Objective Advance care planning (ACP) assists people to identify their goals, values and treatment preferences for future care. Ideally, preferences are documented in an advance care directive (ACD) and used by doctors to guide medical decision-making should the patient subsequently lose their decision-making capacity. However, studies demonstrate that ACDs are not always adhered to by doctors in clinical practice. We aim to describe the attitudes and perspectives of doctors regarding ACD adherence and the utility of ACDs in clinical practice.
Design Face-to-face semistructured interviews were conducted using three case-based vignettes to explore doctors’ decision-making and attitudes towards ACDs. Transcripts were analysed using a thematic analysis.
Setting Doctors from a variety of medical specialties and with varying experience levels were recruited from a large tertiary hospital in Melbourne, Australia.
Participants A total of 21 doctors were interviewed, 48% female (10/21). Most (19/21) reported having experience using ACDs.
Results Four themes were identified: aligning with patient preferences (avoiding unwanted care, prioritising autonomy and anticipating family opposition), advocating best interests (defining futile care, relying on clinical judgement, rejecting unreasonable decisions and disregarding legal consequences), establishing validity (doubting rigour of the decision-making process, questioning patients’ ability to understand treatment decisions, distrusting outdated preferences and seeking confirmation) and translating written preferences into practice (contextualising patient preferences, applying subjective terminology and prioritising emergency medical treatment).
Conclusions ACDs provide doctors with opportunities to align patient preferences with treatment and uphold patient autonomy. However, doctors experience decisional conflict when attempting to adhere to ACDs in practice, especially when they believe that adhering to the ACD is not in the patients’ best interests, or if they doubt the validity of the ACD. Future ACP programmes should consider approaches to improve the validity and applicability of ACDs. In addition, there is a need for ethical and legal education to support doctors’ knowledge and confidence in ACP and enacting ACDs.
- advance care planning
- advance care directives
- semistructured interviews
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Strengths and limitations of this study
Purposive sampling used to promote the inclusion of participants from a variety of medical specialties and experience levels.
Face-to-face semistructured interviews used, providing a deeper understanding of participants attitudes and perspectives about advance care directive (ACD) adherence.
Use of vignettes with hypothetical case scenarios to stimulate doctors’ discussion about the use of ACDs in clinical practice.
Recruitment was undertaken from a single metropolitan health service in Australia, which may limit the transferability of findings to other settings.
Advance care planning (ACP) is a process where an individual makes his or her goals, values and preferences known to guide decision-making for healthcare, in the event he or she cannot make or communicate decisions in the future.1 ACP assists patients’ personal reflection about their values and goals of care, and can be supported by the completion of an advance care directive (ACD), designed to inform medical decision-making in the event the patient loses the capacity to make or communicate decisions.2 The terminology and legal status of ACDs vary considerably within Australia3 and internationally4–7; however, they are commonly used to specify treatment preferences, to express values and/or to appoint a substitute decision-maker2 (see box 1). Completion of an ACD is considered to be a key component of ACP because ACDs provide written support and guidance for clinicians and family members when making medical decisions on behalf of the person.1 8 9
Legal context in Victoria, Australia
Medical Treatment Act 198835
A competent person can refuse treatment in relation to a current condition by completing a legally binding Refusal of Treatment Certificate.
A competent person can appoint an enduring power of medical attorney with the powers to consent, refuse and/or withdraw treatment.
Guardianship and Administration Act 198636
Person responsible, identified as default substitute decision-maker within the Act, can consent to or decline to consent to medical treatment; however, they cannot refuse treatment.
When making a decision, the substitute decision-maker must act in the best interests of the person for whom they are making decisions. The paramount factor in determining the person’s best interests is the wishes of the person, so far as they can be ascertained.3 37
A person can complete a non-statutory advance care directive (inclusive of values and treatment preferences). It should be signed, dated and witnessed.37 The legal standing of such documents has not been tested with Victoria, Australia. However, it was expected that given case law in New South Wales38, common law directives would be upheld in Victoria.
Within the health service, there are policies relating to advance care planning, informed consent (including where the patient lacks capacity) and limitation of life-prolonging treatment. These policies document a clear explanation of the relevant legislation and the legal basis for statutory and non-statutory advance care directives. Information on the identification and role of the substitute decision-maker is also included in hospital policies.
ACP and ACD documentation is part of a broader shift in healthcare to promote person rights and autonomous decision-making. Legal frameworks have supported this evolution, providing greater emphasis on a substituted judgement approach to medical decision-making.10 Substituted judgement requires decisions to be made according to what the patient would have wanted, rather than according to what others assume is in their ‘best interests’. The emergence of ACP can be attributed to the rising value of autonomy in society and increased legal emphasis on informed consent to medical treatment and shared decision-making.11
Research suggests that ACDs are not always adhered to by doctors in clinical practice. A 2007 study found that the presence of an ACD did not influence healthcare providers’ decisions in using life-sustaining treatments or initiating comfort care plans.12 More recently, Australian researchers presented doctors from New South Wales and Victoria with a hypothetical case scenario to investigate compliance with ACDs.13 In responding to the scenario, only 32% of doctors reported that they would adhere to the ACD in accordance with the law. Studies have identified a number of barriers to ACD adherence, including family opposition,14–16 lack of clinical relevance15 17–20 and concerns about the validity of ACDs.15 18 20
The aim of this study is to describe the attitudes and perspectives of doctors involved in the care of patients with chronic disease at an Australian hospital. This study used qualitative interviews and vignettes to explore doctors’ opinions regarding ACDs to gain a deeper understanding about ACD adherence and the utility of ACDs in clinical practice.
Study reporting is based on the Consolidated Criteria for Reporting Qualitative Health Research.21
Participant selection and setting
Participants were doctors involved in the care of patients with chronic disease at Austin Health in Melbourne, Victoria, Australia. Participants were invited to participate through departmental meetings. Purposive sampling was used to ensure participants from a variety of medical specialties and with varying experience levels were included. All doctors who expressed interest in participating went on to participate in interviews. Participants provided informed consent prior to being interviewed.
Face-to-face semistructured interviews were conducted from February to April of 2017 by two investigators (TL and SF) until thematic saturation was reached, defined as the point at which few or no new concepts or themes were arising from subsequent interviews.21 Participants were aware that the interviewers were a medical student and supervisor and some participants were known to the researchers in a collegial capacity, but otherwise, a relationship was not established prior to interview. Interviews were audio-recorded and transcribed verbatim by TL.
An interview guide was developed for this project by the research team (see online supplementary item S1), based on the literature review followed by discussion and consensus among the research team. This guide consisted of three sections:
Demographics and prior experience with ACDs.
Three vignettes (see box 2), each presented with a ‘values’ then an ‘instructional’ ACD.
Questions regarding the perceived usefulness and importance of ACP.
Summary of hypothetical vignettes
Scenario 1: A 73-year-old man with advanced chronic obstructive pulmonary disease and recurrent hospitalisations presents acutely breathless.
Values directive: places high value on his independence and doesn’t want to become a burden to his family.
Instructional directive: not for intubation, intensive care unit intervention or inotropes.
Scenario 2: A 65-year-old previously well man presents with chest pain, then collapses with ventricular fibrillation.
Values directive: he values being an active and productive member of his community. He would ‘rather be dead’ than unable to play golf.
Instructional directive: not for resuscitation, should be ‘allowed to pass away in peace’.
Scenario 3: A 68-year-old man on dialysis, with a history of metastatic pancreatic cancer, presents febrile, tachypneic and hypotensive.
Values directive: he values life above everything—‘God will take me when I’m ready’.
Instructional directive: not for intubation or cardiopulmonary resuscitation (signed 5 years ago).
Additional information: patient’s family disagree with directive and demand full treatment to be given.
Values directive is defined as a general statement about a person’s preferences and values.30 Instructional directive is defined as an express statement in which a person may either consent to or refuse a particular medical treatment, which takes legal effect if the person loses decision-making capacity.30
In section 2, participants were asked to outline their medical management plan for each case scenario. Extracts from ACDs were then used to determine whether the doctor would adhere to the ACDs in each scenario and how, if at all, the ACD would influence their management plan. Probing questions were then used to explore participants’ perspectives with regards to ACD adherence and related issues, such as legal concerns and family input.
Transcripts were entered into HyperRESEARCH (Researchware), a software package used to store and code qualitative data. Drawing from principles of grounded theory and thematic analysis,22 NM inductively identified concepts in the transcripts and then developed and refined the coding structure with assistance from MS. The analysis identified relationships between themes to develop a thematic framework to capture all concepts about participants’ experiences and perspectives about ACD adherence. To further develop and enhance the analytical framework, interpretation of data and preliminary themes were discussed among the research team.
Patient and public involvement
There was no patient or public involvement in the design and conduct of the study.
Participant characteristics are provided in table 1. In total, 21 doctors participated, from a broad range of medical specialties, with a range of experience levels. About half were women (10/21). Almost all participants (19/21) reported having used ACDs in their clinical practice. Interview duration ranged from 12 to 52 (mean: 26 min) min.
We identified four themes: aligning with patient preferences, advocating best interests, establishing validity and translating written preferences into practice. The themes were relevant to all scenarios unless otherwise specified. Selected quotations to illustrate themes are provided in table 2. Conceptual links among themes are presented in figure 1.
Aligning with patient preferences
Avoiding unwanted care
Doctors valued ACDs that supported them to understand ‘what the patient wants and why they want it’ to be able to individualise treatments according to the patients’ preferences. They believed that ACDs can help them to make ‘the right decision for that particular patient’ and avoid unwanted treatment. For instance, some doctors felt that ACDs provided patients with an opportunity to have a say about ‘wanting or not wanting a certain kind of treatment’, after considering past treatment experiences.
Nonetheless, some doctors experienced deep conflict when weighing up decisions to follow ACDs, in scenarios where they disagreed with the patient’s decision. This was because they felt ‘morally tied’ to follow decisions to refuse treatment, because they believed it was important to respect the patients’ autonomy. This was most evident in scenario 2, involving a previously well patient presenting in a sudden cardiac arrest, who had an ACD which refused resuscitation. Even though participants acknowledged that the situation was very difficult given the patient would likely have a good prognosis if treated, some felt obligated to adhere to the ACD, stating that ‘if you’re going to allow people autonomy to decide things … you can’t say, we will sometimes, we won’t sometimes’. These doctors felt that people deserve to maintain their own autonomy, provided the patient has made an informed choice as a competent adult. Thus, despite their reservations, they decided that they would adhere to the ACD and withhold resuscitation, reflecting that it would be ‘unethical to give (the patient) something that (they) never wanted’.
Anticipating family opposition
Doctors described potential difficulties of overcoming family opposition when making decisions based on patient preferences in ACDs. This was particularly evident in scenario 3 in which participants were asked how they would respond if a family member requested more treatment to be given than stated in the ACD. Some doctors felt that although they would ‘try to accommodate’ the preferences of the family, they would draw the line at ‘causing harm to the patient’. One participant argued against permitting family to override the ACD because the patient has ‘the right to make the decision’ and ‘family don’t necessarily play a part in that’. However, other participants reflected on scenarios from their own clinical experience where family members had requested more treatment than the patient wanted and doctors had felt obliged to comply with requests to avoid causing stress to the family.
Advocating best interests
Defining futile care
Doctors considered ACDs to be irrelevant to their decision-making if patients had requested treatment that the doctor perceived to be futile. Doctors felt it was their duty to determine ‘what (was) futile’ to avoid ‘unnecessary and potentially harmful treatment’. Because of this, some doctors refused to follow the treatment requests specified in the ACD, such as for cardiopulmonary resuscitation, intubation or mechanical ventilation, because they believed that treatment would not be in the patient’s ‘best interests’. For example, in scenario 3, a very unwell patient with a poor prognosis had a values ACD that requested all available treatment to be given. Participants decided not to follow the ACD, identifying that ‘any advance care plan for treatment becomes irrelevant, because giving him treatment would be futile’.
Relying on clinical judgement
In contrast to those participants who prioritised patient autonomy in decision-making, some doctors appeared most influenced by their assessment of the potential risks and benefits to patients rather than the preferences specified in the ACDs. They considered factors, such as the patients’ age and comorbidities, to make a judgement about prognosis and whether the condition was potentially reversible. As a result, some doctors refused to follow the ACD and instead proceeded with treatment because they felt it was not ‘right to withhold treatment especially when, based on your medical assessment, there is a great potential for reversibility’. This was best illustrated in scenario 2, where some doctors indicated that they would be unwilling to withhold resuscitation on a relatively young, well patient, despite the patient having a valid ACD which stated that they did not want resuscitation. Even though these doctors were aware of the patient’s preference to ‘pass away in peace’, they felt compelled to override the request because ‘(the patient) has an excellent prognosis … So you would be mischievous not to treat him’.
Rejecting unreasonable decisions
Some doctors were reluctant to implement ACD requests that they considered to be unreasonable, or inconsistent with what a ‘normal’ person might want. Doctors judged whether they thought the preferences specified in the ACD were ‘sensible’ or acceptable for the situation, and some were unwilling to comply with those that they perceived as unreasonable or ‘unusual’. These doctors were willing to respect patients’ treatment choices but only if they conformed within the scope of treatment options that was felt to be reasonable or appropriate in that situation.
Disregarding legal consequences
Overall, doctors appeared to have minimal concerns regarding potential legal consequences of not following ACDs, choosing instead to prioritise what they believed to be the patients’ best interests in their decision-making. One participant explained that ‘the legal stuff never worried me. It’ll be more about the moral thing to do’. Doctors were more motivated to act in the patients’ best interests, rather than uphold the ACD because of a legal obligation. Doctors dismissed the risk of legal consequences, believing that they would be protected if they used ‘reasonable judgement’ that could be justified in a way that their peers would agree, and if ‘the motivation for not following the plan was one of trying to do the best for the patient’.
Doubting rigour of the decision-making process
Some doctors expressed concern because they were unable to verify whether a ‘proper’ ACP discussion had occurred in the formation of an ACD, as they had not been ‘part of the (ACP) process’. Therefore, they could not verify if the patient preferences specified in the ACD were accurate and/or informed. Doctors regarded best practice ACP as a formal discussion with a trained health professional in which issues could be explained, understood and discussed. Some doctors expressed scepticism regarding the rigour of ACP discussions in practice, and therefore felt justified in dismissing any preferences specified in the ACD. For example, these doctors believed that ACP is sometimes completed as a ‘tick box’ where important details about treatments are not discussed.
Questioning patients’ ability to understand treatment decisions
In considering treatment preferences specified in ACDs, some doctors questioned whether patients were capable of ‘fully understand(ing) the consequences’ of their decisions, because they believed that the general population lack required medical knowledge about the risks and benefits of specific treatments. One participant explained that ‘patients aren’t expected to understand all the minutiae or the nuances of what we can do’, especially given the fact that ‘every single sickness they have is going to be individual and different’. As a result, doctors suggested that medicalised instructional ACDs were not appropriate by themselves; instead, these doctors expressed a preference for ACDs that included the patient’s values because these ACDs allowed the medical team to interpret the patients’ preferences in their own words.
Distrusting outdated preferences
Doctors also raised a concern about adhering to ACDs that were created several years ago because they believed that the preferences specified in the ACD were potentially outdated. They questioned whether the patients’ views may have changed over time, especially if their health status had changed. Changes in the patients’ circumstances that doctors believed may influence treatment preferences included receiving a new diagnosis or recovering from ‘acute unwellness’. In contrast, some participants felt comfortable adhering to ACDs that were several years old if they judged it to be appropriate for the current situation. Some felt that if the ACD was already ‘conservative to start with’ and the patients’ prognosis had worsened, then they would still feel confident adhering to the ACD because the patients were unlikely to have changed their mind to opt for more aggressive treatment.
To overcome concerns about the validity of ACDs, doctors believed that they needed to seek confirmation of the ACD to establish its authenticity, such as from family members, before they would be willing to let it influence their management plan. This appeared mostly of concern when doctors disagreed with the preferences specified in the ACD, or if they believed the patient’s treatment preferences were not in line with a reasonable medical course of action. Consequently, if they could not clarify the reasons underlying the preferences specified in the ACD, some doctors decided not to adhere to the ACD.
Translating written preferences into practice
Contextualising patient preferences
While doctors felt that ACDs are useful to help to guide decision-making, some also highlighted the need for each ACD to be considered ‘in the context of what’s happening today’, rather than applied literally in every case. Doctors believed that they should have some leeway to determine whether the ACD was intended to apply to the given clinical scenario, feeling the ACD should be used as ‘a guide, rather than something that is compulsory’—allowing the clinician to have a more ‘nuanced’ interpretation of the patients’ request. Some doctors believed that ACDs are unable to address every possible clinical scenario that might arise and therefore, clinicians should be able to ‘exercise judgement as to whether this will be the situation where values might be important to help to guide (them)’. For this reason, some participants expressed a preference for ACDs that conveyed patients’ preferences in terms of broad values statements because they were ‘more general’, and therefore usually ‘more flexible and more applicable across multiple scenarios’.
Applying subjective terminology
In contrast, doctors sometimes struggled to apply values-based ACDs because they were unsure how to interpret values statements. Doctors felt that values expressions were ‘vague’ and ‘confusing’ because they are subjective. For example, some doctors questioned the meaning of the word ‘independence’, with one participant stating that ‘how I interpret independence is different to how someone else might value his independence’ because ‘values are different from person to person’. Doctors expressed difficulty knowing how to apply values to form a concrete treatment plan and therefore, some participants regarded instructional ACDs as being easier to interpret and apply because they have ‘clear-cut’ directions.
Prioritising emergency medical treatment
Some doctors noted the importance of providing urgent medical treatment in a timely manner, and this was seen as a barrier to ACD adherence because there is not always time available to locate, read and interpret ACD documentation. Doctors reported that aggressive treatments are sometimes administered without consideration of patients’ preferences because of the need to begin treatment immediately. Additionally, some doctors felt ill-equipped to make difficult decisions about end-of-life care, especially in time-pressured situations. Participants explained that in an emergency, they felt obliged to ‘err on the side of … treating’ because ‘you certainly don’t want to do it the wrong way around’.
This qualitative interview study provides evidence regarding the attitudes and perspectives of doctors in relation to ACP and ACDs by exploring vignette scenarios and medical decision-making. The majority of participants valued ACDs that supported them to align with patient preferences in their treatment decisions. However, some doctors experienced conflict in adhering to ACDs when they believed that the ACD did not represent the patients’ best interests. Doctors tended to rely on their own clinical judgement to make treatment decisions, sometimes overriding the ACD on the basis that the treatments requested were ‘unreasonable’, ‘futile’ or that the patients’ condition was potentially ‘reversible’. In addition, there were issues with validity and currency of ACDs, subjective terminology, patients’ ability to understand consequences of preferences, family opposition and time pressures. In this study, consideration of what doctors perceived to be the patients’ best interests was found to be more influential in medical treatment decision-making than legal requirements.
From an ethical perspective, these findings challenge the widely held assumption that autonomy is the most important principle in healthcare decision-making.23 Some participants preferred to follow their clinical judgement to make decisions that were in the patients’ best interests, rather than uphold patient autonomy, although others felt obligated to adhere to patients’ preferences despite this conflict. This is consistent with another vignette interview study conducted in the UK by Thompson et al,24 who also found that some doctors wished to override the ACD and proceed with treatment consistent with the patient’s best interests, whereas others believed that it was most important to respect patient autonomy. Our findings reinforce past Australian and international qualitative studies demonstrating that, while doctors appear to have a shared understanding and respect for patient autonomy in theory, competing values and interests prevent doctors from prioritising patient autonomy in their clinical decisions in practice.17 24 25 Doctors in an interview study by Johnson et al 25 felt that there should be limits to patient autonomy, and patient preferences should not be respected if the treatment choice is ‘unreasonable’, ‘futile’ or ‘medically inappropriate’.
Doctors in this study reported validity concerns to be a key barrier to ACD adherence, in accordance with past Australian and international research findings.15 18 20 24 Doubts about validity, including currency of decisions, were particularly prominent where participants disagreed with the treatment decision specified in the ACD. For instance, when the ACD refused treatment that doctors felt would be beneficial, they tended to question the authenticity of the ACD and wished to establish its validity, such as by confirming with family members. This is consistent with a model of end-of-life decision-making proposed by Willmott and White, suggesting that ‘the further the decision departs from responsible medical opinion, the clearer the evidence must be that this is what the adult wanted’.26 On the other hand, participants in this study did not tend to raise validity concerns if they agreed with the decision in the ACD or thought it was appropriate for the situation. This suggests that doubts about validity were raised as a justification for overriding ACDs. Many participants appeared to only question the validity of ACDs when they felt that the patient’s choices were inappropriate or unreasonable for the situation.
Moreover, in this study, some doctors questioned the ability of patients to accurately predict what they would want in future unknown circumstances. This is consistent with international research evidence suggesting that prognostic uncertainty is a barrier to the creation of useful, clinically relevant ACP documentation.27–29 Some participants appeared to support the idea of using different ACD formats in tandem; instructional ACDs could be used for patients with an existing diagnosis and a predictable disease trajectory to make advance decisions about anticipated situations. Values ACDs could provide a broader context for the decision and act as general guidance for substitute decision-makers. Further to this, legislation governing ACD adherence in Victoria states that doctors do not need to uphold ACDs that are not considered to have been intended to apply to the clinical scenario at hand.30 This provides for some leeway in doctors interpretation of ACDs in clinical practice.
Interestingly, doctors in this study showed minimal concern about the potential legal consequences from overriding a patient’s ACD, because they believed that making an ethical decision was the most important outcome for patients. This is concerning given that legal frameworks are moving towards greater support for patient autonomy, as reflected in current legislation30 and the Medical Board of Australia code of conduct.31 Doctors in this study may not have been aware of relevant law, and the consequences of not following the law, which include civil and criminal liabilities.32 Past research evidence examining compliance with ACDs has been mixed, with some survey studies in Australia33 and overseas34 describing high levels of self-reported adherence of doctors with ACDs. However, unlike the current study, these surveys did not test doctors’ adherence using sample vignette scenarios, and thus may be less reflective of actual clinical practice. In contrast, our study suggests that doctors prioritise ethical considerations and clinical judgement above law in decisions to follow ACDs. In addition, studies have demonstrated varying levels of legal knowledge among health professionals, and have concluded that more training and education is needed to address this gap in knowledge.33
Strengths and limitations
In this study, we used interviews and case vignettes to gain a deeper understanding about doctors’ decision-making. We applied purposive sampling to recruit a cohort of doctors with a diverse range of experience levels and specialty backgrounds, from a hospital with a well-established ACP programme. The majority of participating doctors had personal experience with the use of ACDs in their clinical practice, which strengthened our results by providing a real-world perspective on the utility and application of ACDs in clinical practice. However, doctors were recruited from a single metropolitan health service in Australia, which may limit the transferability of findings to other settings.
Implications for practice, policy and future research
Our research highlights the importance of supporting patients to create relevant, up to date and clear ACDs that are clinically useful for doctors. In addition, this study suggests a need for an integrated approach to ACP, including early involvement of doctors and involvement of substitute decision-makers in discussions. This may minimise the likelihood that doctors will mistrust the validity of the ACD, or support family to override the patients’ preferences. ACP implementation should support increased ethical and legal education and training for health professionals to translate legislation into clinical practice. Future research may seek to track the influence of recent legislative change on ACD adherence; since these interviews were conducted, the law in Victoria was updated to provide greater legal support for ACDs, which legislates substituted judgement decision-making.10 In addition, future research could complement the current findings by characterising clinicians’ perspectives and experiences regarding the role of family and legally appointed substitute decision-makers.
This study provides insight into doctors’ attitudes and perspectives about the clinical utility of ACDs. Doctors found ACDs most useful when they were current, relevant to the scenario, consistent with clinical judgement and contained what doctors considered to be ‘reasonable’ choices. Doctors were less supportive if an ACD conflicted with their treatment plan or if they doubted its validity. Consequently, some doctors decided to override the preferences in the ACD based on judgements that doing so was consistent with the best interests of the patient; whereas others felt morally obligated to respect the patient’s autonomy, even if they disagreed with the decision. Future ACP programmes should consider approaches to improve the validity and applicability of ACDs. In addition, there is an ongoing need for ethical and legal education to support doctors’ knowledge and confidence in ACP and enacting ACDs.
Contributors Research idea and study design: KMD, SF and TL; data acquisition: TL and SF; data analysis/interpretation: NM, MS, TL, SF, KMD and LN; thematic analysis: NM and MS; supervision or mentorship: SF and MS. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval Ethical approval was obtained from the Austin Health Human Research Ethics Committee (LNR/16/Austin/528).
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement No data are available.