareness and acceptance of the clinical relevance of law. We therefore see MLPs as potential vehicles for an increasing acceptance of the constitutive model.

Our central claim is that the continuing separation of medical “facts” from legal and ethical “values” can no longer bear the conceptual or practical weight expected of it, particularly within medicine. The entanglement of facts and values has long been recognized, especially within pragmatist accounts of ethics (Putnam 2002, 28–45), but medicine has been slow to recognize that it is just this entanglement that it so prominently exemplifies in practice. The continuing insistence on scientific fact as medicine’s primary normative authority ignores the pragmatists’ insight that, contrary to the positivist view that ethics (and hence law as a socially endorsed and required distillation of ethical rules) can have nothing to do with matters of fact, ethical judgements are deeply dependent on reasoned argument concerning all kinds of facts. Conversely, the insistence on the “factual” nature of science—an insistence that excludes other sources of meaning and therefore authority—ignores the pragmatist exposure of science as clearly dependent on evaluative markers such as reasonableness, coherence, plausibility, and theory choice (Putnam 2002, 142–143).

The crafting of specific ethical principles and of laws includes reference to facts of different kinds. Dispute about these social, biological, and other facts can effect changes in the specific ethical principle or law, and medical science clearly contributes to the crafting of such principles and laws. These observations imply that while on the one hand so-called “good medical practice” is an evaluative matter, on the other hand ethical principles and laws are in part constituted by different kinds of facts. For clinicians to cleave to “good medical practice” or “appropriate clinical decision-making” in determining what should occur in the belief that these simply reflect scientific knowledge and expertise, while at the same time demoting or ignoring law, fails to recognize the evaluative nature of such scientific “facts,” and that ethics and law also embody and integrate both facts and values as we traditionally distinguish them.

What this analysis points to is that laws crafted to govern a particular area should inform clinical decision-making in as fundamental a way as the medical facts of the case at hand. Law should not follow the medical facts or be applied to the patient’s case, let alone be ignored, even when it seems intuitively right to do so. Nor should law and ethics be applied to an accepted core of medical education. Law and ethics must help constitute the core, together with scientific and clinical knowledge and expertise. That being the case, not only should doctors know the law and comply with it, relevant law should be taught as part of medical knowledge and practice. Doctors utilizing the law as “normative evidence” in their field would be experts to the same extent as they are experts in the traditional medical knowledge of that field, which is just another kind of normative evidence. They may need to appeal to colleagues and other sources of expert legal knowledge in challenging cases, but from such a constitutive base this would be no different from appealing to other experts or consultants for support when needed in the more difficult scientific aspects.

It is helpful here to invoke Foster and Miola’s taxonomy of normative decision-making, including clinical decision-making (Foster and Miola 2015). These authors describe legal, ethical, and moral decision types. In legal decisions affecting medical practice, where the law has intervened and explicitly mandates or prohibits action, they state simply that doctors “have no choice” (Foster and Miola 2015, 507). Their “ethical” decision category refers to those where the law defers to professional ethics or so-called “good medical practice” (Foster and Miola 2015, 507). Moral decisions are stipulated as those that are ungoverned by anything except individual conscience, such as “conscientious objection” decisions, where the rights of the doctor are in view rather than those of the patient, unless patient interests are at risk.

The results of our survey illustrate Foster and Miola’s claim that the most problematic boundary between their three categories is between decisions of the first two types. This is not surprising, since traditionally the law has often deferred to professional medical ethics, while over the last three decades, case law and legislation in the medical and healthcare field have markedly increased, hence expanding law’s purported influence over medical practice. We will remain in a transitional period, en route to a fully accepted constitutive model as long as some doctors perceive the law as just one factor that might be considered (or in some cases might be ignored) while others see the prescriptions of explicit legislation and common law judgements as removing the discretion they were granted and exercised in the past.

We agree with Foster and Miola that the law should determine action in various ethically complex areas including withdrawing and withholding treatment from incapacitated adults as it has via legislation in the Australian jurisdictions included in our survey because law— unlike professional ethics—has in place structures and procedures for examination and adjudication of ethical questions, and because doctors have no unique claim to competence in ethically complex decisions (Foster and Miola 2015, 523–526). While doctors may see the law as thereby reducing their professional independence by weakening their therapeutic discretion, one of the consolations of the constitutive model is that it remains the clinician who would be integrating all the traditionally disparate elements—medical, ethical, legal—in decision-making. If health is best defined in terms of the mixed contribution of social and biophysical factors, medicine and healthcare should incorporate social as well as biophysical “normative facts” in practice and this includes the law as an institutional or social fact (Searle 1995). This integration should have the effect of reducing the resentment that many clinicians currently feel towards the law as an incursion on their territory, once the law is better perceived as part and parcel of the medical armamentarium. This constitutive conception should be engendered from the earliest stages of medical education via fully integrated teaching, encouraging clinicians to view law as a social tool for advancing health (Campbell 2012, 306)— the same way as MLPs view it.

There is a possible “So what?” objection to our argument along the following lines.

What practical difference would it make to conceptualize law as “applied” to, rather than as constituting, core clinical considerations? As applied to the scientific “facts,” law can still be regarded as binding in particular situations—the law is the law. And as for helping to constitute a clinical core, it would be one of a number of considerations to be weighed and might be overridden by other factors.

We suggest that this theoretical symmetry is misleading. The fact that there will be situations where doctors “have no choice” does not mean that the relevant law cannot function as a constitutive part of clinical decision-making; it is more that the incorporation of relevant law into a clinical core does not thereby reduce law’s social status and normative force. However, encouraging doctors to “own” relevant law as part of the factual-evaluative diagnostic and therapeutic evidence to hand should—in time—reduce their unjustified (albeit often unwitting) claim to global competence in complex decisions and the resentment that more overt challenges to this competence produces.

We also do not intend to convey that the constitutive model has not already begun to be employed, at least in individual cases. While we have focused on the more negative attitudes and practices of clinicians in relation to the role of law, a number of clinicians now utilize the law in a routine, constitutive way in areas such as end of life decision-making and consent for treatment by minors. However, as we have indicated via the survey questions and participant responses, this is by no means a systematic, whole-of-profession participation. While our emphasis here has been on the negative responses and comments of participants, the changes we advocate would reflect the attitudes of that group of participants whose comments indicated a positive attitude towards the law.

Another challenge might reside in the fact that law changes over time, and in the contextual nature of law. The temporal and contextual nuances concerning the law that governs different areas of medical practice including the end of life certainly present challenges. But we see no principled difference here between how law is problematized by time and context and how scientific medical knowledge is always provisional, uncertain, and revisable. Just as there will be a current “best evidence” consensus about a medical treatment for a particular condition, there will be a current law that needs to be incorporated into the evidence base. Temporal and contextual changes and qualifications do not extinguish the ever-present need, particularly in medical practice, for action, now. In a case involving the withdrawal of life-sustaining treatment from a patient with severe disease, for example, the current legal rules in the particular jurisdiction, from either legislation or case law or both, should be “front and centre” in the clinician’s thinking. It should, moreover, be just as dispositive for action as the undisputed place for oxygen therapy in a case of suspected heart attack.

Conclusion

In this paper we have argued that the model of applying ethics and law to medical education and medical and healthcare practice is outdated and should be replaced by a constitutive model. This recognizes that values permeate experience, with medicine and healthcare providing a clear example of this. Law should function in no less an authoritative, normative way than medical scientific facts and clinical guidelines, and in many instances where the law is clear, law may be called on to overrule professional medical ethical views, since it is the distillation of current community wisdom on matters about which the medical profession can claim no superior insight. In these instances, this would reverse the priority of professional ethics over law that we suggested underpinned some specific results in our empirical research and that was well illustrated by the participants’ comments included above.

We have provided an argument supporting the implementation of the constitutive model in practice but have not considered the details of how the practice project would best proceed. Implementation would require a number of developments that we can only briefly outline here. Education accreditation bodies would need to agree with and adopt the new approach which we have advocated and that we hope will be further developed. The integration in teaching ethics and law that already occurs in some undergraduate institutions would need to be extended to embrace all teaching across whole programmes in medicine and healthcare—not a small undertaking. Postgraduate education would require considerable development in order to reinforce the idea that clinicians should be medical experts in the constitutive sense, where law ranks in status with scientific knowledge as the raw material for decision-making and in a number of cases will be the primary determinant of action. Overcoming such practical hurdles is never straightforward. We hope our argument provides a starting point for further deliberation.