Article Text

Smokers and non-smokers talk about regulatory options in tobacco control
1. Stacy M Carter1,
2. Simon Chapman2
1. 1Centre for Values, Ethics and the Law in Medicine, Central Clinical School, Faculty of Medicine, The University of Sydney, NSW, Australia
2. 2School of Public Health, The University of Sydney, NSW, Australia
1. Correspondence to:  Dr Stacy M Carter  Centre for Values, Ethics and the Law in Medicine, Central Clinical School, Faculty of Medicine, The University of Sydney, NSW, Australia 2006; carters{at}med.usyd.edu.au

## Abstract

Objective: Community members are occasionally polled about tobacco control policies, but are rarely given opportunities to elaborate on their views. We examined laypeople’s conversations to understand how 11 regulatory options were supported or opposed in interactions.

Design: Qualitative design; purposive quota sampling; data collection via focus groups.

Setting: Three locations in Sydney, Australia.

Participants: 63 smokers and 75 non-smokers, men and women, from three age groups (18–24, 35–44, 55–64 years), recruited primarily via telephone.

Measurements: Semi-structured question route; data managed in NVivo; responses compared between groups.

Results: Laypeople rejected some regulatory proposals and certain arguments about taxation and the cost of cessation treatments. Protecting children and hypothecating tobacco excise for health education and care were highly acceptable. Plain packaging, banning retail displays and youth smoking prevention received qualified support. Bans on political donations from tobacco corporations were popular in principle but considered logistically fraught. Smokers asked for better cessation assistance and were curious about cigarette ingredients. Justice was an important evaluative principle. Support was often conditional and unresolved arguments frequent. We present both sides of these conflicts and the ways in which policies were legitimised or de-legitimised in conversation.

Conclusions: Simple measures of agreement used in polls may obscure the complexity of community responses to tobacco policy. Support was frequently present but contested; some arguments that seem self-evident to advocates were not so to participants. The detailed understanding of laypeople’s responses provided through qualitative methods may help frame proposals and arguments to meet concerns about justice, effectiveness and feasibility.

• ETS, environmental tobacco smoke
• NRT, nicotine replacement therapy
• SES, socioeconomic status
• public policy
• qualitative research
• evaluation
• smoking
• tobacco control

## Statistics from Altmetric.com

Comprehensive tobacco control is advanced in Australia. Mass media cigarette marketing is banned, although below-the-line marketing continues.1 Approximately 55–60% of the retail price of a cigarette is excise2: 25 cigarettes, the most common pack size, cost approximately $A10.00 ($US7.50, €6.50). Smoking is banned in many public indoor spaces. Quit advertising3 was extensive in the late 1990s, although it dwindled in the early 2000s.4 Novel policy proposals have begun emerging from Australian advocates.5–7

However, Australian politicians and bureaucrats have expressed the view that tobacco control is no longer politically salient,8 which led us to question whether similar views are held by lay Australians. Research into community responses to tobacco control has rarely extended beyond opinion polls. These suggest that smoking status influences the attitudes of individuals,9,10 as do broader community attitudes about tobacco control.11 These in turn may vary according to a community’s general political orientation, socioeconomic conditions, cultural composition and legislative context.12 Survey research suggests increasing support for tobacco control, particularly when programmes are framed as preventing youth smoking. Enforcing youth purchase laws is often popular. Other measures such as increased taxes and new smoking bans are less well supported, although support for smoking bans grows after restrictions are introduced. Respondents are mostly negative about the tobacco industry.13–27

However, survey research cannot illuminate how laypeople talk, positively or negatively, about different regulatory options. This paper presents a detailed analysis of conversations between groups of smokers and groups of non-smokers to address this gap, examining the ways in which different regulatory options were discussed as acceptable or unacceptable.

## PARTICIPANTS AND METHODS

We approached this study from a constructivist position, that is, talk was taken to create as well as represent reality and to be a product of the specific interaction and context.28 We did not set out to measure individual attitudes or perceptions: instead our aim was to understand how people negotiated regulatory options in groups.29

A purposive quota sampling strategy was used to identify 138 participants to attend 20 groups (table 1). Participants were invited to attend via telephone calls to their homes (numbers were randomly selected from local telephone directories). This strategy was intended to maximise diversity rather than to produce a representative sample. Participants without tertiary education were recruited from two areas: one a moderately-low socioeconomic status (SES) semi-industrial inner-urban area, another a moderately-low SES outer-metropolitan area. A smaller number of participants with (or enrolled in) tertiary education were recruited from a moderately-high SES inner urban area.30 This was intended to ensure that a variety of socioeconomic experiences were represented in the groups. Groups were held in the area of residence of the participants, to minimise inconvenience to them. Participants were offered \$50 to compensate their time and travel, and informed that the group would discuss “smoking, tobacco policy and the tobacco industry”. In keeping with our ethical responsibilities to provide safe group environments, each group was homogenous for sex, age and smoking status. Self-defined current smokers were kept separate, self-defined non-smokers and ex-smokers were combined. Recent ex-smokers (less than 12 months) were excluded.

Table 1

Group constitution and patterns of attendance

A semi-structured question route was used. The first half of the group was devoted to discussion of social aspects of smoking; the final 45 minutes concentrated on a set of regulatory/policy options, presented in a Likert scale questionnaire, which participants completed alone (table 2). The question route and discussion stimulus were developed through iterative email and face-to-face consultation with local peer researchers and leading tobacco control researchers, advocates and bureaucrats from Australian health organisations. We emphasise that the scale was not designed to be a quantitative data-gathering tool and individuals’ responses on the questionnaire were not analysed. This research aimed to examine interactions, not individual constructs, and our sample was not statistically “representative”. The questionnaire was designed to provide individual thinking time for participants before discussion commenced, a technique that expert focus group researchers recommend to increase participants’ confidence and provide better quality discussion.31 The four main questions put to the participants about the questionnaire are in table 3. As shown, the moderator led a discussion of positive responses first, to ensure that negative responses did not overwhelm the interactions.

Table 2

Discussion stimulus

Table 3

Questions and probes used to investigate participants’ responses to regulatory options

The idea of “saturation”, although somewhat contentious, features in many qualitative research traditions.32 “Saturation” is generally said to be reached when responses become repetitious and no new insights are emerging. Because qualitative research aims to understand, rather than to measure frequency, it is considered inefficient to continue asking the same question beyond the point of saturation. Question 10a stimulated consistent strong negative responses with the same objections raised each time—saturation—early in data collection. For the final six groups, 10a was removed and 10b substituted to allow detailed exploration of an issue raised frequently by earlier participants. This iterative relationship between data analysis and data collection is another common feature of qualitative research.33

The moderator (SMC) corrected and analysed the transcripts and substituted pseudonyms for all names. NVivo was used for data management.34 An adapted, constructivist form of the constant comparative method was used for analysis, first fragmenting the data and then re-connecting subsets of text, both to their original context and to one another by comparing cases.35,36 First, concrete, non-mutually-exclusive codes were developed iteratively from the data, coding whole sections rather than at sentence level. These included a code for each of the regulatory options. The corpus tagged by each code was then examined more closely, comparing (not combining) the talk of different groups (by age, sex, SES and smoking status) and creating an audit trail, using memos, to map and compare constructions and negotiations.

The rigour of this study37 is demonstrated in several ways, including our detailed exposition of methods. Speaking with both smokers and non-smokers allowed for comparison of two groups’ accounts of one another and their experiences. In data collection, the moderator (SMC) repeatedly asked for differences of opinion, avoided value judgements and encouraged all participants to speak. In analysis she recognised that these data were constructed in particular contexts (including the framing of the questions) and constantly sought diversity and outliers as well as recurring patterns, and less forceful as well as dominant voices. Detailed audit trails were kept.38

## RESULTS

Although our sampling strategy included people from a range of educational and economic backgrounds, SES produced little variation in the positions expressed (variation occurred only in cultural and communication conventions—for example, the vernacular used). For this reason, the following discussion generally only contrasts people from different age groups or smokers with non-smokers.

### Option 1: make tobacco companies put all the money they make from teenage smoking into smoking education campaigns run by health groups

Many groups recognised that school education about smoking already existed, and often said it indoctrinated young children to confront adults who smoked. While smokers sometimes resented this pressure, nobody wanted children to smoke, and most wanted “more done” in schools. Two principles supported this option: that “innocent” children deserve protection, and that education is a universal good. However, these were often qualified: in short, it would be nice if education “worked”, but frequently it did not, an argument that resonated with the views of many tobacco control researchers. Participants said that young children were easily turned against smoking, but some would, unavoidably, later commence smoking. Once children started smoking, anti-smoking sentiments would simply inflame their rebellious desire to smoke. Thus education should start early, but some failure was inevitable.

The wording of this regulatory option presumed that experts could quantify profits from teenage smokers. Opponents challenged this, often with derision. Youth smoking was sometimes considered unmeasurable; alternatively, some participants argued that abstract statistical estimates were inherently untrustworthy and could be manipulated, undermining the whole enterprise:

Don (35–44, non-smoker): I don’t believe in statisticians anyway! If you say one said 30%, I can make one say 50%!

Other arguments were also raised. The tobacco industry would refuse to pay, reasoning that underage sales were illegal. The option was illogical, as the industry existed to sell more cigarettes rather than fewer. It might improve tobacco corporations’ public image or give them access to young people via the campaigns their money funded. Some rejected the involvement of health groups, or attributed responsibility to parents rather than schools, or governments rather than corporations: some said the already ample tobacco excise should fund education. Some felt corporations would unfairly pass the cost on to consumers.

### Option 2: ban tobacco company sponsorship of political parties

There was some uncertainty about what this meant, some surprise that tobacco companies donated to political parties, and some confusion about political processes (for example, some said that bans on corporate donations would somehow lead to politicians extricating the money from citizens.) A proportion of participants said they didn’t care: most of these said it was irrelevant to them. Opponents generally argued that all industries were equal, or that politicians should act independently notwithstanding political donations. However, overall, this option was popular. Participants disagreed over whether donations bought favours, but were often cynical, arguing that governments should rescind such avenues of potential influence. Even participants generally opposed to tobacco regulation sometimes condemned donations, framing the combination of regulation and donation as supremely hypocritical. Many said politicians would never implement this option; or companies would circumvent any ban by donating to individual politicians or via their sister companies. However it was generally agreed that if a political party, improbably, was upright enough to ban political donations, the ban would be popular and suggest sincerity. It was sometimes argued that all morally questionable industries should be similarly regulated (alcohol and arms corporations were frequently named.)

### Option 3: stop tobacco companies adding good-tasting things like chocolate and sugar to cigarettes

Smokers, particularly in the middle and older age groups, were fascinated by cigarette ingredients and their effects. Some participants said smokers became sick because of “chemical” additives, not “natural” tobacco; most groups said chemicals were added to cigarettes to increase addiction, harm and/or consumption, especially “jet fuel”, which was commonly said to be added to speed burning. “Chemicals” were a potent touchstone for condemnation of tobacco corporations. However, the idea that cigarettes contained sugar and chocolate (as opposed to “chemicals”) was more surprising. Discussion commonly commenced with amused, incredulous or angry questions about veracity:

Michelle (35–44 smoker): Can we talk about the chocolate thing? [laughing] I mean I can’t believe…  Aamir (18–24, smoker): Chocolate and sugar really?  Luke: Never heard of it.  Aamir: I don’t know of these cigarettes.  Liam: Do they add chocolate and sugar to cigarettes?  Aamir: Personally, I’m shocked.

It was then frequently suggested that flavoured cigarettes were niche products, after which the moderator explained that most manufactured cigarettes have complex flavour formulae added. This was a genuinely novel insight for most smokers, and ensuing arguments contained points for and against removing flavours. Participants argued that flavours made it easier for children to smoke, were a form of marketing to children, or made it easier to smoke too much. Others countered: they enjoyed their cigarettes the way they were and did not want them to taste bad, or conversely, no matter how bad cigarettes tasted, children and smokers would still smoke them. Some smokers supported unflavoured cigarettes because they should logically be cheaper (the additives must cost money) or less harmful or addictive. Perceptible and imperceptible flavours were contrasted: “lolly” (confectionary) cigarettes or overtly fruit-flavoured cigarettes were censured, but for some smokers the revelation that we were discussing their cigarettes—ordinary, “cigarette-tasting” cigarettes, not strange, “chocolate-and-sugar-tasting” cigarettes—diminished interest. Regardless of participants’ positions on flavour removal, most were strongly in favour of disclosure, preferably on or in the pack. No-one knew ingredient information was available on a government website,39 suggesting this is a highly ineffective means of communicating with smokers.

### Options 4, 5, 8: put cigarettes and tobacco under the counter at shops where they cannot be seen; ban cigarette and tobacco sales from supermarkets and convenience stores—only sell at licensed tobacconists; sell cigarettes and tobacco in plain cardboard boxes with only the brand name and the health warning

Participants generally discussed these three options simultaneously. Several groups suggested that none of these options would make much difference alone, but might work in conjunction with other policies. In many conversations, the discursive tussle was over efficacy. Would these measures make any difference? Did marketing and distribution affect uptake or consumption? Arguments were common and rarely resolved.

Some argued that children were attracted to shiny packages; others that plain packages and under-counter storage would produce an illicit, pornographic aura, increasing appeal for young people and stigma for adults. Some argued that inconveniencing smokers would reduce consumption, others that it was a pointless punishment—especially for those who were old, without transport, or in low-density housing and rural areas. Some argued controlled outlets would reduce underage sales, making it easier to catch offending retailers, others that children would always find a way, so further control was pointless. For some, retail display bans and plain packs were a logical extension of existing marketing restrictions; the ready availability of cigarettes was inconsistent with the drive to reduce consumption. Some argued that, as alcohol retailers are licensed and controlled in Australia, so should tobacco retailers be. It was often suggested that too much pressure on supply would expand the black market.

Some smokers said restrictions would reduce their consumption, others said they would stockpile. Smokers argued that shops selling cigarettes should be clearly identified and comparative price information displayed, and worried that reduced competition would increase prices. Some said plain packs should decrease printing costs and thus retail prices. Smokers also wondered how they would purchase cigarettes in the new regimen, as sales assistants frequently did not smoke, and smokers now often had to point and shout out pack colours until the right product was selected.

### Option 6: ban smoking in bars, pubs and clubs

It is well documented that smoking bans are more popular after their introduction than before. Bars, pubs and clubs (henceforth venues) permitted smoking at the time of data collection. Smoking in venues was considered “natural”(SM Carter, unpublished data, 2006). This was used to oppose bans in familiar ways: because venues were “naturally” smoking environments, venues would fight bans, smokers would stay home, ignore bans, and/or become militant, and businesses and the pub-based entertainment industry would fail. This was argued more commonly by smokers and younger participants, although some non-smokers agreed. Smokers more often advocated technical solutions, such as ventilation and smoking rooms, which are rejected by public health advocates.

Bans were only occasionally justified on the grounds of reducing consumption; it was more commonly argued that non-smokers should not be exposed to smoke. Non-smokers and older groups argued this more often, but so did some brave, isolated smokers in middle and older age groups; some smokers also complained about their own environmental tobacco smoke (ETS) exposure in venues. Conversely, some non-smokers argued that excluding smokers from venues was unfair. There was minimal support for outdoor bans. Employees were a common concern, particularly in middle and older age groups, but the issue was rarely resolved. According to some, bar workers should be smokers, accept smoke as a working condition, or sign legal waivers to protect their employers. For others, ETS exposure was an occupational health issue, bar workers often had few employment choices, and even smoking bar workers were affected by ETS. Many groups argued that laws were needed to set out the rules and thus protect bar owners from litigation. Although some said bans were improbable, many groups accepted that the incremental, inexorable strengthening of smoking restrictions would eventually reach venues.

### Option 7: ban smoking in cars carrying children

This option united all but a few renegades in agreement. Children were still growing, thus potentially more vulnerable to smoke; they might be more likely to smoke if adults smoked around them; children were trapped in cars, which were small spaces, unlike houses. Even the most strident anti-regulation smokers generally supported this option. Some, including smokers, argued that smoking in cars should be banned altogether because it reduced drivers’ concentration. While almost everyone agreed with this option in principle, some said it could not be enforced, and occasionally regulation of behaviour inside private space was rejected; these concerns were often successfully answered with existing seatbelt laws and bans on mobile phone conversations while driving. Participants joked that police would welcome the new revenue source resulting from fines on violators.

### Option 9: increase taxes so that cigarettes are more expensive

Increases in tobacco tax were discussed as inevitable and constant, and evoked deep cynicism, being seen as a substantial conflict of interest for governments that undermined their moral right to extend tobacco control (SM Carter, unpublished data, 2006). Some non-smokers, and occasional smokers, said price increases (especially large ones) might prevent uptake by children, stimulate adult quitting, fund smokers’ health care or relieve non-smokers’ tax burden. However, effectiveness was the key concern, and the majority of participants, both smokers and non-smokers, scoffed at suggestions that the programmed incremental six-monthly increases in tobacco taxation were intended to stimulate quitting. Experience told them addicted smokers would complain, perhaps reduce consumption briefly, and then adjust, buying less of something else if necessary. Slow price increases simply created suffering, especially among poorer people: this was “a sin [tax…] a punishment tax” (Edward, 18–24, smoker). Cigarette taxes might become morally justifiable if they were hypothecated to prevention and treatment of smoking-related diseases, but instead they went into the black hole of general revenue with no accountability, demonstrating the insincerity of government rhetoric. Price increases were often said to expand the black market or increase crime.

### Option 10a: only allow registered addicts to buy cigarettes or tobacco

Few participants were in favour of this option. In some groups smokers became so angry they could barely articulate their objections. Labelling smokers as “addicts” would make them junkies and further stigmatise them, and was considered extreme by both smokers and non-smokers. Many participants protested that it was impossible to prove someone was addicted, despite generally agreeing smoking was addictive. Often groups suggested a black market would emerge in response. Even moderate smokers such as Peter and Kerry, who courageously advocated for tougher regulation in their respective groups, objected heartily to the unnecessary storing of information, intrusion on personal freedoms, and connotations regarding their moral character:

Peter (35–44, smoker): Alright, well, I’ve got to be registered now to be a smoker? Who holds the register? How much does that cost? No. No. You know, most responsible people, I think, know that you don’t go and buy kids smokes whether they’ve stood outside the shop—“please, mister, can you get me some of these?”—you don’t do it, but to be registered? I couldn’t see where that was going, personally.  Kerry (35–44, smoker): Registered addicts is a terrible idea. I actually said in terms of discrimination of smokers and ostracisation of smokers by society that would make it worse, that you know you’re kind of this special breed of weirdos who are suddenly being shipped off into a corner.

Those who supported the option said smokers should be made into pariahs, or said it might improve control of sales to children, or provide a focal point for services to smokers. Several older smokers suggested it might be a way of providing old, addicted smokers with cheap cigarettes, so that unregistered smokers, presumably children, could be charged very high prices.

### Option 10b: provide more help and support for individual smokers who want to quit

This analysis is based both on responses to question 10b and spontaneous conversations in earlier groups. Unsurprisingly, support was popular in principle, but discussions of help and support were more nuanced. Smokers traded stories of the problems with existing treatments, sometimes suggesting “cold turkey” was the only way. Some ex-smokers characterised treatments as effective, for many smokers they were “useless,” but often, particularly with women, the focus was on horrible side effects. Bupropion, which has been widely prescribed in Australia,40 was a “magic pill” that took away cravings but could give you terrible rashes or make you “space out”; nicotine replacement therapy (NRT) could give you mood swings, hot flushes, nausea, insomnia, nightmares; some participants said it was illogical to replace one source of nicotine with another. Better treatments were asked for.

Furthermore, these treatments cost money. Tobacco control advocates commonly argue that NRT costs less than typical daily cigarette consumption, thus smokers who complain they cannot afford NRT are poorly motivated to quit. Smokers often recounted another argument that they had heard: that treatment costs were justified by the long-term savings resulting from cessation. The abject failure of these arguments to resonate with smokers cannot be overstated, and non-smokers were also nonplussed. Most smokers reported trying and failing to quit, not once but several times. Cigarettes were characterised as enjoyed, reliable, familiar, cheaper than treatments and able to be made cheaper still—if money was short, cigarettes could be rationed, or a single cigarette put out and re-lit. Treatments were sold in large, expensive packs, threatened horrible side effects, and would probably make no difference. Many participants argued there was not enough money in their household budget for both NRT and cigarettes, so cigarettes were chosen. If NRT was chosen, and the quit attempt failed again, the money would be wasted and there would be no cigarettes until payday. Most said cigarette excise should be at least partly hypothecated for free services or financial incentives to help smokers quit and stay quit (such as refunds after a smoke-free period). Smokers expressed interest in services but were often unaware of available services. Some middle and older aged smokers used the Alcoholics Anonymous or Weight Watchers models to request a group support system run by smokers for smokers. Finally, there were a few smokers who agreed that nothing would make any difference; they were either unwilling or unable to quit.

## DISCUSSION

Focus groups produce a particular kind of data—a record of the way in which a group of strangers interacted around a particular issue. They do not provide access to the beliefs or motivations of their participants. However, they are likely to mirror the interactions people might have about regulatory options if they became topical (for example, if raised in the mass media), and to reflect the social, cultural and political values likely to be attached to these options in such interactions.

Some advocacy arguments appeared not to have convinced these Australians, most notably, arguments about taxation and the cost of quit treatments. The literature on cigarette excise consistently concludes that high prices decrease consumption, but also acknowledges that societal values regarding, for example, the use of taxation revenues, are relevant to policies on cigarette price.41 There is more contention around cessation costs in the literature; however, econometric studies suggest a direct relationship between NRT price and consumption,42 and a recent review concluded that decreased NRT price increases use and cessation success, and that the price of NRT, including the lay-down price due to large pack sizes, is a barrier to NRT use.43 As others have noted, the price of NRT is potentially an inequitable means-barrier to quitting, and the argument “if they can afford to use the substance, they can afford treatment” would be unacceptable in relation to alcohol or illicit drugs.7 The consistent concerns expressed here about the immorality of current taxation and treatment policies suggest the need for more careful policy framing.

In quantitative studies, public opinion of tobacco control has improved over time, particularly among ex-smokers and never-smokers, and particularly when tobacco control is framed as preventing youth smoking. However these studies do not reveal the complexity of public opinion about different regulatory options.

This paper is based on discussions between Australian smokers and non-smokers about 11 regulatory options in tobacco control. We discovered that arguments that seem self-evident to tobacco control advocates were not so for participants, whose general support for tobacco control was frequently qualified. Competing principles and propositions from 20 lay conversations are presented to help tobacco control advocates better understand the complexity of community responses to their proposals.

Some proposals would require reframing to be acceptable to these participants: most obviously registering nicotine addicts, removing flavours from cigarettes, and reducing the number of retail outlets. However, many proposals were discussed positively, despite some important caveats, which generally had to do with efficacy, justice and/or feasibility. Bans on smoking in cars with children were uncomplicatedly just, mirroring the high acceptability reported in local survey research.44 Participants said taxation would be more justifiable if increases were substantial and particularly if profits were hypothecated. These ideas are not new, but it is rarely acknowledged that laypeople talk about non-hypothecation as immoral, cynical and unjustifiable. Plain packaging and removal of retail displays were discussed as a logical, relatively inoffensive, but perhaps also ineffective extension of advertising bans. Similarly these participants, like tobacco control professionals, appeared to doubt the efficacy of youth smoking prevention programmes. In both these instances evidence of efficacy may increase acceptability. Bans on political donations from tobacco corporations were popular in principle because they created moral consistency. However, doubt was expressed about feasibility and credibility, particularly regarding circumvention.

Discussions about ETS mirrored media stories published in this location, which pitted economics and “pub culture” against health concerns and workplace equity.45 Communication about new smoke-free legislation should be mindful of issues of justice and fairness. Outdoor smoking bans were generally considered fundamentally unfair, but for indoor bans, emphasising fairness for workers and the non-smoking majority, and acknowledging, rather than dismissing, the loss of smoker’s freedoms, may create a better match with lay people’s framings. Other communication opportunities relate to smokers’ expressed interest in better treatment and services. This suggested room for pharmaceutical innovation, and increased promotion of the effectiveness and availability of existing services (such as quit lines). Finally, smokers appeared genuinely curious about cigarette ingredients and their effects, affording a unique opportunity to get smokers’ attention and provide better information. The technology already exists for the addition of inserts and outserts to cigarette packages; corporations should be forced to use this technology to disclose not only ingredients, but what they do and how little is known about their health effects.

Finally, this paper demonstrates that qualitative research can elucidate some of the complexities of community responses to policy, and provides a qualitative baseline for further study of community responses—for example, to the introduction of indoor bans in venues that will be introduced in Sydney in mid-2007. The arguments raised by these participants should assist tobacco control advocates to prioritise their energies not just to interventions deemed technically effective, but also to address appropriately the potential responses of their target audiences.

## Acknowledgments

Heartfelt thanks to the participants who generously gave their time and conversations to this project. Thanks also to our reviewers, the experts and our departmental peers who commented on the question route, Inshira Khan for acting as assistant moderator, Interviewing Australia for recruitment services, and Nardia Drayton and Margaret Jackson for transcription.

## Footnotes

• Funding: This study was approved by the University of Sydney Human Research Ethics Committee (Ref 7513) and funded by the Australian National Health and Medical Research Council grant number 253657.

• Competing interests: No competing interests

## Request Permissions

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.