How best to regulate voluntary assisted dying: a qualitative study of perceptions of Australian doctors and regulators

Abstract

It is widely accepted that voluntary assisted dying (VAD) should be regulated but little is known about the most effective way to regulate doctors in this setting. This article reports on empirical research conducted in two Australian states where VAD is lawful (Victoria and Western Australia). Interviews were conducted with 92 participants: one group comprised doctors providing VAD and the other group was regulators in this field. Participants were asked about how best to regulate doctors providing this service. Strikingly, both regulator and doctor participant groups were consistent with each other in their views on what constituted effective regulation. The nature of VAD was perceived by participants to require special regulation, although some felt this was overdone in these states. Reported features of effective regulation included regulators taking an educative approach, regulation being perceived as acceptable by doctors, and it being responsive and nimble to provide the guidance that doctors need. Participants also considered a range of regulatory tools were required to regulate VAD effectively, and some identified a need for these tools to be employed together in a holistic way. This article concludes with a set of principles for effective regulation of VAD, discerned from the views of participants.

I. INTRODUCTION

Effective regulation design is a long-standing challenge across all fields of human endeavour. Challenges include ensuring the regulatory regime achieves its stated policy goals and the standards for behaviour it sets out are understood and applied in a way that reflects such goals.¹ Regulating the field of medicine itself creates additional challenges. Guiding behaviour in a field with clinical complexity and developing generalized rules or norms where individual situations are highly variable is difficult. Compounding this is the fact that doctors have traditionally been recognized as a difficult cohort to regulate. While society has shifted away from the culture of medical dominance characterized by paternalism and self-regulation, some doctors remain reluctant to accept direction that could impinge on professional autonomy or conflict with professional norms.² Indeed, many view regulation as an ‘intrusive and inefficient interference by external authorities that distracts from the importance of clinical care’.³

In this article, we focus on Voluntary Assisted Dying (VAD),⁴ an area of health care where regulation has been extensively considered and debated. This is not surprising given the gravity of the decisions involved. VAD involves intentionally taking steps to end a person’s life, and so it is argued that social control of the practice through regulation is needed.⁵ While not all jurisdictions globally have VAD laws, there is an increasing number of jurisdictions that are passing such laws.⁶ In recent times, VAD laws have been passed in jurisdictions including Spain, New Zealand, and Australia,⁷ and jurisdictions such as Scotland, Ireland, and England and Wales, are actively considering laws.⁸ Conversely, other jurisdictions have a long-standing history of the practice such as the Netherlands, Belgium, and Oregon.

A feature of all VAD laws, and associated debates, is the critical role of regulation. All jurisdictions where VAD is lawful have established systems of regulation to govern VAD. Likewise, all jurisdictions contemplating reform have outlined models of VAD that include specific consideration of its regulation. In most jurisdictions, this regulation includes a statute that makes VAD lawful and outlines how the system will operate.⁹ The legislative models internationally often operate within a medicalized framework,¹⁰ although individual doctors can choose not to participate.¹¹ Such medicalized models require at least two doctors to assess whether a patient meets defined eligibility criteria. Those criteria can include unbearable suffering that cannot be relieved (no terminal illness requirement); being expected to die within a specific time period, such as six months (and not requiring suffering); or requiring both a terminal illness and suffering.¹² VAD systems also generally require the completion of forms or paperwork that attest to the person’s eligibility, which is submitted to an oversight body. While this varies across jurisdictions, some form of oversight has generally been deemed necessary to provide public confidence in the system.¹³ The regulation also governs what type of assistance to die is lawful. This could be via self-administration (patient receives the medication and takes it themselves) or practitioner administration (medication is administered to a patient by a health practitioner).

While legal reform is generally needed to permit VAD (which could otherwise constitute crimes such as murder, manslaughter, or assisting a suicide), regulation in this space goes beyond law and includes a range of other instruments, such as policies, guidelines, ethical codes, training, and funding (including resourcing to enable health services to provide VAD as well as remuneration for individual VAD providers).¹⁴ In recognition of this, it has elsewhere been argued that regulation of end-of-life care requires a ‘holistic approach’ because it is not possible to understand the regulatory impact of single instruments in isolation.¹⁵ For example, while law may permit VAD, there may be ethical guidelines or policies that play a prohibitive role such that, in practice, access to VAD is limited or even not possible in particular circumstances or settings. Other regulatory instruments could also play a facilitative role in law. For example, where a VAD law simply grants permission for VAD to occur in certain circumstances but says little about how the provision of VAD should be operationalized in practice, regulatory guidance about these matters may need to come from local guidelines or training.

This article takes a holistic approach to examining VAD regulation and considers the question of how best to regulate doctors. Very little attention has been given to the most effective way to regulate doctors in VAD systems, which is surprising given their central role.¹⁶ This article reports on the findings of an empirical study of Australian VAD regulation, where VAD is permitted through legislation in all but one jurisdiction, each of which has a slightly different law.¹⁷ This research focuses on the two states that have the longest experience with VAD in Australia, namely Victoria and Western Australia. These jurisdictions’ laws commenced operation in 2019 and 2021, respectively. This study, the first of its kind in Australia, draws on semi-structured interviews with both the creators of regulation (‘regulators’) and their primary regulatory target (doctors providing VAD), exploring how both groups conceptualized optimal regulation¹⁸ of VAD.

Finally, while we recognize that there are different views on whether VAD should be permitted, this article steps past these debates, recognizing that VAD is now lawful in these two states. The focus of this research is on reporting and analysing participants’ perspectives on how best to regulate doctors within these now-established VAD systems.

A. An overview of VAD regulation in Victoria and WA

Before turning to the method of this study, we briefly outline how VAD is regulated in these two Australian states. Both states adopt ‘the Australian model of VAD’, whereby access to VAD is limited to terminally ill adults¹⁹ with decision-making capacity²⁰ who satisfy strict eligibility criteria as assessed by two specially trained and qualified doctors.²¹ In these two states, eligibility criteria include that the person must be experiencing suffering from a disease, illness, or medical condition that is advanced, progressive, and expected to cause death within 6 months (or 12 months for neurodegenerative conditions).²² There are ongoing reporting requirements for doctors involved in assessing eligibility, with dedicated state statutory bodies (the VAD Review Board in Victoria and the VAD Board in WA) charged with system oversight.²³ By international standards, the Australian model of VAD is comparatively highly regulated and very prescriptive in its control of the process.²⁴

While the Victorian and WA regimes are broadly similar, there are some points of difference. WA predominately departed from the Victorian model in response to early learnings from the Victorian experience and to accommodate WA’s vast geography and diverse population.²⁵ Both regimes are regulated by a series of regulatory tools and actors. Tables 1 and 2 provide a brief overview of the most significant tools and regulatory actors involved in VAD regulation.

II. METHOD

This article reports on the findings of semi-structured interviews with two broad categories of participants, namely doctors who provide VAD and regulators. Drawing on Black’s definition above,²⁶ we define ‘regulators’ as those whose role involves ‘altering behaviour’ in this field, which extends to both State and non-State actors (see Table 2 for some examples of regulatory actors). For the purposes of reporting this study (while also ensuring the confidentiality of participants), we break down these regulators further into two main categories: (i) regulators who are part of the State and those who occupy official VAD system roles²⁷ and (ii) health and professional organization representatives.²⁸

A. Sampling and recruitment

Recruitment occurred in four stages (with some overlap), and different strategies were employed depending on participant type and jurisdiction (see Table 3).²⁹ We adopted a purposive sampling approach to aim for diversity in our sample, including obtaining representation across several specialities/regulatory roles and metropolitan/regional representation. Snowball sampling was utilized during each recruitment period to extend the sample.

B. Data collection

Data were collected across four periods (see Table 3). Interviews were carried out with 1–3 participants over the phone or via Zoom by 1–3 interviewers (BPW, CMH, LW, RJ, MS), with one interviewer taking the lead during each interview. The interviews were semi-structured, and interview guides were crafted to suit the participant type. In the case of regulators, each interview guide was tailored to the individual’s regulatory role. While individual interview guides did vary, they generally covered the same domains, including reflections on the VAD regime and how it was operating in practice, the ability of the VAD legislation to meet its broad policy goals (ie providing a choice for eligible patients that is reasonably accessible while ensuring safety so access is limited to those who are eligible), and VAD regulation more broadly (including questions that explored participants’ perspectives on how best to regulate).

Interviews were recorded and transcribed verbatim by a professional transcription service. Transcripts were de-identified and checked for accuracy. Participants were sent their de-identified transcript prior to analysis to provide them with the opportunity to revise their transcript.

C. Data analysis

Transcripts were imported into NVivo 12 and analysed using thematic analysis.³⁰ CMH first familiarized herself with the data by reading through each transcript. CMH then developed a broad coding framework that would capture data relating to regulating doctors in the context of VAD, including sentiments expressed about how to best regulate this cohort. The coding framework was guided by CMH’s familiarization with the data, and a research journal and article plan that BPW and CMH developed during the data collection stage. CMH coded the data, revising the coding framework, as necessary. While most of the coded data relevant to this article was in response to specific questions about how to best regulate doctors, each transcript was analysed entirely by CMH, who employed an open approach to ensure that any potentially relevant data were coded and data were not missed. BPW re-engaged with all of the data and refined the coding framework to define themes (and sub-themes). These themes were then discussed and revised in consultation with CMH, and then further refined by LW based on her recollection of the interviews and expertise in the area. All authors agreed upon the themes defined and reported in this article.

III. RESULTS

Eighty-four interviews were conducted with 93 participants, ranging between 33 and 152 minutes. One participant’s interview was conducted in two parts, two participants participated in multiple interviews (due to different roles), and one participant withdrew their data prior to analysis. The final sample of 92 participants comprised 39 doctors who had provided VAD and 55 regulators (see Table 4 for demographics). ³¹

Thematic analysis led to the description of five themes. The first reports on participants’ views about the special nature of VAD and why they (generally) felt it required specific regulation. The remaining four themes examine participants’ observations about how to regulate VAD most effectively. Figure 1 provides a thematic schema of the themes identified. Significantly, when comparing the views of regulators as a group with the views of doctors, there was striking consistency in the data between both groups across all themes. Where there were points of difference, they are noted below.

A. The special and specific nature of VAD regulation

This theme explored participants’ observations about the nature of VAD regulation with two sub-themes.

1. Specific regulation of VAD is needed, but current regulation is heavy

There was global recognition from participants that VAD warranted specific regulation, namely, its own distinctive regulation that specifically governed VAD and this regulation should be stricter than for other areas of medicine. The nature and gravity of a choice to seek VAD informed this assessment. Many participants perceived VAD to be different from other areas of medicine, hence requiring additional control and oversight.

This is a niche case … a unique thing in medicine. Traditionally medicine has been set up to stop people from dying …This is a big difference … you can't undo it … this is an intervention that … is actively killing patients. So, I think you need that oversight for it and [a] legal approach more than you'd need for other areas of medicine.(#23, VD)

Some participants also identified that having some regulatory framework in place was useful, as it helped to remove ethical complexities and meant that conduct could be assessed according to prescribed rules.

Having the governments [determine] what the standards are and how practitioners should conduct themselves and deal with the situation of assisting others to die ... makes it a lot easier … [because you] don't have those same ethical conundrums to deal with … [regulators can] look at the nature of the conduct and how that aligns with the statutory requirements.(#53, VR)

However, some thought that the chosen level of regulation of VAD was excessive and its enforcement was ‘pedantic’. This was expressed more often in relation to the Victorian system and more often by doctors than regulators. The current approach to regulation was seen as having an adverse impact on the system, including potentially deterring doctors from participating. For others, this level of regulation was perceived to increase participation amongst doctors. Indeed, some participants perceived that the current approach to regulation meant that doctors were protected and, therefore, felt safer being involved.

The more tick boxes that you have and the more that it's prescribed, to a certain extent, the easier it is to get people involved. Because they know that if they do A, B and C, that they're not going to be in trouble.(#32, VD)

Some regulators also regarded this as a positive and, indeed, had intentionally chosen to regulate in a vigorous manner, despite potentially making the VAD process more challenging for doctors (and patients) as a consequence. These participants perceived this was justified to ensure there were no mistakes. A critical consideration underpinning this approach was the need to ‘protect the system’ and retain public trust in the system. Some participants suggested that any perceived problems would make the system vulnerable to attack and compromise public confidence.

My lens is what's the risk I need to consider and how do I manage that particular risk … to protect the programme? The other thing I should say protect the doctor…if they do the wrong thing, they're accountable for a hell of a lot …(#68, VR)

Such sentiments partly reflected the gravity of VAD but also the fact that it is new in Australia, so mistakes or problems with the system in its early days could jeopardize its future.

[Victoria] were [the] first in the country, so [they] were … deliberately trying to put in safeguards that no one could argue against, and that's what we've done over the last …three and-a-half years … if something went wrong on the first ten cases, it would have closed down everything for a decade.(#33, VR)

There was little doubt that the degree of regulation of VAD was unique or at least very rare in medicine: ‘there is no activity in health care, other than a few very small things, that has this degree of oversight and regulation.’ (#91, WAR)

2. There are particular challenges regulating VAD

Participants recognized that while regulating medicine in general is challenging, it is particularly challenging for VAD. Both states have long and complex Acts that make VAD legal, but ‘translating that law, that framework into practice is where … the challenge really lies.’(#43, VR)

Both regulators and doctors noted tension between strict rules that tightly control the process to ensure VAD is done safely and consistently, with the need for the regulatory system to allow for flexibility and discretion to make appropriate judgments for the particular clinical context.

If doctors feel that this regulation is so tough and tight that they are worried about making a mistake or … the wrong thing, then that … might deter them from wanting to get involved. So, I think [good] regulation [is] where they understand the parameters within which they're expected to work, but … it still respects, to some extent, their autonomy as professionals and gives them scope to look after their patients safely.(#38, NR)

… the law just makes it easier to punish those people because they’ve clearly broken a rule. I’d prefer a VAD system based more on … guidelines, … policies and procedures … with maybe some more autonomy and independence and wriggle room.(#29, VD)

B. Effective VAD regulation draws on multiple tools

In addition to acknowledging the need to regulate VAD, many participants expressed views about the most appropriate way to do so, including what regulatory tool(s) to use. Participants identified a wide array of tools needed for effective regulation of VAD and considered their strengths and weaknesses. Most commonly mentioned were law, policies and guidelines, training and education, and peers (including communities of practice).

1. Reflections on individual regulatory tools

a. Law

Many participants discussed the role of law in regulating VAD. This was generally the VAD Act, but sometimes subordinate legal instruments or rules were mentioned, such as health or aged care standards, which are required to be complied with for the purposes of accreditation.

Law was often seen as necessary to set the boundaries for what is possible, establishing a broad framework for VAD.

I think [for] something as serious as this, it probably has to be legislated. I think there has to be a sense of this is a legal requirement.(#27, VD)

Participants rarely reported law directly driving specific behaviour in a particular way but rather as operating at a higher or more abstract level to motivate compliance. Many participants suggested that doctors are keen to follow the law when stakes are high (such as VAD), suggesting ‘the more serious the topic gets, the more law abiding [doctors] become’.(#24, VD)

The adverse consequences of not following the law were seen as significant in promoting compliance by doctors. This included the prospect of sanctions (including public sanctions).

If you want to get doctors to do something, the law is pretty tight. Other than that, doctors think they know better and can do whatever they want.(#28, VD)

My experience with dealing directly with clinicians … is a genuine willingness to be compliant because they don't want to breach the Act… even just that sentence, ‘If you do not do this, it is a breach of the Act,’ … is very motivating in getting people to do what needs to be done.(#76, WAR)

There’s law that you have to abide by, and there’s regulations that you have to follow, or Ahpra will be on your case, or you’ll have complaints, and before you know it, you’ll be in the civil court, or in the coroner’s box. So that stuff is a pretty powerful motivator for doctors …(#46, VR)

However, participants acknowledged limitations on the law’s ability to regulate. First, it cannot assist in promoting good practice; rather, it can only prescribe a ‘minimum standard or a lowest common denominator’(#66, VR). Furthermore, as noted above, the law was perceived to have limited effectiveness in helping doctors understand the steps they should take. Indeed, the law requires translation into clinical practice, and this falls to other regulatory tools such as policy or training (see below). Finally, some participants expressed reservations that the rigidity and prescriptiveness of law constrained professional autonomy in clinical practice, while others queried whether the law should be able to compel doctors to take certain actions (particularly if they are conscientious objectors). Indeed, one suggested ‘you just can’t force somebody to do [something], even if it’s the law … you can just encourage them.’ (#49, NR)

b. Policy and guidelines

Participants identified a broad spectrum of documents within this category (see Table 1 for an overview of regulatory tools including policy and guidelines). At one end of this spectrum were high-level positional statements, including those from medical bodies such as the Australian Medical Association (AMA)³² or medical colleges, and high-level care pathways or decision-making frameworks in relation to VAD (often issued by governments and healthcare institutions). At the other end of the spectrum were more detailed policies and procedures about the practice of VAD, typically issued at the local level (ie institutional level).

In terms of types of policies that were more likely to guide behaviour, both groups of participants identified that policies or guidelines that related specifically to clinical practice and gave concrete direction about how VAD would be operationalized and what processes or actions are needed were of most utility to doctors. Specifically, guidelines that would help doctors answer the question: ‘What can I do for my patient today?’(#84, WAR) were deemed most useful.

Such documents were often perceived as key to translating the law into practice, especially when couched in language that doctors were familiar with (ie in medical terms and/or consistent with clinical and professional standards), as opposed to the law which doctors are less familiar with.

Doctors aren't good at law … there are some advantages in a professional practice document because it starts to set the professional standards as different to the rules.(#91, WAR)

Examples of effective clinical policies or guidelines included Health Pathways,³³ documents with clinically relevant case studies, checklists, flowcharts, and decision-making trees (ie documents that specify required procedural steps).

Those sorts of guidelines and the decision-making tree are probably the thing[s] that the GPs will follow. So that’s probably the most important thing in terms of compliance.(#57, NR)

Some participants also identified that local policies and processes were particularly important in the context of VAD, as they were needed to create permission for things to occur.

I like to just do what I want, but a lot of what I do I can’t do unless everybody's happy with me doing that … [It] comes with this kind of subtle herding … there might not be a written guideline about that, but it’s just what everyone does so that’s what you do. So, if a person comes into a ward with a medication you need to have some sort of hospital [process].(#88, WAD)

Policies perceived as having less direct relevance to clinical practice were generally considered less powerful in steering behaviour (and in contrast to the law, which, although seen as remote, is backed by coercive force). Examples of less influential regulatory tools included position statements of medical organizations and colleges.

This is my personal opinion, but I don’t think the [organisation by stating they] believe doctors should not be involved, is going to change somebody’s individual opinion on the ground.(#49, NR)

While such documents were less likely to directly guide behaviour, making an organization’s position clear was still considered useful by some participants, as it enables the organization’s constituents to assess what lens it uses when issuing advice, which may influence whether the doctor would follow such guidance.

[If the organisation] tried to say anything practical —and [they] didn’t say anything about where [they] stood on the topic–that would always be begging the question … ‘We’re getting advice, but what position is this coming from? Is this coming from a ‘for’ position or an ‘against’ position?’ I mean, I guess you could read between the lines, but I think you're obliged to say what your position is.(#41, NR)

c. Training and education

Training and education were identified as important parts of VAD regulation. This may have been shaped by both states having legislatively mandated training that participating health practitioners need to complete (see Table 1), which was the subject of most of the discussion on this topic. The mandatory training is delivered online.³⁴ The VAD Care Navigators and some health services have hosted in-person training sessions where clinicians were invited to attend to complete the training with other practitioners.

There was a spectrum of views on the utility of this mandatory training. Most felt it equipped doctors with knowledge of the VAD law and their responsibilities under it, and endorsed its role as a safeguard.

I mean the way it’s set up it’s made pretty clear—the training is good, and it’s made quite explicit what are the ironclad rules, law regarding this.(#23, VD)

Other participants from both cohorts were more critical, suggesting the training was too legalistic and long, and a potential barrier to VAD participation. There were mixed views on the assessment component: some perceived it as too hard, while others considered it necessary to ensure careful engagement with the training. Some participants suggested having the opportunity to undertake the online training in person and with peers (as opposed to doing the training alone) was likely to enhance engagement with the material, and willingness to undertake the training.

The very legalistic focus of the training, in the beginning, was probably not well received by doctors. And the 90% pass rate with rather esoteric legal questions wasn’t taken well. But getting [health practitioners] to do training in groups is collegiately good; being together and talking about things is really good. And the pass rate has gone right up as a result.(#48, VR)

There was also a desire for more training about the practice of VAD (rather than the legal framework), whether that was in the mandatory training or generally: ‘It’s doing it practically that is useful. It’s filling in the forms the first time.’(#21, VD) Some participants reflected that some softer process skills (eg communication skills), that were not necessarily VAD-specific and transferable to other aspects of medicine, could be embedded within other forms of education (eg tertiary education and continuing professional development).

The training requirements for this could pretty much fall within the usual approach that we have for professional development in the medical sphere… digging down to what that looks like … well, the universal in health is the communication skills.(#41, NR)

d. Peer opinion and/or professional norms

Perhaps the most striking finding was the frequency with which peer opinion and/or professional norms were mentioned in response to questions about how best to regulate. It is significant that they are not traditionally regarded as ‘regulatory tools’ in the same way as the preceding tools (perhaps a formal community of practice is the closest to being ‘formal regulation’). Yet, both doctors and regulators identified peer opinion and professional norms as powerful factors guiding doctors’ practice.

Some noted this normative power came from a desire to conform to peer opinion as no one wants to be the ‘only person marching out of step’ (#42, NR). Another was that doctors recognize this professional consensus or peer opinion is how clinical practice is regulated generally and resembles how learning usually occurs in medicine.

Everything [doctors] do is peer learning, so we need to adapt that framework to fit into the model that already exists, and that is the primary way that we conduct and learn about medicine.(#1, VR)

A few participants specifically acknowledged the influence not just of peers but of a mentor whose experience or seniority gives credibility. Sentiments also reflected the importance of ‘on-tap’ support when needed(#66, VR), as opposed to training, which is often only done at one point in time.

What you also need is a mentor. So, I personally think, having been through it, that it would have been really great if I had had someone with me the first time who’d done it before.(#21, VD)

The communities of practice formally established in both Victoria and Western Australia and informal communities established by doctors themselves were identified as important sources of VAD regulation. Such communities crystallize a repository of VAD knowledge in an accessible forum which allows doctors to ‘calibrate [conduct] amongst our peers’ (#88, WAD) and, in doing so, provides ‘subtle herding’ (#88, WAD) of doctors towards safe and consistent practice. Doctor participants valued the opportunity to test their thinking generally and in relation to particular cases to support their decision-making. This benefit was also observed by regulator participants.

I think there’s a great value in having those professional communities and networks … being able to engage around practices and learn together in such a new and evolving scheme, to be able to say, ‘What does this education tell us, and what have we learnt from that?’ or, ‘I've had this situation’.(#52, NR)

e. Other regulatory tools

Although the above four regulatory sources were most commonly identified when discussing effective regulation, participants also discussed other regulatory tools. A significant one was the online information management system (or portal) through which forms are submitted (generally by doctors) during the VAD process. Some doctors were critical of this infrastructure, particularly with respect to the time and administrative burden it placed on them. Some participants also criticized the rigidity and useability of this system (especially in Victoria and in the early days before a system upgrade).

However, both doctors and regulators saw the regulatory benefits of the system. First, this infrastructure prompted doctors of the next step in the process, which helped guide them through the correct stages of the complex VAD assessment and approval procedure. Furthermore, the periodic submission of forms during the process allowed for scrutiny of these documents, providing an opportunity to monitor decisions being made informally. This scrutiny of each submitted form meant that potential mistakes or issues could be identified before VAD was provided.

The portal … like it’s a bit foolproof… going through each step. I think that that’s useful, because it means that you’re not sort of going on and thinking, have I stuffed this up … along the way … I guess … [you need] legislation and then technology that means that we can’t deviate from the legislation.(#18, VD)

To help enforce a particular process, you need something that helps guide [health practitioners’] decision-making but also guides automated accountability. And I think the (information management system) … [has] an aspect of guided accountability … in terms of the notification process, the prompts, the underlying education … the reinforcement …I think that's the key thing, in the absence of sort of any other stronger control, in terms of ensuring compliance to legislation.(#85, WAR)

A second significant form of regulation perceived to be effective was the role that key individuals play in guiding behaviour. This is linked with peer opinion above, as this individual was often a health practitioner but is narrower because the guidance was not grounded in wider professional norms but rather came from the concentrated expertise of specific individuals. The most common example was VAD Care Navigators. Their experience with the system meant they were able to guide doctors through the process (and therefore within the law), advise on how Boards and Secretariats require the forms to be completed, and were generally often sought out as trusted sources of reliable information about VAD. Another example participants gave was a lead VAD clinician for a facility or organization.

Other forms of effective regulation participants identified include contracts with health services (which can include conditions that guide behaviour such as a requirement to develop local policies and procedures to manage VAD processes),³⁵ meetings and relationships with health professionals and service providers (as a chance to influence their practice), and research (which can be used to generate evidence to help guide policy and practice). Financial resourcing was also mentioned as a way to increase the participation of doctors in VAD (along with the linked issue of releasing clinicians from existing clinical duties to undertake VAD work accounted for in local policies, discussed above).

Accessible resources and readily available information also helped to regulate the practice of VAD. Public debate (including through the media) and information-sharing forums (eg webinars, conferences) about VAD have also shaped practice, with personal stories about an individual’s VAD experience found to be particularly compelling. Indeed, publicity around suboptimal practice was perceived to encourage compliance with regulation, shape VAD practice or serve as an advocacy tool to help guide future reform. In a different vein, exposure to personal stories was also found to encourage practitioner involvement with VAD.

2. Multi-faceted approaches to regulation are needed

The foregoing discussion shows that participants collectively identified a wide range of regulatory tools when discussing what made for effective regulation of VAD. However, some participants went further and made the specific point that effective regulation of VAD depends on using multiple tools of regulation together.

You need a mix of approaches. I don't think there's a single way of regulating behaviour that is likely to capture all eventualities and to work for all types of practitioners or all types of people … different people are motivated by different things. (#77, WAR)

Other participants demonstrated this view by talking about more than one regulatory tool being needed. Many participants identified that regulation in the form of law was important but should be supplemented with at least one further form of regulation. In particular, and as noted above, this was commonly framed as the law establishing a framework or process that needs translation into clinical practice or guidance, the latter being achieved by other regulatory tools.

C. Educative approaches to VAD regulation are generally more effective

Both doctor and regulator participants generally preferred an educative approach over a punitive approach for a minor error or non-serious breach. Although this theme was present in views from both states, it was particularly so in WA. This difference likely reflects that the fact the WA VAD Board has been recognized to take an educative approach to regulation, compared to the approach by the Victorian VAD Review Board in its early days of operation which has been perceived to be more punitive.³⁶ An educative approach was seen as more effective in guiding practice: it is better to be ‘gently herding’ rather than ‘shaming people’ (#88, WAD). Some participants emphasized the importance of dealing with issues in a constructive and proportional manner.

… things might be a breach of the Act, but then they get notified to the DG [Director-General of the Health Department]. But it's not like … that’s automatic notification to the police. The DG will have a look at it and say ‘what does this actually mean? … Is there something else that needs to be done here? Or is this just evidence of part of the process that we need to improve? In which case, let's improve that part of the process.’(#76, WAR)

An educative approach was seen as appropriate because doctors want to comply with the VAD system, so mistakes or failures generally reflected a lack of understanding, as opposed to a deliberate breach.

I think education is the key … if practitioners fully understand what their obligations are and what steps they need to take to comply, I expect that 99 percent of them will comply… It doesn’t mean there won’t be slipups, and … mistakes… But I genuinely think that if they’re properly educated and they understand their requirements, then they’ll do it.(#53, NR)

Indeed, some regulators identified that it would be inappropriate for them to take action for non-compliance, without first educating their constituents.

[Regulators would] be criticised if [they] hadn’t done [their] bit to try and educate the profession so they know what the requirements are before [they] jump in and start disciplining them.(#52, NR)

Adopting an educative approach was also perceived to be conducive to fostering relationships between regulators and doctors. Maintaining such relationships was particularly important for State-based regulators. From their perspective, educational approaches provided a mechanism to softly correct behaviour without being antagonistic and risk discouraging doctors from continued participation in providing VAD. This was considered imperative to the system’s overall sustainability.

It is about those informal links and relationships … the Board emphasising that they have an educative approach and that if [the Secretariat does] ring a practitioner and tell them they made an error in the form, [they are] not quoting the Act at them, ‘under section blah, blah, blah,’ and ‘if you don’t do this, you could be fined.’ It’s about approaching people with sensitivity, and working with practitioners and if they have made a mistake or an error.(#70, WAR)

Despite the preference for non-punitive approaches, the ability to implement punitive measures was seen as necessary and appropriate in some (exceptional) cases: ‘For the one percenters, deterrence is going to have to be important, and that means that [regulator] will have to take appropriate action’(#53, NR). Punitive measures included risk to medical registration and/or financial sanctions. Furthermore, a more vigorous approach to regulation was also perceived by some regulators to be needed if action was required to ‘protect the system’ (as noted above, given the specific nature of VAD).

D. VAD regulation is effective when it is perceived by doctors as acceptable

This theme explores why doctors might accept or respect VAD regulation as appropriate and choose to follow it. It goes beyond the special nature of VAD warranting particular regulation (theme 1) and how different tools of regulation may be effective, for example, how the law’s coercive power may be effective to promote compliance by doctors with the system’s rules (theme 2). Instead, this theme focuses on how the nature or design of that regulation contributes to doctors perceiving that regulation as legitimate and appropriate in the VAD field and, therefore, worthy of compliance. One participant spoke about the need for such regulation to reach people’s ‘hearts and minds’(#57, NR).

Some participants (both doctors and regulators) noted that doctors are generally not easy to regulate: ‘doctors like to be able to just do what we think is the right thing to do’ (#88, WAD); ‘We’re [an un]ruly bunch, we don’t like rules really’(#27, VD); and ‘a lot of doctors just don’t like being told how to practise medicine’(#49, VR).

Participants suggested that for regulation to be effective, doctors generally needed to recognize (and appreciate) its purpose.

Rules for the sake of rules they’re often very frustrating for doctors who are trying to be very practical … if there’s a clear, logical explanation behind something, then we’ll follow that.(#31, VD)

I hate when somebody tells me that I have to do things in a certain way …I’ll stick to what I think is a sensible guideline…(#88, WAD)

I think sometimes doctors are cut³⁷ more when something seems really much more bureaucratic, but they can't see the logic behind it. Which I don’t think is in the case of VAD.(#24, VD)

The special nature of VAD (theme 1) was identified by some participants as helping make this case, but this was not sufficient, demonstrating that some laws are perceived differently on paper than they are in clinical practice. An example given by several doctors was that state VAD laws made sense and were understandable (particularly the need for eligibility criteria), but federal limits on telehealth³⁸ were not reasonable and, therefore, were less persuasive.

Factors participants identified as contributing to a perception of VAD regulation being justifiable and therefore legitimate included that regulation was of benefit or useful for patients, the public or community, or doctors themselves (eg legal protection).

[Doctors] only care about something if it impacts their patients generally or is going to make them not able to practise medicine anymore.(#84, WAR)

When there’s a lot of bureaucracy and things don't make sense to doctors, then they’re probably, as a group, more willing to push boundaries because … their priority is … their patients’ care rather than bureaucracy.(#24, VD)

… to get medical staff to want to participate in anything there’s got to be a bit of what’s in it for me? … [that] lever is helpful.(#61, VR)

Some participants also pointed out that regulation was more likely to be perceived by doctors as appropriate if it was based on evidence.

I think for medical practitioners, where we try and base most of our decisions on evidence … the rules have to have [a] good rationale behind them, and … be logical. So, I think that if you’re designing a system where [you] want people to do this … it either has to be practical or have a strong degree of evidence there.(#61, VR)

The resonance or ‘fit’ of regulation with clinical practice also had an impact on perceived legitimacy. When regulation is practical and aligns with how doctors understand and practise medicine, it is more likely to be viewed positively. For instance, limiting telehealth for VAD (which is now routinely used in medicine, including end-of-life care) was an example of an impractical law with potentially limited normative force.

One issue I’ve had with the regulation is the whole idea of the telephone conversations or email conversations … The guidance we’ve had is just don’t do it, which is impractical … Whereas when it comes to the state law around it, it’s quite clear what the guidance is, and we know what to do … having that framework, and having it somewhat practical as well, is probably the two things that I’d say to getting doctors, or anybody, to comply with the regulation.(#22, VD)

The process that went into developing the form of regulation also influenced its perceived acceptability. This included who was involved in its development and delivery, with positive observations when key medical leaders were involved. Regulation developed in consultation with doctors and others involved in the system was also seen as having greater legitimacy.

People feel willing to contribute and participate if they feel some sense of co-design or autonomy and it's not unilateral … There’s no point in just saying, ‘All emergency doctors have to do blah, blah, blah. Here’s your directive memo from the general manager, end of story.’ It’s about sitting down with them and saying, ‘OK, these are the things we’re trying to achieve. What’s your role in that and how?’(#60, NR)

This process of participation and inclusion in decision-making was perceived as being especially important for regulators who depend on their medical (and/or broader health) membership for support. Policies or position statements had to juggle reflecting the views (or breadth of views) of members and being representative, while also recognizing the possibility that such statements may go beyond existing views of members and attempt to create leadership in policy directions preferred by specific personnel within those organizations. Support of members for regulatory tools was also perceived to be needed for them to have any weight.

A final observation about the acceptability of VAD regulation related to whether VAD should be conceptualized as different from usual health care or not. Theme 1 identified VAD as warranting special regulation and identified particular challenges with its regulation. This has implications for the legitimacy of VAD regulation, including, as noted above in this theme, providing a rationale for regulation that doctors understood as justified and, therefore, respected. However, by contrast, some regulator participants argued regulation of VAD is more likely to be effective and seen as legitimate if it is treated consistently with how regulation of health care generally occurs.

VAD is being implemented within an existing system where you already have kind of established purposes and relationships… I’ve really tried to ensure that we achieve with VAD is that, wherever possible, we just line up with how [we] normally [do] business. [We]’ve tried really hard not to have an approach that is different and outside of that if we can possibly avoid it.(#77, WAR)

Another regulator participant expressed similar sentiments and even proposed conceptualizing VAD regulation in a way that resonates with how doctors approach clinical work.

The way to approach it is to see if you could frame this as simply being another form of medical treatment, and how would they approach that… In other words, to try and present this as not being something that is, at the moment, seen as this technical legal thing, but more just as another form of treatment and approach it in the same way, and recognising that there are some formal requirements of going through the [eligibility] criteria, but that’s a bit like symptoms. There are some formal requirements for forms, but you’d have to do that if you wanted to prescribe methadone or whatever. So, to try and make it less unfamiliar to them and less seen in a legal framework.(#40, VR)

E. VAD regulation is effective when it is nimble and responsive to doctors’ (and others’) needs

The final theme was that regulation is more effective when it is nimble and able to respond to what doctors need to participate in VAD. This included recognition that the introduction of VAD meant that doctors had to receive and process a large volume of information initially (eg mandatory training) but that ongoing guidance is needed.

I think it’s unreasonable to expect that…a very, very busy clinician who’s got a full normal clinical load, is doing VAD quite literally in their own spare time, who’s done the training … could be expected to memorise the legislation … VAD guidelines, et cetera, and know all those off by heart. We have to have a level of decision-making support, guidance … prompt automation to ensure compliance.(#85, WAR)

Key aspects of effective regulation were right-sizing the volume of information, having it available in the right form, and making it available at the right time.

I think it's about recognising the need for that ongoing resource support and that there will be need for ongoing questions and … reinforcement. And providing that in a chunky way that makes it not so overwhelming and huge, but is accessible … They want to know: what can I do for my patient today? So, providing the information in a way that’s chunked up, that supports that decision making at the right time… is [what doctors] generally need.(#84, WAR)

This includes doctors who might not yet be involved in VAD but nevertheless must respond when it is raised (even if they do not directly participate).

If you think of a GP down at [a regional town], and someone walked in off the street and raised assisted dying with them, it’s going to be the first time they’ve ever turned their mind to it. They’re not going to be sure about what the legislative obligations are. They’re not expected to. So, they really have to know, ‘Who do I speak to now to try and sort out what do I do next?’(#53, NR)

Participants noted it can be challenging to find that information. But some spoke about the vital role that people—as opposed to the system itself or published information—played in guiding people to the right resources or providing advice to assist and being available to troubleshoot where applicable.

One of the things that we’ve tried to achieve is the clear and strict process that needs to be followed, but also the human communication and assistance that can help practitioners to follow that process.(#88, WAR)

… support comes from the Care Navigators, it comes from the Pharmacy, it comes from the Board and it comes from [the Secretariat]. You know, ‘If you need help, give us a phone call. If we don't know the answer, we’ll find the right person.’(#86, WAR)

I think it’s about remembering that doing the training is a particular point in time. So, they get it when they do it, and then they’ll come back to us multiple times with questions: ‘Can I do this? Can I not do this? What does this say? What do you think?’ And we will refer them back to the right bits and bobs. (#84, WAR)

Some participants reported this human support as missing, with one participant responding, when asked what might help them:

Easier technology, the ability to talk with a human. This [online] portal is just bizarre, and you’re not talking to a human. I don’t know why you can’t have a human at the end.(#4, VD)

The layering or multi-faceted nature of the VAD regulation was considered an important feature of nimble and adaptive regulation. These layers of regulation emerge clearly from the above quotes in this theme. They are also reflected in the earlier themes, for example, recognition of the utility of training, but the need to supplement this with mentoring to practically deliver VAD (theme 2).

In addition to responding to individual doctors’ needs at any given point in time, participants also saw the benefit of the system as a whole being responsive to the needs of doctors providing VAD. This included system changes in response to feedback (eg upgrades to the information management system) and issuing generalized advice (eg in the forms of newsletters or practice notes) in response to learnings from the system. This system responsiveness was seen as particularly important because VAD was new and hence required regulators to actively reflect on this evolving practice and initiate timely adjustments.

Participants also gave counter-examples of where regulation, in their view, failed to be nimble and responsive, and so was not effective. One was in relation to the early days of the Victorian system, where the process of reviewing and approving the forms that must be submitted by doctors was slow and ‘pedantic’. Another was where health services organizations failed to develop policies and processes to respond to VAD promptly, resulting in a regulatory vacuum leading to confusion and inaction.

IV. DISCUSSION

This article reported on perspectives about how best to regulate VAD from both doctors who have provided VAD and the regulators who guide this practice (Fig. 1). Strikingly, both regulator and doctor participant groups were consistent with each other in their views on what constituted effective regulation.

The nature of VAD was perceived by participants to require special regulation, although some felt this was overdone in these two Australian states. Reported features of effective regulation included regulators taking an educative approach, regulation being perceived as acceptable by doctors, and it being responsive and nimble to provide the guidance that doctors need. Participants also considered a range of regulatory tools were required to regulate VAD effectively, and some identified a need for these tools to be employed together in a holistic way.

A. Holistic and layered approaches to VAD regulation

Our findings which identify the need for a holistic approach to VAD regulation align with wider regulatory scholarship. As noted above, modern regulatory theory generally recognizes that regulation is decentralized and hence a range of regulatory tools exist, requiring holistic approaches to understanding how behaviour is guided.³⁹ These findings may be unsurprising given how we defined regulation and our purposive approach to sampling, which targeted different regulator types who employed different types of regulatory tools (See Tables 1 and 2). However, the need for multiple regulatory tools, and for them to be used in an integrated way was independently identified by the participants themselves (theme 2). Such findings are congruent with previous calls for holistic approaches to regulating end-of-life care.⁴⁰ Failure to appreciate the interdependence of regulators and their regulatory tools can lead to a ‘mosaic of regulatory arrangements’ which are highly fragmented and result in duplication.⁴¹

Participants’ holistic approaches were also layered in that they generally reported a graduated conception of regulation that begins with a general legal framework which then progresses to increasingly specific guidance at the clinical coalface where actual decision-making about VAD occurs. There was recognition that law was required to permit VAD to occur and establish the boundaries of the regulatory framework. But legal rules, reflecting the challenges of regulation more generally,⁴² are inherently indeterminate in that they can never completely capture the circumstances law and policymakers intend them to be used for, due to their anticipatory nature and their reliance on others to interpret. Further regulatory work is needed to translate those laws into safe and compliant clinical decision-making, and, as is commonly the case in medicine,⁴³ this generally fell to tools such as training, policy, and guidelines. We also note repeated calls from participants, as part of the ‘nimble’ regulation theme, for a further layer of regulation at the coalface, such as ‘on-tap support’ to troubleshoot where necessary, often involving access to a specific person who could advise.

Significantly, participants identified two tools as being prominent in guiding behaviour, beyond what might ordinarily be anticipated in an evaluation of VAD regulation. The first was the State-run online information management system that is used for the submission and management of the forms that doctors (and others) must submit during the VAD approval process. Although information technology is a recognized tool in wider regulatory scholarship,⁴⁴ it was not part of the legislation that established VAD, nor was it one of the considered safeguards.⁴⁵ Rather, an information management system was only conceived of during the implementation phase of the legislation.

Yet participants, consistent with other literature drawing on some of these interviews,⁴⁶ report that this information technology infrastructure exercises very significant practical control over the VAD approval process. This is particularly the case in Victoria, where VAD cannot be accessed until a senior public servant in the Health Department authorizes VAD via a ‘permit’. This permit can only be issued after reviewing all of the submitted online forms for compliance, and hence, this additional final step in Victoria facilitates a gatekeeping function through the information management system.⁴⁷ While the focus of VAD regulation tends to be on tools such as law, policy, and training, this research demonstrates that information technology can exert powerful regulatory control and its potential role needs to be considered in the holistic design of VAD regulation.

A second regulatory tool worthy of specific comment is peer opinion and/or professional norms. It was striking that both doctors and regulators identified how significant these were for guiding behaviour in the VAD field (as well as medicine more generally). While this might be anticipated in the discipline of medicine, which has a long recognition of the power of peer opinion and professional norms in guiding practice, it does sit somewhat awkwardly with the conception of ‘regulation’ proffered by Julia Black. Black’s conception refers to a ‘sustained and focused attempt to alter behaviour’ and does not extend to the operation of market forces, social norms and culture.⁴⁸ In Black’s view, such factors might influence how regulatory systems operate, and regulatory systems might seek to harness and alter them, but they are not in themselves forms of regulation.⁴⁹

Despite this, our participants consistently identified peer opinion and/or professional norms as a critical part of effective regulation of VAD. There is also a link to more formal regulation in that, in both Victoria and WA, the State sought to utilize these forces in regulating VAD through their intentional ‘seeding’ of VAD communities of practice (ie regular meetings of doctors providing VAD).⁵⁰ Resources and infrastructure were initially provided to establish such forums, and some of this State support continues today. Early evidence shows these communities of practice have been valuable in providing a forum for collective expertise, fostering camaraderie, and permitting information-sharing and the development of consistent VAD practice (some of which we would regard as regulation).⁵¹

Regardless of whether peer opinion and/or professional norms can be conceptually regarded as a regulatory tool, their significance for both sets of participants suggests they are a powerful way to guide behaviour in the provision of VAD and should be harnessed, or at least their influence accounted for, when the State is considering its overarching approach to VAD regulation. One issue to monitor is the risk that norms of the medical profession may not align with the intended policy goals of VAD regulation (eg norms which prioritize doctor discretion over compliance with system safeguards). This may require greater State intervention via other regulatory tools to constrain professional norms affecting VAD practice in this way.

B. Why doctors choose to comply with VAD regulation

Other significant findings in this research related to participants’ observations about a regulatory system’s legitimacy, again reflecting wider debates in the regulatory literature.⁵² Many participants believed VAD required regulation due to its ‘special nature’, with some doctors observing this served to protect doctors (as well as patients). This resonates with what has been observed internationally.⁵³ However, beyond acceptance of a threshold point of the need to regulate, the challenge is designing and implementing a regime that commands widespread acceptance.⁵⁴ As identified during the implementation period in Victoria, concerns have been raised by doctors about the inability of a legislative approach to capture the nuances of a clinical environment.⁵⁵ Doctors’ compliance with rules are largely shaped by their values and views on the laws they are governed by, and to what extent they perceive them to be legitimate.⁵⁶ This in turn informs the way in which they resolve tensions in professional obligations.⁵⁷ Participants observed that doctors were more likely to see regulation as legitimate and comply with it when they could appreciate its rationale. This points to the importance of communicating the ‘why’ of VAD regulation to doctors.

Also important for perceptions of legitimacy can be the ‘how’ of regulation in terms of its development. While regulation from the State may be able to depend on enforcement through coercion (which in itself can bring some legitimacy),⁵⁸ non-State actors and tools of regulation may need to construct legitimacy for their regulatory influence to be acknowledged.⁵⁹ Good process in developing regulation can help with this. Participants suggested that regulation was more likely to be acceptable if it was developed through a process that was based on evidence, was consistent with clinical practice, and if doctors (or perceived ‘credible’ or ‘authoritative’ sources) were involved in the design. Although not specifically commented on by participants, this would include legitimacy attached to the original VAD laws due to the credible and inclusive process of developing and passing them.⁶⁰ This resonates with what has been observed internationally. For instance, one study carried out amongst VAD providers in Canada during the COVID-19 pandemic found that in cases of tensions between VAD laws and COVID-19 public health rules, many clinicians appeared to prioritize VAD laws, perceiving them to be more legitimate.⁶¹ The authors surmised that this may be a product of the time, research, debate, and deliberation which lead to their formation.⁶²

A final observation about compliance is to note that regulation is a human enterprise, and decision-making and conduct of individuals were shaped by several ‘motivators’.⁶³ For doctors, this included risk motivators that were either reputational (eg being out of step with colleagues) or professional (eg non-compliance leading to loss of registration and/or financial sanction). In a similar vein, motivators also informed regulators’ approaches to regulation, including preserving the sustainability of the workforce (eg adopting educational approaches to preserve relationships) and protecting the system as a whole (eg by regulating vigorously to protect patient safety and public confidence in the system).

C. A regulatory pyramid for VAD

A widely recognized approach to promoting compliance with regulation is the idea of a regulatory pyramid.⁶⁴ At the bottom of the pyramid, regulators seek to secure compliance with a regulatory framework through persuasion but they have the capacity to progressively escalate up the pyramid to more serious responses should previous approaches not be successful. This includes, at the top of the pyramid, utilizing the coercive power of the State to punish non-compliance.

This reflects the broad approach to regulation preferred by participants. They favoured an educative approach to VAD regulation (sometimes expressly contrasted with a punitive approach) but with scope for punitive measures to be used if needed. Generally, participants considered that in the context of VAD regulation, educative approaches would be sufficient. Perhaps this is because of the significant regulation that surrounds VAD, which already guides behaviour towards compliance with the framework.

There has been, however, one instance in Victoria where the oversight Board has opted for a punitive approach and referred a VAD provider to the Medical Board (a disciplinary body for the medical profession) in relation to a failure to observe legislative procedural requirements (unrelated to patient eligibility).⁶⁵ The practitioner was reprimanded and fined for their conduct.⁶⁶ According to the judgment, the sanction was appropriate for the purposes of deterrence (both general and specific) and to maintain public confidence in the VAD regime.⁶⁷

A specific challenge relevant to regulating doctors providing VAD is that the effective functioning of the VAD system (and its sustainability) depends on having doctors who are willing to participate. In other words, access to VAD is only possible if there are sufficient doctors who will provide this option for patients.⁶⁸ Yet, the limited pool of participating doctors is a known challenge in both states.⁶⁹ This further supports the preference for an educative approach that helps doctors comply with the VAD framework rather than a punitive approach (unless required for system safety), which may drive them away.

D. Exceptionalism of VAD regulation?

The preceding discussion identifies a tension: participants clearly considered that VAD was different and required special regulation, yet they also suggested that part of effective regulation of VAD was to treat it like other aspects of medical practice. Theme 1, which concerned ‘the special and specific nature of VAD regulation’, reflected the commonly held sentiment amongst both doctor and regulator participants that VAD is different. This is what justifies its specific regulatory approach: very detailed and prescriptive legislation with a suite of other tools such as policies and legislatively mandated training for providing doctors. This is also reflected in theme 2’s discussion about the need for multiple tools of regulation and in theme 4, which explained that one reason why doctors accept this high level of regulation in VAD is because it is different from other areas of medicine.

However, participants also made observations suggesting that part of effective regulation of VAD meant not treating it differently from other parts of medicine. For example, some regulators noted that they aimed to use tools of regulation (theme 4) and approaches to regulation (theme 3) that are usually applied in medicine because they are more likely to be acceptable to doctors. A similar point was made in relation to regulation being more effective when the content of regulation aligns with existing clinical practice (theme 4), which may be applicable beyond the empirical research reported on. Indeed, as the above discussion notes, many of the issues and challenges of VAD regulation are not unique and observed in both the regulation of medicine in general as well as other regulatory domains.

We do not seek to resolve this wider conceptual point here. We note that internationally, at least to date, VAD has been treated as an exceptional case in terms of regulation. Even in countries like the Netherlands and Belgium which have over two decades of experience with VAD, and high cultural acceptance of the practice, specific and different regulation governs it.⁷⁰ But further thought is needed about whether VAD is exceptional in requiring special and specific regulation, and, if so, why. Regardless of one’s view on that, we do consider that where possible, as our participants also noted, the benefits of regulating VAD like other fields of medicine should be harnessed. This means that where there are lessons to be learned from the regulation of medicine (or regulation generally), they should be applied and considered in VAD. The converse point can also be made: studies of VAD regulation can inform how to optimally regulate medicine, and perhaps also inform thinking about regulation in other ‘high-stake’ settings.

E. Strengths and Limitations

A strength of this empirical study is that we were able to interview a large number of doctors (as the primary target of regulation) and a wide and comprehensive range of VAD regulators, who regulated VAD directly or indirectly, across two Australian states (as well as national/multi-state regulators) about their views on how best to regulate VAD. The consistency across the views of regulators and doctors gives credence to our findings. A further and linked strength was that some regulators had (or previously had) clinical roles, which means that their insights were informed by first-hand clinical experience, including what might constitute effective regulation.

One study limitation is that interviews with different cohorts were conducted at different points in time (see Table 3), and regulation may have evolved over time. However, the impact of this is partly mitigated by the fact that interviews in Victoria were conducted in both early (doctors) and later (regulators) stages of system development.

Another limitation is that this research is conducted in relation to the Australian model of VAD, which is highly regulated. While findings are likely to have direct relevance for other Australian jurisdictions, or international models which take a similar approach to regulating VAD, including because they have drawn on the Australian model when designing their own system, the findings may be less generalizable to other international models. However, some findings are not directly related to the specific nature of the Australian VAD system, such as findings about the effectiveness of peer regulation and professional norms, and the role of system design in regulating doctors in the context of VAD. As such, these more generic findings may be transferable to other jurisdictions even though they regulate VAD differently from Australia.

V. CONCLUDING REFLECTIONS: EFFECTIVE REGULATION OF VAD

We conclude this article with some reflections, drawing on our participants’ observations, about how best to regulate VAD. Box 1 sets out proposed principles for effective VAD regulation. These principles focus on how regulation should be designed and implemented rather than the content of that regulation; for example, they consider how regulatory tools should be employed rather than what eligibility criteria for VAD should be. These principles also accept as a starting point the current position (expressed by our participants and reflected in Australian and international practice) that VAD warrants specific regulation.

As noted above, we recognize that regulation is a local endeavour that is informed by a range of domestic and contextual factors, and so the utility of these principles may vary for other jurisdictions. Nevertheless, given the paucity of work on effective regulation of VAD to date, we hope the above findings and these principles may contribute to deeper thinking about how this practice is best regulated. As such, this research should be considered when VAD systems are reviewed, for example in the mandatory reviews in Australia and the five-yearly evaluations in the Netherlands. This research also should be considered by jurisdictions that are contemplating legalizing VAD or are implementing their VAD systems. In addition, we note the potential (as mentioned above) for this research about VAD to inform wider debates about the regulation of medicine more generally.

Footnotes

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the participants who supported this research. They also acknowledge Marcus Sellars (MS) for his role in interviewing Victorian doctors and Ruthie Jeanneret (RJ) who was part of one of the Victorian regulator interviews. The authors also acknowledge the thoughtful feedback on part of this article from the Policy and Practice Advisory Group for the research project ‘Optimal Regulation of Voluntary Assisted Dying’.

ETHICS

This research was approved by the Queensland University of Technology Human Research Ethics Committee (approval numbers 2000000033 (26 March 2020) and 20000002700 (13 July 2020)).

FUNDING

This research was supported through an Australian Research Council Future Fellowship (project number FT190100410: Enhancing End-of-Life Decision-Making: Optimal Regulation of Voluntary Assisted Dying) funded by the Australian government. The funder only provided funding and did not participate in this research.

CONFLICTS OF INTEREST

Ben P. White and Lindy Willmott were engaged by the Victorian, Western Australian, and Queensland governments to provide the legislatively required training for doctors (and nurses and nurse practitioners where relevant) involved in voluntary assisted dying. Casey M. Haining was employed as a legal writer on the Queensland training project. Ben P. White and Lindy Willmott were engaged by the Western Australian Government to be involved in the statutory review of the Voluntary Assisted Dying Act 2019 (WA), and Casey M. Haining was appointed as the research fellow on that project. Lindy Willmott is also a member of the Queensland Voluntary Assisted Dying Review Board.