Article Text

How do New Zealand youth perceive the smoke-free generation policy? A qualitative analysis
  1. Janet Hoek1,
  2. Ell Lee2,
  3. Lani Teddy1,
  4. Elizabeth Fenton3,
  5. Jude Ball1,
  6. Richard Edwards1
  1. 1 Department of Public Health, University of Otago, Wellington, New Zealand
  2. 2 University of Otago Medical School, Dunedin, New Zealand
  3. 3 Bioethics Centre, University of Otago Bioethics Centre, Dunedin, New Zealand
  1. Correspondence to Professor Janet Hoek, Department of Public Health, University of Otago, Wellington, New Zealand; janet.hoek{at}


Introduction Aotearoa New Zealand (NZ) plans to introduce a smoke-free generation (SFG) policy, alongside denicotinisation and reducing the availability of tobacco products. The SFG has a clear rationale, yet we know little about how young people, those the policy targets, perceive it. To inform policy design, communication and implementation, we explored how NZ youth perceived the SFG.

Methods We undertook in-depth interviews with a sample of 20 youth aged 17 or 18 and explored their knowledge of the SFG, and how they perceived its individual and societal implications. We interpreted the data using a reflexive thematic analysis approach.

Results We identified two overarching themes. The first theme, ‘societal good and protection from harm’, reflected benefits participants associated with the SFG, which outweighed perceptions of lost freedoms. The second theme, ‘privileging personal choice’, corresponded to two small groups within the sample. The first preferred measures they considered less restrictive, such as increasing the purchase age, and some came to support the SFG as they rationalised their views. The second subgroup expressed more entrenched opposition and felt the SFG deprived them of a choice.

Conclusions Young people’s deep reflection on the SFG led most to view it as liberating rather than restrictive. Communications that avoid prompting heuristic-based responses could encourage youth to reflect on the policy and elicit strong support from the group the SFG aims to benefit.

  • public policy
  • denormalization
  • human rights
  • priority/special populations
  • tobacco industry

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. Not applicable.

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:

Statistics from

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.

Video abstract

Disclaimer: this video summarises a scientific article published by BMJ Publishing Group Limited (BMJ). The content of this video has not been peer-reviewed and does not constitute medical advice. Any opinions expressed are solely those of the contributors. Viewers should be aware that professionals in the field may have different opinions. BMJ does not endorse any opinions expressed or recommendations discussed. Viewers should not use the content of the video as the basis for any medical treatment. BMJ disclaims all liability and responsibility arising from any reliance placed on the content.


  • The smoke-free generation (SFG) policy goes beyond purchase age restrictions and frames tobacco as an unambiguously harmful product.

  • Tobacco companies have drawn on freedom and choice arguments to foment opposition to the SFG policy.


  • Young people endorsed positive freedom; most supported the SFG and the priority it places on collective well-being.

  • A minority rejected the loss they saw as implicit in the SFG and saw it as restricting rather than liberating, though some came to affirm the SFG after reflecting deeply on their views.


  • Communications that prompt reflective rather than heuristic processing may avoid instinctive loss-aversion responses and elicit stronger support for the SFG.

  • The SFG aligns with many tobacco companies’ purported goal of ‘transforming’ and supporting a smoke-free world; any companies offering less than full support to the SFG should be held to account for their inconsistency.


To achieve tobacco endgame goals, which aim to decrease smoking prevalence to minimal levels by a specified date, some countries have proposed reducing the addictiveness and widespread availability of tobacco products.1 2 While these measures will accelerate smoking cessation, governments must also prevent tobacco companies from recruiting youth as ‘replacement smokers’ who ensure their future profitability.3–7 Mass media campaigns,8 reducing exposure and access to tobacco products9 10 and tobacco’s decreasing affordability11 have seen youth smoking prevalence fall to historically low levels.12 However, evidence that prevalence varies inequitably across population groups and rises as youth become young adults has prompted calls for measures that will end smoking among all young people.12–15

The smoke-free generation (SFG) policy stops the sale of tobacco products to people born after a specified year and creates a cohort that may never legally be sold tobacco.16 It addresses a limitation of age restrictions (eg, R18 or R21 measures), where each year young people ‘graduate’ beyond the protection these afford.16–18 Over time, and in combination with other endgame measures (eg, denicotinised tobacco), the SFG would see smoking uptake virtually end.16 17 We summarise the SFG’s history in online supplemental file 1.

Supplemental material

The SFG policy addresses two important criticisms levelled at age restriction policies: first, that these measures implicitly frame tobacco use as an adult behaviour, which may create ‘rite of passage perceptions’16 18 19; second, that age restrictions incorrectly imply tobacco use after a specified age is accepted or safe, given the protections no longer apply.16 19 Furthermore, the SFG challenges industry depictions of smoking as a right available to people over a certain age who can allegedly make ‘informed choices’ to smoke.20 21 The SFG recognises that tobacco use is never safe, dismantles perceptions of smoking as a rite or right, and creates what van der Eijk described as a ‘norm cascade’ that establishes smoking as an outdated practice.16 17 19 22 23 By recognising tobacco as innately harmful, the SFG also draws on consumer safety arguments and advances young people’s right to protection from a product that kills two-thirds of long-term users prematurely.24

To date, only Balanga city (Philippines) and Brookline (USA) have introduced SFG policies prohibiting the sale of tobacco products to anyone born after 1 January 2000, though both have faced challenges.25 26 Courts have upheld pre-emption objections raised by the Philippine Tobacco Institute, which receives industry support; despite city advocates exploring further options, they may lack the financial resources to fund litigation.26 The Massachusetts Attorney General rejected similar arguments from Brookline retailers in July 2021, and policy implementation began in September 2021; however, a further case is pending in the Massachusetts Superior Court.25

Tasmania considered an SFG policy, but the proposal lapsed when Parliament was prorogued.27 Finland did not proceed with the SFG because of concerns the measure may be perceived as inequitable and thus represent a constitutional breach.28 Despite these setbacks, other nations have recently announced SFG plans. New Zealand (NZ) proposes implementing an SFG policy that will apply to young people born after 2008,29 30 the Khan review proposed England consider developing an SFG,31 and Malaysia and Denmark plan to end sales of all nicotine products to anyone born after 2005 and 2010, respectively.32 33

NZ modelling research predicted the SFG could halve smoking prevalence within 14 years among people aged 45 and under (relative to those over 45) and reduce inequities in smoking prevalence, although more slowly than denicotinisation.34 35 This work also estimated that the SFG will have a greater impact on projected adult smoking prevalence than reducing tobacco availability by 95% overall or increasing excise taxes by 10% annually until 2025.34 A Singaporean study that compared increasing the minimum age, introducing a SFG policy, increasing taxes and legalising e-cigarettes found the SFG and large tax increases the most effective individual policies for reducing smoking prevalence, though noted the effects were longer term.36

Public support for the SFG has been high, irrespective of age or smoking status. In Tasmania, 75% of the public and 66% of young people (ie, those who would be affected), supported a SFG policy.37 Public opinion surveys in Singapore found both smokers and non-smokers (60.0% and 72.7%, respectively) strongly supported an SFG18; NZ surveys reported strong support (>75%) for an SFG policy among people who smoke or who had recently quit.38 Furthermore, public support often increases as measures are socialised and implemented, as occurred with policies mandating smoke-free bars and restaurants.39

However, tobacco companies have claimed the SFG would restrict young people’s ‘personal freedoms, their right to autonomy in their private lives’ and argued it ‘amounts to age discrimination’.40 Using ‘freedom of choice’ arguments to frame smoking as an ‘inalienable right’ and oppose new policies is a common industry tactic,41–43 but bioethicists and public health researchers have challenged definitions of freedom that depend on maximum choice and minimum regulation.44 They note truly free choice requires autonomy and explain how peer pressure, alcohol use and addiction itself compromise young people’s choices.45–47 Rather than adopting a ‘negative’ view of freedom, which views regulation as removing rights, they promote ‘positive’ freedom and argue that protecting young people from addiction would promote autonomy.44 48

However, cognitive heuristics that lead people to privilege loss over gain, and short-term over longer-term outcomes,49 may further complicate perceptions of the SFG, which supports future freedoms over immediate gains. Evidence young people’s brain development continues into their 20s may make them particularly prone to such cognitive heuristics.50 We know little about how loss-aversion influences young people’s perceptions of the SFG in relation to freedom and choice, or how they prioritise individual and collective well-being. To address these questions, which could shape policy implementation, we explored how NZ youth perceived the proposed SFG policy’s likely impact on freedoms and well-being and probed how they rationalised their views.


Study design

We used a qualitative approach to explore how New Zealanders aged 17–18 years viewed the SFG, which we asked them to assume would come into effect in 2022. Those aged 17 would not be able to purchase tobacco, while those aged 18 would, thus enabling us to probe how those on either side of the ‘boundary’ perceived the policy. We drew on a social constructionist epistemology, which aligned with our interest in participants’ perceptions and anticipated experience of the policy.51


JH and EL recruited participants using social media, community and personal networks, and snowball sampling (ie, where participants passed on study details within their networks). People registered their interest by phone or email or directly via an online survey (see online supplemental file 2), where we collected details of their ethnicity, gender, smoking susceptibility and history, which we used to structure the sample. As recruitment proceeded, we used snowball sampling to promote ethnic diversity and recruit people with varied smoking experience. We offered participants a $30 voucher to recognise any costs they incurred participating in the study.

Supplemental material


JH and EL undertook in-depth interviews using a semistructured guide to explore participants’ knowledge and perceptions of the SFG policy, and their views on how an SFG policy would affect freedom, choices and protection from harm (see online supplemental file 3). We conducted online or in-person interviews (depending on participants’ preference) between November 2021 and January 2022; these lasted between 58 and 74 min. Discussions began to converge after 17 interviews, and we conducted three further interviews to establish information sufficiency (ie, when additional participants did not raise new ideas). With participants’ permission, we recorded the interviews and used an online service ( to transcribe these into anonymous verbatim records that we checked for accuracy.

Supplemental material

Data analysis

JH and EL used a reflexive thematic analysis approach to interpret the transcripts; we followed the six-stage process Braun and Clarke outlined,52 53 including reading and rereading the transcripts several times, and using a line-by-line approach to review and code the data, and identify common metaphors. We wrote field notes memos to record participant and process-related reflections, and used analytical memos to draw comparisons between interviews, record recurring metaphors, and review ideas expressed within and across interviews.

JH and EL first independently reviewed and then compared three transcripts and relevant notes to develop an initial coding framework, which we reviewed while coding the remaining transcripts. We met regularly to contrast our interpretations and develop abstracted themes. LT reviewed a sample of transcripts to consider differences between Māori and non-Māori participants, and how these should be reported. We also reflected on our own positions as non-smokers working in public health and the difference between our world experiences and those of our participants.


We outline our sample’s characteristics before presenting overarching themes we identified: striving for societal well-being and privileging personal choices. We use participants’ comments to illustrate these themes and provide additional quotations in online supplemental file 4; we have assigned all participants pseudonyms and report their age, ethnicity (M=Māori; p=Pacific ethnicity; NZE=NZ European; A=Asian ethnicity; O=other ethnicity) and smoking status.

Supplemental material

Our sample comprised 20 participants; 10 identified as male; eight as Māori and five as Pacific, and six reported currently (five) or formerly (one) smoking (online supplemental file 5 outlines the recruitment process). Table 1 outlines the sample’s demographic profile. We did not observe differences in participants’ views according to their gender or smoking status, but report some differences by ethnicity.

Supplemental material

Table 1

Participant characteristics

Societal good and protection from harm

Most participants, particularly those who identified as Māori, supported government intervention to promote community well-being. They valued societal protection above preserving ‘choices’ some thought illusory, and envisaged future generations free from addiction and exposure to secondhand smoke. Several prioritised a longer-term perspective that considered the ‘bigger picture’ and looked beyond their own interests. Marama argued:

…It’s better for everyone in the long run on… the wider scale and for the future… people should be… open to thinking about those… wider implications. Marama, 17 (M/P, NS)

These participants recognised that freedom had (or should have) limits and noted that functional societies relied on rules and were willing to forgo some choices in pursuit of a greater good. Eddy explained:

I don't really think that’s a really important freedom at all, because I mean, okay, it is important, but banning, I feel like banning smoking has many more benefits, rather than saying, “Oh, it’s freedom, so we should have it, we should keep it. Eddy, 18 (A, NS)

Several contrasted government intervention with addiction and the long-term effects of smoking and concluded that, because addiction removed choice, the government should intervene. Tipene explained the government had good reason to take away the choice to smoke:

It is kind of taking their choice away from them, but it is for a good cause, it is a good reason to do it. Tipene, 18 (M/NZE, NS)

Some went further and saw the ‘good cause’ as a government responsibility. Pera noted:

The government essentially is supposed to keep you safe, and they're not supposed to… make things readily available that are gonna actively harm you. Pera, 18 (M/NZE, NS)

These participants’ focus on the greater good, deep dissatisfaction with the slow declines in smoking prevalence and belief that governments had a responsibility to protect people led some to demand new approaches to address health inequities and protect young people. Hinemoa stated:

I think we kinda owe it to our young people, especially our Māori and Pasifika ones. We need better health outcomes… Like 40% of our Māori and Pasifika smoke…. Hinemoa, 17 (M/NZE, NS)

Drawing on their own experiences, these participants imagined how the SFG could offer others better futures. Leila saw ending the inter-generational transmission of smoking as transformational:

…parents teach you not to smoke when you grow up. But then, yeah… you go and watch them smoking. So, you're like, ‘I want to be like Mum… or Dad.’ One of my sisters goes, ‘I want to be like my parents and…’ Yeah, I feel bad for my sister 'cause… I know she smokes 'cause we all smoke in the house. Leila, 18 (M/NZE, S)

By protecting young people from pressures inherent in their daily lives, Amy believed the SFG could prevent smoking uptake:

…I really got pressured into it, 'cause people that were around my age had a hold of it. And that would stop all of the opportunities for people to have it and be pressured into it, 'cause I feel like majority of the people that smoke were… pressured into it…. Amy, 18 (NZE, S)

Participants also recognised how cognitive heuristics undermined free choices. Anahera noted the challenge of comprehending long-term outcomes when making in-the-moment decisions; she explained:

not many people know the disadvantages so it'll take them a really long time… to find out for themselves. So I think definitely stopping them is a big option just because it takes [so] long to realise the disadvantages of it. Anahera, 17 (P/NZE, NS)

Rather than view smoking as a human right that merited protection, they saw it as antithetical to freedom and requiring regulation:

…this is not freedom. This is giving them a product that damages their health… It’s not something good to do. Ivan, 18 (A, NS)

Viewing the SFG unambiguously as ‘solely about protection’, these participants argued it was not‘really taking away any human rights or anything’ (June, 17, A, NS). Those with lived experience of smoking rejected claims it was a ‘choice’:

Whether it’s the government taking the choice or you being addicted to smokes. You've got no choice either way. If you're addicted to smoking it’s not like you are choosing to go buy smokes, you're going, ‘Oh, I needed a packet of smokes this week’. Leila, 18 (M/NZE, S)

Offering protection from a product that removed choice led some participants to envisage multiple societal benefits. Fetu, who initially saw the SFG as unfair, revised his views, saw benefits amplifying across each new generation, and came to view the measure as utopian:

…it [the SFG] would just change the world in a few years. Like, if you stop… the young, the next generation will stop. Then when they're the leaders of their generation, or generations below them… it just will get better and better and better, the younger they go. Fetu, 17 (P, S)

Privileging personal choices

Participants who presented personal choice arguments either proposed what they saw as more proportionate measures or fundamentally opposed the SFG. The former argued that raising the minimum purchase age, increasing excise taxes or providing more education would allow people to make decisions when they were less vulnerable to social pressures, and thus balancing protection and freedom more appropriately. Kevin supported limiting rather than removing choice:

I think limiting people’s freedom is good. Taking it away isn't…. less choice would be better than no choice at all… it would still give people a choice to do it later in life, but earlier in life… more people are around you and trying to pressure you to do things, and you trying to fit in with, like, groups and stuff… Kevin, 17 (NZE, NS)

Participants who favoured ‘middle ground’ measures thought these would reduce the risk of maladaptive responses and foresaw a risk of ‘forbidden fruit’ behaviours as people asserted their agency:

It’s just sometimes the more you restrict something the more people want to try it. Ben, 17 (NZE, NS)

A minority complained the SFG would reduce their autonomy and rationalised smoking as an individual choice and entitlement, not a societal problem requiring government intervention. While they thought governments should protect people from harms inflicted by others, they believed individuals should remain free to make choices about themselves, even when these could lead to harmful outcomes. Mikaere explained:

I think we should be protected… We shouldn't… worry about people… coming in our house and killing us, that kinda thing… But I don't think we should be protected from… ourselves by laws… [or that] you should be protected from yourself, from addiction, from all that by the government… That’s up to you. Mikaere, 18 (M/NZE, S).

Because they believed young people had access to information about smoking’s risks, these participants argued choices should reside with them:

you shouldn’t be protected from your environment. You should have the information and the knowledge behind you, but you should still experience it yourself. Mikey, 18 (O, FS)

All participants reflected deeply on how to balance individual rights and societal well-being, and some who initially opposed the SFG moderated their views, acknowledged collective rights and looked beyond themselves. Sara illustrates these internal debates:

I think it’s fine, as long as it’s… reasonable, well balanced; but that’s also a very big statement. I think if it’s … communal health, if it’s something that could affect others, then I think it’s okay for the government to take that into their own hands… but people who've formed opinions about wanting to try it, they would feel more restricted…. I feel like we should be considerate of them …. it’s really hard to balance their freedoms with what the majority of people agree with. Sara, 17 (NZE, NS)

Only a small minority consistently and strongly opposed the SFG. These participants thought people had access to information about smoking’s risks and resented the Government ‘making a choice for them’ (Teuila, 17, M/P, NS). They felt the SFG drew arbitrary distinctions, risked infantilising those affected and replaced individual agency with central authority. They prioritised loss avoidance and envisioned a world without this choice as dystopian:

…it comes down to the idea it’s a utopia (but) the utopia is nothing if you can't live your life the way you wanna live. Ben, 18, (NZE, NS)


Most participants supported the SFG policy and some drew on their own experiences to imagine a future where others would never face the addiction they battled. They thought eliminating access to harmful products would preserve choices and freedoms, and explained how young people’s predisposition to heuristic-based thinking made truly informed choices impossible. They prioritised collective gain and societal well-being above any personal inconvenience the SFG could cause.

Others’ views evolved as the interviews proceeded. Some moved from opposing to supporting the SFG, while others created middle-ground positions as they reflected on how to balance individual and collective rights. Only a small minority valued the ‘freedom’ to purchase tobacco above the future benefit of freedom from addiction. These participants argued that people could easily access information about smoking’s harms and felt they should source and use this information to assess the short-term gains and longer-term costs of smoking.

Freedom and choice metaphors featured prominently in participants’ discourse; most drew on positive freedom concepts and demonstrated a sophisticated ability to weigh up individual loss and collective gain. Although prospect theory suggests rapid processing will emphasise loss aversion,49 thus predisposing a negative freedom perspective, few participants consistently privileged the status quo. Instead, most engaged in an internal debate, with some revising their initial position, and later embracing a positive freedom perspective and endorsing the SFG. Our findings illustrate young people’s capacity to reflect deeply on policies that will profoundly change their environment. Involving young people as study participants helps bring their voices to the policy table, and incorporating their expertise in policy design, implementation and communication may help enhance the SFG’s acceptability.

We anticipated that younger participants (ie, those aged 17 and who we asked to imagine being affected by the policy) would be less supportive than their older counterparts, but did not find age-related differences, although a larger survey may detect these. Future research could also explore perceptions held by those aged 16 and under, particularly whether and how they would obtain tobacco products if an SFG policy applied. In general, we found Māori participants more likely than non-Māori to perceive and value the collective benefits an SFG policy could deliver. This finding may reflect important Māori cultural concepts and highlights the need for kaupapa Māori research to explore perspectives that may be unique to rangatahi (youth) Māori.

Participants’ evolving views may reflect the COVID-19 context and interview experience; data collection took place shortly after intensive COVID-19 restrictions were lifted, and our intensive probing encouraged deep reflection. Our findings suggest the SFG’s framing and communication should stimulate reflective thinking that avoids instinctive, heuristic-based responses. Given brain development continues into adulthood,50 communications could promote longer-term outcomes that young people may not otherwise consider. Outlining how the SFG will create intergenerational benefits, enhance well-being and support a more equitable society could encourage deep reflection; positive freedom and liberation metaphors may pre-empt industry claims, such as nanny-state interference.42 43

Our findings highlight a growing tension between tobacco companies’ transformation rhetoric, including statements that young people should not smoke, and the SFG.54 Logically, the tobacco industry should endorse a policy to end youth smoking, which would demonstrate their commitment to ‘unsmoking’ the world.55 However, BAT(NZ) opposed the SFG and claimed access to tobacco is an ‘important freedom’ that supports young people’s ‘right to autonomy’.40 Given the constraint and harm that nicotine addiction causes, these richly ironic claims question the sincerity of transformation rhetoric.56 Policy makers and civil society should hold tobacco companies to account for discrepancies between their past statements and current behaviour, and rigorously implement FCTC Article 5.3 guidelines to limit their influence on policy making.57

While we cannot generalise our findings, they provide few grounds to support tobacco companies’ arguments that youth will find the SFG unacceptable. Nor should arguments the SFG is not feasible deter policy makers. Despite concerns over retailer non-compliance and increased social supply (eg, via peers or family networks),16 18 58 countries that raised the minimum purchase age for tobacco report good compliance, particularly when accompanied by strong surveillance and enforcement, and penalties.59 Furthermore, as the age gap between those who have or do not have legal access to tobacco increases, social supply opportunities will decrease; denicotinisation, which will rapidly decrease smoking prevalence, will further reduce social supply.1 16 30 Nonetheless, monitoring smoking prevalence and access remains important to assess whether social access routes persist and detect policy loopholes.60–62

Like all studies, ours has strengths and limitations. We recruited a diverse rather than representative sample; while diversity is a known strength of qualitative work, estimating the prevalence of perceptions identified requires quantitative studies. Although we enhanced external validity by asking participants to imagine the SFG would come into effect in 2022, the actual policy will not come into effect for several years (and will affect a more diverse age range than the group sampled). Nonetheless, comments on perceived loss suggest participants interpreted the policy timing as requested.

Our findings also raise new questions. Future work could explore the increased age restrictions some participants proposed, and more far-reaching measures, such as the nicotine-free generation policy called for by Māori leaders63 and proposed by Denmark and Malaysia.32 33 Rising vaping prevalence, particularly among rangatahi Māori,12 13 suggests NZ should also examine this approach.

Despite the limitations noted, our findings provide rich insights into how young people perceive the SFG and its impact, and could guide the policy’s implementation. In particular, challenging industry depictions of the SFG as a ban and reframing it as a conduit to freedom could foster reflection, clarify the policy’s benefits, and elicit even stronger support from the cohort the SFG aims to benefit.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. Not applicable.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by a departmental reviewer (assigned delegated authority to review low risk projects) on behalf of the University of Otago Human Ethics Committee (reference D21/393). Participants gave informed consent to participate in the study before taking part.


We thank the study participants who shared their views with us. We also thank Anaru Waa, who offered advice on the study protocol, and Emeritus Professor Phil Gendall and Mei-Ling Blank, who commented on the penultimate manuscript version.


Supplementary materials


  • X @JudeBall7

  • Contributors JH conceptualised the study, developed the study protocol, including the sampling strategy and interview guide, and obtained ethics approval. JH is guarantor of the article. EL worked as a summer scholarship student under JH’s supervision. EL and JH conducted the interviews, undertook the initial data analyses and prepared a preliminary report outlining the study findings. JH developed and wrote the manuscript and managed the revisions; EL and EF commented on early manuscript drafts; LT reviewed transcripts from some Māori participants and advised on the interpretation of these. JB and RE commented on advanced drafts. Authors are listed in descending order of contribution; all authors approved the final manuscript.

  • Funding EL received a student scholarship funded by the Health Research Council of New Zealand via the Whakahā o te Pā Harakeke programme (contract 19/641). The funders had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.