Control

Contents April 2015 Volume 24 Supplement 2 Implementation and evaluation of the Australian tobacco plain packaging policy Editors: G B Hastings and C Moodie

Special communication ii58 “You’re made to feel like a dirty filthy smoker ii1 Death of a salesman G B Hastings, C Moodie when you’re not, is another thing all together.” Responses of Australian cigar and ii3 Plain packaging: a logical progression for cigarillo smokers to plain packaging C L Miller, in one of the world’s ‘darkest K A Ettridge, M A Wakefield markets’ M Scollo, M Bayly, M Wakefield Cover credit: Quit Victoria collection, 2012. ii66 Changes in use of products by ii9 Standardised packaging and new enlarged pack sizes and price segments, prices paid and graphic health warnings for tobacco products in consumption following the introduction of plain Australia—legislative requirements and packaging in Australia M Scollo, M Zacher, implementation of the Tobacco Plain Packaging K Coomber, M Bayly, M Wakefield Act 2011 and the Competition and Consumer (Tobacco) Information Standard, 2011 M Scollo, ii76 Use of illicit tobacco following introduction of K Lindorff, K Coomber, M Bayly, M Wakefield standardised packaging of tobacco products in Australia: results from a national cross- Research papers sectional survey M Scollo, M Zacher, K Coomber, ii17 Australian adult smokers’ responses to plain M Wakefield packaging with larger graphic health warnings ii82 The advertised price of packs in retail 1 year after implementation: results from a outlets across Australia before and after the national cross-sectional tracking survey implementation of plain packaging: a repeated M Wakefield, K Coomber, M Zacher, S Durkin, measures observational study OPEN ACCESS M Scollo, E Brennan, M Scollo M Bayly, M Wakefield This issue has been made freely available online under the BMJ Journals Open Access scheme. ii26 Short-term changes in quitting-related cognitions See http://tobaccocontrol.bmj.com/site/about/ and behaviours after the implementation of plain Brief reports guidelines.xhtml#open packaging with larger health warnings: findings ii90 Did the recommended retail price of tobacco from a national cohort study with Australian products fall in Australia following the adult smokers S Durkin, E Brennan, K Coomber, implementation of plain packaging? M Scollo, M Zacher, M Scollo, M Wakefield M Bayly, M Wakefield This journal is a member of and subscribes to the principles of the Committee on Publication ii33 Are quitting-related cognitions and behaviours ii94 Personal pack display and active smoking at Ethics outdoor café strips: assessing the impact of www.publicationethics.org.uk predicted by proximal responses to plain packaging with larger health warnings? Findings plain packaging 1 year postimplementation from a national cohort study with Australian M Zacher, M Bayly, E Brennan, J Dono, C Miller, adult smokers E Brennan, S Durkin, K Coomber, S Durkin, M Scollo, M Wakefield M Zacher, M Scollo, M Wakefield Research letter ii42 Has the introduction of plain packaging with ii98 Did smokers shift from small mixed businesses larger graphic health warnings changed to discount outlets following the introduction of Receive regular table of contents by email. adolescents’ perceptions of cigarette packs and Register using this QR code. plain packaging in Australia? A national cross- brands? V White, T Williams, M Wakefield sectional survey M Scollo, K Coomber, M Zacher, ii50 Do larger graphic health warnings on M Wakefield standardised cigarette packs increase adolescents’ cognitive processing of consumer PostScript health information and beliefs about smoking- ii25, ii32 Side related harms? V White, T Williams, A Faulkner, M Wakefield tobaccocontrol.bmj.com

ttobaccocontrol_24_S2_TOC.inddobaccocontrol_24_S2_TOC.indd 1 223/02/153/02/15 3:453:45 PMPM Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication Death of a salesman

Gerard B Hastings,1 Crawford Moodie2

1Department of Marketing, TAKING THE LEAD THIS ISSUE UNPACKED University of Stirling and the In December 2012 Australia became the first juris- The evidence suggests that plain packaging is Open University, Stirling, UK 2Department of Marketing, diction to mandate plain (or standardised) pack- severely restricting the ability of the pack to com- Institute for Social Marketing, aging for tobacco products. Many governments municate and create appeal with young people and – University of Stirling, Stirling, have been looking on with interest, anxious to adults.2 4 For instance, school-based surveys with UK learn if this is also the next step forward in their students aged 12–17 year in 2011 and 2013 show own efforts to tackle the harms caused by tobacco. that the removal of branding and uniformity of Correspondence to Professor Gerard B Hastings, This special issue begins to answer this question by pack appearance has increased negative pack ratings 4 Department of Marketing, presenting a series of studies which provide an and decreased positive ones. Cognitive processing University of Stirling and the initial evaluation of the policy. What have been the of the health warnings did not change, however, Open University, Stirling early impacts of plain packaging on young people suggesting that pack appearance is more relevant to FK9 4LA, UK; 5 [email protected] and adult smokers? And are there any unintended young people than are the warnings. consequences—has it reduced prices, for instance, A number of studies with adult smokers point to or encouraged illicit tobacco use? plain packaging fulfilling its core aims of reducing appeal, particularly among young adults, and increasing warning salience.3 In a cross-sectional THE LAST WORD IN TOBACCO MARKETING? tracking survey of cigarette smokers, plain pack- Plain packaging is the latest of many moves by aging was associated with increased thinking about policy makers to constrain tobacco marketing. It is quitting and quit attempts.6 In addition, dislike of over 50 years since the first tentative steps were the pack, lower satisfaction from and taken, typically picking off TV advertising, but it attributing motivation to quit to the warnings pre- was not until this century that the need for controls dicted daily thoughts of quitting.7 These findings to be comprehensive was fully appreciated. Market may also help to explain why smokers were more forces, it became clear, dictate that any gaps will be likely to conceal their packs in outdoor venues exploited and promotional budgets moved to after the introduction of plain packaging.8 Research unregulated channels. The pack itself is one of the with adults is not confined to cigarette smokers last of these promotional fall-backs. In recent years either, with a mixed methods study suggesting that the literature has documented a seemingly endless the benefits of plain packaging may also extend to flow of elaborate packaging innovations. It has also cigar and cigarillo smokers.2 A lesson perhaps for demonstrated how tobacco companies have used those governments (eg, Ireland, UK) planning to the pack to promote their products, mislead consu- exclude such products from plain packaging mers about the harmfulness of smoking, and under- legislation. mine the legally mandated health warnings. So These findings highlight the potential impacts on even in markets that are otherwise dark, packaging adult smokers, a group often overlooked in the continues to offer tobacco companies a glimmer of plain packaging debate. Clearly, prevention is better light, providing a last channel for conveying image, than cure, which is why stopping new generations symbolism and brand meaning.1 from starting to smoke is a key objective of plain packaging, but adult smokers are also an important FROM THEORY INTO PRACTICE potential target. That they are amenable to change Until recently the evidence base for plain packaging is beyond question: in Australia (as in many other has, perforce, been hypothetical. Australia’s lead markets) there are more ex-smokers than current has now delivered a real world natural experiment. smokers. These studies suggest that plain packaging The real world, of course, is messy and natural can play a role in encouraging this transition. experiments less neat than custom-designed ones. This issue also examines whether plain packaging Open Access Scan to access more As such, it remains a challenge for researchers, par- has had any unintended consequences. Did it lower free content ticularly as plain packaging has been part of a prices for licit tobacco or increase the use of illicit wider package of measures introduced by the tobacco, both of which might be expected to Australian Government, including larger health encourage smoking? There is no evidence for – warnings, mass media campaigns and tax increases. either effect.9 12 A review of retail magazines, for Separating out the effects of each can prove diffi- instance, shows that following the introduction of cult—it was ever so. Nonetheless, the picture to plain packaging, average inflation-adjusted recom- emerge from the papers in this special issue sug- mended retail prices actually increased for cigar- gests that plain packaging is delivering on its hypo- ettes in all price segments (value, mainstream and 10 To cite: Hastings GB, thetical promise, and the potential downsides, premium). The rise in cost was greatest for cigar- Moodie C. Tob Control much vaunted by its opponents, are not ettes in the mainstream and premium segments, – 2015;24:ii1 ii2. materialising. which may, in part, help explain the shift to value

Hastings GB, et al. Tob Control 2015;24:ii1–ii2. doi:10.1136/tobaccocontrol-2015-052285 ii1 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication brands.11 As for the source of retail purchase, there was no why it has one of the lowest smoking prevalence rates in the decline in the percentage of smokers purchasing from conveni- world. ence stores and no indication of increased purchase from over- seas, online or duty-free.13 And in contrast to the findings from Competing interests None. the commissioned KPMG report,14 there was Provenance and peer review Not commissioned; internally peer reviewed. no evidence of an increase in the consumption of illicit ‘cheap Open Access This is an Open Access article distributed in accordance with the white’ cigarettes.12 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 WHAT WILL HAPPEN IF GOVERNMENTS DO NOT ACT? properly cited and the use is non-commercial. See: http://creativecommons.org/ This issue is focused on what happens when plain packaging is licenses/by-nc/4.0/ introduced, but it is instructive to consider what is likely to occur in markets where government does not act and the status quo prevails. The pack will continue to be used as a marketing REFERENCES channel and innovations will proliferate. New pack structures— 1 Underwood RL, Ozanne JL. Is your package an effective communicator? A sizes, openings, construction materials—are inevitable. Beyond normative framework for increasing the communicative competence of packaging. J Mark Commun 1998;4:207–20. visual appeal we are also likely to see: packs with special foils, 2 Miller C, Ettridge K, Wakefield M. Research paper: “You’re made to feel like a dirty varnishes or coatings that create distinct tactile experiences; filthy smoker when you’re not, cigar smoking is another thing all together.” audio packs that play pre-recorded messages, music or other Responses of Australian cigar and cigarillo smokers to plain packaging. Tob Control noises, which are already available for some other products15; 2015;24:ii58–65. fi et al ’ and packs which release fragrances (which are already patented 3 Wake eld M, Coomber K, Zacher M, . Australian adult smokers responses to plain packaging with larger graphic health warnings 1 year after implementation: by tobacco companies). Inks will also have an important role to results from a national cross-sectional tracking survey. Tob Control 2015;24:ii17– play in future on-pack marketing, with phosphorescent ‘glow in 25. the dark’ inks having already appeared on cigarette packs in 4 White V, Williams T, Wakefield M. Has the introduction of plain packaging with some markets, and packs with photochromic (light-sensitive), larger graphic health warnings changed adolescents’ perceptions of cigarette packs and brands? Tob Control 2015;24:ii42–9. thermochromic (heat-sensitive) or oxygen-sensitive inks likely to 5 White V, Williams T, Faulkner A, et al. Do larger graphic health warnings on follow. Although only for display purposes, photochromic standardised cigarette packs increase adolescents’ cognitive processing of consumer Camel packs which change colour appeared in retailers in the health information and beliefs about smoking-related harms? Tob Control Netherlands in 2014. Conductive inks, which enable printed 2015;24:ii50–7. et al electronics technology and the incorporation of cheap electrical 6 Durkin S, Brennan E, Coomber K, . Short-term changes in quitting-related cognitions and behaviours after the implementation of plain packaging with larger circuits into cardboard are also opening up a world of market- health warnings: findings from a national cohort study with Australian adult 16 ing opportunities. Cartons of Kent cigarettes, with circuit smokers. Tob Control 2015;24:ii26–32. boards and touchpads enabling scrolling on-pack messages, have 7 Brennan E, Durkin S, Coomber K, et al. Are quitting-related cognitions and already appeared in duty-free outlets in Asia. As technology behaviours predicted by proximal responses to plain packaging with larger health warnings? Findings from a national cohort study with Australian adult smokers. advances, it is possible that printed electronics may be applied Tob Control 2015;24:ii33–41. to deliver moving images on tobacco packaging. Similarly, we 8 Zacher M, Bayly M, Brennan E, et al. Personal pack display and active smoking at may see cigarette packs capable of digitally communicating with outdoor café strips: assessing the impact of plain packaging 1 year consumers and allowing connectivity to their virtual worlds via postimplementation. Tob Control 2015;24:ii94–7. fi smart devices (smartphones, smart watches, smart glasses). 9 Scollo M, Bayly M, Wake eld M. The real world advertised prices of cigarette packs in retailers before and after the implementation of plain packaging: a repeated Whatever directions these innovations take, it is clear that the measures observational study. Tob Control 2015;24:ii82–9. marketing power of the pack is only going to increase. So gov- 10 Scollo M, Bayly M, Wakefield M. Did the recommended retail price of tobacco ernments which do not act on plain packaging today will have a products fall in Australia following the implementation of plain packaging? Tob bigger problem to tackle tomorrow. Control 2015;24:ii90–3. 11 Scollo M, Zacher M, Durkin S, et al. Changes in use of types of tobacco products by pack sizes and price segments, prices paid and consumption following the A CASEBOOK EXAMPLE introduction of plain packaging in Australia. Tob Control 2015;24:ii66–75. Plain packaging in Australia has been a casebook example of 12 Scollo M, Zacher M, Durkin S, et al. Use of illicit tobacco following introduction of effective tobacco control—a policy measure driven by evidence, standardised packaging of tobacco products in Australia: results from a national Tob Control – carefully designed and implemented, and now rigorously cross-sectional survey. 2015;24:ii76 81. 13 Scollo M, Coomber K, Zacher M, et al. Did smokers shift from small mixed assessed. Further, it is set within the context of wider Australian businesses to discount outlets following the introduction of plain packaging in tobacco control, reinforcing the most basic lesson learned over Australia? A national cross-sectional survey. Tob Control 2015;24:ii98–100. the last half century: action has to be strategic and comprehen- 14 KPMG. Illicit tobacco in Australia. London: KPMG LLP, 2014. sive. There are no silver bullets. This issue demonstrates that 15 Moodie C, Angus K, Ford A. The importance of cigarette packaging in a dark market: the ‘Silk Cut’ experience. Tob Control 2014;23:274–8. plain packaging is beginning to deliver on its promise, and an 16 Mance H. Could printed electronics be getting a cigarette break? Financial Times important step forward, but it is still only part of the solution. 12 July 2012. http://www.ft.com/cms/s/0/247561ba-cc09-11e1-839a-00144 Australia has learned and applied this lesson well and that is feabdc0.html#axzz3PkyiJ8DN

ii2 Hastings GB, et al. Tob Control 2015;24:ii1–ii2. doi:10.1136/tobaccocontrol-2015-052285 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Death of a salesman

Gerard B Hastings and Crawford Moodie

Tob Control 2015 24: ii1-ii2 doi: 10.1136/tobaccocontrol-2015-052285

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii1

These include: Supplementary Supplementary material can be found at: Material http://tobaccocontrol.bmj.com/content/suppl/2015/03/18/tobaccocon trol-2015-052285.DC1.html References This article cites 2 articles, 1 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii1#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication Plain packaging: a logical progression for tobacco control in one of the world’s ‘darkest markets’ Michelle Scollo, Megan Bayly, Melanie Wakefield

Centre for Behavioural ABSTRACT ‘Quit’ campaigns established in each state from Research in Cancer, Cancer The Australian approach to tobacco control has been a 1983 used mass media to educate the community Council Victoria, Melbourne, 8 Victoria, Australia comprehensive one, encompassing mass media about the dangers of smoking. Government campaigns, consumer information, taxation policy, access funding was secured to place advertisements during Correspondence to for smokers to advice and prime-time television rather than merely in late Dr Michelle Scollo, pharmaceutical treatments, protection from exposure to night ‘community service’ spots.9 Professional Centre for Behavioural and regulation of promotion. World-first public relations activities encouraged media cover- Research in Cancer, Cancer Council Victoria, 615 St Kilda legislation to standardise the packaging of tobacco was age and used celebrities and high-rating television Rd, Melbourne, VIC 3004, a logical next step to further reduce misleadingly and radio programmes to popularise the ‘Quit’ Australia; reassuring promotion of a product known for the past message.9 Public support for the ‘Quit’ initiative [email protected] 50 years to kill a high proportion of its long-term users. helped to encourage governments to seriously con- Received 26 September 2014 Similarly, refreshed, larger pack warnings which started sider, and then start to enact, recommendations Revised 20 November 2014 appearing on packs at the end of 2012 were a logical from international health agencies to ban all forms Accepted 13 January 2015 progression of efforts to ensure that consumers are of promotion of tobacco products,10 and to raise better informed about the health risks associated with taxes on tobacco products with the dual objectives smoking. Regardless of the immediate effects of of making smoking less affordable, generating add- legislation, further progress will continue to require a itional funds for expanded public education cam- comprehensive approach to maintain momentum and paigns and replacing tobacco sponsorship of ensure that government efforts on one front are not sport.11 undermined by more vigorous efforts and greater During the late 1980s and early 1990s, concerns investment by tobacco companies elsewhere. about the health effects of exposure to other people’s smoking12 saw the progressive restriction of smoking in more and more workplaces,13 14 15 16 A COMPREHENSIVE APPROACH which then generalised elsewhere. The result- fi ant ever-expanding restrictions on smoking in hos- Australia was not the rst country in the world to 17 attempt to discourage smoking. It was not until pitality venues and public places combined with 1973 that a discreet, faint gold-lettered warning the ever-growing evidence about the health effects about smoking being a health hazard appeared on and social costs of smoking, all contributed to cigarette packs,1 almost a decade after similar warn- growing antismoking sentiment. Antismoking 2 norms have been demonstrated to have a profound ings were required in the USA. Televised cigarette 18 19 advertisements continued until the mid-1970s, effect on the frequency and uptake of smoking. about 10 years after they had disappeared from television screens in the UK,3 the USA4 and New SUSTAINED INVESTMENT YIELDS RESULTS Zealand.5 Despite this tentative beginning, since Over the past four decades of intense activity, con- the early 1980s Australian Governments of all per- sumption of tobacco products declined substantially suasions have pursued the tobacco control agenda in Australia, reducing from a high of more than with vigour and determination. An early achiever 3500 g of tobacco per person (15 years and older) of international best practice on many different in 1961 to less than an estimated 875 g per capita fronts since that time, Australia was one of the first in recent years (see figure 2). Prevalence of 40 countries to ratify the WHO Framework smoking has also declined substantially. Convention on Tobacco Control (FCTC),6 and so A stall in the decline in prevalence occurred in became a full Party on 27 February 2005, the date the mid-1990s, corresponding with reduced on which the FCTC came into force. The FCTC expenditure on public campaigns20 and less media requires Parties to adopt a systematic and broadly interest following a decade of intense political cam- Open Access 6 Scan to access more encompassing approach to tobacco control, paigning; however, a major injection of funds free content including numerous measures to reduce the through the National Tobacco Campaign in – demand and supply of tobacco products. Since the 199721 23 kick-started the decline again in that – early 1980s, the Australian approach to tobacco year.24 26 Campaigns over the late 2000s were control has been just such a comprehensive one, funded at more commercially realistic levels in encompassing mass media campaigns, consumer most states. This has allowed Australian smokers to information, taxation policy, access for smokers to be consistently exposed to television advertisements smoking cessation advice and pharmaceutical treat- about the health effects of smoking (see figure 3 ments, protection from exposure to tobacco smoke for an indication of the annual reach and frequency To cite: Scollo M, Bayly M, and regulation of promotion.7 The timeline of that advertising in Australia since 2001). Wakefield M. Tob Control depicted in figure 1 shows some of the major mile- Tax policy has always been a crucial part of – 2015;24:ii3 ii8. stones and activities on these fronts. Australia’s comprehensive approach to discouraging

Scollo M, et al. Tob Control 2015;24:ii3–ii8. doi:10.1136/tobaccocontrol-2014-052048 ii3 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication

Figure 1 Timeline of major tobacco control policies in Australia, 1970s to 2014. Source: Scollo M and Winstanley M, eds. Tobacco in Australia: facts and issues 2012 using material from Chapters 1, 2 3, 4, 5, 7, 11, 12A, 13, 14, 15 and 16. Notes: Health warnings were determined at state level prior to 1996, nationally consistent by agreement. Restrictions on smoking in public places other than in aircraft and airports are covered by state legislation. Antismoking education campaigns have been run at the state level and at the national level commencing with the NHMRC poster/leaflet campaign in the early 1980s. NRT, replacement therapy; OTC, over-the-counter; NHMRC, National Health and Medical Research Council; GHW, graphic health warning. smoking.27 28 Frequent increases in state fees on tobacco from extended from subsidies limited to war veterans and Indigenous the early 1980s until their abolition in 1997 carried through to smokers to all Australian smokers). Since 2001, almost three frequent increases in the price of tobacco products, though the million prescriptions for treatments have been dispensed (see effects were somewhat blunted by tobacco companies’ develop- blue bars in figure 3). ment of large pack sizes which attracted much less tax than smaller packs. The tax on large packets of cigarettes increased RECENT INITIATIVES substantially following tax reforms adopted in 1999, with The social costs of smoking and the case for investment in further increases associated with the implementation of smoking cessation and tobacco control more generally have 29 Australia’s Goods and Services Tax in 2000–2001. Taxes been widely accepted in Australia since the late 1990s.32 33 30 increased again substantially in April 2010, December 2013 Faced with an ageing cohort of postwar baby boomers and the and September 2014, with further increases scheduled for prospect of a shrinking workforce to support rising healthcare September 2015 and 2016. The recommended retail price over costs, recent Australian governments have looked to tobacco 31 time of Winfield 25s, Australia’s leading brand, is shown in control for continuing returns for their investment in disease figure 3. prevention. In 2008 and then again in 2012, all governments in Consistent with a long-standing commitment to a comprehen- Australia—state, territory and federal—signed a national health- sive approach, Australian governments have not relied on tax care agreement34 with the ambitious goal of reducing adult alone. A variety of telephone, internet, SMS programmes and daily smoking prevalence to 10% and halving the adult daily smartphone applications have been put in place across the smoking rate among Aboriginal and Torres Strait Islanders by country to support and encourage smokers in their quit 2018 (clause 18.3).35 In May 2010, the Government released its attempts. Smoking cessation aids were listed on the national response to a far-reaching and detailed set of recommendations Pharmaceutical Benefits Scheme in 2001 (bupropion), 2008 formulated by a national taskforce on preventive health.36 This (varenicline) and 2011 (nicotine replacement therapies, document affirmed the Government’s intention to implement ii4 Scollo M, et al. Tob Control 2015;24:ii3–ii8. doi:10.1136/tobaccocontrol-2014-052048 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication

Figure 2 Prevalence of smoking and per capita consumption of tobacco products, Australia 1910 to 1960 (5-yearly), 1960 to 2011. Source: Scollo M. Figure I.2 Major tobacco promotion and tobacco control policies and regular smoking & per capita consumption of tobacco products, Australia 1910 to 1960 (5-yrly), 1960 to 2010 in Introduction. Scollo M and Winstanley M, eds. Tobacco in Australia: facts and issues 2012. plain packaging and an immediate 25% increase in customs/ response document also outlined the Government’s commit- excise duty on tobacco (announced on the 29 April 2010—see ment to adopting numerous other recommended measures Scollo et al (this volume) for a timeline of events).37 The including enlarged graphic health warnings, tightening of

Figure 3 Affordability of tobacco and investment in tobacco control in Australia—media TARPs, recommended price of Winfield 25s (leading Australian brand) and number of prescriptions for pharmaceutical treatments, 2001 to 2013. Sources: TARPs* Metropolitan TV Monthly TARP Flighter data, 2001–2013. Prepared for Cancer Council Victoria. Macquarie Park, Australia: Nielsen Company Pty Ltd. Prices Australian Retail August editions, 2001 to 2013. Prescriptions Health Insurance Commission, Pharmaceutical Benefits Schedule Item reports for nicotine replacement therapy, bupropion and varenicline. https://www.medicareaustralia.gov.au/statistics/pbs_item. shtml *Note: TARPs are a product of the percentage of the target audience (aged 18+ years) potentially exposed to an advertisement (reach) and the average number of times the target audience is exposed (frequency). 1000 TARPs represents 100% of the target audience receiving 10 advertisements per year, or 50% receiving 20 advertisements per year. NRT, nicotine replacement therapy; TARP, Target Audience Rating Points.

Scollo M, et al. Tob Control 2015;24:ii3–ii8. doi:10.1136/tobaccocontrol-2014-052048 ii5 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication restrictions on advertising of tobacco products in particular on warnings in 1995.46 47 Australia was one of the first countries in the internet, increased funding for mass media campaigns and the world to follow Canada’s lead with graphic health warnings additional programmes for Indigenous smokers and people complemented by comprehensive back-of-pack information living with mental illness. In 2012 the Australian Government explicating the warning statement implemented in 2006.48 Once and state and territory governments approved a new national again, the refreshed, larger pack warnings which started appear- tobacco strategy, the NTS 2012–2018,38 which is much more ing on packs at the end of 201249 were a logical progression of – far-reaching than its predecessors39 41 and aims to strengthen efforts to ensure that consumers are better informed about the and extend activities in all the major streams of tobacco control health risks associated with smoking (see figure 4). over the 6 years to 2018. EVALUATING IMPACT TOBACCO PLAIN PACKAGING—A LOGICAL PROGRESSION The decline in smoking in Australia since the late 1990s resulted Australia has been described by the tobacco industry as the from more people quitting, and fewer young people taking up 42 – world’s ‘darkest market’. Tobacco advertising has been banned smoking.50 52 In line with the findings of research throughout in virtually every form of media—on TV and radio through the the rest of the world,53 studies measuring short-term effects 1970s, on billboards and outside shops during the 1980s, in the have been able to attribute reductions in smoking prevalence in print media and through sports sponsorship during the 1990s Australia to increasing taxes,27 28 greater expenditure on social and at point of sale from the early 2000s, with retail display of marketing campaigns27 28 54 55 and smoke-free policies.28 56 43 products banned altogether in most states from about 2010. Multivariate analysis of the effects of policy on prevalence of By the mid-2000s, attractive design of packs was one of the few smoking among adolescents in various Australian states from ways that Australian tobacco companies could continue to 1990 to 2005 also indicates strong effects for increases in the promote their products. World-first legislation to standardise the price of tobacco products, expenditure on social marketing and 44 packaging of tobacco was both a response to this marketing comprehensiveness of smoke-free policies in public places.57 strategy and a logical next step to further reduce the mislead- However, such studies tell only part of the story. ingly reassuring promotion of a product known to cause the As illustrated in US Surgeon General’s reports, which have 45 death of more than half of its long-term users. exhaustively reviewed the evidence about the effectiveness of tobacco control over the past five decades,58 59 smoking is a ENLARGED GRAPHIC HEALTH WARNINGS—ANOTHER multi-factorial problem—a tug-of-war between the forces which LOGICAL PROGRESSION promote and facilitate the use of tobacco products and the In addition to some of the earliest and strongest television-led forces which discourage and inhibit its use; a tug-of-war played antismoking campaigns,8 clear and direct information for out at the individual, household and community levels as well tobacco consumers on product packaging has also been an as in the wider culture. Each of the regulatory, educational and important part of the Australian approach, with four rotating clinical factors highlighted in figure 1 vary widely in their tech- warnings introduced on cigarette packs in 19871 and bold text niques and effects, some of which are contributory rather than

Figure 4 Marlboro cigarettes displaying changing consumer product information—as they appeared in Australia from 1987 (rear), the late 1990s, the late 2000s (second from front) and from December 2012 after the introduction of plain packaging (front). Source: Quit Victoria pack collection. ii6 Scollo M, et al. Tob Control 2015;24:ii3–ii8. doi:10.1136/tobaccocontrol-2014-052048 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication independent,58 and difficult to capture at the population level 2 Centres for Disease Control. Selected actions of the U.S. Government regarding the – through standard statistical analysis.60 63 However, it seems regulation of tobacco sales, marketing, and use (excluding laws pertaining to agriculture or excise tax). Atlanta, GA: CDC, 2012 [updated 15 Nov 2012; Aug likely that each would have contributed in some way to reduce 2014]. http://www.cdc.gov/tobacco/data_statistics/by_topic/policy/legislation/ —either directly or indirectly—by having: 3 Television Act, 1964, (1964). http://www.legislation.gov.uk/ukpga/1964/26 reduced the glamour and appeal of tobacco products; increased 4 Public Health Cigarette Smoking Act, (1969). http://www.law.cornell.edu/uscode/ knowledge about health effects; reduced cues and opportunities text/15/1331 for smoking; reduced the social acceptability and other rewards 5 The Smokefree Coalition. The control in New Zealand. 2012. ’ http://www.sfc.org.nz/infohistory.php of smoking and increased its costs; increased smokers knowl- 6 WHO Framework Convention on Tobacco Control. New York: United Nations, 2003. edge about how to manage the quitting process; or reduced 2302. http://www.who.int/tobacco/framework/WHO_FCTC_english.pdf withdrawal symptoms during quitting. 7 Scollo M, Winstanley M. Introduction. In: Scollo M, Winstanley M., eds. Tobacco in The studies in this volume examine the impact of Australia’s Australia: facts and issues. Melbourne: Quit Victoria, 2012. http://www. tobaccoinaustralia.org.au tobacco plain packaging legislation and the simultaneously intro- ‘ ’ 37 8 Pierce J, Dwyer T, Frape G, et al. Evaluation of the Sydney Quit For Life duced enlarged graphic health warnings not on smoking anti-smoking campaign. Part 1. Achievement of intermediate goals. Med J Aust prevalence, which is affected by a variety of demographic, mar- 1986;144:341–4. keting and policy factors over time, but rather on the perceived 9 Centre for Behavioural Research in Cancer. Quit evaluation studies. Melbourne: appeal of tobacco packaging, the effectiveness of health warn- Victorian Smoking and Health Program, 1985 to 2001. http://www.quit.org.au/ 64–66 67 browse.asp?ContainerID=1755 ings and consumer misperceptions of harm. Downstream 10 Powles JW, Gifford S. Health of nations: lessons from Victoria, Australia. Br Med J 68 69 effects on attitudes, beliefs and intentions are also examined, 1993;306:125–6. as are tobacco industry claims about possible and unintended 11 Borland R, Winstanley M, Reading D. Legislation to institutionalize – consequences.70 73 Regardless of the immediate effects, further resources for tobacco control: the 1987 Victorian Tobacco Act. Addiction 2009;104:1623–9. progress in Australia will continue to require a comprehensive 12 Makkai T, McAllister I. Public opinion towards drug policies in Australia 1985 to approach to maintain momentum and ensure that government 95. Canberra: Australian Government, 1998. http://www.aic.gov.au/publications/ efforts on one front are not undermined by more vigorous previous20series/other/21-40/public20opinion20towards20drug20policies efforts and greater investment by tobacco companies elsewhere. 20in20australia201985-95.html 13 Wakefield M, Roberts L, Owen N. Trends in prevalence and acceptance of workplace smoking bans among indoor workers in South Australia. Tob Control 1996;5:205–8. What this paper adds 14 Borland R, Morand M, Mullins R. Prevalence of workplace smoking bans in Victoria. Aust N Z J Public Health 1997;21:694–8. 15 Walsh R, Tzelepis F. Support for smoking restrictions in bars and gaming areas: What is already known on this topic review of Australian studies. Aust N Z J Public Health 2003;27:310–22. ▸ Australia has been an early achiever on many different fronts 16 Walsh R, Tzelepis F, Paul C, et al. Environmental tobacco smoke in homes, motor vehicles and licensed premises: community attitudes and practices. Aust N Z J Public in tobacco control. Health 2002;26:536–42. ▸ It was the first country in the world to standardise the 17 Merom D, Rissel C. Factors associated with smoke-free homes in NSW: results from packaging of tobacco products. the 1998 NSW Health Survey. Aust N Z J Public Health 2001;25:339–45. 18 Alamar B, Glantz S. Effect of increased social unacceptability of cigarette smoking What this paper adds on reduction in cigarette consumption. Am J Public Health 2006;96:1359–63. ▸ This paper provides a brief history of Australia’s 19 DeCicca P, Kenkel D, Mathios A, et al. Youth smoking, cigarette prices, and – comprehensive approach to tobacco control and explains the anti-smoking sentiment. Health Econ 2008;17:733 49. 20 Hill DJ, White VM, Scollo MM. Smoking behaviours of Australian adults in 1995: rationale for adoption of plain packaging legislation and trends and concerns. Med J Aust 1998;168:209–13. enhanced graphic health warnings. 21 Hill D, Alcock J. Background to campaign. In: Hassard K., ed. Australia’s National Tobacco Campaign: evaluation report. Vol 1. Canberra: Commonwealth Department of Health and Aged Care, 1999. http://www.health.gov.au/internet/main/publishing. nsf/Content/health-pubhlth-publicat-document-metadata-tobccamp.htm Contributors MW and MS conceived of this paper. MS and MB coordinated 22 Hill D, Borland R, Carroll T, et al. Perspectives of the Australian National Tobacco collection of data and undertook data analysis. MS drafted the manuscript and all Campaign. In: Hassard K., ed. Australia’s National Tobacco Campaign: evaluation authors contributed to the finalisation of the manuscript. report. Vol 2. Canberra: Commonwealth Department of Health and Aged Care, – Funding Production of this paper was supported by Cancer Council Victoria. 2000:1 9. http://www.health.gov.au/internet/main/publishing.nsf/Content/ health-pubhlth-publicat-document-tobccamp_2-cnt.htm Competing interests The authors wish to advise that MS was a technical writer 23 Hill D, Carroll T. Australia’s National Tobacco Campaign. Tob Control 2003;12:ii9–14. for and MW a member of the Tobacco Working Group of the Australian National 24 Wakefield M, Freeman J, Boulter J. Changes associated with the National Tobacco Preventive Health Task Force and MW was a member of the Expert Advisory Campaign: pre and post campaign surveys compared. In: Hassard K., ed. Australia’s Committee on Plain Packaging that advised the Australian Department of Health on National Tobacco Campaign: evaluation report. Vol 1. Canberra: Commonwealth research pertaining to the plain packaging legislation. MW holds competitive grant Department of Health and Aged Care, 1999. http://www.health.gov.au/internet/main/ funding from the Australian National Health and Medical Research Council, US publishing.nsf/Content/health-pubhlth-publicat-document-metadata-tobccamp.htm National Institutes of Health, Australian National Preventive Health Agency and 25 Tan N, Wakefield M, Freeman J. Changes associated with the National Tobacco BUPA Health Foundation. Campaign: results of the second follow-up survey. In: Hassard K., ed. Australia’s Provenance and peer review Not commissioned; externally peer reviewed. National Tobacco Campaign: evaluation report. Vol 2. Canberra: Commonwealth Department of Health and Aged Care, 2000:21–75. http://www.health.gov.au/internet/ Open Access This is an Open Access article distributed in accordance with the main/publishing.nsf/Content/health-pubhlth-publicat-document-tobccamp_2-cnt.htm Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which 26 Wakefield M, Freeman J, Inglis G. Ch 5 Changes associated with the National permits others to distribute, remix, adapt, build upon this work non-commercially, Tobacco Campaign: results of the third and fourth follow-up surveys. In: Hassard K., and license their derivative works on different terms, provided the original work is ed. Australia’s National Tobacco Campaign; evaluation report. Vol 3. Canberra: properly cited and the use is non-commercial. See: http://creativecommons.org/ Department of Health and Ageing, 2004. http://www.health.gov.au/internet/main/ licenses/by-nc/4.0/ publishing.nsf/Content/health-pubhlth-publicat-document-tobccamp_3-cnt 27 Wakefield M, Durkin S, Spittal M, et al. Impact of tobacco control policies and mass media campaigns on monthly adult smoking prevalence: time series analysis. REFERENCES Am J Public Health 2008;98:1443–50. 1 Scollo M. Chapter 12.1 Health warnings. In: Scollo M, Winstanley M., eds. Tobacco 28 Wakefield MA, Coomber K, Durkin SJ, et al. Time series analysis of the impact of in Australia: facts and issues. Melbourne: Cancer Council Victoria, 2013. http:// tobacco control policies on smoking prevalence among Australian adults, www.tobaccoinaustralia.org.au/a12-1-1-history-health-warnings 2001–2011. Bull World Health Organ 2014;92:413–22.

Scollo M, et al. Tob Control 2015;24:ii3–ii8. doi:10.1136/tobaccocontrol-2014-052048 ii7 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication

29 Scollo M, Younie S, Wakefield M, et al. Impact of tobacco tax reforms on tobacco 53 Wilson LM, Avila Tang E, Chander G, et al. Impact of tobacco control interventions prices and tobacco use in Australia. Tob Control 2003;12:ii59–66. on smoking initiation, cessation, and prevalence: a systematic review. J Environ 30 Rudd K, Swan W, Roxon N. Minister, Treasurer, Minister for Health. Anti-Smoking Public Health 2012;2012:961724. Action. Canberra 29 April 2010. Available from: http://pmtranscripts.dpmc.gov.au/ 54 Durkin S, Brennan E, Wakefield M. Mass media campaigns to promote smoking browse.php?did=17255 cessation among adults: an integrative review. Tob Control 2012;21:127–38. 31 NSW Retail Tobacco Traders’ Association. Cigarette price lists. Australian Retail 55 National Cancer Institute. Part 4—Tobacco control and media interventions. The Role Tobacconist 2001 to 2013;60 to 69. of the Media. Smoking and Tobacco Control Monograph no. 19. Bethesda, MD: US 32 Abelson P. Applied Economics. Returns on investment in public health. Canberra: Department of Health and Human Services, National Institutes of Health, National Department of Health and Ageing, 2003. http://www.health.gov.au/internet/main/ Cancer Institute, 2008. http://cancercontrol.cancer.gov/tcrb/monographs/19/index.html publishing.nsf/Content/health-pubhlth-publicat-document-roi_eea-cnt.htm 56 International Agency for Research on Cancer. Evaluating the effectiveness of 33 Hurley S. Chapter 17. The economics of tobacco control. In: Scollo M, Winstanley smoke-free policies. Lyon, France: IARC, 2009. 13. http://com.iarc.fr/en/publications/ M., eds. Tobacco in Australia: facts and issues. Melbourne: Cancer Council Victoria, pdfs-online/prev/handbook13/ 2013. http://www.tobaccoinaustralia.org.au/chapter-17-economics 57 White V, Warne C, Spittal M, et al. What impact have tobacco control policies, 34 COAG Reform Council. National Healthcare Agreement: baseline performance report cigarette price and tobacco control program funding had on Australian adolescents’ for 2008–09. Sydney: COAG Reform Council, 2010. http://www. smoking? Findings over a 15-year period. Addiction 2011;106:1493–502. federalfinancialrelations.gov.au/content/national_agreements.aspx 58 US Department of Health and Human Services. Reducing tobacco use: a report of the 35 Excise Tariff Amendment (Tobacco) Act 2014, Stat. No 9 (2014). http://www. Surgeon General. Atlanta, GA: US Department of Health and Human Services, Public comlaw.gov.au/Details/C2014A00009 Health Service, Centers for Disease Control and Prevention, National Center for Chronic 36 Australian Government. Taking preventative action: government’s response to Disease Prevention and Health Promotion, Office on Smoking and Health, 2000. http:// Australia: the healthiest country by 2020. Canberra: Department of Health and www.cdc.gov/tobacco/data_statistics/sgr/2000/complete_report/index.htm Ageing, 2010. http://yourhealth.gov.au/internet/yourhealth/publishing.nsf/Content/ 59 US Department of Health and Human Services. Preventing tobacco use among report-preventativehealthcare youth and young adults: a report of the Surgeon General. Atlanta, GA: US 37 Scollo M, Lindorff K, Coomber K, et al. Standardised packaging and new enlarged Department of Health and Human Services, Public Health Service, Centers for graphic health warnings for tobacco products in Australia—legislative requirements Disease Control and Prevention, National Center for Chronic Disease Prevention and and implementation of the Tobacco Plain Packaging Act 2011 and the Competition Health Promotion, Office on Smoking and Health, 2012. http://www.cdc.gov/ and Consumer (Tobacco) Information Standard, 2011. Tob Control 2015;24:ii9–16. tobacco/data_statistics/sgr/2012/index.htm 38 Intergovernmental Committee on Drugs Standing Committee on Tobacco. National 60 Chapman S. Commentary: if you can’t count it...it doesn’t count: the poverty of Tobacco Strategy 2012–2018. Canberra: Australian Government, 2013. http://www. econometrics in explaining complex social and behavioural change. Health Promot J nationaldrugstrategy.gov.au/internet/drugstrategy/publishing.nsf/Content/national_ Austr 1999;9:206–7. ts_2012_2018 61 Borland R. On apparent consumption and what goes up in smoke: a commentary 39 Commonwealth Department of Human Services and Health. National Health Policy on Bardsley & Olekalns. Health Promot J Austr 1999;9:208–9. on Tobacco in Australia and examples of strategies for implementation. Canberra: 62 Bardsley P, Olekalns N. The impact of anti-smoking policies on tobacco Commonwealth Department of Human Services and Health, 1991. 1 p. consumption in Australia. Health Promot J Austr 1999;9:202–5. 40 National Expert Advisory Committee on Tobacco. National Tobacco Strategy 1999 to 63 Chapman S. Unravelling gossamer with boxing gloves: problems in explaining the 2002–03, prepared for the Ministerial Council on Drug Strategy. Canberra: Department decline in smoking. BMJ 1993;307:429–32. of Health and Aged Services, 1999. http://www.health.gov.au/internet/wcms/ 64 Wakefield M, Coomber K, Zacher M, et al. Australian adult smokers’ responses to Publishing.nsf/Content/health-pubhlth-publicat-document-metadata-tobccstrat.htm plain packaging with larger graphic health warnings 1 year after implementation: 41 Ministerial Council on Drug Strategy. Australian National Tobacco Strategy 2004– results from a national cross-sectional tracking survey. Tob Control 2015;24:ii17–25. 2009. Canberra: Department of Health and Ageing, 2005. http://www.health.gov. 65 White V, Williams T, Faulkner A, et al. Do larger graphic health warnings on au/internet/main/publishing.nsf/Content/tobacco-strat standardised cigarette packs increase adolescents' cognitive processing of consumer 42 Carter SM. Going below the line: creating transportable brands for Australia’s dark health information and beliefs about smoking-related harms? Tob Control 2015;24: market. Tob Control 2003;12:iii87–94. ii50–7. 43 Freeman B. Chapter 11. Advertising and promotion of tobacco. In: Scollo M., ed. 66 White V, Williams T, Wakefield M, et al. Has the introduction of plain packaging with Tobacco in Australia: facts and issues. Melbourne: Cancer Council Victoria, 2013. larger graphic health warnings changed adolescents’ perceptions of cigarette packs http://www.TobaccoInAustralia.org.au and brands? Tob Control 2015;24:ii42–9. 44 Tobacco Plain Packaging Act. http://www.comlaw.gov.au/Details/C2011A00148 67 Miller C, Ettridge KA, Wakefield MA. “You’re made to feel like a dirty filthy smoker 45 Preventative Health Taskforce. Australia: the healthiest country by 2020. National when you’re not, cigar smoking is another thing all together.” Responses of Preventative Health Strategy. Canberra: Commonwealth of Australia, 2009. http:// Australian cigar and cigarillo smokers to plain packaging. Tob Control 2015;24: www.preventativehealth.org.au/internet/preventativehealth/publishing.nsf/Content/ ii58–65. national-preventative-health-strategy-1lp 68 Brennan E, Durkin S, Coomber K, et al. Are quitting-related cognitions and 46 Trade Practices (Consumer Product Information Standards) (Tobacco) Regulations— behaviours predicted by proximal responses to plain packaging with larger health repealed, 83 and 408 (1994 (Cth)). http://www.austlii.edu.au/au/legis/cth/num_reg/ warnings? Findings from a national cohort study with Australian adult smokers. tppisr1994n83741/ Tob Control 2015;24:ii33–41. 47 Borland R, Hill D. The path to Australia’s tobacco health warnings. Addiction 69 Durkin S, Brennan E, Coomber K, et al. Short-term changes in quitting-related 1997;92:1151–7. cognitions and behaviours after the implementation of plain packaging with larger 48 Trade Practices (Consumer Product Information Standards) (Tobacco) Regulations health warnings: findings from a national cohort study with Australian adult 2004, no.264 (2004 (Cth)). http://www.austlii.edu.au/au/legis/cth/num_reg/ smokers. Tob Control 2015;24:ii26–32. tppisr20042004n264741/ 70 Scollo M, Bayly M, Wakefield M. Did the recommended retail price of tobacco 49 Competition and Consumer (Tobacco) Information Standard,—F2011L02766 products fall in Australia following the implementation of plain packaging? (2011). http://www.comlaw.gov.au/Details/F2011L02766 Tob Control 2015;24:ii90-93. 50 Australian Data Archive Social Sciences collection. National Drug Strategy Household 71 Scollo M, Bayly M, Wakefi eld M. The advertised price of cigarette packs in retail Survey, 2010 [database on the internet]. Australian National University. 2011. [cited outlets across Australia before and after the implementation of plain packaging: Dec 2011]. http://www.ada.edu.au/social-science/01237 a repeated measures observational study. Tob Control 2015;24:ii82–9. 51 Germain D, Durkin S, Scollo M, et al. The long-term decline of adult tobacco use in 72 Scollo M, Zacher M, Coomber K, et al. Changes in use of types of tobacco products Victoria: changes in smoking initiation and quitting over a quarter of a century of by pack sizes and price segments, prices paid and consumption following the tobacco control. Aust N Z J Public Health 2012;36:17–23. introduction of plain packaging in Australia. Tob Control 2015;24:ii66–75. 52 Australian Institute of Health and Welfare. Highlights from the 2013 survey: tobacco 73 Scollo M, Zacher M, Coomber K, et al. Use of illicit tobacco following introduction of smoking. Canberra: AIHW, 2014. [updated 27 Jul; AIHW cat. no. PHE 145]. http:// standardised packaging of tobacco products in Australia: results from a national cross- www.aihw.gov.au/alcohol-and-other-drugs/ndshs/2013/data-and-references/ sectional survey. Tob Control 2015;24:ii76–81.

ii8 Scollo M, et al. Tob Control 2015;24:ii3–ii8. doi:10.1136/tobaccocontrol-2014-052048 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Plain packaging: a logical progression for tobacco control in one of the world's 'darkest markets' Michelle Scollo, Megan Bayly and Melanie Wakefield

Tob Control 2015 24: ii3-ii8 doi: 10.1136/tobaccocontrol-2014-052048

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii3

These include: References This article cites 27 articles, 7 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii3#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication Standardised packaging and new enlarged graphic health warnings for tobacco products in Australia— legislative requirements and implementation of the Tobacco Plain Packaging Act 2011 and the Competition and Consumer (Tobacco) Information Standard, 2011 Michelle Scollo,1,2 Kylie Lindorff,2 Kerri Coomber,1 Megan Bayly,1 Melanie Wakefield1

1Centre for Behavioural ABSTRACT response to the Taskforce released several days Research in Cancer, Cancer This paper describes the development, content and later.7 Council Victoria, Melbourne, implementation of two pieces of Australian tobacco The Department of Health and Ageing commis- Victoria, Australia 8 2Cancer Council Victoria, control legislation: one to standardise the packaging of sioned a set of studies to establish the impact of Melbourne, Victoria, Australia tobacco products and the other to introduce new, existing packaging design on appeal and attractive- enlarged graphic health warnings. It describes the ness and to establish optimal colour for packaging Correspondence to process of legislative drafting, public consultation and and the optimal size and colour for standardised Dr Michelle Scollo, 910 Centre for Behavioural parliamentary consideration. It summarises exactly how font indicating brand and variant names. These Research in Cancer, tobacco products have been required to look since late and related studies on GHWs were conducted Cancer Council Victoria, 2012. Finally, it describes implementation, most between December 2010 and March 2011 and 615 St Kilda Rd, Melbourne, particularly, the extent to which packs compliant with involved regular smokers aged 18–65 in two VIC 3004, Australia; [email protected] the legislation became available to consumers over time. face-to-face focus group studies (38 groups in total) and four online quantitative surveys (ranging from Received 29 September 2014 n=205 to n=455).11 In addition, face-to-face inter- Accepted 13 January 2015 DEVELOPMENT OF LEGISLATION views were conducted with retailers aged 40 years In 2008, the Australian Government established a and older to establish legibility of brand names in national taskforce to recommend measures to the retail environment.11 reduce the burden of disease caused by the prevent- An ‘exposure draft’ of the Tobacco Plain able risk factors of smoking, alcohol abuse and Packaging Bill was released in April 2011.12 A obesity.1 Among a large number of educational and Consultation Paper was released at the same time regulatory measures, and based on consideration of outlining the Government’s response to design a growing body of international research, plain testing and targeted consultations,13 and explaining packaging of tobacco products and enlarged the most important provisions of the draft Bill graphic health warnings (GHWs) were proposed (2011) and associated regulations.14 later that year in a technical paper on tobacco2 pre- The Explanatory Memorandum that accompan- pared for the National Preventative Health ied the Bill tabled in Parliament on the 6 July Taskforce. After consultation on a discussion 201115 stated that the legislation aimed to prevent paper3 and consideration of submissions,4 the pro- tobacco advertising and/or promotion on tobacco posals were included as recommendations in a draft product packaging in order to: reduce the attractive- preventive health strategy released by the Taskforce ness and appeal of tobacco products to consumers, in September 2009.5 The proposal for plain pack- particularly young people; increase the noticeability aging was vigorously opposed by the tobacco indus- and effectiveness of mandated health warnings; try in Australia6: accounts of the tobacco industry’s reduce the ability of the tobacco product packaging campaign strategies and of activities by health to mislead consumers about the harms of smoking; groups over the late 2000s are available elsewhere, and through the long-term achievement of these for instance (see http://www.tobaccoinaustralia.org. Open Access objectives as part of a comprehensive suite of Scan to access more au/chapter-11-advertising/11–10-tobacco-display-as- tobacco control measures, contribute to efforts to free content advertising1). Table 1 provides a timeline of improve public health by reducing smoking rates.15 legislative milestones in the development and The Government conducted further consultation implementation of plain packaging and enlarged on the detail of the proposed design features for the GHWs in Australia. plain packaging of non-cigarette tobacco products in October 201116 and also on the Tobacco Plain Plain packaging Packaging Regulations. The Tobacco Plain Packaging On 29 April 2010, the Prime Minister announced Act 2011 (the Act)17 18 and associated Trade Marks 19 To cite: Scollo M, that the Government would adopt tobacco plain Amendment (Tobacco Plain Packaging) Act 2011 Lindorff K, Coomber K, et al. packaging and immediately increase tobacco excise received Royal Assent on 1 December 2011. The Tob Control 2015;24: and customs duty by 25%. This position was Tobacco Plain Packaging Regulations 2011 – ii9 ii16. detailed in full in the Government’s formal (Regulations) approved by the Government’s

Scollo M, et al. Tob Control 2015;24:ii9–ii16. doi:10.1136/tobaccocontrol-2014-052073 ii9 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication

Table 1 Timeline of adoption and implementation of tobacco plain packaging legislation in Australia 2008 9 April Health Minister Nicola Roxon announces establishment of the National Preventative Health Taskforce 10 October Release of the draft report of the National Preventative Health Taskforce recommending tobacco plain packaging 17–22 November Parties to the Framework Convention on Tobacco Control adopt guidelines on advertising and package labelling that recommend the use of plain packaging 2009 30 June National Preventative Health Taskforce final report to Australian Government for consideration 1 September Health Minister releases final report 2010 29 April Prime Minister Kevin Rudd announces Australia will adopt tobacco plain packaging 11 May Release of Australian Government’s formal response to recommendations 21 August Australian general election 2011 7 April Release of exposure draft of legislation for consultation Tobacco Plain Packaging Bill 2011 6 July Trade Marks Amendment (Tobacco Plain Packaging) Bill 2011 and Tobacco Plain Packaging Bill introduced and read in House of Representatives. House refers Bills to Standing Committee on Health and Ageing 4 August Hearings of the House of Representatives Standing Committee on Health and Ageing 18 August Senate refers Trade Marks Amendment (Tobacco Plain Packaging) Bill 2011 to Legal and Constitutional Affairs Committee which calls for submissions (by 2 September) 22 August House of Representatives Standing Committee on Health and Aged Care tables the report on its Inquiry into Tobacco Plain Packaging 13 September Hearings of the Senate’s Legal and Constitutional Affairs Committee 17 September Release of consultation paper on new graphic health warnings for tobacco products 19 September Legal and Constitutional Affairs Committee provides report to Senate 26 October Release of Proposed mandatory standard for tobacco labelling (graphic health warnings): second round consultation paper 2 November Health Minister, announces that the implementation of tobacco plain packaging will be delayed until 1 December, 2012 as a result of delays in the Senate review of the bill 14 November Release of Proposed mandatory standard for tobacco labelling (graphic health warnings): draft regulation for comment 21 November Final passage of amended Bill through House of Representatives 1 December Bill received royal assent and became law—Tobacco Plain Packaging Act 2011 1 December British American Tobacco initiates proceedings in Australia’s High Court to test the validity of the law as it relates to property rights of two brands, Winfield and Dunhill. Press release ▸Writ of summons (British American Tobacco Australasia Limited) 6 December Imperial Tobacco similarly initiates proceedings ▸Writ of summons (Imperial Tobacco Australia Limited) 15 December Japan Tobacco files a Writ of Summons against the Australian Government alleging that the plain packaging law infringes its trademark rights ▸Writ of summons ( Japan Tobacco International SA) 20 December Philip Morris Australia files a Writ of Summons in the High Court against the Australian Government alleging that the plain packaging law infringes its trademark and property rights ▸Writ of summons (Philip Morris Limited) and Annexures 22 December Release of new Information Standard specifying enlarged graphic health warnings for tobacco products Competition and Consumer (Tobacco) Information Standard 2011 to start phase in from 1 January 2012 2012 17–19 April High Court case hearing of Full Court opens in Canberra 27 June Introduction of Customs Amendment (Smuggled Tobacco) Bill 2012 15 August High Court of Australia announces tobacco industry’s case against Australian Government has failed and awards costs to the government. Judgement to be released at later date. Further information 1 October All packs manufactured to be plainly packaged 6 November Assent of Customs Amendment (Smuggled Tobacco) Act 2012 1 December All packs sold to be in plain packaging with prescribed enlarged health warning. Only the first set of graphic health warnings are to be displayed on retail packaging of cigarettes and most other smoked tobacco products (excluding and bidis). Cigars and bidis are required to display different product specific warnings 2013 1 February Automatic (biannual since 1985) indexation of tobacco excise/customs duty in line with increase in Consumer Price Index June 2012 to December 2013 1 August Automatic (biannual since 1985) indexation of tobacco excise/customs duty, in line with increase in Consumer Price Index December 2012 to June 2013 andGraphic health warnings from the second set of warnings start to be displayed on relevant packaging; packs with warnings from the first set of warnings may continue to be manufactured until the end of November 2013 1 December 12.5% increase in excise/customs duty. Only packs bearing one of the second set of graphic health warnings may be manufactured from this date 2014 1 March First of automatic indexation increases in excise and customs duty in line with changes in Average Weekly Earnings Source: a more detailed timeline can be viewed on the Plain Facts website, Cancer Council Victoria, at https://www.cancervic.org.au/plainfacts/browse.asp? ContainerID=timelineandinternationaldevelopments.

ii10 Scollo M, et al. Tob Control 2015;24:ii9–ii16. doi:10.1136/tobaccocontrol-2014-052073 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication

Executive Council on 7 December 201120 21 set out require- tobacco packaging.21 For retail packaging of cigarettes, Regulation ments for the retail packaging and appearance of cigarettes and 2.4.1 specifies they must be in Lucida Sans font type and that the were amended in early 201222 to incorporate the additional speci- brand name can be no larger than 14 points in size and the variant fications for plain packaging of non-cigarette tobacco products. name no larger than 10 points in size. The font must be normal The legislation was challenged by tobacco companies in the High weighted regular font and in a light grey colour known as Pantone Court of Australia which heard the case in April 2012 and upheld Cool Gray 2C. Only the first letter in each word may be capitalised the legislation in its ruling released in August 2012.23 and no other upper case letters may be used. Regulation 2.4.2 sti- pulates similar provisions for non-cigarette tobacco products. New enlarged GHWs The placement of the brand and variant name on the pack- GHWs have been required on most tobacco products in aging is also tightly regulated under Sections 21 (2) and (3) of Australia since 2006. Prior to late 2012, cigarette packs and the Tobacco Plain Packaging Act 2011. The brand name may cartons had displayed GHWs on 30% of the front of pack and appear once only on each of the front, top and bottom surfaces 90% of the back in two rotating sets of seven warnings. An of cigarette packs, must appear across one line only, oriented information message was also required on the side of the horizontally and centred in the relevant space beneath the pack.24 Evaluation of these health warnings conducted in health warning. The variant name must appear horizontally 200825 suggested that while the new graphic warnings had and immediately below the brand name—see front of pack in greater impact with consumers than the previous text-only figure 1B. Regulations 2.4.3 and 2.4.4 set out corresponding warnings, effectiveness could be further improved with clearer requirements for non-cigarette packaging. and larger, front-of-pack warnings, updated and refreshed images and simple and clear text messages in larger, clearer font Cigarette packaging features and simplified side-of-pack text. The evaluation report also Market research determined that innovative packaging shape, recommended extension of GHWs to other tobacco products size and opening affected the level of appeal and brand person- and introduction of plain packaging to remove conflict and ality of cigarettes9 (confirmed by experimental research31 32) competition between the pack design and health warnings.25 and, as a result, the dimensions, shape and opening of cigarette During 2010 and 2011—the same period that plain pack- packs were restricted. ‘Soft packs’—demonstrated in the aging legislation was being drafted—the Department of Health research to be perceived as more masculine—were also prohib- and Ageing commissioned further market research to guide ited. Under Section 18 of the Act, cigarette packs must also be – development of new, larger GHWs.26 29 This research examined made of rigid material and must not have any decorative ridges, potential new images and warning statements, more detailed embossing or embellishments. Regulation 2.1.1 specifies explanatory messages to appear on the back of pack and infor- minimum and maximum height, width and depth for cigarette mation messages to appear on the side of packs. This and packaging and that packs must have a flip-top lid. related research also tested different colours and layouts for warning messages and options for various sizes of GHWs.82729 Allowable information on cigarette sticks This research demonstrated that the proposed 75% In light of results of market research which showed that brand- front-of-pack GHWs were more immediately noticeable, serious ing and decorative features are strongly associated with level of and difficult to avoid than the previous 30% warnings.811 appeal and brand personalities,9 Section 26 of the Act bans tra- Prior to finalising the new health warnings, the Australian demarks or other marks from appearing on tobacco products, Competition and Consumer Commission undertook three thus outlawing brand or variant names and other embellish- public consultations between September 2011 and December ments from appearing on cigarette sticks (see figure 2). 2011 and the new Competition and Consumer (Tobacco) Regulations require the paper casing for cigarette sticks to be all Information Standard 2011 prescribed under the Australian white or white with a cork tip (Regulation 3.1.1). While they Consumer Law in Schedule 2 to the Competition and Consumer allow an alphanumeric code in a specified font style, size, colour Act 2010 was made on 22 December 2011.30 and location to appear on sticks, the type of code is restricted, including a requirement that it must not constitute tobacco REQUIREMENTS FOR PLAIN PACKAGING advertising or promotion (Regulation 3.1.2). Figure 1 depicts Australia’s leading brand of cigarettes (A) before and (B) after the introduction of tobacco plain packaging Other information allowable in Australia. The Tobacco Plain Packaging Regulations also allow for a ‘meas- urement mark’ as required by Division 4.4 of the National Packaging colour Trade Measurement Regulations 200933 to be placed on the Eight different colours were market tested for plain packaging side of the pack along with the name, address and phone and a drab dark brown was found to be the least appealing and number of the person who packed the product or on whose attractive, had the lowest potential for the pack to mislead con- behalf it was packed (see Regulations 2.3.4, 2.3.8 and 2.3.9). sumers about the harms of tobacco, and increased the impact of This information must be in the typeface known as Lucida Sans, GHWs.9 Regulation 2.2.1 (2) specifies that Pantone 448C is the no larger than 10 points in size, normal weighted regular font exact shade of drab dark brown which must be used on the and in a light grey colour known as Pantone Cool Gray 2C. outer surfaces of tobacco packaging. A barcode is also allowed on the side of packs under regulation 2.3.5. The barcode must be coloured either black and white or Specifications for display of brand names the colour Pantone 448C (drab dark brown) and white—see Market research determined the font style, size and colour to be example here. It must not form a picture, symbol or design. used for brand and variant names as well as their location on the An ‘origin mark’ is also allowed to assist with strategies to packaging that provided for legibility in the retail setting while prevent counterfeiting and illicit trade. Regulation 2.3.2 states maintaining the public health objectives of the Act.9 The that the origin mark must be either covert and not visible to the Regulations detail how brand and variant names must appear on naked eye or an alphanumeric code—see example here.

Scollo M, et al. Tob Control 2015;24:ii9–ii16. doi:10.1136/tobaccocontrol-2014-052073 ii11 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication

Figure 1 (A) Typical Australian tobacco packs, back and front before…(B)…and after the introduction of tobacco plain packaging. Source: Quit Victoria collection, 2012.

REQUIREMENTS FOR ENLARGED GHWS Health warning size and elements Similar to the 2004 Regulations, the Competition and Consumer (Tobacco) Information Standard 2011 (the Standard) requires that retail packaging of most tobacco products contain a combination of warning statement, graphic image, explanatory message and information message. The size of GHWs on cigar- ette packaging increased from 30% to 75% of the front surface —see figure 1. The 2011 Standard maintained 90% of the back surface for warnings including text, image and explanatory mes- sages on cigarette packs and cartons—see figure 1. An increase to 75% of the front and back surface was required of most other tobacco products including roll-your-own (RYO) and pipe tobacco pouches, cigar packs and shisha (tobacco for use in water pipes). Figure 2 Horizon cigarette sticks before and after introduction of The side-of-pack Information Message (previously a single legislation. Source: Quit Victoria pack collection, December 2012. message required to be white text on a black background) from

ii12 Scollo M, et al. Tob Control 2015;24:ii9–ii16. doi:10.1136/tobaccocontrol-2014-052073 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication

Figure 3 Winfield Blue 25s: packs showing the first set of seven health warnings. Source: Quit Victoria pack collection, December 2012.

1 December 2012 onwards now has to be one of several rotat- The five graphic warnings for retail packaging for cigars ing messages (with information about chemicals in tobacco (other than cigar tubes) as detailed in Part 5 (items 5.2–5.6) of smoke, paired with, and relevant to, the particular GHW on the Standard are ‘Cigar smoking causes mouth cancer’, ‘Cigar each pack) presented in black text on a yellow background—see smoking causes lung cancer’, ‘Cigar smoking is not a safe alter- here for image. This requirement was in line with market native’, ‘Cigar smoking causes throat cancer’ and ‘Cigar research which determined this updated format to be more smoking damages your teeth and gums’. Part 6 of the Standard noticeable.826 states that the same five cigar warning statements are required on single cigar tubes as a text-only warning—see example here. The warning statements in Part 6, item 6.2 can be viewed at: Quitline logo http://www.comlaw.gov.au/Details/F2013C00598/Html/Text#_ ‘ ’ The Quitline logo and phone number, Quitline 137 848 ,is Toc363806264. The text-only warnings for bidis (Part 7) and required to appear overlaid on the graphic on the back of for smokeless tobacco (Part 8) can be viewed at: http://www. — tobacco packages that have them (Part 1, Section 1.3 (6)) see comlaw.gov.au/Details/F2013C00598. fi — back-of-pack images in gure 1 and on the graphic on the Part 9, Division 4 of the Standard provides diagrams of how front of cigar packs. the individual requirements for the health warnings should be set out for different package formats. Health warnings for different types of tobacco products The new GHWs developed as a result of market testing covered LEGISLATED TIMING OF TRANSITION TO PLAIN a broader range of topics and mix of image styles than the previ- PACKAGING AND ROLL-OUT OF NEW HEALTH WARNINGS ous warnings, including a stronger emotional component and a Plain packaging greater emphasis on morbidity than mortality. Fourteen warning All tobacco packs manufactured or packaged in Australia for statements with corresponding graphic images, explanatory mes- domestic consumption from 1 October 2012 and all packs sold sages and information messages are required on cigarettes and from 1 December 2012 were required to be in plain packaging 17 smoking tobacco (RYO tobacco and pipe tobacco) as specified in as described in detail above. the Standard in Part 3 (items 3.2–3.8; the first set of health warnings; figure 3) and Part 4 (items 4.2–4.8; the second set of Health warnings health warnings; figure 4). Each of the warnings can be viewed The Competition and Consumer (Tobacco) Information at: http://www.comlaw.gov.au/Details/F2013C00598/Html/Text#_ Standard 2011 (the Standard)30 allowed a ‘phase-in’ period. Toc363806248. Between 1 January 2012 and 30 November 2012 inclusive, sup- All tobacco products including packs of cigars and single pliers could choose to comply with the Standard30 or with the cigars, bidis and smokeless tobacco are required to bear health 2004 Regulations24 previously in force. All tobacco products warnings. Hence, the Australian Standard is aligned with Article supplied to consumers had to be in packaging that complied 11 Guidelines for the WHO Framework Convention on with the Standard from 1 December 2012. Tobacco Control which state that “there should be no exemp- Rotation requirements for the health warnings are set out in tions for small volume companies or brands or for different Part 9 of the Standard for cigarettes and other smoked tobacco types of tobacco products” (see WHO FCTC Article 11 products. The first set of warnings (described in Part 3 of the Guidelines,34 Product Category Considerations, Section 46). Standard) could be displayed on packs sold in retail outlets

Figure 4 Winfield Blue 25s: packs showing the second set of seven health warnings. Source: Quit Victoria, December 2013.

Scollo M, et al. Tob Control 2015;24:ii9–ii16. doi:10.1136/tobaccocontrol-2014-052073 ii13 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication

Figure 5 Adjusted proportion of respondents with packs purchased in Australia which were self-reported as plain and with Set 2 warnings on their current . Notes: Transition to plain packs: Cigarette smokers (total n=8679; analysed and plotted n=7659). We excluded n=49 who currently smoked unbranded ‘chop-chop’ tobacco and n=234 who did not report their current brand name and thus did not get asked the plain packaging question. We also excluded those who purchased duty free or overseas (n=74), or did not know where the pack was purchased (n=305), in order to limit the analysis to packs purchased in Australia. Finally, it excludes those who were not using their original pack (n=24) or who refused or responded not applicable or did not know (n=162), and those with missing demographic information (n=172). Rollout of Set 2 warnings: Cigarette smokers who had their current pack of cigarettes/tobacco with them (total n=1726; analysed and plotted n=1553). We excluded n=13 who currently smoke unbranded ‘chop-chop’ tobacco, and thus did not get asked to identify their pack’s health warning, those who purchased duty free or overseas (n=8), did not know where the pack was purchased (n=23), who reported a non-Australian or no health warning (n=52), those who refused or did not know (n=21), and with missing demographic information (n=56). between 1 January 2012 and 30 November 2012 (prior to the not correct or exclude obvious reporting errors occurring in a full implementation of tobacco plain packaging) and were the small minority of respondents, such as packs being reported as only warnings that could be displayed between 1 December plain many months prior to October 2012. In order to monitor 2012 and 31 July 2013, with near-as-possible-to-equal fre- the roll-out of the second set of health warnings, from August quency required for all seven warnings in this period (and for 2013 to the end of the survey period, cigarette smokers were 8 months starting 1 December for each subsequent even- asked to report the health warning on their current pack using numbered year). two questions. The first asked current smokers if they have their The second set of health warnings (described in full in Part 4 current pack with them, with those answering ‘yes’ asked to of the Standard) could be displayed on packs manufactured read out the bold white warning statement on the front of their from 1 August 2013 to 30 November 2014 and each had to be pack. displayed with as near-as-possible-to-equal frequency for the As described in detail in Scollo et al36 in this volume, 8 months starting 1 December 2013 (and for 8 months starting smokers were also asked about current brand smoked. Brands 1 December of each odd-numbered year). were then coded by manufacturer and market segment Either set of warnings could be displayed on packs manufac- (premium, mainstream or value, based on classifications tured from 1 August to 30 November 2013. This pattern— adopted by the Retail World magazine). 8 months of the first set only, 4 months of both, 8 months of Analyses of the proportion of packs that were plain and the the second set only—will continue for the life of the Standard. proportion of packs with Set 2 warnings were undertaken by month or week using logistic regression, adjusting for age REAL-WORLD TRANSITION TO PLAIN PACKAGING AND (18–29; 30–49; 50+ years), sex, education (less than high ROLL-OUT OF NEW HEALTH WARNINGS school; high school or some tertiary; tertiary and above), and The transition to plain packaging and the roll-out of Set 2 warn- area socioeconomic status (low, mid and high using the 2011 ings was monitored as part of a national continuous cross- index of relative disadvantage37) and Heaviness of Smoking sectional survey with approximately 100 interviews completed Index (0: low to 6: high38). per week between April 2012 and March 2014, detailed in Figure 5 shows the adjusted proportion of current tobacco Wa k e field et al35 in this volume. Telephone interviews with packages (including cigarette packs and RYO pouches) pur- adult (18–69 years) smokers and recent quitters (quit in the last chased in Australia which were self-reported by cigarette year) were conducted using a dual-frame sample design, with smokers to be plain, by month. On average, 5.1% of packs were half of respondents approached via landline random digit dial- reported to be plain between April and September 2012. There ling (RDD) and half by mobile phone RDD (overall response was a rapid increase in the percentage of packs reported to be rate 57%). All participants who smoked daily or weekly were plain between October and December 2012, with 14.3% classified as smokers and those who smoked less often (monthly reported as plain in October, 57.8% in November and 94.5% or less-than-monthly) were allowed to self-identify as either a in December. From January 2013, 95.1% of packs were current smoker or recent quitter. Transition to plain packaging reported as plain. was assessed among 7659 cigarette smokers. To provide a more detailed analysis of the rate of transition to Over the entire survey period, cigarette smokers—smokers of plain packaging, we examined the percentage of cigarette packs either factory-made (FM) cigarettes or RYO cigarettes—were which were self-reported as plain over each of the weeks in asked whether or not their current cigarette pack or RYO pouch October, November and December 2012. Plain packs began to was a plain dark colour with all of its logos removed. We did be used by smokers from mid-October, so that in the last half of ii14 Scollo M, et al. Tob Control 2015;24:ii9–ii16. doi:10.1136/tobaccocontrol-2014-052073 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication

October 2012, a total of 21.2% of packs were reported as plain. take account of the very gradual roll-out of the Set 2 health This accelerated rapidly, increasing to 57.4% in the week begin- warnings. ning 11 November, 73.4% in the last half of November and then 91.9% in week beginning 2 December. In interaction ana- lyses, the rate of transition over each of the weeks in October, What this paper adds November and December 2012 did not differ by state of resi- dence (F=0.94, p=0.565), cigarette type (FM to RYO; F=1.17, ▸ p=0.313), or market segment (premium, mainstream, value; The Tobacco Plain Packaging Act 2011 and the Competition F=1.00, p=0.455) or tobacco manufacturer (British American and Consumer (Tobacco) Information Standard 2011 have Tobacco Australia (BATA), Philip Morris (PM), Imperial Tobacco, comprehensively altered the appearance of tobacco products other; F=1.06, p=0.386) of the current brand, suggesting that in Australia. ▸ plain packs emerged at a similar rate throughout the market. Manufacturers implemented enlarged graphic health The roll-out of Set 2 GHWs on plain packs was also exam- warnings at the same time as plain packaging though they ined among the 1553 cigarette smokers interviewed from were free to introduce them earlier. ▸ August 2013 to March 2014. In August 2013, 17.8% of packs Overall, plain packaging transitioned as intended over a depicted a Set 2 warning, compared with 30.3% in September, relatively short period between the required dates of 64.3% in October and 76.8% in November. From 1 December manufacture and retail sale in late 2012. The second set of 2013, on average, 89.8% of packs were reported to have a Set 2 enlarged graphic health warnings rolled out gradually from warning, with 96.2% of packs having a Set 2 warning by March August 2013 with almost all packs bearing the second set of 2014. Over the period from August 2013 through December warnings by the end of February 2014. ▸ 2013, the proportion of packs per week with the second set of The new standardised packs were available and likely health warnings did not differ by state of residence (F=0.93, already exerting an impact in the Australian market from p=0.628), cigarette type (FM or RYO; F=1.23, p=0.235), October 2012 onwards, well before the 1 December market segment (premium, mainstream, value; F=1.00, mandated introduction date. p=0.465) or tobacco manufacturer (BATA, PM, Imperial Tobacco, other; F=1.34, p=0.107), suggesting that Set 2 warn- ings emerged at a similar rate throughout the market. Contributors MW and MS conceived of this paper. MS and KL drafted the It was notable that, well after full implementation of plain manuscript with assistance from MB. KC undertook analysis of data on fi packaging (on 1 December 2012), about 5% of cigarette implementation. All authors contributed to nalisation of the manuscript. smokers consistently reported their tobacco packs were not ‘a Funding Production of this paper was supported by Cancer Council Victoria. The plain dark colour with all logos removed’. This could not be National Plain Packaging survey was funded under a contract with the Australian Government Department of Health and Ageing. explained by cigarette smokers buying their packs from overseas as such smokers were excluded, so some might interpret this as Competing interests The authors wish to advise that MS was a technical writer for and MW was a member of the Tobacco Working Group of the Australian an indication of use of non-compliant packs or even use of illicit National Preventive Health Task Force and MW and KL were members of the Expert tobacco, an issue explored in more detail in Scollo et al39 in this Advisory Committee on Plain Packaging that advised the Australian Department of volume. However, we consider it more likely that a large com- Health on research pertaining to the plain packaging legislation. MW holds ponent of this 5% would be misreporting (not completely competitive grant funding from the Australian National Health and Medical Research “ Council, US National Institutes of Health, Australian National Preventive Health understanding the question). Respondents were asked is the Agency and BUPA Health Foundation. pack/package you are currently smoking from a plain dark ” Ethics approval The survey was approved by the Cancer Council Victoria Human colour with all of its logos removed? Interviewers were Ethics Committee (HREC 0018). instructed to code ‘no’ if the respondent indicated that some Provenance and peer review Not commissioned; externally peer reviewed. logos remained. It is probable that some smokers interpreted the brand or variant name (still allowed under the plain pack- Open Access This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which aging Act) to be a logo. It should be noted that after February permits others to distribute, remix, adapt, build upon this work non-commercially, 2013 (by which time older style packs might be expected to and license their derivative works on different terms, provided the original work is have been consumed) of the 180 smokers interviewed who said properly cited and the use is non-commercial. See: http://creativecommons.org/ that their packs were purchased in Australia but were ‘not licenses/by-nc/4.0/ plain’, 137 also reported warnings 50% or larger. Fifteen reported GHWs less than 50% (most likely packs that complied REFERENCES with legislation in force in 2012) and 10 reported that their 1 Preventative Health Taskforce. Preventative Health Taskforce. Canberra: pack did not have a GHW on the front of the pack (these may Commonwealth of Australia, 2008. [updated 21 May 2008]. http://www. have been illicit). The level of apparent non-compliance after preventativehealth.org.au/internet/preventativehealth/publishing.nsf/Content/home-1 (accessed Jul 2014). implementation of plain packaging would appear to be similar 2 Tobacco Working Group. Technical report no. 2. Tobacco in Australia: making to the level of misreporting before implementation: 5.1% of smoking history. Canberra: National Preventative Health Taskforce, 2008. http:// respondents during April to September 2012—well before plain www.preventativehealth.org.au/internet/preventativehealth/publishing.nsf/Content/ packaging—conversely reported that they did possess a pack tech-tobacco (accessed Jul 2014). that was plain dark brown with all logos removed despite plain 3 Preventative Health Taskforce. Australia: the healthiest country by 2020: A discussion paper. Preventative Health Taskforce: Commonwealth of Australia, packs not being on the market at this time. 2008. http://www.preventativehealth.org.au/internet/preventativehealth/publishing. Studies evaluating the impact of plain packaging and enlarged nsf/Content/discussion-healthiest (accessed Jul 2014). GHWs need to take account of the fact that the new standar- 4 Preventative Health Taskforce. Engagement and consultation. Canberra: dised packs were available and likely already exerting an impact Commonwealth of Australia, 2008. http://www.preventativehealth.org.au/internet/ preventativehealth/publishing.nsf/Content/engagement-and-consultation-1lp in the Australian market from October 2012 onwards, well (accessed Jul 2014). before the 1 December mandated introduction date. Similarly, 5 Preventative Health Taskforce. Australia: the healthiest country by 2020. National studies examining the effects of specific health warnings need to Preventative Health Strategy. Canberra: Commonwealth of Australia, 2009. http://

Scollo M, et al. Tob Control 2015;24:ii9–ii16. doi:10.1136/tobaccocontrol-2014-052073 ii15 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Special communication

www.preventativehealth.org.au/internet/preventativehealth/publishing.nsf/Content/ 20140211204134/http:/www.yourhealth.gov.au/internet/yourhealth/publishing.nsf/ national-preventative-health-strategy-1lp (accessed 10 Jun 2008). Content/plainpack-tobacco-non-cig (accessed Jul 2014). 6 Liberman J, Scollo M, Freeman B, et al. in Australia: the 23 British American Tobacco Australasia Limited and Ors v. The Commonwealth of historical and social context. In: Voon T, Mitchell AD, Liberman J, et al., eds. Public Australia. High Court of Australia, 2012. health and plain packaging of cigarettes. Cheltenham, UK: Edward Elgar Publishing, 24 Trade Practices (Consumer Product Information Standards) (Tobacco) Regulations, 2013. Stat. Statutory Rules 2004 no.264. (29 March 1994, 2004 (Cth)). http://www. 7 Australian Government. Taking preventative action: government’s response to comlaw.gov.au/Details/F2007C00131 Australia: the healthiest country by 2020. Canberra: Department of Health and 25 Elliot and Shanahan Research. Evaluation of the effectiveness of the graphic health Ageing, 2010. http://www.health.gov.au/internet/preventativehealth/publishing.nsf/ warnings ontobacco product packaging 2008. Canberra: Department of Health and Content/taking-preventative-action (accessed 21 May 2010). Ageing, 2009. http://webarchive.nla.gov.au/gov/20140801094920/http://www. 8 Gfk Blue Moon. Market research reports on tobacco plain packaging and graphic health.gov.au/internet/main/publishing.nsf/Content/phd-tobacco-eval-graphic-health- health warnings. Canberra: Department of Health and Ageing, 2011. http://www. warnings-full-report (accessed Jul 2014). health.gov.au/internet/main/publishing.nsf/Content/mr-plainpack (accessed May 26 Gfk Blue Moon. Market testing of new health warnings and information messages 2014). for tobacco product packaging: phase 2 front and back of pack messages. 9 Gfk Blue Moon. Market research to determine effective plain packaging of Canberra: Department of Health and Ageing, 2011. http://www.health.gov.au/ tobacco products. Canberra: Department of Health and Ageing, 2011. http://www. internet/main/publishing.nsf/Content/mr-plainpack (accessed May 2014). health.gov.au/internet/main/publishing.nsf/Content/mr-plainpack (accessed May 27 Gfk Blue Moon. Market testing of new health warnings and information messages 2014). for tobacco product packaging: phase 1 side of pack messages. Canberra: 10 Gfk Blue Moon. Market research to determine impact of plain packaging on other Department of Health and Ageing, 2011. http://www.health.gov.au/internet/main/ tobacco products. Canberra: Department of Health and Ageing, 2011. http://www. publishing.nsf/Content/mr-plainpack (accessed May 2014). health. gov.au/internet/main/publishing.nsf/Content/mr-plainpack (accessed May 2014). 28 Gfk Blue Moon. Market testing of new health warnings and information messages 11 Gfk Blue Moon. Market research to determine effective plain packaging of tobacco for tobacco product packaging: phase 3 refinement of health messages. Canberra: products. Canberra: Department of Health and Ageing, 2011. http://www.health. Department of Health and Ageing, 2011. http://www.health.gov.au/internet/main/ gov.au/internet/main/publishing.nsf/Content/mr-plainpack (accessed May 2014). publishing.nsf/Content/mr-plainpack (accessed May 2014). 12 Tobacco Plain Packaging Bill 2011—Exposure draft, http://webarchive.nla.gov.au/ 29 Gfk Blue Moon. Market testing of new health warnings and information messages gov/20140211201716/http://www.yourhealth.gov.au/internet/yourhealth/publishing. for tobacco product packaging: premium cigars, cigarillos/little cigars and roll your nsf/Content/tpp-bill2011 own. Canberra: Department of Health and Ageing, 2011. http://www.health.gov.au/ 13 Department of Health and Ageing. Public consultation on plain packaging of internet/main/publishing.nsf/Content/mr-plainpack (accessed May 2014). cigarette products. Canberra: Australian Government, 2011. [updated 9 February 30 Competition and Consumer (Tobacco) Information Standard, (2011). http://www. 2012]. http://webarchive.nla.gov.au/gov/20140211204139/http:/www.yourhealth. comlaw.gov.au/Details/F2013C00598 gov.au/internet/yourhealth/publishing.nsf/Content/plainpack-non-cig-tob-products 31 Hammond D, Dockrell M, Arnott D, et al. Cigarette pack design and perceptions of (accessed Jul 2014). risk among UK adults and youth. Eur J Public Health 2009;19:631–7. 14 Department of Health and Ageing. Public consultation on plain packaging of 32 Borland R, Savvas S, Sharkie F, et al. The impact of structural packaging design on cigarette products: consultation paper. Canberra: Australian Government, 2011. young adult smokers’ perceptions of tobacco products. Tob Control 2013;22:97–102. http://webarchive.nla.gov.au/gov/20140211201705/http://www.yourhealth.gov.au/ 33 National Trade Measurement Regulations (2009). http://www.comlaw.gov.au/Series/ internet/yourhealth/publishing.nsf/Content/tpp-paper (accessed Jul 2014). F2009L03479 15 Tobacco Plain Packaging Bill 2011—Exposure draft: Explanatory memorandum, http:// 34 Conference of the Parties to the World Health Organization Framework Convention parlinfo.aph.gov.au/parlInfo/search/display/display.w3p;query=Id%3A%22legislation% on Tobacco Control. Elaboration of guidelines for implementation of Article 11 of 2Fems%2Fr4613_ems_d7b0bff9-4a09-4107-8b69-002903b648af%22 the Convention. World Health Organization, 2008. http://www.who.int/gb/fctc/PDF/ 16 Department of Health and Ageing. Consultation Paper and the submissions received cop3/FCTC_COP3_7-en.pdf for the consultation on plain packaging of non-cigarette tobacco products. 35 Wakefield M, Coomber K, Zacher M, et al. Australian adult smokers’ responses to Canberra: Australian Government, 2011. http://webarchive.nla.gov.au/gov/ plain packaging with larger graphic health warnings 1 year after implementation: 20140211204134/http://www.yourhealth.gov.au/internet/yourhealth/publishing.nsf/ results from a national cross-sectional tracking survey. Tob Control 2015;24:ii17– Content/plainpack-tobacco-non-cig (accessed Jul 2014). 25. 17 Tobacco Plain Packaging Act (since amended), No. 148 (2011). http://www. 36 Scollo M, Zacher M, Coomber K, et al. Changes in use of types of tobacco products comlaw.gov.au/Details/C2011A00148 by pack sizes and price segments, prices paid and consumption following the 18 Tobacco Plain Packaging Act, No. 148 (2011). http://www.comlaw.gov.au/Details/ introduction of plain packaging in Australia. Tob Control 2015;24:ii66–75. C2013C00190 37 Australian Bureau of Statistics. Technical paper: socio-economic indexes for areas 19 Trade Marks Amendment (Tobacco Plain Packaging) Act, No. 149, 2011 (2011). (SEIFA) 2011. Canberra, Australian Capital Territory: ABS, 2013. http://www.abs. http://www.comlaw.gov.au/Details/C2011A00149 gov.au/websitedbs/censushome.nsf/home/seifa2011?opendocument&navpos=260 20 Tobacco Plain Packaging Regulations (since amended), SLI 2011 No. 263 (2011). (accessed Jul 2014). http://www.comlaw.gov.au/Details/F2011L02644/Download 38 Heatherton TF, Kozlowski LT, Frecker RC, et al. Measuring the heaviness of 21 Tobacco Plain Packaging Regulations, SLI 2011 No. 263 (2011). http://www. smoking: using self-reported time to the first cigarette of the day and number of comlaw.gov.au/Details/F2013C00801 cigarettes smoked per day. Br J Addict 1989;84:791–9. 22 Department of Health and Ageing. Consultation on draft Regulations incorporating 39 Scollo M, Zacher M, Coomber K, et al. Use of illicit tobacco following introduction requirements for non-cigarette tobacco products. Canberra: Australian Government, of standardised packaging of tobacco products in Australia: results from a national 2011. [updated 9 February 2012]. http://webarchive.nla.gov.au/gov/ cross-sectional survey. Tob Control 2015;24:ii76–81.

ii16 Scollo M, et al. Tob Control 2015;24:ii9–ii16. doi:10.1136/tobaccocontrol-2014-052073 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Standardised packaging and new enlarged graphic health warnings for tobacco products in Australia −−legislative requirements and implementation of the Tobacco Plain Packaging Act 2011 and the Competition and Consumer (Tobacco) Information Standard , 2011 Michelle Scollo, Kylie Lindorff, Kerri Coomber, Megan Bayly and Melanie Wakefield

Tob Control 2015 24: ii9-ii16 doi: 10.1136/tobaccocontrol-2014-052073

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii9

These include: References This article cites 2 articles, 1 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii9#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper Australian adult smokers’ responses to plain packaging with larger graphic health warnings 1 year after implementation: results from a national cross-sectional tracking survey Melanie Wakefield, Kerri Coomber, Meghan Zacher, Sarah Durkin, Emily Brennan, Michelle Scollo

▸ Additional material is ABSTRACT set of 14 new graphic health warnings (GHWs)— published online only. To view Background We assessed whether the Australian plain enlarged from 30% to 75% of the front of tobacco please visit the journal online — (http://dx.doi.org/10.1136/ packs with larger graphic health warnings (GHWs) packs while maintaining 90% of the back seven tobaccocontrol-2014-052050). achieved three specific objectives of reducing the appeal of which appeared on new plain packs as they of tobacco, increasing health warning effectiveness and rolled out onto the market. The selection of the Centre for Behavioural Research in Cancer, Cancer reducing the ability of packaging to mislead about colour for Australian plain packs and the exact Council Victoria, Melbourne, smoking harms. content, size and placement of the refreshed and Victoria, Australia Methods We compared responses from continuous enlarged GHWs were determined after a series of cross-sectional telephone surveys of n=2176 cigarette experimental and qualitative studies to test which Correspondence to – Professor Melanie Wakefield, smokers during pre-plain packaging (April September pack colour consumers found least appealing, and Centre for Behavioural 2012, pre-PP) with n=759 surveyed in the transition the nature and format of GHWs on plain packs – Research in Cancer, Cancer period (October–November 2012) and n=4240 during that were most salient, relevant and believable.4 7 Council Victoria, 615 St Kilda the first year of implementation (December 2012– Most smokers reported using plain packs with Rd, Melbourne, VIC 3004, November 2013, PP year 1), using multivariate logistic enlarged GHWs during November 2012.28Plain Australia; melanie.wakefield@ cancervic.org.au regression analyses. packs with the second set of seven larger GHWs Results From pre-PP to PP year 1, more smokers rolled out from August 2013 and were used by Received 25 September 2014 disliked their pack (p<0.001), perceived lower pack most during October 2013.2 Accepted 16 December 2014 appeal (p<0.001), lower cigarette quality (p<0.001), The specific aims of the Tobacco Plain Packaging lower satisfaction (p<0.001) and lower value (p<0.001) Act 20111 were to: (A) reduce the appeal of and disagreed brands differed in prestige (p=0.003). tobacco products to consumers; (B) increase the There was no change in perceived differences in taste of effectiveness of health warnings on retail packaging different brands. More smokers noticed GHWs of tobacco products and (C) reduce the ability of (p<0.001), attributed much motivation to quit to GHWs retail packaging of tobacco products to mislead (p<0.001), avoided specific GHWs when purchasing consumers about the harmful effects of smoking. (p<0.001), and covered packs (p<0.001), with no Through the achievement of these aims in the change in perceived exaggeration of harms. PP year 1 longer term, as part of a comprehensive suite of saw an increased proportion believing that brands do tobacco control methods, it was intended that the not differ in harmfulness (p=0.004), but no change in legislation would contribute to efforts to reduce the belief that variants do not differ in strength or the smoking rates by discouraging people from taking perceived harmfulness of cigarettes compared with a up smoking, encouraging smokers to quit and dis- year ago. Interactions signified greater change for four couraging relapse.1 outcomes assessing aspects of appeal among young The evidence base for the plain packaging legisla- adults and two appeal outcomes among mid-aged tion was supported not just by many qualitative adults. and experimental research studies comparing plain – Conclusions The specific objectives of plain packaging packs with increasingly or fully branded packs,9 11 were achieved and generally sustained among adult but by several decades of research that had conclu- smokers up to 12 months after implementation. sively established the causal role of tobacco market- ing in the initiation of tobacco use and the beneficial impact of tobacco marketing restrictions Open Access – Scan to access more INTRODUCTION on reducing tobacco uptake and use.12 14 The legis- free content In late 2012, Australia became the first nation to lation represented the next logical progression for implement plain packaging of tobacco products.1 tighter marketing restrictions in a nation where As described further in Scollo et al2 (this supple- packaging had become one of the few remaining ment), Australian tobacco manufacturers were avenues for continued tobacco promotion15 (this required to produce only plain packs from 1 supplement). October 2012 and retailers were prohibited from A large number of qualitative and experimental – selling anything but these from 1 December 2012. studies summarised in recent reviews,9 11 including To cite: Wakefield M, Packs were required to be a standard drab dark several naturalistic studies where smokers were Coomber K, Zacher M, et al. brown, with brand and variant names appearing in assigned to carry and smoke from fully branded or – Tob Control 2015;24: a standard font and position. At the same time, a plain packs over several weeks,16 18 have demon- – ii17 ii25. complementary Information Standard3 mandated a strated plain packaging can reduce elements of the

Wakefield M, et al. Tob Control 2015;24:ii17–ii25. doi:10.1136/tobaccocontrol-2014-052050 ii17 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper appeal of tobacco products, as judged by measures of reduced female at-home respondents,37 interviewers asked to speak to the attractiveness, liking, prestige, stylishness, quality, satisfaction youngest male aged 18–69 years, and if not available, the young- – and taste,16 28 and that smokers of plain packs are associated est female. The mean monthly response rate, adjusted for those with less desirable personality attributes.19 20 24 25 28 29 who declined to be formally screened but may have been Reductions in measures of appeal have been repeatedly evident in-scope for the survey, was 57% and this did not vary over the in studies of youth and young adults,10 11 consistent with the survey period. Further detail on survey methods is available in a critical role played by brand image during youth and young Technical Report (see Supplementary Technical report).38 adulthood, the period when smoking behaviour develops and The current study used responses from smokers of factory- brand preferences are established.12 13 Experimental and natur- made or roll-your-own cigarettes interviewed between April alistic studies also suggest plain packaging makes health warn- 2012 and November 2013 inclusive, corresponding to 6 months ings more salient, with warnings being better recalled and pre-plain packaging (pre-PP, April–September 2012), 2 months considered more serious and believable.10 11 16 30 31 Studies of transition (October/November 2012) and 1 year after full also suggest that plain packaging can redress smokers’ miscon- implementation (December 2012–November 2013). After ceptions about product harm, although findings have been more excluding n=149 respondents who did not provide valid data variable, depending on the plain pack colour used and presence on all demographic covariates, we analysed responses from of descriptive variant names.10 11 A common feature of the 7175 cigarette smokers (weighted n; unweighted n=7133). qualitative and experimental studies of plain packaging con- ducted to date is that most have involved only a short period of Outcome measures exposure, so responses may be attributable to pack novelty. Questionnaire measures were adapted from other population Population-based study of consumer responses where plain surveys such as the Australian arm of the International Tobacco packaging has become the norm such as in Australia is import- Control survey (http://www.itcproject.org/surveys) and tapped ant for understanding ‘real world’ effects. similar constructs to those used in past studies to assess appeal, While other papers in this volume have examined early health warning effectiveness and perceived harm. Full details of responses of youth to the new packaging,32 33 the current study the questions asked, response options and the way in which aimed to evaluate the effectiveness of plain packaging with responses were aggregated for analysis are provided in the larger GHWs among Australian adult cigarette smokers in Technical Report (see Supplementary Technical report).38 achieving the three specific aims of reducing the appeal of tobacco, increasing GHW effectiveness and correcting misper- Appeal-related outcomes ceptions of harm. Since plain packaging was implemented at the As a general measure of overall appeal, the extent to which same time as larger and refreshed GHWs, this study necessarily respondents liked the look of their current pack was rated on a examines outcomes of these two packaging changes in 5-point scale from ‘strongly agree’ to ‘strongly disagree’ (dichot- aggregate. omised into disagree/strongly disagree vs neither/agree/strongly It is expected that elements of packaging appeal should agree). Smokers were also asked to rate their current cigarettes reduce after the introduction of the packaging changes, since or tobacco as ‘higher’, ‘lower’ or ‘about the same’ compared brand colours and designs on packs were replaced with the with a year ago, in terms of quality,16 19 22 26 satisfaction,19 same unattractive colour and the GHWs took up a greater pro- value for money19 and appeal of the packaging (coded as lower – portion of the pack. We also expected perceived differences than a year ago vs higher/about the same).19 23 Additionally, between brands and variants to be undermined once all packs smokers were asked whether or not there were differences are plain and their distinctive brand-associated colour and between brands in prestige (no vs yes/do not know)19 20 25 and design connotations are replaced with a standard drab brown. on a 4-point scale from ‘not at all different’ to ‘very different’, Further, the larger refreshed GHWs would be expected to be how different cigarette brands were in taste (not at all different noticed more because of their increased size and imagery and vs a little/somewhat/very different/do not know).21 25 27 presentation against a drab background. If credible, these GHWs should then provide more motivation to quit.34 Health warning effectiveness outcomes Achievement of change in these kinds of proximal outcomes in Noticeability of GHWs was assessed by asking “When you look the first instance would provide confidence that the packaging at a cigarette/tobacco pack, what do you usually notice first: the changes are working as intended,34 subsequent to other investi- warning labels, or other aspects of the pack, such as branding?” gations of more distal outcomes pertaining to reduced smoking (noticed the warnings first vs other responses).34 39 Respondents uptake or progress towards quitting35 36 (this supplement). were also asked whether they agreed/disagreed that the dangers of smoking had been exaggerated (somewhat/strongly disagree METHODS vs neutral/agree responses) as an indicator of perceived credibil- Study design and participants ity of the GHWs.34 39 Cognitive response to warnings was mea- The National Plain Packaging Tracking Survey employed a con- sured by asking smokers to what extent the health warnings on tinuous cross-sectional design, with an average of 100 interviews packs made them more motivated to quit (much more vs not at – completed per week from April 2012–March 2014. Data collec- all/a little/somewhat).39 41 Avoidant responses were measured tion was suspended for 2 weeks over each December–January with two questions: whether in the past month they had asked holiday period. Telephone interviews with adult smokers for a pack with a different health warning on it (yes vs no)39 42 (respondents who smoked daily or weekly, or who smoked and how often in the past month they had covered up or con- monthly or less-than monthly and self-identified as smokers) and cealed their pack or put their cigarettes in another container recent quitters (quit in the last year) were conducted in English. (several or many times vs other responses).16 22 39 40 Respondents were recruited using a dual-frame sample design, with half of all respondents approached via landline random Perceived harm outcomes digit dialling (RDD) and half by mobile phone RDD. For the Smokers were asked whether or not there were differences landline sample, to correct for over-representation of older between brands in harm (no vs yes/do not know)20 21 27 and as ii18 Wakefield M, et al. Tob Control 2015;24:ii17–ii25. doi:10.1136/tobaccocontrol-2014-052050 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper for the appeal-related outcomes, to rate their current cigarettes to an advertisement (reach) and the average number of times the or tobacco as ‘higher’, ‘lower’ or ‘about the same’ compared target audience is exposed (frequency). Thus, 100 TARPs would with a year ago, in terms of harmfulness (higher vs other represent 100% of the target audience receiving one advertise- responses). Although descriptive brand variant names persisted ment per month, or 50% reached twice. TARPs for each media on plain packs, the pack colours that had previously signified market were sourced from AC Nielsen/OZTAM Australian televi- different tar or strength levels were standardised to dark drab sion audience measurements for adults aged 18+years. We used brown, so it was of interest to examine whether consumers con- advertising exposure over the 3 months prior to interview, based tinued to perceive differences in strength between brand var- on research demonstrating advertising effects within this time iants. Smokers were therefore asked to rate on a 4-point scale frame46 47 (mean 3-month TARPs=1548; SE=11.04). Cigarette from ‘not at all different’ to ‘very different’, how different costliness in the month of interview was calculated as the ratio of brand variants are in strength (not at all different vs other the average recommended retail price of the top 10 brands responses).19 26 43 (obtained from Australian Retail Tobacconist, February–April 2012 to January–March 2014 and weighted by market share) to the average weekly earnings for each respondents’ jurisdiction of Covariates residence.48 Similar to the method of Dunlop et al,38 49 we used Data were collected on sex, age and highest educational attain- percentage change in costliness over the past 3 months as the cov- ment. The Heaviness of Smoking Index (HSI) assessed nicotine ariate (mean percentage costliness change for total survey dependence using a 0 (low) to 6 (high) score based on the sum period=0.173, SE=0.010; months of excise/customs duty index- of daily cigarette consumption and minutes to first cigarette of ation; August 2012=−0.095, SE=0.034; February 2013=3.37, the day,44 with HSI for non-daily smokers coded as zero. We SE=0.044; August 2013=0.327, SE=0.65). used the Socio-Economic Indices for Areas index of relative dis- advantage45 to measure socioeconomic status (SES), which was Statistical analyses based on each respondent’s residential postcode (low=quintiles Data were weighted to account for telephony status (landline or 1–2; mid=quintiles 3–4; high=quintile 5). mobile phone), gender, age by education and state of residence Smokers’ responses might have been influenced by (see Supplementary Technical report).38 All statistical analyses tobacco-related media activity and tobacco prices. Potential were conducted using Stata V.12.150 using weighted data (using exposure to televised antismoking advertising campaigns was the svy command with ‘p’ weights). In addition, an uncondi- estimated using monthly Target Audience Rating Points (TARPs). tional approach (ie, the ‘subpopulation’ command in Stata TARPs are a product of the percentage of the audience exposed V.12.1) was used to limit the sample as appropriate for each set of analyses, ensuring correct estimation of the SEs. To test whether outcomes differed between three phases of plain pack- Table 1 Characteristics of current smokers, by study period aging implementation (pre-PP (referent), transition and during Pre-PP Transition PP year 1 plain packaging (PP year 1)), we used a series of unadjusted and 9 April–30 1 October–30 1 December adjusted logistic regression analysis. All adjusted models September November 2012–30 included HSI, demographic characteristics, recent antismoking 2012 2012 November 2013 campaign activity and change in cigarette costliness as covari- N=2176 N=759 N=4240 ates. We tested for interactions between phase (pre-PP vs PP Total n (%) n (%) n (%) year 1) and age (18–29; 30–49; 50–69 years), sex and SES (low, Sex mid and high) for all outcomes. In tables we present unadjusted Male 1191 (55) 416 (55) 2326 (55) and adjusted models, and adjusted models for interactions. For Female 985 (45) 342 (45) 1914 (45) brevity, we refer only to adjusted percentages and models in Age text. For the PP year 1 phase, we also examined the form of 18–29 641 (29) 213 (28) 1171 (28) change over time by examining adjusted regression models that 30–49 963 (44) 352 (46) 1973 (47) included linear and quadratic terms for month. If the quadratic fi 50–69 572 (26) 193 (25) 1096 (26) term was not signi cant, we present results from a model with fi Education the linear term for month only. The presence of a signi cant fi Less than high 648 (30) 281 (37) 1381 (33) linear term within PP year 1 indicates a signi cant linear school increase (or decrease) within the year, while the additional pres- Completed high 1248 (57) 382 (50) 2318 (55) ence of a significant quadratic term signifies the increase (or school/some decrease) was curvilinear over the months, that is, that it tertiary reached a peak (or trough) and then declined (increased) again. Tertiary or above 280 (13) 96 (13) 541 (13) Finally, sensitivity testing examined whether the inclusion of Socioeconomic status December/January season variables influenced the pattern of Low 825 (38) 297 (39) 1724 (41) observed findings in adjusted models. On the basis of studies Mid 912 (42) 329 (43) 1778 (42) examining quitting interest and behaviours,51 we constructed High 439 (20) 132 (17) 738 (17) indicator measures of season for the last two survey weeks in Mean (SE) Mean (SE) Mean (SE) December prior to Christmas (less interest and quitting) and the fi Heaviness of 2.07 (0.04) 2.12 (0.07) 2.17 (0.03) rst 2 weeks of January (more interest and quitting). Smoking Index (0–6) Antismoking 1816 (20.7) 1237 (29.4) 1467 (13.9) RESULTS advertising in past 3 months (TARPs) Table 1 shows the characteristics of the cigarette smokers who were interviewed—2176 pre-PP, 759 during transition and 4240 PP, plain packaging; TARPs, Target Audience Rating Points. during PP year 1, including distribution of covariates.

Wakefield M, et al. Tob Control 2015;24:ii17–ii25. doi:10.1136/tobaccocontrol-2014-052050 ii19 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Appeal-related outcomes There were significant interactions between age and phase Compared with the pre-PP period (58.6%), a significantly (pre-PP vs PP year 1) for the pack disliking (F=3.92(2, 6016), greater adjusted proportion of smokers in the PP year 1 period p=0.020), quality (F=6.83(2, 6217), p=0.001), satisfaction reported they disliked their pack (85.0%), with this higher pro- (F=4.52(2, 6217), p=0.011), prestige (F=3.37(2, 6178), portion being maintained throughout PP year 1 (table 2). p=0.034) and taste (F=7.60(2, 6129), p<0.001) outcomes. Compared with pre-PP when 12.3% of smokers rated their While all age groups showed change in the expected direction current pack as less appealing than a year ago, significantly for pack disliking, quality and satisfaction, younger smokers more smokers indicated this was the case in PP year 1 (56.2%). experienced relatively more change than older smokers (figure Both of these outcomes changed rapidly, with significant 1A–C). Mid-aged (30–49 years) smokers showed significant increases also evident in the transition period. From the begin- change in the expected direction for the prestige outcome ning of PP year 1, odds of reporting lower appeal compared while younger and older smokers remained stable (figure 1D). with a year ago declined linearly over each month, although For the taste outcome, the younger and mid-aged groups evi- always remaining higher than pre-PP. Odds of reporting that denced significant increases in the proportion who believed one’s cigarettes were of lower quality, lower satisfaction and brands did not differ in taste, while the older age group lower value for money than a year ago were also significantly showed a significant decrease (figure 1E). There were no greater in PP year 1 than in pre-PP (table 2) and these increased interactions by sex and phase, or SES and phase for any proportions remained stable over time. appeal-related outcomes. Compared with pre-PP, in PP year 1 a significantly higher and stable proportion of smokers believed that brands do not differ Health warning effectiveness outcomes in prestige. Overall, there was no change in the proportions Ta b l e 3 shows that GHWs became more noticeable during who believed brands do not differ in taste. the transition and PP year 1 periods and this higher

Table 2 Unadjusted and adjusted percentages, ORs and 95% CIs from logistic regression models predicting appeal-related outcomes* PP year 1 trend—adjusted Comparing phases—unadjusted models Comparing phases—adjusted models† models Variable Per cent OR (95% CI) p Value Per cent OR (95% CI) p Value Form OR p Value

Dislikes pack‡ (n=6728) Pre-PP 59.1 1.00 58.6 1.00 Transition 65.0 1.28 (1.05 to 1.56) 0.013 65.1 1.32 (1.08 to 1.62) 0.007 PP year 1 84.9 3.90 (3.39 to 4.46) <0.001 85.0 4.06 (3.52 to 4.69) <0.001 Linear 1.02 0.158 Lower pack appeal than a year ago‡ (n=6179) Pre-PP 12.7 1.00 12.3 1.00 Transition 26.0 2.40 (1.86 to 3.10) <0.001 26.5 2.59 (1.99 to 3.37) <0.001 PP year 1 55.8 8.63 (7.27 to 10.24) <0.001 56.2 9.29 (7.79 to 11.09) <0.001 Linear 0.97 0.004 Lower quality than a year ago§ (n=6954) Pre-PP 13.9 1.00 14.0 1.00 Transition 17.2 1.29 (1.01 to 1.66) 0.045 17.2 1.28 (0.99 to 1.65) 0.063 PP year 1 26.7 2.26 (1.93 to 2.64) <0.001 26.6 2.24 (1.91 to 2.64) <0.001 Linear 0.99 0.554 Lower satisfaction than a year ago§ (n=6954) Pre-PP 12.2 1.00 12.3 1.00 Transition 13.8 1.15 (0.87 to 1.51) 0.319 13.9 1.15 (0.87 to 1.51) 0.334 PP year 1 20.7 1.88 (1.59 to 2.22) <0.001 20.6 1.85 (1.56 to 2.19) <0.001 Linear 0.99 0.690 Lower value than a year ago§ (n=6901) Pre-PP 50.9 1.00 50.4 1.00 Transition 50.7 1.02 (0.84 to 1.22) 0.870 51.6 1.05 (0.87 to 1.27) 0.622 PP year 1 56.7 1.27 (1.13 to 1.43) <0.001 56.8 1.30 (1.15 to 1.46) <0.001 Linear 1.02 0.058 Believes brands do not differ in prestige‡ (n=6904) Pre-PP 44.7 1.00 45.1 1.00 Transition 42.1 0.90 (0.75 to 1.09) 0.273 43.1 0.91 (0.75 to 1.11) 0.373 PP year 1 49.9 1.23 (1.10 to 1.38) <.001 49.5 1.21 (1.07 to 1.37) 0.003 Linear 1.01 0.310 Believes brands do not differ in taste‡ (n=6840) Pre-PP 6.7 1.00 6.7 1.00 Transition 8.1 1.23 (0.87 to 1.72) 0.240 8.4 1.27 (0.90 to 1.80) 0.174 PP year 1 7.7 1.16 (0.93 to 1.45) 0.190 7.7 1.17 (0.93 to 1.47) 0.189 Linear 1.02 0.250 Bold text indicates significant findings at p<0.05. *Valid n’s for each model vary because of differences in the number of respondents who were eligible for each question (see below) and the number of missing cases on each outcome. The pack appeal model excluded 15.2% of cases that were missing, while for all other models missing cases ranged from 5.1% to 7.6%. †Multivariate models adjusted for age, gender, educational attainment, SES, Heaviness of Smoking Index, past 3-month Target Audience Rating Points for antismoking mass media campaigns and change in cigarette costliness. ‡Items were only asked of cigarette smokers who, when asked to name which brand of FM cigarettes/RYO tobacco they were currently smoking, provided a valid brand name. §Items were asked of all cigarette smokers. FM, factory-made; PP, plain packaging; RYO, roll-your-own; SES, socioeconomic status. ii20 Wakefield M, et al. Tob Control 2015;24:ii17–ii25. doi:10.1136/tobaccocontrol-2014-052050 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Figure 1 (A–E) Significant interactions by age group for appeal-related outcomes in adjusted models. (A) Dislikes pack: 18–29 years: pre-plain packaging (pre-PP) 53.3%, PP year 1 84.9% (OR=5.07 (95% CI 3.78 to 6.81)), p<0.001. 30–49 years: pre-PP 61.2%, PP year 1 84.4% (OR=3.47 (2.77 to 4.33)), p<0.001. 50–69 years: pre-PP 60.7%, PP year 1 86.2% (OR=4.14 (3.26 to 5.26)), p<0.001. (B) Lower quality: 18–29 years: pre-PP 10.2%, PP year 1 29.3% (OR=3.76 (2.63 to 5.38)), p<0.001. 30–49 years: pre-PP 14.5%, PP year 1 24.7% (OR=1.94 (1.50 to 2.50)), p<0.001. 50–69 years: pre-PP 17.3%, PP year 1 27.2% (OR=1.79 (1.40 to 2.30)), p<0.001. (C) Lower satisfaction: 18–29 years: pre-PP 10.4%, PP year 1 21.1% (OR=2.32 (1.61 to 3.33)), p<0.001. 30–49 years: pre-PP 12.2%, PP year 1 21.1% (OR=1.94 (1.49 to 2.54)), p<0.001. 50–69 years: pre-PP 14.7%, PP year 1 19.1% (OR=1.37 (1.05 to 1.78)), p=0.020. (D) Brands do not differ in prestige: 18–29 years: pre-PP 26.9%, PP year 1 30.3% (OR=1.18 (0.90 to 1.55)), p=0.237. 30–49 years: pre-PP 45.5%, PP year 1 53.5% (OR=1.38 (1.14 to 1.66)), p<0.001. 50–69 years: pre-PP 63.6%, PP year 1 63.3% (OR=0.99 (0.80 to 1.21)), p=0.899. (E) Brands do not differ in taste: 18–29 years: pre-PP 3.3%, PP year 1 6.8% (OR=2.14 (1.20 to 3.81)), p=0.010. 30–49 years: pre-PP 5.5%, PP year 1 8.1% (OR=1.53 (1.04 to 2.23)), p=0.029. 50–69 years: pre-PP 12.5%, PP year 1 8.5% (OR=0.65 (0.47 to 0.91)), p=0.013.

noticeability was maintained throughout PP year 1. Overall, 1. There was a sustained increase from pre-PP to PP year 1 in there was no change by phase in perceived exaggeration of the proportion of smokers who reported concealing their packs the dangers of smoking, although a significant quadratic or using a different container for their cigarettes, and an imme- trend and an examination of the month-by-month data (not diate and sustained increase in the proportion who had presented) pointed to a slight increase in the middle of PP requested a different GHW when purchasing a pack in the past year 1 in the proportion of smokers believing health effects month (table 3). No interactions were found between phase and were not exaggerated, followed by a return to its initial level, age, sex or SES for any of the health warning effectiveness coinciding with the roll-out of the second set of new GHWs outcomes. from August 2013.2 Compared with pre-PP a greater proportion of smokers in PP Perceived harm outcomes year 1 attributed the GHWs with providing much more motiv- Table 4 shows there was a significant increase in the adjusted ation to quit. The significant linear decline within PP year 1 proportion of smokers who believed that brands do not differ reflected an early sharp peak followed by a gradual decline to in harmfulness in PP year 1 (69.8%) compared with pre-PP slightly above the baseline level by the end of PP year (65.7%). There was no change in the belief that one’s cigar- 1. Avoidant responses increased from pre-PP to PP year ettes were higher in harmfulness compared with a year ago and

Wakefield M, et al. Tob Control 2015;24:ii17–ii25. doi:10.1136/tobaccocontrol-2014-052050 ii21 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 3 Unadjusted and adjusted percentages, ORs and 95% CIs from logistic regression models predicting health warning effectiveness outcomes* PP year 1 trend—adjusted Comparing phases—unadjusted models Comparing phases—adjusted models† models Variable Per cent OR (95% CI) p Value Per cent OR (95% CI) p Value Form OR p Value

Notices GHW first when looking at pack‡ (n=7007) Pre-PP 34.4 1.00 34.0 1.00 Transition 44.9 1.55 (1.29 to 1.88) <0.001 44.9 1.60 (1.32 to 1.95) <0.001 PP year 1 67.5 3.96 (3.51 to 4.48) <0.001 67.7 4.26 (3.74 to 4.85) <0.001 Linear 1.00 0.844 Does not believe dangers of smoking are exaggerated‡ (n=7163) Pre-PP 65.6 1.00 65.1 1.00 Transition 65.7 1.00 (0.83 to 1.22) 0.969 65.7 1.03 (0.84 to 1.25) 0.788 PP year 1 63.8 0.92 (0.82 to 1.04) 0.194 64.0 0.96 (0.84 to 1.08) 0.470 Linear 1.25 0.028 Quad 0.99 0.021 Attributes much more motivation to quit to GHWs‡ (n=7123) Pre-PP 9.4 1.00 9.3 1.00 Transition 11.8 1.28 (0.96 to 1.72) 0.097 11.6 1.29 (0.95 to 1.75) 0.101 PP year 1 13.3 1.47 (1.21 to 1.77) <0.001 13.4 1.52 (1.25 to 1. 86) <0.001 Linear 0.96 0.013 Concealed pack in past month‡ (n=7096) Pre-PP 17.3 1.00 17.1 1.00 Transition 19.0 1.12 (0.89 to 1.42) 0.325 19.1 1.15 (0.90 to 1.46) 0.252 PP year 1 23.1 1.44 (1.24 to 1.66) <0.001 23.1 1.47 (1.26 to 1.71) <0.001 Linear 0.99 0.533 Requested different GHW in past month‡ (n=7126) Pre-PP 3.9 1.00 3.9 1.00 Transition 8.2 2.19 (1.47 to 3.28) <0.001 8.5 2.28 (1.52 to 3.44) <0.001 PP year 1 9.2 2.48 (1.90 to 3.24) <0.001 9.1 2.49 (1.89 to 3.29) <0.001 Linear 0.99 0.756 Bold text indicates significant findings at p<0.05. *Valid n’s for each model vary because of the number of missing cases for each outcome. Missing cases ranged from 2.2% to 4.3%. †Multivariate models adjusted for age, gender, educational attainment, SES, Heaviness of Smoking Index, past 3-month Target Audience Rating Points for antismoking mass media campaigns and change in cigarette costliness. ‡Items were asked of all cigarette smokers. GHW, graphic health warning; PP, plain packaging; SES, socioeconomic status. no change in the proportions who believed variants do not Sensitivity testing differ in strength. No significant interactions between phase Inclusion of two season variables (last 2 weeks of December; and age, sex or SES were found for any perceived harm first 2 weeks of January) did not change the pattern of results outcome. (see online supplementary appendix A).

Table 4 Unadjusted and adjusted percentages, ORs and 95% CIs from logistic regression models predicting perceived harm outcomes* PP year 1 trend—adjusted Comparing phases—unadjusted models Comparing phases—adjusted models† models

Per cent OR (95% CI) p Value Per cent OR (95% CI) p Value Form OR p Value

Believes brands do not differ in harmfulness‡ (n=6924) Pre-PP 65.8 1.00 65.7 1.00 Transition 67.0 1.06 (0.86 to 1.29) 0.598 67.6 1.09 (0.89 to 1.35) 0.405 PP year 1 69.8 1.20 (1.06 to 1.36) 0.004 69.8 1.21 (1.06 to 1.38) 0.004 Linear 0.97 0.012 Higher harmfulness than a year ago§ (n=6838) Pre-PP 24.2 1.00 23.8 1.00 Transition 22.1 0.89 (0.70 to 1.12) 0.323 21.8 0.89 (0.70 to 1.13) 0.349 PP year 1 23.4 0.96 (0.83 to 1.10) 0.532 23.6 0.99 (0.86 to 1.14) 0.877 Linear 1.00 0.811 Believes variants do not differ in strength‡ (n=6894) Pre-PP 5.2 1.00 5.3 1.00 Transition 5.9 1.16 (0.78 to 1.72) 0.477 5.8 1.09 (0.72 to 1.64) 0.683 PP year 1 6.1 1.19 (0.92 to 1.55) 0.178 6.1 1.15 (0.88 to 1.51) 0.303 Linear 1.01 0.812 Bold text indicates significant findings at p<0.05. *Valid n’s for each model vary because of differences in the number of respondents who were eligible for each question (see below) and the number of missing cases for each outcome. Missing cases ranged from 5.4% to 6.6%. †Multivariate models adjusted for age, gender, educational attainment, SES, Heaviness of Smoking Index, past 3-month Target Audience Rating Points for antismoking mass media campaigns and change in cigarette costliness. ‡Items were only asked of cigarette smokers who, when asked to name which brand of FM cigarettes/RYO tobacco they were currently smoking, provided a valid brand name. §Items were asked of all cigarette smokers. FM, factory-made; PP, plain packaging; RYO, roll-your-own; SES, socioeconomic status. ii22 Wakefield M, et al. Tob Control 2015;24:ii17–ii25. doi:10.1136/tobaccocontrol-2014-052050 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

DISCUSSION and wanting to hide or cover packs from the view of others, our This national study examined the extent to which all three spe- study additionally found increased prevalence of reported behav- cific objectives of the plain packaging legislation—to reduce ioural actions taken to attempt to avoid warnings (covering up appeal of tobacco, increase effectiveness of health warnings and the pack at least several times in the past month; asking for a reduce ability of packaging to mislead consumers about smoking pack with a different warning) up to 1 year after implementation. harms—were met among adult smokers in the first year after Those outcomes are important, since cohort studies suggest that full implementation of the packaging changes. warning avoidance predicts subsequent quitting attempts,40 42 54 Compared with a 6-month baseline period, the study found most likely via the association of warning avoidance with the new packs to be associated with rapid and substantial increased thinking about smoking harms,54 which in turn pre- changes in the direction of reduced appeal which were sustained dicts quit intentions and quit attempts. Avoidant behaviours such throughout the first year of full implementation. Of the four as pack concealment can have benefits beyond smokers them- appeal variables that required smokers to compare the experi- selves by reducing exposure of others to tobacco promotion.55 ence of their current cigarettes with those smoked the year Relatedly, an observational Australian study of patrons of beforehand, the observed increase in effect in the PP year 1 outdoor cafes and bars up to a year after plain packaging found a period weakened towards the end of this phase for one variable sustained reduction in the visibility of tobacco packs, mostly only (lower pack appeal). As this period late in 2013 required attributable to fewer packs being left out on display, especially in smokers to compare to late 2012 when plain packs were already venues where children were present56 57 (this supplement). common,2 this regression may reflect a reduction in the specifi- In relation to the third objective of plain packaging, there was city of comparison between plain and branded packs, rather increased appreciation after plain packaging that brands do not than a diminution of the impact of plain packaging per se. differ in harm. However, there was no change over time in the The emergence of lower perceived quality and satisfaction and perceived harm of one’s tobacco compared with a year ago, and the increase in beliefs that brands do not differ in prestige among no change over time in smokers’ beliefs about variants being dif- smokers overall (and do not differ in taste for the young-aged and ferent in strength. Concepts of ‘taste’ (measured in the appeal mid-aged population) occurred despite efforts from tobacco com- outcomes), ‘strength’ and ‘harm’ are strongly related in the – panies to reassure consumers that tobacco products would ‘remain minds of smokers.43 58 60 Objectively speaking, cigarette brands the same’ with the changed packaging.52 These findings signifying and variants do not differ in harmfulness,61 62 although an reduced appeal are consistent with and reinforce the results of past extensive body of research shows many smokers believe they experimental research,10 11 18 24 53 government-funded pretesting do.10 27 43 29 63 While plain packaging removed the pack colour studies to design the Australian packs,7 an early study of how associated with different brands and variants and replaced it smokers responded to the new packaging at the time they emerged with a consistent drab brown, it placed no limitations on the onto the market8 and a recently published study from one naming of brands and continued to permit the use of brand Australian state reporting increases in negative pack perceptions up variant names including colour names. An observational study to 6 months after plain packaging implementation.49 of tobacco product developments coinciding with implementa- It was notable that for three appeal-related variables (pack dis- tion of plain packaging reported incorporation of colour names liking, quality, satisfaction), young adults evidenced more change into longer variant descriptors (eg, ‘Dunhill Distinct’ became than others. Young-aged and middle-aged adults also showed ‘Dunhill Distinct Blue’) and more expressive brand variant more movement towards believing brands did not differ in taste, names to evoke the sensation or feature previously connoted and middle-aged smokers showed more movement in believing only by colour (eg, ‘Pall Mall Blue’ became ‘Pall Mall Rich brands did not differ in prestige than older smokers. This overall Blue’).52 These developments are clearly designed to reinforce pattern of age-related effects for the appeal outcomes was not variant differences. Beliefs about strength differences may observed for the health warning or harm perception outcomes. persist due to the differing perceptions of strength resulting Such specificity of change for the young adult population concurs from the sensory experience created by diluted inhaled smoke with the experimental studies of plain packaging which found (from filter ventilation) and strength expectations generated by stronger effects on appeal-related outcomes for younger than variant names.21 29 64 older respondents,10 11 and with the findings of White et al32 Study strengths included its national focus and a large sample (this supplement) who found reduced pack and brand appeal size, sufficient to compare responses overall and by major demo- among adolescents following the introduction of plain packaging graphic subgroups. The continuous tracking survey design in Australia. Furthermore, it is consistent with strong evidence enabled comparison of smokers’ responses during a preimple- that tobacco branding and image-related factors are very import- mentation period, as the new packs transitioned onto the market, ant for young people,12 13 leading to the conclusion in the 2012 and over the first year of full implementation. Observed changes US Surgeon-General’s Report that “advertising and promotional in outcomes were independent of exposure to antismoking mass activities by the tobacco companies cause the onset and continu- media campaigns and change in cigarette costliness, and insensi- ation of smoking among adolescents and young adults.”13 tive to seasonal adjustment for the December/January period. As Our study also found consistent improvements in health the sample was limited to November 2013, the study further warning effectiveness outcomes, including more noticing of excluded impacts following the 12.5% December 2013 tobacco warnings, motivation to quit attributed to the warnings, and avoi- tax increase. dant behavioural responses. These findings extend those of A potential study limitation was that since the survey started Dunlop et al49 who found increased noticing of warnings and in April 2012, we did not have responses from one or more attribution of encouragement to quit to warnings up to 6 months years prior to plain packaging implementation to control for the after implementation of plain packaging with larger GHWs in possibility of seasonal effects other than the December–January one Australian state, by demonstrating that such responses were holiday period. However, it seems unlikely these particular sustained nationally up to 1 year after implementation. While beliefs and behaviours would have been sensitive to season per Dunlop et al found increased emotional responses to warnings se. It does however seem likely that the variable presence of 6 months after implementation in terms of worry about smoking educational interventions would affect such beliefs, and for this

Wakefield M, et al. Tob Control 2015;24:ii17–ii25. doi:10.1136/tobaccocontrol-2014-052050 ii23 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper reason we adjusted for antismoking media campaign activity. We and implementation of the Tobacco Plain Packaging Act 2011 and the Competition further note the findings of Dunlop et al49 who found similar and Consumer (Tobacco) Information Standard 2011. Tob Control 2015;24:ii9–16. 3 Competition and Consumer (Tobacco) Information Standard. 2011. http://www. increases for a subset of variables similar to ours up to 6 months comlaw.gov.au/Details/F2013C00598 into plain packaging compared with 1–6 years pre-PP using time 4 Gfk Blue Moon. Market testing of new health warnings and information messages series analyses which adjusted for seasonality. for tobacco product packaging: phase 1 side of pack messages. Canberra: On the basis of the findings of this study, we conclude that Department of Health and Ageing, 2011. http://www.health.gov.au/internet/main/ the first two objectives of the legislation relating to reducing publishing.nsf/Content/mr-plainpack 5 Gfk Blue Moon. Market testing of new health warnings and information messages appeal and increasing GHW effectiveness have been achieved for tobacco product packaging: phase 2 front and back of pack messages. among adult smokers. This study suggests that the third object- Canberra: Department of Health and Ageing, 2011. http://www.health.gov.au/ ive of reducing the extent to which smokers are misled about internet/main/publishing.nsf/Content/mr-plainpack the harms of smoking has been partially met. Perceptions of dif- 6 Gfk Blue Moon. Market testing of new health warnings and information messages for tobacco product packaging: phase 3 refinement of health messages. Canberra: ferential harmfulness may be maintained by the naming of Department of Health and Ageing, 2011. http://www.health.gov.au/internet/main/ brand variants which are still permitted to emphasise colour and publishing.nsf/Content/mr-plainpack other product characteristics, suggesting future standardised 7 Gfk Blue Moon. Market research reports on tobacco plain packaging of tobacco packaging laws should further limit such variant names. products. Canberra: Department of Health and Ageing, 2011. http://www.health. gov.au/internet/main/publishing.nsf/Content/mr-plainpack 8 Wakefield MA, Hayes L, Durkin S, et al. Introduction effects of the Australian plain packaging policy on adult smokers: a cross-sectional study. BMJ Open 2013;3:7. What this paper adds 9 Quit, Cancer Council Victoria. Plain packaging of tobacco products: a review of the evidence. Melbourne, Australia: Cancer Council Victoria, May (updated August 2011) 2011. http://www.cancer.org.au/content/pdf/CancerControlPolicy/ ▸ Compared with a pre-plain packaging period, the first year PositionStatements/TCUCCVBkgrndResrchPlainPak270511ReEnd_FINAL_May27.pdf of Australian implementation of plain packaging with larger 10 Stead M, Moodie C, Angus K, et al. Is consumer response to plain/standardised tobacco packaging consistent with framework convention on tobacco control graphic health warnings saw reduced appeal and increased guidelines? A systematic review of quantitative studies. PLoS ONE 2013;8:e75919. health warning effectiveness, and some improvement in 11 Hammond D. Standardized packaging of tobacco products: evidence review. correcting misperceptions of harms in a national sample of Prepared on behalf of the Irish Department of Health. Waterloo, Ontario, Canada: adult cigarette smokers. University of Waterloo, March 2014. http://www.tri.ie/news/2014/ ▸ new-irish-focused-review-on-plain-packaging.aspx There was greater change for four appeal-related outcomes 12 National Cancer Institute. The role of the media in promoting and reducing tobacco for young (18–29 years) adults and two appeal-related use. Tobacco Control Monograph No 19. Bethesda, MD: US Department of Health outcomes for mid-aged (30–49 years) adults, with no and Human Services, National Institutes of Health, National Cancer Institute, June differences by age, sex or socioeconomic status for other 2008. NIH Pub No 07-6242. http://cancercontrol.cancer.gov/tcrb/monographs/19/ outcomes. index.html ▸ fi 13 US Department of Health and Human Services. Preventing Tobacco use among The speci c objectives of plain packaging have been youth and young adults: a report of the surgeon general. Report. Atlanta, Georgia: achieved among adult smokers and were generally sustained U.S. Department of Health and Human Services, Public Health Service, Office of the up to 12 months after implementation. Surgeon General, Rockville, MD, 2012. http://www.cdc.gov/tobacco/data_statistics/ sgr/2012/index.htm 14 United States v. Philip Morris USA. 449 F. Supp. 2d 1, (D.D.C., 2006) 980. fi — Acknowledgements The authors thank Professor Ron Borland for sharing 15 Scollo M, Bayly M, Wake eld M. Plain packaging a logical progression for ’ ‘ ’ questions from the ITC-survey for inclusion in the telephone tracking survey, the tobacco control in one of the world s darkest markets . Tob Control. ’ Social Research Centre for data collection and Megan Bayly for data checking. 16 Moodie CS, Mackintosh AM. Young adult women smokers response to using plain cigarette packaging: a naturalistic approach. BMJ Open 2013;3:e002402. Contributors MW and MS designed the study. KC and MZ cleaned data files, 17 Moodie C, Bauld L, Ford A, et al. Young women smokers’ response to using plain undertook analysis and reported results. SD and EB provided additional advice on cigarette packaging: qualitative findings from a naturalistic study. BMC Public analysis and interpretation of results. MW drafted the manuscript with contributions Health 2014;14:812. from all authors. All authors approved the final manuscript. 18 Gallopel-Morvan K, Moodie C, Eker F, et al. Perceptions of plain packaging among Funding The National Plain Packaging Tracking survey was funded under a contract young adult roll-your-own smoker in France: a naturalistic approach. Tob Control with the Australian Government Department of Health and Ageing. Published Online First: 11 Jun 2014 doi:10.1136/tobaccocontrol-2013-051513 19 Wakefield M, Germain D, Durkin S. How does increasingly plainer cigarette Competing interests The authors wish to advise that MW was a member and packaging influence adult smokers’ perceptions about brand image? An MS a technical writer for the Tobacco Working Group of the Australian National experimental study. Tob Control 2008;17:416–21. Preventive Health Task Force and MW was a member of the Expert Advisory 20 Hammond D, Doxey J, Daniel S, et al. Impact of female-oriented cigarette Committee on Plain Packaging that advised the Australian Department of Health on packaging in the United States. Nicotine Tob Res 2011;13:579–88. research pertaining to the plain packaging legislation. MW, SD and EB hold 21 Doxey J, Hammond D. Deadly in pink: the impact of cigarette packaging among competitive grant funding from the Australian National Health and Medical Research young women. Tob Control 2011;20:353–60. Council and MW holds competitive grant funding from the US National Institutes of 22 Moodie C, Mackintosh AM, Hastings G, et al. Young adult smokers’ perceptions of Health, Australian National Preventive Health Agency and BUPA Health Foundation. plain packaging: a pilot naturalistic study. Tob Control 2011;20:367–73. Ethics approval The survey was approved by the Cancer Council Victoria Human 23 Gallopel-Morvan K, Moodie C, Hammond D, et al. Consumer perceptions of Ethics Committee (HREC 0018). cigarette pack design in France: a comparison of regular, limited edition and plain packaging. Tob Control 2012;21:502–6. Provenance and peer review Not commissioned; externally peer reviewed. 24 Guillaumier A, Bonevski B, Paul C, et al. Socioeconomically disadvantaged smokers’ Open Access This is an Open Access article distributed in accordance with the ratings of plain and branded cigarette packaging: an experimental study. BMJ Open Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which 2014;4:e004078. permits others to distribute, remix, adapt, build upon this work non-commercially, 25 Germain D, Wakefield M, Durkin S. Adolescents’ perceptions of cigarette brand image: and license their derivative works on different terms, provided the original work is does plain packaging make a difference? J Adolesc Health 2010;46:385–92. properly cited and the use is non-commercial. See: http://creativecommons.org/ 26 Bansal-Travers M, Hammond D, Smith P, et al. The impact of cigarette pack design, licenses/by-nc/4.0/ descriptors, and warning labels on risk perception in the U.S. Am J Prev Med 2011;40:674–82. REFERENCES 27 Hammond D, Dockrell M, Arnott D, et al. Cigarette pack design and perceptions of 1 Tobacco Plain Packaging Act 2011. No. 148, 2011 as amended. 2011. http://www. risk among UK adults and youth. Eur J Public Health 2009;19:631–7. comlaw.gov.au/Details/C2013C00190 28 White CM, Hammond D, Thrasher JF, et al. The potential impact of plain packaging 2 Scollo M, Lindorff K, Coomber K, et al. Standardised packaging and new enlarged of cigarette products among Brazilian young women: an experimental study. BMC graphic health warnings for tobacco products in Australia—legislative requirements Public Health 2012;12:737. ii24 Wakefield M, et al. Tob Control 2015;24:ii17–ii25. doi:10.1136/tobaccocontrol-2014-052050 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

29 Hammond D, Daniel S, White CM. The effect of cigarette branding and plain packaging 47 Dunlop S, Cotter T, Perez D, et al. Televised antismoking advertising: effects of level on female youth in the United Kingdom. J Adolesc Health 2013;52:151–7. and duration of exposure. Am J Public Health 2013;103:e66–73. 30 Goldberg M, Liefeld J, Madil J, et al. The effect of plain packaging on response to 48 Australian Bureau of Statistics. Average weekly earnings, Australia. Cat No 6302.0. health warnings. Am J Public Health 1999;89:1434–5. November 2013. http://www.abs.gov.au/ausstats/[email protected]/mf/6302.0 31 Beede P, Lawson R. The effect of plain packages on the perception of cigarette 49 Dunlop SM, Dobbins T, Young JM, et al. Impact of Australia’s introduction of health warnings. Public Health 1992;106:315–22. tobacco plain packs on adult smokers’ pack-related perceptions and responses: 32 White V, Williams T, Wakefield M. Has the introduction of plain packaging with results from a continuous tracking survey. BMJ Open 2014;4:e005836. larger graphic health warnings changed adolescents’ perceptions of cigarette packs 50 StataCorp. Stata Statistical Software: release 12. College Station, TX: StataCorp LP, and brands? Tob Control 2015;24:ii42–9. 2011. 33 White V, Williams T, Faulkner A, et al. Do larger graphic health warnings on standardised 51 Durkin S, Wakefield M, Spittal M. Which types of televised anti-tobacco campaigns cigarette packs increase adolescents’ cognitive processing of consumer health prompt more quitline calls from disadvantaged groups? Health Educ Res information and beliefs about smoking related harms? Tob Control 2015;24:ii50–7. 2011;26:998–1009. 34 IARC Handbooks of Cancer Prevention. Tobacco control, Volume 12: Methods for 52 Scollo M, Occleston J, Bayly M, et al. Tobacco product developments coinciding evaluating tobacco control policies. Lyon, France: International Agency for Research with the implementation of plain packaging in Australia. Tob Control Published on Cancer, 2008. Online First: 30 Apr 2014 doi:10.1136/tobaccocontrol-2013-051509 35 Brennan E, Durkin S, Coomber K, et al. Are quitting-related cognitions and behaviours 53 Brose LS, Chong CB, Aspinall E, et al. Effects of standardised cigarette packaging predicted by proximal responses to plain packaging with larger health warnings? Findings on craving, motivation to stop and perceptions of cigarettes and packs. Psychol from a national cohort study with Australian adult smokers. Tob Control 2015;24:ii33–41. Health 2014;29:849–60. 36 Durkin S, Brennan E, Coomber K, et al. Short-term changes in quitting-related 54 Yong HH, Borland R, Thrasher JF, et al. Mediational pathways of the impact cognitions and behaviours after the implementation of plain packaging with larger of cigarette warning labels on quit attempts. Health Psychol 2014;33:1410–20. health warnings: findings from a national cohort study with Australian adult 55 Ferraro R, Bettman J, Chartrand TL. The power of strangers: the effect of smokers. Tob Control 2015;24:ii26–32. incidental consumer brand encounters on brand choice. J Consum Res 37 Lavrakas PJ. Within-household respondent selection: how best to reduce total survey 2009;35:729–41. error? Media Rating Council Inc., November. 2008. http://www.mediaratingcouncil. 56 Zacher M, Bayly M, Brennan E, et al. Personal tobacco pack display before and org/MRC%20Point%20of%20View%20-%20Within%20HH%20Respondent% after the introduction of plain packaging with larger pictorial health warnings in 20Selection%20Methods.pdf Australia: an observational study of outdoor café strips. Addiction 2014;109: 38 Coomber K, Zacher M, Durkin S, et al. Australian National Tobacco Plain Packaging 653–62. Tracking Survey: Technical Report. Melbourne: Cancer Council Victoria and Social 57 Zacher M, Bayly M, Brennan E, et al. Personal pack display and active smoking at Research Centre, December 2014. http://tobaccocontrol.bmj.com/lookup/suppl/ outdoor café strips: assessing the impact of plain packaging 1 year doi:10.1136/tobaccocontrol-2014-052050/-/DC1 postimplementation. Tob Control 2015;24:ii94–7. 39 Shanahan P, Elliott D. Evaluation of the effectiveness of the graphic health warnings 58 Mutti S, Hammond D, Borland R, et al. Beyond light and mild: cigarette brand on tobacco product packaging 2008. Canberra, Australia: Australian Government descriptors and perceptions of risk in the International Tobacco Control (ITC) Four Department of Health and Ageing, 2009. http://www.health.gov.au/internet/main/ Country Survey. Addiction 2011;106:1166–75. publishing.nsf/Content/tobacco-warn-eval08 59 Brown A, McNeill A, Mons U, et al. Do smokers in Europe think all cigarettes are 40 BorlandR,YongHH,WilsonN,et al. How reactions to cigarette packet health warnings equally harmful? Eur J Public Health 2012;22:35–40. influence quitting: findings from the ITC Four-Country survey. Addiction 2009;104:669–75. 60 Elton-Marshall T, Fong GT, Yong HH, et al. Smokers’ sensory beliefs mediate the 41 Hammond D, Fong GT, Borland R, et al. Text and graphic warnings on cigarette relation between smoking a ‘light/low tar’ cigarette and perceptions of harm. packages: findings from the international tobacco control four country study. Am J Tob Control Published Online First: 4 Nov 2014 doi:10.1136/tobaccocontrol-2014- Prev Med 2007;32:202–9. 051977 42 Hammond D, Fong GT, McDonald PW, et al. Graphic Canadian cigarette warning 61 National Cancer Institute Expert Committee. NCI monograph no. 8: changes in labels and adverse outcomes: evidence from Canadian smokers. Am J Public Health cigarette-related disease risks and their implication for prevention and control. 2004;94:1442–5. Bethesda, MD: National Institutes of Health, 1997. http://onlinebooks.library.upenn. 43 Hammond D, Parkinson C. The impact of cigarette package design on perceptions edu/webbin/book/lookupid?key=olbp24454 of risk. J Public Health (Oxf) 2009;31:345–53. 62 National Cancer Institute. Risks associated with smoking cigarettes with low 44 Heatherton TF, Kozlowski LT, Frecker RC, et al. Measuring the heaviness of machine-measured yield of tar and nicotine. Bethseda, MD: US Department of smoking: using self-reported time to the first cigarette of the day and number of Health and Human Services National Institutes of Health, National Cancer Institute, cigarettes smoked per day. Br J Addict 1989;84:791–9. October 2001. NIH Pub. No 02–5074. http://cancercontrol.cancer.gov/TCRB/ 45 Australian Bureau of Statistics. Technical paper: socio-economic indexes for areas monographs/13/m13_preface.pdf (SEIFA) 2011. Canberra, Australia: Australian Bureau of Statistics, 2013. Catalogue 63 O’Connor RJ, Caruso RV, Borland R, et al. Relationship of cigarette-related no. 2033.0.55.001. http://www.abs.gov.au/ausstats/[email protected]/mf/2033.0.55.001/ perceptions to cigarette design features: findings from the 2009 ITC U.S. Survey. 46 Wakefield M, Spittal M, Durkin S, et al. Effects of mass media campaign exposure Nicotine Tob Res 2013;15:1943–7. intensity and durability on quit attempts in a population-based cohort study. Health 64 Wakefield M, Morley C, Horan J, et al. The cigarette pack as image: new evidence Educ Res 2011;26:988–97. from tobacco industry documents. Tob Control 2002;11:i73–80.

Wakefield M, et al. Tob Control 2015;24:ii17–ii25. doi:10.1136/tobaccocontrol-2014-052050 ii25 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Australian adult smokers' responses to plain packaging with larger graphic health warnings 1 year after implementation: results from a national cross-sectional tracking survey Melanie Wakefield, Kerri Coomber, Meghan Zacher, Sarah Durkin, Emily Brennan and Michelle Scollo

Tob Control 2015 24: ii17-ii25 doi: 10.1136/tobaccocontrol-2014-052050

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii17

These include: Supplementary Supplementary material can be found at: Material http://tobaccocontrol.bmj.com/content/suppl/2015/02/16/tobaccocon trol-2014-052050.DC1.html References This article cites 34 articles, 15 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii17#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper Short-term changes in quitting-related cognitions and behaviours after the implementation of plain packaging with larger health warnings: findings from a national cohort study with Australian adult smokers Sarah Durkin, Emily Brennan, Kerri Coomber, Meghan Zacher, Michelle Scollo, Melanie Wakefield

▸ Additional material is ABSTRACT measures, the government also anticipated that this published online only. To view Background Plain packaging (PP) with larger graphic legislation would contribute to efforts to reduce please visit the journal online 12 (http://dx.doi.org/10.1136/ health warnings (GHWs) was implemented in Australia in Australian smoking rates over the longer term. tobaccocontrol-2014-052058) late 2012. This study examined effects of these packaging At the same time, new and larger GHWs covering changes on short-term changes in quitting-related 75% of the front of cigarette packs (up from 30% Centre for Behavioural Research in Cancer, Cancer cognitions and behaviours. previously) and maintaining coverage of 90% of 23 Council Victoria, Melbourne, Methods We used a series of cohorts of Australian the back, were also introduced. This paper aims Victoria, Australia adult cigarette smokers originally sourced from a to examine the effects of these packaging changes nationally representative cross-sectional tracking survey, on short-term changes in quitting-related cogni- Correspondence to Professor Melanie Wakefield, followed up approximately 1 month after their baseline tions and behaviours 1 year after the full implemen- Centre for Behavioural interview (n(weighted)=5441). Logistic regression tation of this legislation. Research in Cancer, Cancer analyses compared changes in seven quitting-related Experimental and naturalistic studies have pro- Council Victoria, 615 St Kilda outcomes over this 1-month follow-up period for the vided some indications that PP with large GHWs Rd, Melbourne, VIC 3004, fi cohorts surveyed before PP, over the period of transition may increase quitting-related cognitions and beha- Australia; melanie.wake eld@ fi cancervic.org.au to PP, and during the rst year of PP, adjusting for viours. Experimental studies found that compared baseline levels of the outcome and covariates. to fully branded packs, plain packaged cigarettes Received 25 September 2014 Results Compared to the referent group of smokers and packages with fewer branded elements were Revised 12 December 2014 who completed their follow-up survey pre-PP, those who associated with lower consumer demand and Accepted 13 January 2015 were followed-up in the early transition period showed greater self-reported likelihood of quitting among – significantly greater increases in rates of stopping adults,4 8 and reduced demand, lower intentions to themselves from smoking (OR=1.51, 95% CI (1.08 to try smoking and greater cessation behaviour among – 2.10)) and higher quit attempt rates (OR=1.43, 95% CI adolescents.9 11 Another recent experimental study (1.00 to 2.03)), those followed-up in the late transition using plain packs modelled after those introduced period showed greater increases in intentions to quit in Australia, found that smokers who had been (OR=1.42, 95% CI (1.06 to 1.92)) and pack concealment abstinent for 12 h and who were randomised to (OR=1.55, 95% CI (1.05 to 2.31)), and those followed- view and then describe a plain pack, subsequently up in the first year of PP showed higher levels of pack reported lower craving levels than those rando- concealment (OR=1.65, 95% CI (1.01 to 2.72)), more mised to view fully branded packs, although after premature stubbing out of cigarettes (OR=1.55, 95% CI this brief exposure, reported motivation to stop (1.01 to 2.36)), and higher quit attempt rates (OR=1.52, smoking did not differ.12 95% CI (1.01 to 2.30)). One naturalistic study of young adult smokers Conclusions These findings provide some of the who used plain packs for 2 weeks reported reduced strongest evidence to date that implementation of PP consumption, less smoking around others, more with larger GHWs was associated with increased rates of thoughts about quitting, more avoidant behaviours quitting cognitions, microindicators of concern and quit such as concealing the pack, and more microindica- attempts among adult cigarette smokers. tors of concern such as forgoing cigarettes, than 13 Open Access when using fully branded packs. Another natural- Scan to access more istic study of young adult roll-your-own smokers free content INTRODUCTION found that smoking from plain packs with large Through the introduction in late 2012 of tobacco GHWs for 10 days was associated with greater plain packaging (PP), the Australian government avoidant behaviours (hiding the pack and not explicitly aimed to: reduce the attractiveness and smoking in front of others) and cessation-related appeal of tobacco products, particularly among thoughts (greater interest in reducing consumption young people; increase the effectiveness of the and quitting).14 graphic health warnings (GHWs) on packs; and To date, cross-sectional tracking surveys of To cite: Durkin S, reduce the ability of the packaging to mislead con- smokers at the state and national level have demon- Brennan E, Coomber K, sumers about the harms of smoking.12Ultimately, strated that implementation of the new Australian et al. Tob Control 2015;24: through the achievement of these objectives, and as packaging reduced the appeal of tobacco products – ii26 ii32. part of a comprehensive suite of tobacco control and increased the effectiveness of health warnings ii26 Durkin S, et al. Tob Control 2015;24:ii26–ii32. doi:10.1136/tobaccocontrol-2014-052058 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper up to 615 and 12 months after implementation,16 as well as to be formally screened, but may have been eligible for the finding some reduction in the extent to which smokers are study, was 57% (range 51–63%). misled about the harms of smoking.16 All survey participants who agreed to be recontacted were fol- Early effects of the implementation of packaging changes in lowed up approximately 1 month later (median=29 days, range Australia on adult quitting cognitions and behaviours have been 18–64 days), thereby creating an ongoing series of 1-month indicated in two published studies. A population survey con- cohort samples. All participants who completed the follow-up ducted during the transition to the new packaging17 found that had the chance to win one of five $100 retail vouchers per smokers using plain packs were more likely than those using month. For the current study, we restricted the sample to base- fully branded packs to have at least daily thoughts about quitting line current smokers of factory-made or roll-your-own cigarettes in the past week, and to rate quitting as a higher priority in (ie, we excluded n(unweighted)=158 cigar, pipe, ‘something their life, but did not differ in quitting intentions. A second else’ smokers and n(unweighted)=1553 recent quitters). Of the study examining ‘Quitline’ calls in the Australian state of New eligible baseline cigarette smokers (n(unweighted)=8597), 95% South Wales found, similar to the pattern after the initial intro- agreed to be recontacted (n(unweighted)=8144) and of these, duction of GHWs in 2006, calls increased by 78%, peaking 83% were successfully recontacted and completed the follow-up 4 weeks after the start of the transition to the new packaging. survey (n(unweighted)=6775). Notably though, calls remained elevated for 6 months, a longer We further restricted the sample to those who completed time than that observed after the introduction of the 2006 their follow-up survey before the implementation of the 12.5% GHWs.18 tax increase for tobacco products that occurred in Australia on 1 Another study of patrons seated in outdoor areas of cafes, res- December 2013 (excluding n(unweighted)=1288 eligible cigar- taurants and bars found that observed rates of active smoking ette smokers who completed follow-up after this date). Limiting and pack display reduced from preimplementation to postimple- the study period in this way allowed us to better isolate the mentation,19 with lower rates maintained 1 year later.20 These effects of the packaging policy from those of the tax increase. reductions in active smoking rates could reflect reduced preva- Finally, we also limited the sample to those respondents who lence of population smoking over time, consistent with recently provided valid data on all covariates (excluding n(unweighted) released national survey data,21 while another interpretation is =61 cigarette smokers, <2% of remaining sample). that smokers may have chosen to light up less often in public This sample of ‘baseline cigarette smokers’ (n(unweighted) after plain packs were introduced to avoid being judged by =5426; (weighted)=5441) was used in analyses predicting quit others. attempts. For analyses predicting quitting cognitions and micro- Building on this previous work, we used a series of cohorts of indicators of concern, the sample was restricted to those base- Australian adult smokers originally sourced from a nationally line cigarette smokers who continued to be a cigarette smoker at representative cross-sectional tracking survey, followed up follow-up (‘continuing cigarette smokers’; n(unweighted) approximately 1 month after their baseline interview to further =5145; (weighted)=5137), given that these questions were only examine the effects of the new packaging on quitting-related able to be asked of those who currently smoked. cognitions and behaviours. We compared change in these out- comes over this 1-month follow-up period for the cohorts sur- Outcome measures veyed before the packaging changes, over the 2-month period of Quitting-related cognitions at follow-up transition to the new packaging, and during the first year of the Substantial research has demonstrated that thoughts about quit- new packaging. ting and quit intentions prospectively predict making quit – attempts.23 25 Frequency of thoughts about quitting was METHODS assessed by asking ‘During the past week, how often have you Study design and participants thought about quitting?’ with response options: ‘several times a Between 9 April 2012 and 30 March 2014, 10 308 respondents day’; ‘once a day’; ‘once every few days’; ‘once’;or‘not at all’. completed a continuous cross-sectional telephone baseline Consistent with previous research which found daily thoughts survey (except between 21 December 2012 and 2 January of quitting increased with antismoking advertising,26 responses 2013, and 20 December 2013 and 6 January 2014, when the were dichotomised into those who had thought about quitting interview centre was closed over the Christmas-New Year at least once a day in the past week versus those who had period). Telephone interviews were conducted using a dual- thought about quitting less often. Quit intentions were mea- frame sample design, with half of baseline participants recruited sured using two questions: ‘Do you intend to quit in the next via landline random digit dialling (RDD) and half by mobile month?’ and ‘Have you set a firm date to quit in the next phone RDD. Details of sampling methodology are described month?’ elsewhere.16 22 Respondents were aged 18–69 years and were either smokers (n(unweighted)=8755) or recent quitters (quit in Pack concealment and microindicators of concern at follow-up the past 12 months, n(unweighted)=1553). Those who smoked Covering up or concealing cigarette packs, prematurely stubbing daily or weekly were classified as smokers, while those who out cigarettes in response to thoughts about the harms, and for- smoked monthly or less-than-monthly were allowed to self- going cigarettes, have all been found to prospectively predict – identify as a current or ex-smoker. quit attempts over the long-term in population-based studies.27 The mean monthly baseline survey response rate was defined 32 Respondents were asked ‘In the last month, how often did as completed baseline interviews as a proportion of ‘estimated you deliberately cover up or conceal your pack, or put your in-scope contacts’ that could be interviewed within the survey cigarettes in another container?’, ‘In the last month, how often period. This is a conservative assessment of the response rate, to did you stub out a cigarette before you finished it because you take account of the fact that some households/respondents that thought about the harms of smoking?’ and ‘In the last month, refused the screening process would in fact be in-scope (see how often did you stop yourself from having a cigarette when technical report for detail of calculations22). The mean monthly you had an urge to smoke?’ For each behaviour, response baseline survey response rate, adjusted for those who declined options were, ‘never’; ‘once or twice’, ‘several times’ or ‘many

Durkin S, et al. Tob Control 2015;24:ii26–ii32. doi:10.1136/tobaccocontrol-2014-052058 ii27 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper times’. Responses were dichotomised into those who reported pre-PP). The early transition phase included those surveyed at the behaviour ‘several times’ or ‘many times’ (in the past baseline in the pre-packaging changes period (20 August–28 month) versus those who reported the behaviour less frequently. September 2012) and followed-up during the transition to the new packaging (1 October–11 November 2012; Baseline pre-PP Quit attempts at follow-up —Follow-up Transition). The late transition phase included Those who were smoking at baseline were asked at follow-up if those first surveyed during the transition to the new packaging they were still smoking and if so, whether they had made any (1 October–30 November 2012) and followed-up either during attempts to quit smoking over the past month. We created a the transition or soon after the full implementation of the new binary variable that compared those who had not made a quit packaging (29 October 2012–20 January 2013; Baseline attempt between baseline and follow-up, with those who had Transition—Follow-up Transition/Early PP). The PP year 1 made a quit attempt and were still smoking at follow-up (ie, phase included those who completed both surveys in the first made an unsuccessful quit attempt), or were quit at follow-up year of full implementation of the new packaging (baseline (ie, made a successful quit attempt). Thus, this variable allowed surveys: 1 December 2012 to 4 November 2013; follow-up us to predict the proportion of smokers making quit attempts, surveys: 2 January 2013–30 November 2013; Baseline PP year irrespective of whether these attempts were successful. 1—Follow-up PP year 1). We included each individual’s baseline level of each outcome variable as a predictor of that particular Covariates outcome variable at follow-up (eg, baseline frequency of quit- All cohort analyses controlled for sex, age, education and socio- ting thoughts was included as a predictor in the models where economic status (measured using a 2011 index of relative disad- follow-up frequency of quitting thoughts was the outcome vari- vantage based on respondents’ postcode).33 Analyses also able). This enabled the use of the phase variable as a predictor controlled for baseline addiction level using the Heaviness of of the variance in follow-up quitting cognitions and behaviour Smoking Index (HSI),34 and for past 3-month antismoking that remained unexplained by an individual’s baseline levels. So, advertising activity using Target Audience Rating Points in effect the phase variable acted as a predictor of the difference (TARPs,26 35 based on date of follow-up interview). Cigarette in an individual’s quitting cognitions and behaviours between costliness in the month of follow-up interview was calculated as the baseline and follow-up surveys (approximately 1 month the ratio of the average recommended retail price (RRP) of the apart). In all logistic regression analyses, the pre-PP phase was top 10 brands (weighted by market share), to the average used as the referent category. We conducted models that were weekly earnings in the respondents’ state of residence.36 We unadjusted and adjusted for covariates. Preliminary logistic controlled for percentage change in costliness over the past regression analyses (unadjusted and adjusted) were first con- 3 months, in those survey months in which RRPs increased due ducted to examine if there were any differences between phases to excise/customs duty indexation:22 percentage change in cost- on the baseline levels of each outcome variable (see online sup- liness in August 2012, M=−0.00, SE=0.03; February 2013, plementary appendix A). M=3.29, SE=0.05; August 2013, M=0.42, SE=0.08. All ana- In analyses examining whether PP phase was associated with lyses also controlled for the number of days between the base- quit attempts at follow-up, we used the recency of previous quit line and follow-up surveys, and the date of follow-up interview. attempts at baseline as the baseline level of the outcome variable Table 1 provides the weighted number of participants within (as described above), given that Partos et al37 found that the each covariate subgroup and provides associated percentages recency of previous quit attempts is an important predictor of overall and by each analytic phase. subsequent quit attempts. We used a modified version of the variable created by Partos et al,37 with five categories indicating Weighting whether smokers had never previously tried to quit, had tried to A baseline design weight and post-stratification weight were quit more than 12 months ago, had tried to quit between six used. As detailed elsewhere, the design weight accounted for the and 12 months ago, had tried to quit 2–6 months ago or had probability of each respondent participating in the survey, based tried to quit within the past month. on landline or mobile phone usage.22 The post-stratification Additionally, as previous research indicates that increased weight accounted for telephony status, and sex, age by educa- quitting cognitions and behaviours can occur in the lead up to tion, and state of residence. tax increases that have been announced ahead of their imple- The longitudinal weight for the follow-up data (used in these mentation time,38 39 we conducted sensitivity testing to explore analyses) was derived from an adjustment to the baseline weight- effects (in adjusted models) with and without inclusion of data ing variable. The probability of response to the follow-up survey collected in November 2013, the month prior to the tax was modelled for all baseline cases using separate logistic regres- increase on 1 December 2013. In addition, previous research sions for smokers and recent quitters. The baseline weight was has indicated that interest in quitting tends to vary in reliable divided by the predicted probability from this regression so that ways in the last 3 weeks of December (less interest, less quitting) low probability cases were weighted up relative to higher prob- and the first 2 weeks of January (more interest, more quitting).40 ability cases. Finally, the weights were calibrated to meet the ori- Therefore, we also repeated all adjusted analyses including two ginal baseline rim weighting targets for sex, age by education indicator variables to capture these seasonality effects. status, and state of residence.22 All analyses were conducted in Stata V.12.1,41 adjusting for the effects of sample weighting on parameter estimates and SEs. Statistical analysis In addition, an unconditional approach (ie, the ‘subpopulation’ Proportions of those reporting quitting-related cognitions and command in Stata V.12.1)was used to limit the sample as appro- behaviours in the follow-up survey were compared across four priate for each set of analyses, ensuring correct estimation of the distinct phases. The pre-PP phase included those who completed SEs. All reported adjusted proportions and ORs were adjusted both baseline (10 April–1 September 2012) and follow-up for age, sex, education, socioeconomic status, HSI, antismoking surveys (7 May 2012–30 September 2012) prior to implementa- advertising activity, change in cigarette costliness, number of tion of the packaging changes (Baseline pre-PP—Follow-up days between the baseline and follow-up surveys and date of ii28 Durkin S, et al. Tob Control 2015;24:ii26–ii32. doi:10.1136/tobaccocontrol-2014-052058 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 1 Weighted number of cases and percentages of demographic and other covariates among continuing smokers (CS)* and among baseline smokers (BS)†, overall and by plain packaging phase Total Pre-PP Early transition Late transition PP year 1

CS BS CS BS CS BS CS BS CS BS

Weighted, n 5137 5441 1339 1423 254 276 595 617 2948 3125 Demographic covariates, % Sex Males 54.6 54.9 55.1 55.0 53.1 54.3 53.5 54.1 54.7 55.1 Females 45.4 45.1 44.9 45.0 46.9 45.7 46.5 45.9 45.3 44.9 Age 18-29 27.8 28.6 29.1 30.1 25.4 27.4 28.5 28.4 27.3 28.1 30-49 45.9 45.5 43.9 43.4 46.3 45.3 46.3 46.2 46.7 46.4 50-69 26.2 25.9 26.9 26.5 28.3 27.3 25.2 25.4 25.9 25.6 Education Less than high school 33.1 32.5 30.9 30.3 29.3 29.2 38.0 37.2 33.4 32.9 Completed high school/ 54.5 54.7 57.0 57.2 56.8 57.8 48.4 49.5 54.4 54.3 some tertiary Tertiary or above 12.5 12.8 12.1 12.6 14.0 13.0 13.6 13.3 12.2 12.8 Socio-economic status Low 40.2 39.9 38.5 38.4 36.4 35.8 40.0 39.2 41.4 41.0 Mid 41.8 42.0 41.7 41.2 42.8 43.9 43.1 43.2 41.6 42.0 High 17.9 18.1 19.8 20.4 20.8 20.3 17.0 17.6 17.0 17.0 Other covariates, mean (SE) Heaviness of Smoking Index 2.19 (0.03) 2.13 (0.03) 2.17 (0.05) 2.09 (0.05) 2.22 (0.12) 2.14 (0.12) 2.08 (0.08) 2.06 (0.08) 2.23 (0.04) 2.17 (0.04) Days between baseline and 29.92 (0.08) 29.97 (0.08) 29.15 (0.13) 29.19 (0.13) 30.03 (0.38) 30.45 (0.39) 30.80 (0.29) 30.80 (0.28) 30.08 (0.10) 30.11 (0.10) follow-up Anti-smoking advertising in 1483 (12.8) 1485 (12.5) 1662 (25.4) 1667 (24.5) 1102 (56.7) 1125 (56.1) 1421 (28.2) 1425 (27.5) 1447 (17.2) 14.46 (16.8) past 3 months All data are weighted using longitudinal weights. Due to rounding, Ns may not sum to total and percentages may not sum to 100. Descriptive data for the covariate capturing percentage change in cigarette costliness is presented in text. *Eligible sample limited to respondents who were cigarette smokers at baseline and at follow-up. This sample was used in analyses predicting quitting thoughts, intentions, pack concealment, and micro-indicators of concern. †Eligible sample limited to respondents who were cigarette smokers at baseline (i.e., either still smoking or not smoking at follow-up). This sample was used in analyses predicting attempts to quit in the past month. follow-up survey. In Tables, we present results from the rate of intentions to quit at follow-up during late transition, unadjusted and adjusted models. For brevity, we refer only to although the rate returned to baseline levels among continuing results from the adjusted models in text. smokers during PP year 1. A similar pattern emerged for the proportion of those reporting a firm date to quit within the RESULTS next month although the difference between those surveyed Sample characteristics and preliminary analyses during pre-PP and late transition was not significant. Table 1 shows the weighted sample at each phase and the Continuing smokers were significantly more likely to report characteristics of the sample who were smoking cigarettes at increased follow-up rates of concealing their pack in both baseline and follow-up interviews (continuing smokers), late transition and PP year 1 compared to pre-PP (table 2). and those who were smoking at baseline and either continued Continuing smokers were also significantly more likely to report to smoke or had quit at follow-up interview (baseline smokers). increases in stubbing out their cigarette in PP year 1 compared Preliminary analyses indicated some significant differences in to pre-PP. Continuing smokers surveyed during early transition quitting cognitions and behaviours measured in the baseline were significantly more likely to report increased rates of stop- cross-sectional surveys between the pre-PP phase and subsequent ping smoking when they had an urge at follow-up as compared phases (see online supplementary appendix A). Compared to to those surveyed in pre-PP. pre-PP, there was a significantly lower level of baseline daily Importantly, there were significantly greater increases in quit thoughts of quitting in those surveyed during the late transition attempt rates in the early transition and PP year 1 phases, as and PP year 1 phases. There was also a significantly greater pro- compared to the pre-PP phase (table 2). portion setting a baseline firm date to quit and concealing packs Sensitivity testing excluding data from November 2013 and in those surveyed during PP year 1 as compared to pre-PP. including additional covariates to adjust for seasonal factors showed a similar pattern of effects for all findings (available in Analyses of the impact of PP on short-term within-person online supplementary appendices B and C). change in quitting cognitions and behaviours Table 2 shows that compared to the pre-PP phase there were no DISCUSSION significant differences in the proportion of continuing smokers Overall, we found evidence that smokers were more likely to who changed their thoughts about quitting to at-least-daily at show short-term increases in quitting intentions, and to engage follow-up across each subsequent phase. Compared to those sur- in quitting behaviours after implementation of the packaging veyed in pre-PP, there were significantly greater increases in the changes. After adjusting for important covariates and baseline

Durkin S, et al. Tob Control 2015;24:ii26–ii32. doi:10.1136/tobaccocontrol-2014-052058 ii29 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 2 Association between plain packaging (PP) phases and follow-up levels of quitting-related cognitions and behaviours—logistic regression results including baseline levels of each outcome variable in all models Weighted N Pre-PP Early transition Late transition PP year 1

Daily thoughts about quitting in the past week† 5079 Unadjusted % 36.1 37.4 36.8 36.6 OR (95% CI) 1.00 1.08 (0.77 to 1.53) 1.04 (0.81 to 1.34) 1.03 (0.86 to 1.23) Adjusted % 33.2 36.2 36.1 38.2 OR (95% CI) 1.00 1.20 (0.83 to 1.73) 1.20 (0.89 to 1.60) 1.36 (0.90 to 2.04) Intend to quit in next month† 5137 Unadjusted % 34.8 35.5 41.7** 36.3 OR (95% CI) 1.00 1.04 (0.74 to 1.47) 1.50 (1.16 to 1.95) 1.09 (0.92 to 1.30) Adjusted % 36.1 36.5 42.0* 35.6 OR (95% CI) 1.00 1.03 (0.72 to 1.47) 1.42 (1.06 to 1.92) 0.98 (0.64 to 1.48) Firm date to quit in next month† 5137 Unadjusted % 8.4 7.1 10.8 6.6 OR (95% CI) 1.00 0.82 (0.48 to 1.39) 1.36 (0.92 to 2.02) 0.76 (0.57 to 1.00) Adjusted % 8.6 7.1 11.0 6.5 OR (95% CI) 1.00 0.80 (0.46 to 1.39) 1.35 (0.86 to 2.13) 0.73 (0.36 to 1.47) Concealed or covered pack several or many times 5065 Unadjusted % 16.1 17.9 20.0* 18.2 in past month† OR (95% CI) 1.00 1.18 (0.78 to 1.79) 1.42 (1.00 to 2.02) 1.22 (0.97 to 1.53) Adjusted % 14.3 16.0 18.8* 19.5* OR (95% CI) 1.00 1.19 (0.77 to 1.84) 1.55 (1.05 to 2.31) 1.65 (1.01 to 2.72) Stubbed out several or many times in past month† 5103 Unadjusted % 24.3 23.1 23.2 26.9 OR (95% CI) 1.00 0.92 (0.62 to 1.37) 0.93 (0.70 to 1.23) 1.18 (0.98 to 1.43) Adjusted % 21.9 22.2 22.5 28.4* OR (95% CI) 1.00 1.02 (0.68 to 1.54) 1.04 (0.76 to 1.43) 1.55 (1.01 to 2.36) Stopped from smoking several or many times 5080 Unadjusted % 36.8 44.7* 39.3 38.2 in past month† OR (95% CI) 1.00 1.48 (1.08 to 2.03) 1.14 (0.89 to 1.45) 1.07 (0.91 to 1.26) Adjusted % 36.7 44.9* 39.4 38.1 OR (95% CI) 1.00 1.51 (1.08 to 2.10) 1.15 (0.87 to 1.52) 1.08 (0.74 to 1.57) Attempted to quit in past month‡ 5411 Unadjusted % 23.6 28.3 18.7* 24.1 OR (95% CI) 1.00 1.32 (0.94 to 1.85) 0.71 (0.54 to 0.95) 1.03 (0.86 to 1.23) Adjusted % 20.2 25.5* 17.4 26.6* OR (95% CI) 1.00 1.43 (1.00 to 2.03) 0.81 (0.58 to 1.12) 1.52 (1.01 to 2.30) Unadjusted=unadjusted for covariates, adjustment for baseline level of each outcome variable was retained for comparability with adjusted findings. Adjusted=adjusted for baseline level of each outcome variable and covariates: date of the follow-up survey; number of days between baseline and follow-up survey; cumulative Target Audience Rating Points (antismoking television advertising) in the 3 months prior to the follow-up survey; change in cigarette costliness (based on month of follow-up survey); sex; age; education; socioeconomic status; and Heaviness of Smoking Index. *p<0.05, **p<0.01. †Eligible sample limited to respondents who were cigarette smokers at baseline and at follow-up. ‡Eligible sample limited to respondents who were cigarette smokers at baseline (ie, either still smoking or not smoking at follow-up). levels of each outcome variable, compared to the pre-PP period, was linked to reduced pack display and active smoking at we found greater increases in rates of smokers stopping them- outdoor dining areas immediately following19 and around selves from smoking when they had an urge to smoke and 1 year after implementation.20 higher quit attempt rates in the early transition period, and Results indicated that short-term changes in quit intentions were greater intentions to quit in the next month and greater avoid- no different after the full implementation of the new packaging ance of GHWs through covering or concealing their pack than in the pre-PP phase. The absence of these effects on next- during the late transition phase. Also, we observed higher levels month quit intentions and having a firm date to quit, in the presence of pack concealment, more stubbing out of cigarettes prema- of effects on microindicators of concern and quit attempts, may turely, and higher quit attempt rates in the first full year of the indicate that plain packs with larger GHWs continue to trigger new packaging. increased concern in the smoking moment—reflectedinincreased Given the observed reductions in pack appeal and greater stubbing out before finishing a cigarette and pack concealment—but GHW effectiveness in cross-sectional surveys by Dunlop et al15 do not continue to increase deliberate plans to quit well after the and Wakefield et al,16 and the ability of these appeal and GHW shock of the transition phase. This would be consistent with previ- effectiveness variables to prospectively predict quitting-related ous research that has indicated that larger and stronger GHWs are cognitions and behaviours,42 it is likely that at least some of the associated with increased microindicators of concern which in turn effects observed in this study were driven by changes in these prospectively predict quit attempts independent of quit inten- – ‘upstream’ factors more closely tied to the introduction of these tions,27 30 43 suggesting the presence of a less deliberative pathway policies. The findings from this study support previous experi- of effects. Alternatively, it is also possible that we were unable to mental and naturalistic studies,81213indicating the potential detect short-term increases in having a firm date to quit because the for the packaging changes to influence quitting cognitions, baseline levels of such quit intentions were already elevated in those intentions and behaviours. They are also broadly consistent with surveyed in the first year after the packaging changes, as revealed by previous studies demonstrating that larger and more graphic our preliminary analyses of the baseline levels of quitting cognitions warning labels stimulate greater quitting thoughts, avoidance and behaviours (see online supplementary appendix A). behaviour, quit intentions and quit attempts.24 28 32 43 The find- The survey used a relatively short follow-up period in order ings also accord with studies suggesting these policies had a to focus on follow-up quitting cognitions and ensure accurate positive impact on quitting thoughts during the transition recall of quit attempts,44 and so was not designed or powered to phase,17 was associated with increased calls to a quitting help- examine quitting success. Therefore, the extent to which the line up to 6 months after the introduction of plain packs,18 and positive outcomes we observed may be maintained and translate ii30 Durkin S, et al. Tob Control 2015;24:ii26–ii32. doi:10.1136/tobaccocontrol-2014-052058 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper into longer-term reductions in smoking prevalence still needs to quitting cognitions, microindicators of concern and increased be determined. An additional limitation of this study was that quit attempts among adult smokers. the PP year 1 period had many more participants than the other transition phases, and so it is possible that we were unable to Acknowledgements The authors thank Professor Ron Borland for sharing fi questions from the ITC survey for inclusion in the telephone tracking survey, the detect small effects in the transition phases, whereas in the nal Social Research Centre for data collection and Megan Bayly for data checking. phase we may have had greater ability to detect smaller effects. Contributors MW designed the study. KC and MZ managed data collection and Strengths of this study include the examination of these cogni- cleaned data files. SD, EB and KC analysed the data and interpreted the results. SD tions and behaviours in a national sample with a rigorous sampling drafted the manuscript with contributions from all authors. All authors approved the method, using consistent questions from prior to after the pack- final manuscript. aging changes, with good follow-up rates. By adjusting for individ- Funding The National Plain Packaging Tracking Survey was funded under a ual differences and levels of each outcome variable at baseline, we contract with the Australian Government Department of Health and Ageing. were able to minimise the influence of any month-to-month cross- Competing interests The authors wish to advise that MW was a member and sectional sampling variation and permit a clear examination of the MS a technical writer for the Tobacco Working Group of the Australian National influence of the packaging changes on monthly changes in quitting Preventive Health Task Force and MW was a member of the Expert Advisory Committee on Plain Packaging that advised the Australian Department of Health on cognitions and behaviours within the same individuals. research pertaining to the plain packaging legislation. MW, SD and EB hold Adjustment for the amount of antismoking television advertising competitive grant funding from the Australian National Health and Medical Research and changes in cigarette costliness limited the viability of these Council and MW holds competitive grant funding from the US National Institutes of interventions as competing explanations for observed effects. Health, Australian National Preventive Health Agency and BUPA Health Foundation. Furthermore, the sensitivity testing removing November 2013 Ethics approval The survey was approved by the Cancer Council Victoria Human from the analyses and including adjustment for the impact of sea- Ethics Committee (HREC 0018). sonal factors indicated that the observed effects were not unduly Provenance and peer review Not commissioned; externally peer reviewed. influenced by anticipation of the December 2013 preannounced Data sharing statement No further data are available. tax increase, nor the Christmas and New Year periods. Open Access This is an Open Access article distributed in accordance with the The observed effects were generally modest in size. It is not Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which unexpected that changes in these ‘downstream’ quitting out- permits others to distribute, remix, adapt, build upon this work non-commercially, comes would be small, as the introduction of the new packaging and license their derivative works on different terms, provided the original work is is one factor among many other personal, interpersonal and properly cited and the use is non-commercial. See: http://creativecommons.org/ licenses/by-nc/4.0/ external (mass media, policy and tobacco marketing) factors likely to combine to influence individual quitting cognitions and – behaviours..25 45 47 The overall impact on quitting and smoking REFERENCES 1 Tobacco Plain Packaging Act. No. 148, 2011 as amended. 2011. http://www. prevalence will need to be examined in cohort studies with comlaw.gov.au/Details/C2013C00190 longer-term follow-up and repeated national surveys using stan- 2 Scollo M, Lindorff K, Coomber K, et al. Standardised packaging and new enlarged dardised methods. In this respect, there have been some recent graphic health warnings for tobacco products in Australia—legislative requirements promising reductions in smoking prevalence from the latest and implementation of the Tobacco Plain Packaging Act 2011 and the Competition and Consumer (Tobacco) Information Standard. Tob Control 2015;24:ii9–16. national Australian household surveys conducted before and 21 3 Competition and Consumer (Tobacco) Information Standard. F2013C00598. 2011. after the introduction of the new packaging. Since reductions http://www.comlaw.gov.au/Details/F2013C00598 in overall smoking rates are driven not only by adult smokers 4 Thrasher JF, Rousu MC, Hammond D, et al. Estimating the impact of pictorial health quitting, but by adolescents avoiding taking up smoking, warnings and “plain” cigarette packaging: evidence from experimental auctions – national surveys of schoolchildren will also be important to among adult smokers in the United States. Health Policy 2011;102:41 8. 5 Hoek J, Wong C, Gendall P, et al. Effects of dissuasive packaging on young adult examine, including in the years to come when more complete smokers. Tob Control 2011;20:183–8. 48 49 quantification of effects might be possible. Nonetheless, 6 Hammond D, Doxey J, Daniel S, et al. Impact of female-oriented cigarette these data provide some of the strongest evidence to date that packaging in the United States. Nicotine Tob Res 2011;13:579–88. implementation of PP and larger GHWs was associated with 7 White CM, Hammond D, Thrasher JF, et al. The potential impact of plain packaging of cigarette products among Brazilian young women: an experimental study. BMC Public Health 2012;12:737. 8 Hammond D. Standardized packaging of tobacco products: evidence review. Prepared on behalf of the Irish Department of Health; March 2014. http://www. What this paper adds drugsandalcohol.ie/22106/1/Standardized-Packaging-of-Tobacco-Products-Evidence- Review.pdf (accessed 19 Sep 2014). 9 Hammond D, Daniel S, White CM. The effect of cigarette branding and plain ▸ Experimental and naturalistic studies suggested that plain packaging on female youth in the United Kingdom. J Adolesc Health 2013;52:151–7. — packaging with large graphic health warnings may increase 10 Rootman I, Flay BR. A study on youth smoking plain packaging, health warnings, event marketing and price reductions. Toronto: University of Toronto, University of quitting-related cognitions and behaviours among adult Illinois at Chicago, York University, Ontario, Tobacco Research Unit, Addiction smokers. Research Foundation, 1995. ▸ Australian studies conducted since plain packaging 11 Kotnowski K, Hammond D, Fong GT, et al. The efficacy of standardized cigarette implementation demonstrated reduced appeal of tobacco packaging among young women in Canada: a discrete choice experiment. Seattle, WA: Society for Research on Nicotine & Tobacco Annual Meeting, 2014. products, increased effectiveness of health warnings and an 12 Brose LS, Chong CB, Aspinall E, et al. Effects of standardised cigarette packaging increase in Quitline calls. on craving, motivation to stop and perceptions of cigarettes and packs. ▸ This national cohort study of Australian adult cigarette Psychol Health 2014;29:849–60. smokers provides evidence that plain packaging with larger 13 Moodie C, Mackintosh AM, Hastings G, et al. Young adult smokers’ perceptions of – graphic health warnings increased rates of quit intentions, plain packaging: a pilot naturalistic study. Tob Control 2011;20:367 73. 14 Gallopel-Morvan K, Moodie C, Eker F, et al. Perceptions of plain packaging among pack avoidance, stopping oneself from smoking and quit young adult roll-your-own smokers in France: a naturalistic approach. Tob Control attempts during the transition period, as well as increased Published Online First: 11 Jun 2014. doi:10.1136/tobaccocontrol-2013-051513 levels of pack avoidance, stubbing out prematurely and quit 15 Dunlop SM, Dobbins T, Young JM, et al. Impact of Australia’s introduction of attempts in the first year of the packaging changes. tobacco plain packs on adult smokers’ pack-related perceptions and responses: results from a continuous tracking survey. BMJ Open 2014;4:e005836.

Durkin S, et al. Tob Control 2015;24:ii26–ii32. doi:10.1136/tobaccocontrol-2014-052058 ii31 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

16 Wakefield M, Coomber K, Zacher M, et al. Australian adult smokers’ responses to www.ausstats.abs.gov.au/ausstats/subscriber.nsf/0/22CEDA8038AF7A0DCA257B3 plain packaging with larger graphic health warnings 1 year after implementation: B00116E34/$File/2033.0.55.001%20seifa%202011%20technical%20paper.pdf results from a national cross-sectional tracking survey. Tob Control 2015;24:ii17–25. (accessed 19 Sep 2014). 17 Wakefield M, Hayes L, Durkin S, et al. Introduction effects of the Australian plain 34 Heatherton TF, Kozlowski LT, Frecker RC, et al. Measuring the heaviness of packaging policy on adult smokers: a cross-sectional study. BMJ Open 2013;3:e003175. smoking: using self-reported time to the first cigarette of the day and number of 18 Young JM, Stacey I, Dobbins TA, et al. Association between tobacco plain cigarettes smoked per day. Br J Addict 1989;84:791–9. packaging and Quitline calls: a population-based, interrupted time-series analysis. 35 Wakefield M, Spittal M, Durkin S, et al. Effects of mass media campaign exposure Med J Aust 2014;200:29–32. intensity and durability on quit attempts in a population-based cohort study. 19 Zacher M, Bayly M, Brennan E, et al. Personal tobacco pack display before and after Health Educ Res 2011;26:988–97. the introduction of plain packaging with larger pictorial health warnings in Australia: 36 Australian Bureau of Statistics. Average Weekly Earnings, Australia. Cat No 6302.0. an observational study of outdoor café strips. Addiction 2014;109:653–62. November 2013. http://www.abs.gov.au/ausstats/[email protected]/mf/6302.0 (accessed 20 Zacher M, Bayly M, Brennan E, et al. Personal pack display and active smoking at 19 Sep 2014). outdoor café strips: assessing the impact of plain packaging 1 year 37 Partos TR, Borland R, Yong HH, et al. The quitting rollercoaster: how recent quitting postimplementation. Tob Control 2015;24:ii94–7. history affects future cessation outcomes (data from the International Tobacco 21 Australian Institute of Health and Welfare. Highlights from the 2013 survey: tobacco Control 4-country cohort study). Nicotine Tob Res 2013;15:1578–87. smoking. Canberra: Australian Institute of Health and Welfare, 2014. http://www. 38 Harwell TS, Lee L, Haugland C, et al. Utilization of a tobacco quit line prior to and aihw.gov.au/alcohol-and-other-drugs/ndshs/ (accessed 19 September 2014). after a tobacco tax increase. J Public Health Manag Pract 2007;13:637–41. 22 Coomber K, Zacher M, Durkin S, et al. Australian National Tobacco Plain Packaging 39 Bush T, Zbikowski S, Mahoney L, et al. The 2009 US federal cigarette tax increase Tracking Survey: Technical Report. Melbourne: Cancer Council Victoria and Social and Quitline utilization in 16 states. J Environ Public Health 2012;314740. Research Centre, 2014. http://tobaccocontrol.bmj.com/lookup/suppl/doi:10.1136/ 40 Durkin S, Wakefield MA, Spittal MJ. Which types of televised anti-tobacco campaigns tobaccocontrol-2014-052050/-/DC1. prompt more quitline calls from disadvantaged groups? Health Educ Res 23 Hyland A, Borland R, Li Q, et al. Individual-level predictors of cessation behaviours 2011;26:998–1009. among participants in the International Tobacco Control (ITC) Four Country Survey. 41 StataCorp. Stata Statistical Software: Release 12. College Station, TX: StataCorp LP, 2011. Tob Control 2006;15:iii83–94. 42 Brennan E, Durkin S, Coomber K, et al. Are quitting-related cognitions and 24 Borland R, Yong HH, Balmford J, et al. Motivational factors predict quit attempts behaviours predicted by proximal responses to plain packaging with larger health but not maintenance of smoking cessation: findings from the International Tobacco warnings? Findings from a national cohort study with Australian adult smokers. Tob Control Four country project. Nicotine Tob Res 2010;12:S4–11. Control 2015;24:ii33–41. 25 Vangeli E, Stapleton J, Smit ES, et al. Predictors of attempts to stop smoking and 43 Borland R, Wilson N, Fong GT, et al. Impact of graphic and text warnings on their success in adult general population samples: a systematic review. Addiction cigarette packs: findings from four countries over five years. Tob Control 2011;106:2110–21. 2009;18:358–64. 26 Dunlop S, Cotter T, Perez D, et al. Televised antismoking advertising: effects of level 44 Borland R, Partos TR, Yong HH, et al. How much unsuccessful quitting activity is and duration of exposure. Am J Public Health 2013;103:e66–73. going on among adult smokers? Data from the International Tobacco Control Four 27 Borland R. Tobacco health warnings and smoking-related cognitions and Country cohort survey. Addiction 2012;107:673–82. behaviours. Addiction 1997;92:1427–35. 45 Wakefield MA, Durkin S, Spittal MJ, et al. Impact of tobacco control policies and 28 Borland R, Yong HH, Wilson N, et al. How reactions to cigarette packet health warnings mass media campaigns on monthly adult smoking prevalence. Am J Public Health influence quitting: findings from the ITC Four-Country survey. Addiction 2009;104:669–75. 2008;98:1443–50. 29 Partos TR, Borland R, Thrasher JF, et al. The predictive utility of micro indicators of 46 Zhou X, Nonnemaker J, Sherrill B, et al. Attempts to quit smoking and relapse: concern about smoking: findings from the International Tobacco Control Four factors associated with success or failure from the ATTEMPT cohort study. Addict Country study. Addict Behav 2014;39:1235–42. Behav 2009;34:365–73. 30 Li L, Borland R, Fong GT, et al. Smoking-related thoughts and microbehaviours, and 47 Wakefield MA, Coomber K, Durkin SJ, et al. Time series analysis of the impact of their predictive power for quitting: findings from the International Tobacco Control tobacco control policies on smoking prevalence among Australian adults, 2001– (ITC) Survey. Tob Control Published Online First: 25 Feb 2014. doi:10.1136/ 2011. Bull World Health Organ 2014;92:413–22. tobaccocontrol-2013-051384 48 White VM, Warne CD, Spittal MJ, et al.Whatimpacthavetobaccocontrol 31 Hammond D, Fong GT, McDonald PW, et al. Graphic Canadian cigarette warning policies, cigarette price and tobacco control programme funding had on Australian labels and adverse outcomes: evidence from Canadian smokers. Am J Public Health adolescents’ smoking? Findings over a 15-year period. Addiction 2011;106: 2004;94:1442–5. 1493–502. 32 Yong H-H, Borland R, Thrasher JF, et al. Mediational pathways of the impact 49 White V, Durkin S, Coomber K, et al. What is the role of tobacco control advertising of cigarette warning labels on quit attempts. Health Psychol 2014;33:1410–20. intensity and duration in reducing adolescent smoking prevalence? Findings from 33 Australian Bureau of Statistics. Technical paper: socio-economic indexes for areas 16-years of tobacco control mass media advertising in Australia. Tob Control (SEIFA) 2011. Canberra, Australia: Australian Bureau of Statistics, 2013. http:// Published Online First: 29 Aug 2013. doi:10.1136/tobaccocontrol-2012-050945

ii32 Durkin S, et al. Tob Control 2015;24:ii26–ii32. doi:10.1136/tobaccocontrol-2014-052058 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Short-term changes in quitting-related cognitions and behaviours after the implementation of plain packaging with larger health warnings: findings from a national cohort study with Australian adult smokers Sarah Durkin, Emily Brennan, Kerri Coomber, Meghan Zacher, Michelle Scollo and Melanie Wakefield

Tob Control 2015 24: ii26-ii32 doi: 10.1136/tobaccocontrol-2014-052058

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii26

These include: Supplementary Supplementary material can be found at: Material http://tobaccocontrol.bmj.com/content/suppl/2015/02/25/tobaccocon trol-2014-052058.DC1.html http://tobaccocontrol.bmj.com/content/suppl/2015/03/18/tobaccocon trol-2014-052058.DC2.html References This article cites 31 articles, 11 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii26#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper Are quitting-related cognitions and behaviours predicted by proximal responses to plain packaging with larger health warnings? Findings from a national cohort study with Australian adult smokers Emily Brennan, Sarah Durkin, Kerri Coomber, Meghan Zacher, Michelle Scollo, Melanie Wakefield

▸ Additional material is ABSTRACT ability of packaging to mislead consumers about published online only. To view Background Implementation of tobacco plain the harmful effects of smoking. More broadly please visit the journal online (http://dx.doi.org/10.1136/ packaging (PP) with larger graphic health warnings though, and as part of a comprehensive tobacco tobaccocontrol-2014-052057). (GHWs) in Australia had positive effects on responses control programme, the PP legislation aimed to reflecting the specific objectives of the PP policy and on reduce smoking rates by discouraging people from Centre for Behavioural Research in Cancer, Cancer follow-up quitting-related cognitions and behaviours. The taking up smoking, encouraging smokers to quit Council Victoria, Melbourne, aim of this study was to examine predictive relationships and discouraging relapse. In a complementary way, Victoria, Australia between these proximal and distal outcomes. the Information Standard that prescribed the new Methods A nationally representative sample of GHWs23aimed to increase consumer knowledge Correspondence to Professor Melanie Wakefield, Australian adult cigarette smokers completed a baseline of the health effects resulting from tobacco use, Centre for Behavioural survey and a 1-month follow-up survey within the first ensure the continuing effectiveness of health warn- Research in Cancer, Cancer year of policy implementation (n(weighted)=3125). ings on the retail packaging of tobacco products, Council Victoria, 615 St Kilda Logistic regression analyses tested whether baseline and by ensuring their effectiveness, encourage ces- Rd, Melbourne Vic 3004, measures of cigarette appeal, GHW effectiveness, sation and discourage initiation and relapse.3 Australia; melanie.wakefield@cancervic. perceived harm and concern/enjoyment predicted each of Using data collected in continuous cross-sectional org.au seven follow-up measures of quitting-related cognitions surveys conducted between April 2012 and and behaviours, adjusting for baseline levels of the November 2013 with a nationally representative Received 25 September 2014 outcome and covariates. sample of over 7000 Australian adult cigarette Accepted 6 January 2015 Results In multivariable models, we found consistent smokers (the National Plain Packaging Tracking evidence that several baseline measures of GHW Survey), Wakefield et al4 (this volume) demonstrated effectiveness positively and significantly predicted the that the three specific objectives of the PP policy were likelihood that smokers at follow-up reported thinking largely achieved within the first year of implementa- about quitting at least daily, intending to quit, having a tion. Cigarette and cigarette packaging appeal firm date to quit, stubbing out cigarettes prematurely, reduced after implementation of the packaging stopping oneself from smoking and having attempted to changes. Smokers were more likely to report disliking quit. Two of the quitting-related outcomes were also the look of their pack, that their pack had lower predicted by feeling more smoking-related concern than appeal compared to a year ago, their cigarettes were enjoyment. A smaller number of the appeal variables of lower quality, less satisfying, and of lower value were prospectively associated with quitting-related compared to a year ago, and that cigarette brands do outcomes, while believing that brands do not differ in not differ in prestige.4 Postimplementation, smokers harmfulness did not positively predict any outcomes. were also more likely to notice the GHW first when Conclusions These findings provide an initial insight they looked at their pack, to claim the GHWs had into the pathways through which PP with larger GHWs made them feel motivated to quit, and to report that may lead to changes in smoking behaviour. Future they had tried to avoid the GHWs by deliberately research should examine whether the effects are concealing their pack and requesting a pack with a conditional on individual demographic and smoking different health warning. Demonstrating the credibil- characteristics. ity of the GHWs, the proportion of smokers who did not believe that the dangers of smoking had been

Open Access exaggerated remained stable at around two-thirds. Scan to access more INTRODUCTION Finally, there was some evidence that the packaging free content Drab dark brown plain packaging (PP) for tobacco changes reduced the ability of tobacco packaging to products was first introduced in Australia from 1 mislead consumers. The proportion of smokers who October 2012 and fully implemented from 1 believed brands do not differ in harm increased from December 2012.12At the same time, new and before to after implementation. However, there was larger graphic health warnings (GHWs), covering no change over time in the perception that cigarette 75% of the front of cigarette packs (up from 30% brands differ in strength or that one’s own cigarettes previously) and maintaining a coverage of 90% of were more harmful compared to a year ago.4 23 5 To cite: Brennan E, the back, were also introduced onto packs. The Supplementing this evidence, Durkin et al (this Durkin S, Coomber K, et al. specific objectives of the PP legislation were to: (A) volume) found increases in quitting-related cogni- Tob Control 2015;24: reduce the appeal of tobacco products; (B) increase tions and behaviours among adult cigarette – ii33 ii41. the effectiveness of the GHWs and (C) reduce the smokers following implementation of PP with

Brennan E, et al. Tob Control 2015;24:ii33–ii41. doi:10.1136/tobaccocontrol-2014-052057 ii33 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper larger GHWs. The National Plain Packaging Tracking Survey (unweighted)=2907) were used in models predicting all out- was extended to a cohort study by recontacting smokers comes except for quit attempts. Second, cigarette smokers at 1 month later, which allowed for comparisons in the rate of baseline who completed the follow-up survey (‘baseline cigarette 1-month change in quitting-related outcomes for cohorts sur- smokers’; n(weighted)=3125; n(unweighted)=3081) were used veyed during the preimplementation, transition and postimple- in models predicting the likelihood that smokers had attempted mentation periods. Most notably, compared with smokers to quit in the month between the baseline and follow-up surveyed before the packaging changes, those who completed surveys. their follow-up survey after full implementation were more likely to report that in the past month they had concealed their Outcome measures pack, stubbed out cigarettes prematurely and attempted to quit.5 Quitting-related cognitions Although no research has yet specifically examined predictive As described by Durkin et al,514in the baseline and follow-up relationships between proximal responses to the Australian pack- interviews we measured the occurrence of thoughts about quit- aging changes and subsequent quitting-related outcomes, it is ting in the past week (thought about quitting once or several reasonable to expect that these measures will be prospectively times a day vs thought about quitting only once every few days, related. Cigarette packaging—as one of many forms of cigarette once or not at all). Intentions to quit were measured by asking marketing—is known to play an important role in driving the respondents whether they intended to quit in the next month – initiation, continuation and cessation of smoking behaviour,6 11 (yes vs no or do not know/cannot say), and by further asking and exposure to GHWs has been shown to increase quitting those who intended to quit if they had set a firm date to quit in cognitions and attempts12 and even to reduce smoking preva- the next month (yes vs no or do not know/cannot say (including lence.13 In the light of evidence that PP with larger GHWs had those who did not intend to quit)). the desired effect on both the specific objectives of the PP legis- lation4 and also on a number of more distal outcomes,5 the aim Pack concealment and microindicators of concern of the current study was to investigate associations between At baseline and follow-up, we asked respondents to report how these proximal and distal measures of effectiveness of the pack- frequently in the past month they had concealed or covered their aging changes. Specifically, we examined whether each of the cigarette packs, stubbed out a cigarette because they thought variables that Wakefield et al4 used to measure the effectiveness about the harms of smoking, and had stopped themselves from of PP with larger GHWs and that were found to change in the smoking when they had an urge to smoke. Response options for expected direction following implementation prospectively pre- all three measures were dichotomised to allow us to compare dicted each of the quitting-related cognitions and behaviours those who had engaged in the behaviour several or many times examined by Durkin et al5 (thoughts about quitting, intentions in the past month with those who had performed the behaviour to quit, pack concealment, stubbing out, stopping oneself from never or just once or twice.514 smoking and attempting to quit). Evidence of prospective rela- tionships between these proximal and distal measures would Quit attempts shed light on the mechanisms through which PP with larger At baseline, respondents were asked whether they had ever GHWs may contribute to quitting cognitions and behaviours. attempted to quit smoking and, if so, how long it had been since their last quit attempt. At follow-up, those who had been METHODS current smokers at baseline were asked if they were still smoking Study design and procedure and, if so, whether they had made any attempts to quit smoking Using dual frame random digit dialling telephone surveys over the past month. Our measure of attempts to quit between (response rate 57%), we conducted a prospective cohort study baseline and follow-up included all baseline smokers who had in which respondents completed a follow-up interview approxi- made a quit attempt, including those who were still smoking mately 1 month after baseline (median time to follow-up=29 and those who were quit at follow-up (ie, had made a successful days, range=18–64 days; mean retention rate per month=83%, quit attempt). Given the evidence that the recency of previous range=78%–87%). The study procedure is described in more quit attempts is an important predictor of subsequent quit detail elsewhere.4514 attempts,15 we controlled for smokers’ recent quitting history using a five-category measure: never tried to quit; tried to quit Sample more than 12 months ago; tried to quit between 6 and For the purposes of this study, the sample was restricted to current 12 months ago; tried to quit between 2 and 6 months ago or smokers of factory-made or roll-your-own cigarettes (currently tried to quit within the past month. smoked daily or weekly, or smoked monthly or less-than-monthly but self-identified as a smoker rather than as an ex-smoker) who Predictor variables completed their baseline and follow-up surveys during the first Appeal variables year of implementation of the packaging changes; that is, after 1 As predictor variables, we used six appeal-related variables that December 2012 and before 30 November 2013 (prior to imple- changed from preimplementation to postimplementation in a mentation of the 12.5% tax increase for tobacco products that direction consistent with the aims of the PP legislation.4 At base- occurred in Australia on 1 December 2013). We further limited line, respondents: (1) indicated whether they liked the look of the sample to those who had valid data on all covariates (>98% of their current cigarette pack (somewhat or strongly disagreed cases). In total, within this period, 4240 (weighted; n(unweighted) that they liked their pack vs all other responses); rated their =4229) cigarette smokers completed their baseline survey and current cigarettes as being higher, lower or about the same com- 3125 (weighted; n(unweighted)=3081) of these also completed pared to a year ago in terms of (2) appeal of the packaging, (3) the follow-up survey. quality, (4) satisfaction and (5) value for money (lower com- We used two analytical samples. First, cigarette smokers at pared to a year ago vs all others); and (6) reported whether they baseline who continued to be cigarette smokers at follow-up believed there are differences between brands in prestige (no vs (‘continuing cigarette smokers’; n(weighted)=2948; n yes/do not know).14 ii34 Brennan E, et al. Tob Control 2015;24:ii33–ii41. doi:10.1136/tobaccocontrol-2014-052057 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Health warning effectiveness variables the ratio of the average recommended retail price (RRP) of the Five GHW effectiveness variables changed in a direction consist- top 10 brands (weighted by market share) to the average weekly ent with the aims of the PP legislation.4 At baseline, respondents earnings in the respondents’ state of residence.22 We controlled indicated: (1) whether they usually noticed the GHW first when for percentage change in costliness over the past 3 months, in looking at a pack of cigarettes (vs noticing other aspects of the those survey months in which RRPs increased due to excise/ pack first, such as branding); (2) whether they agreed or dis- customs duty indexation14: percentage change in costliness in agreed that the dangers of smoking have been exaggerated (some- February 2013, M=3.29, SE=0.05; August 2013, M=0.42, what/strongly disagree vs neutral and agree responses); (3) SE=0.08. whether the health warnings made them feel more motivated to quit (much more motivated vs all others); and whether they had Statistical analysis tried to avoid the GHWs in the past month by (4) covering up or The baseline sample was weighted using a design weight and a concealing their pack or putting their cigarettes in another con- poststratification weight, accounting for telephony status (land- tainer (several or many times vs other responses) or (5) request- line or mobile), gender, age by education, and state of resi- ing a pack with a different health warning on it (yes vs no).14 dence.414The follow-up sample was weighted using a longitudinal weight, derived from an adjustment to the baseline weighting variable, which accounted for each participant’s prob- Perceived harm variables 514 The one harm perception variable that changed in a direction ability of being retained in the follow-up sample. consistent with the aims of PP4 was also included. Smokers were We conducted a series of initial logistic regression models to asked whether they believed that some cigarette brands are examine the association between each predictor and each more harmful than others (no vs yes/do not know).14 outcome (ie, one model per predictor/outcome). When more than one significant predictor (at p<0.05) of an outcome was identified, we then conducted a multivariable model that Balance between enjoyment and concern included all predictors associated with the outcome at p<0.05, Deciding to give up smoking can be conceptualised as a struggle so as to identify the strongest independent predictors. We con- between factors that enhance the enjoyment of smoking and 16 ducted initial and multivariable models that were unadjusted factors that increase concern about the impact of smoking. (presented in appendix A of the online supplementary material) Although reducing enjoyment and increasing concern were not and adjusted for the covariates described above as well as the the primary objectives of the PP legislation, it is likely that these date of the follow-up survey and the number of days between fl two constructs would be in uenced by cigarette appeal, GHW surveys (results from adjusted models are presented here in effectiveness and perceived harm, and in combination, they have 17 tables and text). Unadjusted and adjusted models both con- been found to predict quitting-related outcomes. We created a trolled for the baseline level of the outcome variable. baseline measure of the balance between enjoyment and We conducted two sets of sensitivity testing. First, to examine concern from items measuring how often respondents had the possibility that associations between the predictors and the thought about how much they enjoy smoking in the past outcomes were influenced by the anticipation of the 12.5% tax month: 1=never; 2=once or twice; 3=several times or increase on 1 December 2013 rather than the packaging 4=many times, and how concerned they were that smoking changes, we repeated all adjusted analyses excluding respon- may affect or has already affected their own personal health: dents who were followed up in November 2013 (n(weighted) 1=not at all; 2=a little concerned; 3=somewhat concerned; =297 baseline cigarette smokers and 284 continuing cigarette 4=very concerned and 5=extremely concerned. We then stan- smokers). Second, previous research has indicated that interest dardised these two variables using z-scores and subtracted the in quitting tends to be lower in the last 3 weeks of December ‘ ’ ‘ ’ standardised concern score from the standardised enjoyment and higher in the first 2 weeks of January.23 As none of the score. After inspecting a histogram of resulting scores to identify respondents in this study were followed up in December, we natural cut points, we constructed a categorical variable with repeated all adjusted analyses including an indicator variable to three levels: more enjoyment (<1 SD below the mean); capture the January seasonality effect. balanced between enjoyment and concern; and more concern All analyses were conducted in Stata V.12.1,24 adjusting for (>1 SD above the mean). the effects of sample weighting on parameter estimates and SEs. In addition, an unconditional approach was used to limit the Covariates sample as appropriate for each set of analyses, ensuring correct At baseline, respondents reported their sex, age and highest level estimation of the SEs. Cases that had missing data on outcome of education. Socioeconomic status was measured using the variables, the baseline versions of these variables and predictor Australian Bureau of Statistics’ Index of Socio-Economic variables (typically <5% combined) were deleted listwise from Disadvantage, using the 2011 census data of the postcode area in each model. which respondents resided.18 Respondents also reported the number of cigarettes smoked per day and the time to first cigarette, RESULTS which were combined into the Heaviness of Smoking Index.19 Table 1 presents sample characteristics, and descriptive statistics We sourced records from ACNielsen of monthly Target for the predictor variables and outcome variables. Audience Rating Points (TARPs) for adults aged 18 and above for all antismoking campaigns aired on television over the Predicting quitting-related cognitions survey period. TARPs represent potential exposure to advertis- Among continuing cigarette smokers at follow-up, 36.5% ing. Consistent with previous research which has shown that reported that they had thought about quitting at least daily in effects of antismoking advertising occur within 3-month the past week. Table 2 shows that, in the initial models, cigarette periods,20 21 we used a cumulative sum of the previous smokers who at baseline disliked the look of their pack were sig- 3 months’ TARPs, based on the date of follow-up. Cigarette nificantly more likely to report thinking about quitting at least costliness in the month of follow-up interview was calculated as daily at follow-up, compared with those who did not dislike the

Brennan E, et al. Tob Control 2015;24:ii33–ii41. doi:10.1136/tobaccocontrol-2014-052057 ii35 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 1 Sample characteristics, predictor variables at baseline, and outcome variables at follow-up for continuing and baseline cigarette smokers at follow-up Continuing cigarette smokers Baseline cigarette smokers at follow-up* at follow-up† (N=2948) (N=3125)

Sample characteristics at baseline and covariates %% Sex Males 54.7 55.1 Females 45.3 44.9 Age (years) 18–29 27.3 28.1 30–49 46.7 46.4 50–69 25.9 25.6 Highest level of education Less than high school 33.4 32.9 Completed high school/some tertiary 54.4 54.3 Tertiary or above 12.2 12.8 Socioeconomic status Low 41.4 41.0 Mid 41.6 42.0 High 17.0 17.0 Mean (SE) Mean (SE) Heaviness of smoking index (0–6) 2.23 (0.04) 2.17 (0.04) Days between baseline and follow-up survey 30.08 (0.10) 30.11 (0.10) Antismoking advertising in past 3 months 1447 (17.2) 1446 (16.8) Predictor variables at baseline %% Appeal variables Dislikes pack 79.9 79.5 Lower pack appeal 48.9 48.7 Lower quality 26.3 26.1 Lower satisfaction 20.6 20.5 Lower value for money 55.1 55.0 Believes brands do not differ in prestige 48.2 48.0 Health warning effectiveness variables Notices GHW first when looking at pack 65.9 66.1 Does not believe dangers of smoking are exaggerated 62.1 62.9 Attributes much more motivation to quit to GHWs 12.4 13.3 Concealed or covered pack in past month 23.0 23.0 Requested different GHW in past month 8.9 9.1 Perceived harm variables Believes brands do not differ in harmfulness 67.8 67.3 Balance between enjoyment and concern More enjoyment 16.4 21.0 Balance 58.5 55.2 More concern 23.0 21.7 Outcome variables at follow-up Quitting-related cognitions Daily thoughts about quitting in past week 36.5 – Intend to quit in next month 36.5 – Firm date to quit in next month 6.7 – Pack concealment and microindicators of concern Concealed or covered pack several or many times in past month 19.5 – Stubbed out several or many times in past month 26.7 – Stopped oneself from smoking several or many times in past month 37.8 – Quitting behaviours Attempted to quit in past month – 23.6 All data are weighted using longitudinal weights. Owing to rounding and missing data on outcome variables (including do not know, not applicable and refused responses), percentages may not sum to 100. Descriptive data for the covariate capturing percentage change in cigarette costliness are presented in text. *Sample used in analyses predicting daily thoughts about quitting, intentions to quit in next month, firm date to quit in next month, pack concealment, stubbing out and stopping oneself from smoking. †Sample used in analyses predicting attempts to quit in past month. GHW, graphic health warning.

ii36 Brennan E, et al. Tob Control 2015;24:ii33–ii41. doi:10.1136/tobaccocontrol-2014-052057 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper look of their pack. Similarly, smokers who at baseline reported not differ in prestige were significantly less likely to have less satisfaction from their cigarettes compared to a year ago attempted to quit, although this effect did not remain statistic- were more likely to report daily thoughts about quitting. ally significant in the multivariable model. In contrast, smokers Quitting thoughts were also significantly and positively pre- were more than twice as likely to have attempted if they attribu- dicted by noticing GHWs first, disagreeing that the dangers of ted motivation to quit to GHWs, and if they had requested a smoking have been exaggerated, attributing motivation to quit pack with a different GHW, and both of these effects remained to the GHWs and pack concealment. Quitting thoughts were significant in the multivariable model. Finally, quit attempts unrelated to believing brands do not differ in harm, but were were also predicted by the perceived harm variable. positively predicted by feeling more concern than enjoyment. Unexpectedly, smokers who believed that brands do not differ Entering all of these significant predictors into a multivariable in harmfulness were less likely to have attempted to quit at model changed the pattern of results only slightly: the predictive follow-up, although again this effect did not remain statistically effects of noticing GHWs first and feeling more concern than significant in the multivariable model (table 4). enjoyment were no longer statistically significant (table 2). Among continuing cigarette smokers at follow-up, 36.5% Sensitivity testing intended to quit within the next month. No appeal variables Appendix A in the online supplementary material presents predicted intentions, but in the initial models, smokers were sig- results for initial and multivariable models that were unadjusted nificantly more likely to intend to quit if at baseline they for covariates, and shows that there was very little difference in reported noticing GHWs first, disagreeing that the dangers of the overall pattern of findings. Online supplementary appendi- smoking have been exaggerated and attributing motivation to ces B and C show that the two sets of sensitivity testing also did quit to the GHWs. Intentions were also significantly and posi- not change the overall pattern of results. tively predicted by feeling more concern than enjoyment. In the multivariable model, the predictive effect of all variables except DISCUSSION noticing GHWs first remained statistically significant (table 2). We found consistent evidence in the multivariable models that Smokers who intended to quit were further asked if they had set indicators of GHW effectiveness were prospectively related to a firm date to quit, and 6.7% of all continuing smokers reported 1-month changes in several of the quitting-related outcomes. In that they had. Having a firm date to quit was significantly and posi- particular, smokers were more likely to have engaged in four of tively predicted in the initial models by GHW noticeability, dis- the seven quitting cognitions and behaviours if they did not agreeing that the dangers of smoking have been exaggerated and believe that the dangers of smoking are exaggerated (an indica- attributing motivation to quit to the GHWs. The pattern and sig- tor of GHW credibility), and they were significantly more likely nificance of findings remained in the multivariable model (table 2). to have engaged in five of the seven outcomes if they attributed motivation to quit to GHWs. Consistent with evidence of the Predicting pack concealment and microindicators of concern role that GHWs play in promoting smoking cessation12 13 25 26 Among continuing cigarette smokers at follow-up, 19.5% and experimental research demonstrating that GHWs are more – reported that they had concealed or covered up their cigarette effective when they appear on plain packs,27 31 these findings pack several or many times in the past month. None of the pre- suggest that the short-term increases in quitting-related inten- dictor variables were significantly associated with pack conceal- tions and behaviours observed in Australia following implemen- ment in the initial models, so no multivariable model was tation of PP with larger GHWs5 are most likely explained by conducted (table 3). smokers’ responses to the new and larger GHWs. Just over one-quarter (26.7%) of continuing smokers reported In the multivariable models, we also found that smokers who that they had stubbed out a cigarette at least several times in the experienced more concern than enjoyment from their smoking past month. In the initial models, stubbing out was significantly were significantly more likely to have an intention to quit, and and positively predicted by lower satisfaction compared to a to report stubbing out. In contrast, the belief that brands do not year ago; by GHW noticeability, attributing motivation to quit, differ in harmfulness was not significantly associated with any pack concealment and requesting a different GHW; and by of the outcomes, and only some of the appeal variables were feeling more concern than enjoyment. The multivariable model prospectively related to quitting-related cognitions and beha- showed a similar pattern of results (table 3). viours: smokers who at baseline disliked the look of their cigar- More than two-thirds (37.8%) of continuing cigarette smokers ette pack were more likely to report daily thoughts about reported that they had stopped themselves from having a cigarette quitting, and those who reported lower satisfaction were more at least several times in the past month. In the initial models, stop- likely to report daily thoughts about quitting, stubbing out and ping smoking was significantly and positively predicted by lower stopping oneself from smoking. satisfaction compared to a year ago, disagreeing that the dangers Providing further evidence that GHW effectiveness may be of smoking have been exaggerated, attributing motivation to quit the primary driver of the observed increases in quitting cogni- to GHWs and feeling more concern than enjoyment. In the multi- tions and behaviours among adult smokers, the initial models variable model, the predictive effects of attributing motivation to revealed an additional four significant effects of appeal, per- quit to GHWs and feeling more concern than enjoyment were no ceived harm and enjoyment/concern variables that did not longer statistically significant (table 3). remain statistically significant once entered into the multivari- able model with the GHW effectiveness variables. Besides this Predicting quit attempts small number of effects though, the pattern of results (including Overall, 23.6% of baseline cigarette smokers had made an the magnitude of effects) was largely the same across the initial attempt to quit in the month between their baseline and and multivariable models—particularly for the GHW effective- follow-up survey (including those who had relapsed to smoking ness variables—indicating that the various predictors explain and those who were still quit at follow-up). In the initial unique variance in the outcomes. In particular, results from the models, quit attempts were significantly predicted by one appeal multivariable models predicting quitting thoughts and the two variable: unexpectedly, smokers who believed that brands do microindicators of concern suggest that the effect of lower

Brennan E, et al. Tob Control 2015;24:ii33–ii41. doi:10.1136/tobaccocontrol-2014-052057 ii37 ii38 eerhpaper Research

Table 2 Associations between appeal, health warning effectiveness, perceived harm and enjoyment/concern variables measured at baseline, and quitting-related cognitions measured at 1-month Downloaded from follow-up among continuing cigarette smokers Daily thoughts about quitting in past week Intend to quit in next month Firm date to quit in next month

Initial models Multivariable model Initial models Multivariable model Initial models Multivariable model (N=2565 to 2915) (N=2571) (N=2584 to 2995) (N=2831) (N=2584 to 2995) (N=2883) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) http://tobaccocontrol.bmj.com/

Appeal variables Dislikes pack 1.59** (1.17 to 2.16) 1.38* (1.00 to 1.89) 1.16 (0.85 to 1.57) – 1.01 (0.62 to 1.63) – Lower pack appeal 1.01 (0.81 to 1.26) – 1.00 (0.80 to 1.25) – 1.19 (0.82 to 1.73) – Lower quality 1.26 (1.00 to 1.59) – 1.00 (0.78 to 1.27) – 0.96 (0.65 to 1.41) – Lower satisfaction 1.46** (1.13 to 1.88) 1.49** (1.14 to 1.95) 1.19 (0.92 to 1.53) – 1.33 (0.90 to 1.97) – Lower value for money 1.12 (0.91 to 1.38) – 1.01 (0.81 to 1.24) – 1.06 (0.76 to 1.48) – Believes brands do not differ in prestige 0.90 (0.72 to 1.11) – 1.08 (0.87 to 1.33) – 0.93 (0.65 to 1.32) – Health warning effectiveness variables rna E, Brennan Notices GHW first when looking at pack 1.37** (1.10 to 1.72) 1.07 (0.84 to 1.35) 1.32* (1.06 to 1.64) 1.20 (0.95 to 1.50) 1.82** (1.23 to 2.71) 1.60* (1.06 to 2.40) onMarch18,2015-Publishedby Does not believe dangers of smoking are exaggerated 1.72*** (1.38 to 2.14) 1.69*** (1.33 to 2.15) 1.48*** (1.19 to 1.84) 1.39** (1.11 to 1.75) 1.64* (1.12 to 2.40) 1.50* (1.02 to 2.23) Attributes much more motivation to quit to GHWs 2.72*** (1.90 to 3.90) 2.52*** (1.71 to 3.73) 1.71*** (1.24 to 2.36) 1.48* (1.06 to 2.07) 2.00** (1.29 to 3.11) 1.81** (1.17 to 2.78) tal et Concealed or covered pack in past month 1.61*** (1.28 to 2.03) 1.46** (1.14 to 1.88) 1.03 (0.82 to 1.30) – 0.91 (0.61 to 1.36) –

. Requested different GHW in past month 1.00 (0.68 to 1.48) – 1.26 (0.86 to 1.85) – 0.95 (0.53 to 1.68) – o Control Tob Perceived harm variables Believes brands do not differ in harmfulness 0.83 (0.66 to 1.04) – 0.86 (0.68 to 1.08) – 0.89 (0.61 to 1.29) – Balance between enjoyment and concern 2015; More enjoyment 1.00 1.00 1.00 1.00 1.00 –

24:ii33 Balance 1.19 (0.88 to 1.60) 1.02 (0.74 to 1.39) 1.34 (0.98 to 1.82) 1.21 (0.89 to 1.65) 0.93 (0.57 to 1.52) More concern 1.59** (1.13 to 2.24) 1.25 (0.87 to 1.81) 2.29*** (1.61 to 3.25) 1.96*** (1.37 to 2.82) 1.30 (0.76 to 2.23)

– Bolded results are statistically significant at p<0.05. The weighted N per initial model varies due to missing data on outcome variables and predictor variables. The weighted N for each multivariable model includes only those cases with valid data on the i1 doi:10.1136/tobaccocontrol-2014-052057 ii41. outcome variable and all of the predictor variables included in the model. All models adjust for the outcome variable measured at baseline and for: date of the follow-up survey; number of days between baseline and follow-up survey; cumulative Target group.bmj.com Audience Rating Points (antismoking television advertising) in the 3 months prior to the follow-up survey; change in cigarette costliness (based on month of the follow-up survey); sex; age; education; socioeconomic status and Heaviness of Smoking Index (measured at baseline). – Predictor variable not included in multivariable model due to non-significant (at p<0.05) association with outcome variable in initial model. ***p≤0.001, **p<0.01, *p<0.05. GHW, graphic health warning. Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 3 Associations between appeal, health warning effectiveness, perceived harm and enjoyment/concern variables measured at baseline, and pack concealment and microindicators of concern measured at 1-month follow-up among continuing cigarette smokers Concealed or covered Stubbed out several or many times in past Stopped oneself from smoking several or many pack several or many month times in past month times in past month Initial models (N=2568 to Initial models (N=2566 Multivariable model Initial models (N=2567 Multivariable model 2899) to 2919) (N=2559) to 2913) (N=2770) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)

Appeal variables Dislikes pack 1.05 (0.69 to 1.61) 1.40 (1.00 to 1.95) – 1.13 (0.85 to 1.50) – Lower pack appeal 1.25 (0.94 to 1.66) 1.11 (0.88 to 1.40) – 1.17 (0.95 to 1.44) – Lower quality 1.07 (0.80 to 1.41) 1.12 (0.88 to 1.43) – 1.07 (0.86 to 1.33) – Lower satisfaction 1.07 (0.78 to 1.45) 1.36* (1.05 to 1.77) 1.41* (1.07 to 1.84) 1.36* (1.07 to 1.72) 1.38** (1.09 to 1.76) Lower value for money 0.94 (0.71 to 1.23) 0.94 (0.76 to 1.18) – 1.10 (0.90 to 1.35) – Believes brands do not 0.93 (0.71 to 1.21) 1.08 (0.86 to 1.35) – 0.99 (0.82 to 1.21) – differ in prestige Health warning effectiveness variables Notices GHW first when 0.95 (0.72 to 1.26) 1.51*** (1.19 to 1.91) 1.46** (1.13 to 1.88) 1.09 (0.89 to 1.33) – looking at pack Does not believe dangers 0.91 (0.70 to 1.18) 1.14 (0.91 to 1.42) – 1.31** (1.07 to 1.60) 1.35** (1.09 to 1.67) of smoking are exaggerated Attributes much more 1.26 (0.84 to 1.89) 1.93*** (1.38 to 2.71) 1.75** (1.23 to 2.48) 1.39* (1.02 to 1.90) 1.32 (0.95 to 1.83) motivation to quit to GHWs Concealed or covered pack – 1.37** (1.08 to 1.74) 1.32* (1.03 to 1.69) 1.21 (0.97 to 1.50) – in past month Requested different GHW 1.46 (0.93 to 2.28) 1.57* (1.07 to 2.30) 1.49 (0.99 to 2.22) 1.27 (0.89 to 1.82) – in past month Perceived harm variables Believes brands do not 0.96 (0.72 to 1.27) 0.86 (0.68 to 1.10) – 0.98 (0.79 to 1.22) – differ in harmfulness Balance between enjoyment and concern More enjoyment 1.00 1.00 1.00 1.00 1.00 Balance 1.01 (0.70 to 1.46) 1.18 (0.86 to 1.62) 1.15 (0.82 to 1.62) 1.18 (0.90 to 1.56) 1.14 (0.85 to 1.52) More concern 0.91 (0.60 to 1.38) 1.79*** (1.27 to 2.52) 1.62* (1.12 to 2.34) 1.43* (1.04 to 1.96) 1.31 (0.94 to 1.82) Bolded results are statistically significant at p<0.05. The weighted N per initial model varies due to missing data on outcome variables and predictor variables. The weighted N for each multivariable model includes only those cases with valid data on the outcome variable and all of the predictor variables included in the model. All models adjust for the outcome variable measured at baseline and for: date of the follow-up survey; number of days between baseline and follow-up survey; cumulative Target Audience Rating Points (antismoking television advertising) in the 3 months prior to the follow-up survey; change in cigarette costliness (based on month of the follow-up survey); sex; age; education; socioeconomic status and Heaviness of Smoking Index (measured at baseline). – Predictor variable not included in multivariable model due to non-significant (at p<0.05) association with outcome variable in initial model. ***p≤0.001, **p<0.01, *p<0.05. GHW, graphic health warning. perceived satisfaction after the packaging changes contributed the possibility that the predictive effects observed in this study to these quitting-related cognitions and behaviours independ- may be conditional on demographic and smoker characteristics. ently of responses that were more closely tied to the new It is possible that the variables measured in this study are GHWs. related to each other in more complex ways.32 For example, Although caution must be exercised when interpreting find- one recent study identified a series of mediational pathways ings that were not part of a systematic pattern of effects and through which health warnings lead to quit attempts,12 while were not significant in the multivariable models, we do note another used structural equation models to explore associations that two of the findings in the initial models were in an unex- between aspects of cigarette and cigarette packaging appeal, and pected direction and are difficult to explain (negative associa- cigarette consumption and quit attempts.33 In addition, longitu- tions between believing that brands do not differ in prestige/ dinal studies have established that quitting-related cognitions – harmfulness and quit attempts). It may be that these appeal and and microindicators of concern predict attempts to quit.17 34 37 perceived harm variables operate in different ways for different Studies that more thoroughly investigate the pathways through smokers, influenced by factors such as the brand and variant which PP and GHW policies lead to changes in smoking beha- they smoke (eg, cigarettes from premium, mainstream or value viours are critical. brand segments) and their history with that brand. It is also pos- A particular strength of this study is the use of a cohort sible that the impact of these beliefs may be moderated by design, with a good retention rate and a nationally representa- responses to the GHWs. Furthermore, it is possible that the pre- tive sample. In these analyses, respondents acted as their own dictive effects observed in this study are different for different control, allowing us to minimise the influence of unobserved demographic subgroups, particularly given that Wakefield et al4 individual differences and any time-invariant response biases. found greater change on the appeal variables for younger than While these analyses provide some confidence that the proposed older adult smokers. Further research is required to investigate causal order between variables is the correct one, further

Brennan E, et al. Tob Control 2015;24:ii33–ii41. doi:10.1136/tobaccocontrol-2014-052057 ii39 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 4 Associations between appeal, health warning effectiveness, perceived harm and enjoyment/concern variables measured at baseline, and quit attempts measured at 1-month follow-up among baseline cigarette smokers Initial models Multivariable model (N=2726 to 3116) (N=2964) OR (95% CI) OR (95% CI)

Appeal variables Dislikes pack 0.89 (0.66 to 1.20) – Lower pack appeal 1.04 (0.83 to 1.30) – Lower quality 1.22 (0.96 to 1.56) – Lower satisfaction 1.12 (0.87 to 1.45) – Lower value for money 1.03 (0.83 to 1.29) – Believes brands do not differ in prestige 0.79* (0.64 to 0.98) 0.88 (0.70 to 1.10) Health warning effectiveness variables Notices GHW first when looking at pack 1.05 (0.83 to 1.33) – Does not believe dangers of smoking are exaggerated 0.98 (0.78 to 1.23) – Attributes much more motivation to quit to GHWs 2.31*** (1.73 to 3.09) 2.15*** (1.59 to 2.91) Concealed pack in past month 1.20 (0.94 to 1.52) – Requested different GHW in past month 2.04*** (1.43 to 2.89) 1.88*** (1.30 to 2.71) Perceived harm variables Believes brands do not differ in harmfulness 0.78* (0.62 to 0.99) 0.86 (0.68 to 1.09) Balance between enjoyment and concern More enjoyment 1.00 – Balance 0.98 (0.72 to 1.34) More concern 1.40 (0.99 to 1.99) Bolded results are statistically significant at p<0.05. The weighted N per initial model varies due to missing data on the outcome variable and predictor variables. The weighted N for the multivariable model includes only those cases with valid data on the outcome variable and all of the predictor variables included in the model. All models adjust for recency of the last quit attempt made at baseline and for: date of the follow-up survey; number of days between baseline and follow-up survey; cumulative Target Audience Rating Points (antismoking television advertising) in the 3 months prior to the follow-up survey; change in cigarette costliness (based on month of the follow-up survey); sex; age; education; socioeconomic status and Heaviness of Smoking Index (measured at baseline). –Predictor variable not included in multivariable model due to non-significant (at p<0.05) association with outcome variable in initial model. ***p≤0.001, **p<0.01, *p<0.05. GHW, graphic health warning. mediation analyses and controlled experimental studies are interactive relationships between the appeal, GHW effectiveness required to establish if the appeal, GHW effectiveness, perceived and perceived harm variables, and by testing whether these harm and enjoyment/concern variables are causally responsible effects are conditional on smoker characteristics. for the observed changes in quitting-related outcomes.5 Nonetheless, the policy relevance of these findings is reinforced by the fact that we controlled for other factors known to predict quitting (antismoking television advertising and changes in cig- What this paper adds arette costliness) and also found that the effects were robust to adjustments for seasonality effects and any anticipatory effects ▸ This is the first study to examine whether the proximal related to the December 2013 tax increase. The short duration measures of plain packaging with larger and refreshed of the 1-month follow-up period meant that we were unable to graphic health warning effectiveness—measures of appeal, fairly assess rates of sustained quitting and engagement in other graphic health warning effectiveness and perceived harm— behaviours that smokers may use to circumvent heightened feel- predict subsequent changes in quitting-related cognitions ings of risk, including switching products or reducing daily cig- and behaviours. arette consumption. These low incidence outcomes would be ▸ Using a prospective cohort study with a nationally better assessed in longer term cohort studies. representative sample of Australian adult cigarette smokers fi Overall, the current ndings strengthen the evidence base for surveyed within the first year of plain packaging, we found PP with larger GHWs by demonstrating that quitting-related increased graphic health warning effectiveness to be cognitions and behaviours are prospectively predicted by the particularly influential in driving adult quitting cognitions more proximal beliefs and perceptions widely used as outcome and behaviour, some influence for select appeal variables measures in the experimental and naturalistic studies that helped and no contribution from accurate harm perceptions. 10 11 38 fi make the case for PP. Our ndings suggest that, among ▸ These findings provide initial insight into the mechanisms adults, increased GHW effectiveness is likely to be particularly through which plain packaging with larger and refreshed fl in uential in driving quitting behaviour. As such, these results graphic health warnings is likely to bring about changes in fi contrast with study ndings of adolescents, where implementa- smoking behaviours. tion of the new packaging was found to lead adolescents to have fewer favourable and more unfavourable perceptions of cigarette packs,39 with little evidence that the new GHWs had an effect.40 However, our understanding of the role that the appeal Acknowledgements The authors thank Professor Ron Borland for sharing fl questions from the ITC survey for inclusion in the telephone tracking survey, the of cigarettes and cigarette packaging plays in in uencing adult Social Research Centre for data collection and Megan Bayly and Emily Bain for data smoking behaviour may be improved by further examining checking. ii40 Brennan E, et al. Tob Control 2015;24:ii33–ii41. doi:10.1136/tobaccocontrol-2014-052057 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Contributors MW designed the study. KC and MZ managed the data collection 15 Partos TR, Borland R, Yong H-H, et al. The quitting rollercoaster: how recent and cleaned the data files. EB and SD analysed the data and interpreted the results. quitting history affects future cessation outcomes (data from the International EB drafted the manuscript with contributions from all authors. All authors approved Tobacco Control 4-Country Cohort Study). Nicotine Tob Res 2013;15:1578–87. the final manuscript. 16 Stillman S. Smoking cessation. Melbourne: Cancer Council Victoria, 2011. http:// www.tobaccoinaustralia.org.au/chapter-7-cessation/7-13-methods-services-and- Funding The National Plain Packaging Tracking Survey was funded under a products-for-quitting-co (accessed 18 Sep 2014). contract with the Australian Government Department of Health and Ageing. 17 Vangeli E, Stapleton J, Smit ES, et al. Predictors of attempts to stop smoking and Competing interests The authors wish to advise that MW was a member and their success in adult general population samples: a systematic review. Addiction MS a technical writer for the Tobacco Working Group of the Australian National 2011;106:2110–21. Preventive Health Task Force and MW was a member of the Expert Advisory 18 Australian Bureau of Statistics. Technical paper: socio-economic indexes for areas (SEIFA) Committee on Plain Packaging that advised the Australian Department of Health on 2011. Canberra, Australia: Australian Bureau of Statistics, 2013. http://www.ausstats.abs. research pertaining to the plain packaging legislation. MW, SD and EB hold gov.au/ausstats/subscriber.nsf/0/22CEDA8038AF7A0DCA257B3B00116E34/$File/2033. competitive grant funding from the Australian National Health and Medical Research 0.55.001%20seifa%202011%20technical%20paper.pdf (accessed 18 Sept 2014). Council and MW holds competitive grant funding from the US National Institutes of 19 Heatherton TF, Kozlowski LT, Frecker RC, et al. Measuring the heaviness of Health, Australian National Preventive Health Agency and BUPA Health Foundation. smoking: using self-reported time to the first cigarette of the day and number of – Ethics approval The survey was approved by the Cancer Council Victoria Human cigarettes smoked per day. Br J Addict 1989;84:791 9. fi Ethics Committee (HREC 0018). 20 Wake eld MA, Spittal MJ, Yong HH, et al. Effects of mass media campaign exposure intensity and durability on quit attempts in a population-based cohort Provenance and peer review Not commissioned; externally peer reviewed. study. Health Educ Res 2011;26:988–97. Open Access This is an Open Access article distributed in accordance with the 21 Dunlop S, Cotter T, Perez D, et al. Televised antismoking advertising: effects of level Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which and duration of exposure. Am J Public Health 2013;103:e66–73. permits others to distribute, remix, adapt, build upon this work non-commercially, 22 Australian Bureau of Statistics. Average Weekly Earnings, Australia. Canberra, and license their derivative works on different terms, provided the original work is Australia: Australian Bureau of Statistics, 2013. 6302.0. http://www.abs.gov.au/ properly cited and the use is non-commercial. See: http://creativecommons.org/ ausstats/[email protected]/mf/6302.0 (accessed 18 Sep 2014). licenses/by-nc/4.0/ 23 Durkin S, Wakefield MA, Spittal MJ. Which types of televised anti-tobacco campaigns prompt more quitline calls from disadvantaged groups? Health Educ Res 2011;26:998–1009. 24 StataCorp. Stata Statistical Software: Release 12. College Station, TX: StataCorp LP; REFERENCES 2011. 1 Tobacco Plain Packaging Act. No. 148, 2011 as amended. 2011. http://www. 25 Hammond D. Health warning messages on tobacco products: a review. Tob Control comlaw.gov.au/Details/C2013C00190 2011;20:327–37. 2 Scollo M, Lindorff K, Coomber K, et al. Standardised packaging and new enlarged 26 Institute for Global Tobacco Control. State of evidence review: health warning labels graphic health warnings for tobacco products in Australia––legislative requirements on tobacco products. Baltimore, MD: Johns Hopkins Bloomberg School of Public and implementation of the Tobacco Plain Packaging Act 2011 and the Competition Health, 2013. http://globaltobaccocontrol.org/sites/default/files/HealthWarnings_ and Consumer (Tobacco) Information Standard, 2011. Tob Control 2015;24:ii9–16. state_of_evidence_final_11_18_2013_web_0.pdf (accessed 18 Sep 2014). 3 Competition and Consumer (Tobacco) Information Standard. F2013C00598. 2011. 27 Munafo MR, Roberts N, Bauld L, et al. Plain packaging increases visual attention to http://www.comlaw.gov.au/Details/F2013C00598 health warnings on cigarette packs in non-smokers and weekly smokers but not 4 Wakefield M, Coomber K, Zacher M, et al. Australian adult smokers’ responses to daily smokers. Addiction 2011;106:1505–10. plain packaging with larger graphic health warnings 1 year after implementation: 28 Hoek J, Wong C, Gendall P, et al. Effects of dissuasive packaging on young adult results from a national cross-sectional tracking survey. Tob Control 2015;24:ii17–25. smokers. Tob Control 2011;20:183–8. 5 Durkin S, Brennan E, Coomber K, et al. Short-term changes in quitting-related 29 Wakefield M, Germain D, Durkin S, et al. Do larger pictorial health warnings cognitions and behaviours after the implementation of plain packaging with larger diminish the need for plain packaging of cigarettes? Addiction 2012;107:1159–67. health warnings: findings from a national cohort study with Australian adult 30 Al-Hamdani M. The effect of cigarette plain packaging on individuals’ health smokers. Tob Control 2015;24:ii26–32. warning recall. Healthcare Policy 2013;8:68–77. 6 Slade J. The pack as advertisement. Tob Control 1997;6:169–70. 31 Moodie CS, Mackintosh AM. Young adult women smokers’ response to using plain 7 Wakefield M, Morley C, Horan JK, et al. The cigarette pack as image: new evidence cigarette packaging: a naturalistic approach. BMJ Open 2013;3:e002402. from tobacco industry documents. Tob Control 2002;11:i73–80. 32 International Agency for Research on Cancer. Methods for evaluating tobacco 8 National Cancer Institute. The role of the media in promoting and reducing tobacco control policies. Lyon, France: International Agency for Research on Cancer, 2008. use. Bethesda, MD: United States Department of Health and Human Services, 33 Ayo-Yusuf OA, Agaku IT. The association between smokers’ perceived importance of National Institutes of Health, National Cancer Institute. NIH Pub. No 07–6242, the appearance of cigarettes/cigarette packs and “Smoking Sensory Experience”: 2008. http://www.cancercontrol.cancer.gov/tcrb/monographs/19/index.html a structural equation model. Nicotine Tob Res 2015;17:91–7. (accessed 18 Sep 2014). 34 Borland R, Yong HH, Wilson N, et al. How reactions to cigarette packet health 9 U.S. Department of Health and Human Services. Preventing tobacco use among warnings influence quitting: findings from the ITC Four-Country survey. Addiction youth and young adults: a report of the Surgeon General. Atlanta, GA: US 2009;104:669–75. Department of Health and Human Services, Centers for Disease Control and 35 Borland R, Yong HH, Balmford J, et al. Motivational factors predict quit attempts Prevention, National Center for Chronic Disease Prevention and Health Promotion, but not maintenance of smoking cessation: findings from the International Tobacco Office on Smoking and Health, 2012. http://www.surgeongeneral.gov/library/reports/ Control Four country project. Nicotine Tob Res 2010;12:S4–S11. preventing-youth-tobacco-use/index.html (accessed 18 Sep 2014). 36 Hyland A, Borland R, Li Q, et al. Individual-level predictors of cessation behaviours 10 Hammond D. Standardized packaging of tobacco products: evidence review. Prepared among participants in the International Tobacco Control (ITC) Four Country Survey. on behalf of the Irish Department of Health: March 2014. http://www.drugsandalcohol. Tob Control 2006;15:iii83–94. ie/22106/1/Standardized-Packaging-of-Tobacco-Products-Evidence-Review.pdf 37 Li L, Borland R, Fong GT, et al. Smoking-related thoughts and microbehaviours, and (accessed 18 Sep 2014). their predictive power for quitting: findings from the International Tobacco Control 11 Chantler C. Standardised packaging of tobacco. Report of the independent review (ITC) China Survey. Tob Control Published Online First: 25 Feb 2014. doi:10.1136/ undertaken by Sir Cyril Chantler. April 2014. http://www.kcl.ac.uk/health/ tobaccocontrol-2013-051384 packaging-review.aspx (accessed 18 Sep 2014). 38 Stead M, Moodie C, Angus K, et al. Is consumer response to plain/standardised 12 Yong HH, Borland R, Thrasher JF, et al. Mediational pathways of the impact of tobacco packaging consistent with Framework Convention on Tobacco Control cigarette warning labels on quit attempts. Health Psychol 2014;33:1410–20. guidelines? A systematic review of quantitative studies. PLoS ONE 2013;8:e75919. 13 Huang J, Chaloupka FJ, Fong GT. Cigarette graphic warning labels and smoking 39 White V, Williams T, Wakefield M. Has the introduction of plain packaging with prevalence in Canada: a critical examination and reformulation of the FDA larger graphic health warnings changed adolescents’ perceptions of cigarette packs regulatory impact analysis. Tob Control 2014;23:7–12. and brands? Tob Control 2015;24:ii42–9. 14 Coomber K, Zacher M, Durkin S, et al. Australian National Tobacco Plain Packaging 40 White V, Williams T, Faulkner A, et al. Do larger graphic health warnings on Tracking Survey: Technical Report. Melbourne: Cancer Council Victoria and Social standardised cigarette packs increase adolescents’ cognitive processing of consumer Research Centre, December 2014. http://tobaccocontrol.bmj.com/lookup/suppl/ health information and beliefs about smoking-related harms? Tob Control 2015;24: doi:10.1136/tobaccocontrol-2014-052050/-/DC1 ii50–7.

Brennan E, et al. Tob Control 2015;24:ii33–ii41. doi:10.1136/tobaccocontrol-2014-052057 ii41 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Are quitting-related cognitions and behaviours predicted by proximal responses to plain packaging with larger health warnings? Findings from a national cohort study with Australian adult smokers Emily Brennan, Sarah Durkin, Kerri Coomber, Meghan Zacher, Michelle Scollo and Melanie Wakefield

Tob Control 2015 24: ii33-ii41 doi: 10.1136/tobaccocontrol-2014-052057

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii33

These include: Supplementary Supplementary material can be found at: Material http://tobaccocontrol.bmj.com/content/suppl/2015/02/25/tobaccocon trol-2014-052057.DC1.html References This article cites 20 articles, 11 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii33#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper Has the introduction of plain packaging with larger graphic health warnings changed adolescents’ perceptions of cigarette packs and brands? Victoria White, Tahlia Williams, Melanie Wakefield

Centre for Behavioural ABSTRACT controls on packaging and labelling to better Research in Cancer, Cancer Objective To examine the impact of plain packaging of inform consumers about health risks (Article 11) as Council Victoria, Melbourne, Victoria, Australia cigarettes with enhanced graphic health warnings on well as comprehensive bans on the promotion of adolescents’ perceptions of pack image and perceived tobacco products (Article 13). Australia, a FCTC Correspondence to brand differences. signatory, became the first country to implement Associate Professor Victoria Methods Cross-sectional school-based surveys these elements of the guidelines, with all tobacco White, Centre for Behavioural conducted in 2011 (prior to introduction of new cigarette products sold in Australia from 1 December 2012 Research in Cancer, Cancer – ‘ ’ Council Victoria, 615 St Kilda packaging) and in 2013 (7 12 months afterwards). required to use the same drab dark brown col- Road, Melbourne, VIC 3004, Students aged 12–17 years (2011 n=6338; 2013 oured packaging, with brand and variant names in Australia; n=5915) indicated whether they had seen a cigarette a standard font and position on the pack. At the [email protected] pack in previous 6 months. Students rated the character same time, through separate legislation, the size of Received 6 October 2014 of four popular cigarette brands, indicated level of the graphic health warnings (GHWs) on the front Accepted 13 January 2015 agreement regarding differences between brands in ease of cigarette packs and cartons increased from 30% of smoking, quitting, addictiveness, harmfulness and look to 75%, while remaining at 90% of the back (for of pack; and indicated positive and negative perceptions legislated requirements and images of tobacco pack- of pack image. Changes in responses of students seeing aging in Australia preplain and postplain packaging cigarette packs in the previous 6 months (2011: 60%; see Scollo et al6 (this volume) or http://www. 2013: 65%) were examined. cancervic.org.au/plainfacts/default.asp). – Results Positive character ratings for each brand Findings from experimental7 12 and qualitative reduced significantly between 2011 and 2013. Changes studies413suggest that removing pack design ele- were found for four of five statements reflecting brand ments reduces the perceived attractiveness of cigar- differences. Significantly fewer students in 2013 than ette packs and perceived quality of cigarettes 2011 agreed that ‘some brands have better looking packs while increasing negative perceptions of smokers than others’ (2011: 43%; 2013: 25%, p<0.001), with using those cigarette packs, among young people. larger decreases found among smokers (interaction Reviews have demonstrated relative consistency in – p<0.001). Packs were rated less positively and more these findings.14 16 Most of these studies have used negatively in 2013 than in 2011 (p<0.001). The decrease mocked-up versions of plain packages designed by in positive image ratings was greater among smokers. the researchers. No study has examined adolescents’ Conclusions The introduction of standardised responses to cigarette packaging after plain pack- packaging has reduced the appeal of cigarette packs. aging of tobacco products has become the norm in a Further research could determine if continued exposure to country. It is possible that responses found in experi- standardised packs creates more uncertainty or mental studies were influenced by the novelty of the disagreement regarding brand differences in ease of plain packs. Additionally, given the controlled envir- smoking and quitting, perceived addictiveness and harms. onment of experimental studies, where individuals provide an immediate response to an image shown for a standard length of time,7 9 12 17 estimates of INTRODUCTION pack design effects on adolescents’ responses from Easily purchased and frequently displayed in these studies may be optimal. public, cigarette packs are powerful communication The specific aims of the Tobacco Plain Packaging – tools.1 3 Their combined portability and visibility Act are to: reduce the appeal of tobacco products; allows them to function as ‘badge’ products with increase effectiveness of health warnings and the smoker taking on characteristics associated with reduce the ability of packaging to mislead consu- Open Access 18 Scan to access more their preferred brand.3 The tobacco industry has mers about the harms of smoking. Through the free content investigated all aspects of pack design to ensure the achievement of these aims and as part of a compre- messages it wants communicated are understood by hensive suite of tobacco control strategies, it was consumers.2 These messages have broad reach, intended that the legislation would contribute to including to children and adolescents who have efforts to reduce smoking uptake in the long been found to describe some brands as ‘exclusive’, term.18 Achievement of change in indictors of the ‘elegant’ and ‘sophisticated’ while others are seen legislation’s specific aims in the short term would as ‘common’ or ‘cheap’.4 suggest that the legislation is working as intended. To cite: White V, As part of a comprehensive approach to reducing Studies assessing the contribution of the legislation Williams T, Wakefield M. death and disease caused by smoking, the WHOs’ on reduced smoking uptake would need to be con- Tob Control 2015;24: Framework Convention on Tobacco Control ducted subsequently, using several years of postim- – ii42 ii49. (FCTC)5 requires signatory countries to implement plementation data and controlling for the effect of ii42 White V, et al. Tob Control 2015;24:ii42–ii49. doi:10.1136/tobaccocontrol-2014-052084 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper other tobacco policy (eg, tax increases, smoking bans) and November 2013. The procedure for selecting students to survey tobacco industry efforts to promote smoking. was the same as in 2011. In this study, we use cross-sectional data from surveys of sec- ondary school students conducted before (2011) and after Parental consent procedures (2013) the full implementation of the new packaging of tobacco In both states in 2011 and in one state in 2013, parents were products in Australia in December 2012, to examine whether informed about the study and asked to let the school know if adolescents’ perceptions of cigarette packs changed. While a they did not want their child to participate. Owing to require- companion paper in this volume investigates change over time ments stipulated by the education authorities governing govern- in health warning effectiveness among adolescents (the second ment and Catholic schools in the second state in 2013, an active specific aim of plain packaging),19 the current study assesses the parental consent procedure was used. In this procedure, parents extent to which the first and third specific aims of the Tobacco were informed about the study and provided written consent to Plain Packaging Act have been achieved in the short term among the school for the student’s participation. While active parental adolescents who had seen a cigarette pack in the previous six consent procedures reduce student participation numbers and months. Specifically, the study aims to determine whether there increase the statistical intraclass correlation, substance use esti- had been (1) a reduction in the perceived appeal of four of the mates are similar to those found with passive parental consent.21 five most popular cigarette brands; (2) a decrease in positive image, and an increase in negative image, ratings of cigarette Procedure for all surveys packs and (3) decreases in the perceived differences between On an agreed day, members of the research team attended the brands in their harmfulness, addictiveness, ease of smoking and school to administer the pencil-and-paper questionnaire to quitting and attractiveness of their packs, 7–12 months after the classes of students during school time. After working through a introduction of the new tobacco packaging. practice survey with research staff, students were given a description of the main survey and asked for study consent. METHODS Students were told they could withdraw from the survey at any Study overview and design time. Consenting students completed the survey independently The current study used data from school-based surveys of ado- and anonymously. lescents conducted in two Australian states in 2011 and 2013. Both surveys had institutional ethics approval and approvals Measures from the appropriate education authorities. Recency of seeing cigarette packs Students indicated how long ago they had seen a cigarette pack School samples by choosing a response from: in the past 6 months, more than 2011 Survey 6 months ago but less than 12 month ago, more than 12 months Data were obtained from each state’s component of a national tri- ago. ennial survey that aimed to determine the prevalence of sub- stance use among a representative sample of Australian students Cigarette brand character ratings (appeal outcome) in year levels 7–12 (age 12–17 years). The procedure for the tri- In both survey years, students were presented with a photo- ennial survey is explained more fully elsewhere.20 In brief, within graphic image of each of four brands of Australian cigarettes and each state, schools were randomly selected from the three main asked to indicate their level of agreement with three statements Australian education sectors (government, Catholic and inde- about the brand and the pack (‘this brand appeals to me’, ‘the pendent) to ensure a proportional representation of students pack looks good’, ‘the pack looks ugly’) and three statements across sectors. Principals of selected schools were approached about people who smoke the brand (‘are cool’, ‘are successful’, and permission was sought to conduct the survey. If a school ‘are daggy (uncool)’). These outcomes were adapted from studies refused, a replacement school (the geographically closest school assessing branded and plain packaging appeal among adolescents within the same education sector) selected at the same time as the or young adults.812Responses were made on a five-point scale original school was approached. In 2011, we aimed to survey stu- ranging from ‘strongly agree’ (1) to ‘strongly disagree’ (5) with dents from 117 schools across two states. A total of 324 schools ‘not sure’ in the middle (3). The brands assessed were the three were approached regarding study participation and 97 agreed most commonly smoked by Australian adolescents (Winfield (30% response rate). Schools were requested to allow one class (used by 44% of adolescent smokers); Peter Jackson (25%); of students from each of year levels 7–10 (average age range 12– Longbeach (10%) and a premium brand (Benson & Hedges) that 15) or two classes of students from each of years 11–12 (average was smoked by 8% of adolescent smokers making it the fifth age range 16 and 17). Researchers worked with school staff to most commonly smoked brand in 2011).22 All images included a ensure that selected classes provided a cross section of students GHW as mandated at that time, with the same health warning (ie, special education classes were excluded). Data collection used for each pack image within a survey year (eg, ‘Smoking occurred between June and December 2011. causes mouth and throat cancer’ in 2011; ‘Smoking causes mouth cancer’ in 2013). For each brand, responses for the six 2013 Survey items were summed with items recoded where necessary such We aimed to survey students from the 97 secondary schools par- that higher scores indicated a positive view (range 6–30). ticipating in the 2011 survey. Principals of schools were (Cronbach’s α for each brand in each year was adequate: 2011 approached by letter and approval for participation was sought. range: 0.77–0.78; 2013 range: 0.73–0.75). Fifty-eight schools agreed to participate (60%). When a school refused, a replacement school from the list of replacement Attraction of cigarette packs (appeal outcome) schools selected for the 2011 survey was approached. Sixty- Based on our previous research,23 24 students who had seen a three replacement schools were approached and 24 agreed cigarette pack in the previous 6 months indicated their level of (response 38%), giving a total of 82 schools participating in the agreement to four positive (‘cool’, ‘good’, ‘interesting’, ‘excit- 2013 survey. Data collection occurred between June and ing’), and four negative (‘ugly’, ‘daggy (uncool)’, ‘gross’,

White V, et al. Tob Control 2015;24:ii42–ii49. doi:10.1136/tobaccocontrol-2014-052084 ii43 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

‘disgusting’) descriptions of cigarette packs using a five-point assessing brand differences. Smoking status, age, sex, school scale ((1) ‘strongly disagree’ to (5) ‘strongly agree’). Students education sector and state were included as covariates in ana- could also respond that they ‘cannot comment’ with these lyses examining effect of year. When the effect of year was sig- responses coded as missing. Positive and negative subscale scores nificant, its interaction with smoking status was fitted to were created by taking the average of the five-point ratings for determine if the effect was consistent across smoking status the items on each scale. Both scales have good internal reliabil- groups. For significant interactions, means adjusted for covari- ity23 with internal reliability for the current study high (positive ates are reported in the text. Students with missing data on vari- pack image scale: α=0.85; negative pack image scale: α=0.78). ables were excluded from relevant analyses.

Brand differences (harm and appeal outcomes) Sensitivity testing Based on Hammond et al,10 we assessed the extent to which The analyses described above were repeated using data only plain packaging may be associated with a reduction in perceived from the state where there had been no change in parental differences in brands in harm and harm-related outcomes, as consent procedures, to examine whether this change influenced well as one appeal outcome. Students indicated their level of findings. We also examined whether adjusting for parental and agreement (‘strongly disagree’ (1) to ‘strongly agree’ (4)) to five friend smoking altered the pattern of results by repeating all statements reflecting that some brands of cigarettes are: ‘easier analyses controlling for these variables. to smoke than others’, ‘more addictive than others’, ‘easier to quit than others’, ‘have more harmful substances in them than RESULTS others’ and ‘have better looking packs than others’. Students Table 1 shows the characteristics of the entire sample, the pro- could also give a ‘don’t know’ (5) response. Items were recoded portion of students who had seen a cigarette pack in the previ- into three categories: ‘strongly agree/agree’, ‘strongly disagree/ ous 6 months and student’s smoking status for the whole sample disagree’ and ‘don’t know’. and for those who had seen a cigarette pack in previous 6 months. Weighting the data removed differences between Student variables years in sex, age and education sector of school. Among all stu- Smoking status dents, in both survey years, most of the sample were NSNS with Students indicated their lifetime history of smoking (never this proportion increasing between 2011 and 2013 (p<0.01). smoked; smoked a few puffs; smoked less than 10 cigarettes; In both the weighted and unweighted data, a greater proportion smoke more than 10 but less than 100 cigarettes; smoked 100 of students reported seeing a cigarette pack in the previous or more cigarettes), whether they had smoked in the past month 6 months in 2013 than in 2011 (p<0.01). Among students (yes or no) and their intention to smoke in the next 12 months seeing a cigarette pack in the previous 6 months, there was a sig- (seven-point scale ranging from certain not to smoke to certain nificant increase in the proportion of NSNS and a decrease in to smoke). Students’ responses to these questions were used to the proportion of Triers and past MS between 2011 and 2013 classify them into four smoking status groups that drew on the (table 1). Weighted data are reported for the rest of the paper. concept of smoking susceptibility25 and stage models of smoking uptake.26 Non-susceptible never-smokers (NSNS) had Brand character ratings never smoked a cigarette and were certain they would not Among students seeing a cigarette pack in the previous smoke in the next 12 months. Susceptible never-smokers (SNS) 6 months, those with some smoking experience gave more posi- were never-smokers, who were not certain they would not tive ratings for each brand in both 2011 and 2013 (table 2). smoke in the next 12 months. Triers (Triers) had had at least a However, mean character ratings for each brand reduced signifi- puff of a cigarette, but had not smoked in the previous 4 weeks. cantly between 2011 and 2013 (table 2). The reduction in mean Past month smokers (MS) were those who had had a cigarette in scores was generally similar among smoking status group for the the previous 4 weeks. four brands (no interaction between survey year and smoking status for Winfield, Peter Jackson, Benson and Hedges, border- Demographic characteristics line significance for Longbeach (p=0.054)). Students indicated their sex and age, whether their mother and father smoked (yes or no), and indicated how many of their five Attraction of cigarette packs closest friends smoked (none through 5). The school education Among students who had seen a cigarette pack in the previous sector (government, Catholic and independent) was recorded. 6 months, negative pack image ratings increased (p<0.001), while positive image ratings decreased between 2011 and 2013 Statistical analyses (p<0.001; table 3). There was a significant interaction between Analyses focus on data from students aged 12–17 years as this is year and smoking status for positive image ratings (p=0.01), the typical age range for secondary students in Australia. To with adjusted mean scores indicating that the decrease was correct for any oversampling or undersampling of students within greater for MS (2011: 2.56; 2013: 2.18) than NSNS (2011: age, sex and education sector groups, data were weighted to reflect 1.44; 2013: 1.36) and SNS (2011: 1.84; 2013: 1.78). the number of male and female students in each age enrolled in each education section in each state in each survey year.27 Stata Brand differences V.11.228 was used for all analyses. Analyses adjusted for clustering Table 4 shows, for students who had seen a pack of cigarettes in of students within schools and robust SEs were used. the previous 6 months, the proportion agreeing, disagreeing or Students who had seen a cigarette pack in the previous indicating ‘don’t know’ to the five brand differences statements. 6 months formed the sample for this paper. Generalised linear There was a significant change in the distribution of responses regression models tested the change in scores across survey years for four of the five statements between 2011 and 2013. For the for brand character ratings and positive and negative pack image statement ‘some cigarette brands are easier to smoke than ratings. Multinomial logistic regression examined change in the others’, there was a significant increase in the proportion distribution of responses for the three-level categorical variables responding ‘don’t know’ (p=0.006) and a significant decrease ii44 White V, et al. Tob Control 2015;24:ii42–ii49. doi:10.1136/tobaccocontrol-2014-052084 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 1 Characteristics of all students participating in each survey, the proportion of students seeing cigarette packs in the previous 6 months and the proportion of these students in each of the smoking status groups (unweighted and weighted data) Unweighted data Weighted data

2011 2013 p Value 2011 2013 p Value

(Total n) (6338) (5915) (6338) (5984) Sex (% males) 46% 50% <0.001 49% 51% 0.67 Age (mean) 14.6 14.5 <0.001 14.4 14.4 0.99 Attend Government school 68% 68% 0.016 61% 59% 0.95 Catholic school 18% 16% 23% 23% Independent school 14% 15% 16% 18% Smoking status NSNS 64% 70% <0.001 66% 71% 0.009 SNS 8% 8% 0.38 8% 8% 0.82 TS 17% 14% <0.001 16% 13% 0.01 MS 11% 8% <0.001 10% 8% 0.08 Percentage with no friends who smoke 68% 75% <0.001 71% 77% 0.016 Mother smokes (% yes) 22% 20% 0.07 21% 19% 0.28 Father smokers (% yes) 27% 26% 0.29 26% 24% 0.27 Saw cigarette packs in previous 6 months 61% 65% <0.001 60% 65% 0.004 Among students who had seen a cigarette pack in previous 6 months (N) (3838) (3819) (3738) (3856) Smoking status NSNS 56% 64% <0.001 58% 66% <0.001 SNS 9% 9% 0.48 9% 9% 0.87 TS 21% 16% <0.001 20% 15% 0.003 MS 14% 10% <0.001 13% 9% 0.01 MS, past month smoker; NSNS, non-susceptible never-smoker; SNS, susceptible never-smoker; TS, tried smoking.

in the proportion ‘disagreeing’ (p<0.001). For the statement statement ‘some cigarette brands contain more harmful sub- ‘some brands are more addictive than others’, there was a stances than others’, there was a significant decrease in the pro- decrease in the proportion disagreeing (p=0.02). For the portion of students disagreeing between 2011 and 2013

Table 2 Unadjusted mean scores (SEs) on brand character ratings for students who had seen a cigarette pack in the previous 6 months in 2011 and 2013 by smoking status (weighted data) Smoking status F-statistics and p values*

Interaction: year NSNS SNS Triers MS Total Year Smoking status and smoking status

2011 weighted n 2150 349 757 482 3738 2013 weighted n 2558 353 586 359 3856

Mean (SE) Mean (SE) Mean (SE) Mean (SE) Mean (SE)

Winfield

2011 9.11 (0.10) 10.91 (0.20) 11.32 (0.16) 15.04 (0.21) 10.49 (0.09) F(1,184)=61.09 F(3,182)=280.43 F(3,182)=1.79 2013 8.25 (0.09) 10.05 (0.20) 10.45 (0.14) 14.18 (0.20) 9.30 (0.08) p<0.001 p<0.001 p=0.150 Peter Jackson

2011 9.14 (0.10) 11.07 (0.22) 11.35 (0.16) 15.13 (0.22) 10.55 (0.09) F(1,184)=49.59 F(3,182)=245.03 F(3,183)=1.46 2013 8.37 (0.09) 10.30 (0.20) 10.58 (0.16) 14.36 (0.23) 9.43 (0.09) p<0.001 p<0.001 p=0.227 Longbeach

2011 9.23 (0.11) 11.17 (0.19) 11.67 (0.20) 15.10 (0.23) 10.66 (0.11) F(1,184)=56.52 F(3,182)=238.52 F(3,182)=2.60 2013 8.31 (0.09) 10.25 (0.18) 10.75 (0.17) 14.17 (0.23) 9.39 (0.08) p<0.001 p<0.001 p=0.054 Benson & Hedges

2011 9.34 (0.11) 11.24 (0.22) 11.63 (0.17) 14.81 (0.21) 10.69 (0.10) F(1,184)=74.72 F(3,182)=274.25 F(3,182)=1.29 2013 8.34 (0.09) 10.25 (0.21) 10.63 (0.16) 13.82 (0.22) 9.37 (0.08) p<0.001 p<0.001 p=0.278 Ninety-four students with missing data on smoking status excluded from analyses. Across the two surveys for the sample, missing data for brand characteristic variables never exceeded 2% (Winfield: n=42; Peter Jackson: n=63; Longbeach: n=63; Benson & Hedges: n=59). Minimum=6 (least positive character) to maximum=30 (most positive character). *Analyses included smoking status, sex, age, school denomination and state as covariates. MS, past month smokers; NSNS, non-susceptible never-smokers; SNS, susceptible never-smokers; Triers, tried smoking.

White V, et al. Tob Control 2015;24:ii42–ii49. doi:10.1136/tobaccocontrol-2014-052084 ii45 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 3 Unadjusted mean scores (SEs) on positive and negative pack image ratings for 2011 and 2013 students by smoking status for students who have seen a pack in previous 6 months (weighted data) Year F-statistics and p values*

NSNS SNS Triers MS Total Interaction: year Mean (SE) Mean (SE) Mean (SE) Mean (SE) Mean (SE) Year Smoking status and smoking status

Negative image†

2011 4.18 (0.02) 3.87 (0.04) 3.80 (0.03) 3.28 (0.04) 3.96 (0.02) F(1,184)=28.80 F(3,182)=158.83 F(3,182)=1.01 2013 4.31 (0.02) 4.00 (0.04) 3.93 (0.04) 3.41 (0.04) 4.15 (0.02) p<0.001 p<0.001 p=0.39 Positive image†

2011 1.47 (0.02) 1.88 (0.03) 1.90 (0.03) 2.46 (0.05) 1.72 (0.02) F(1,184)=40.26 F(3,182)=160.57 F(3,182)=3.92 2013 1.34 (0.02) 1.75 (0.04) 1.77 (0.03) 2.33 (0.05) 1.52 (0.02) p<0.001 p<0.001 p=0.01 Ninety-four students had missing data on the smoking status variable and were excluded from analyses. Across the two survey years, 558 students had missing data (no response or response ‘cannot comment’) for a positive pack image items and 559 had missing data (no response or ‘cannot comment’) for the negative pack image items and were excluded from analyses. Therefore, analyses based on: positive image: 2001: 3553; 2013: 3484; negative image: 2011: 3555; 2013: 3479. *Analyses included smoking status, sex, age, school denomination and state as covariates. †Measure on scale ranging from 1 ‘strongly disagree’ to 5 ‘strongly agree’. MS, past month smokers; NSNS, non-susceptible never-smokers; SNS, susceptible never-smokers; Triers, tried smoking.

(p<0.01). There was no change over time in responses to the brands are more addictive than others’ (interaction: p<0.003). statement ‘some cigarette brands are easier to quit than others’. Figure 1 shows the adjusted proportions by smoking status and The largest change was found for the statement ‘Some brands year for these two items. For the item ‘some cigarette brands have better looking packs than other brands’ (p<0.001) with have better looking packs than others’, the decrease in the pro- fewer students agreeing with this statement in 2013 (25%) than portion agreeing with this statement was greatest among those in 2011 (43%). involved with smoking. For the statement ‘some cigarette brands There was a significant interaction between year and smoking are more addictive than others’, change was greatest among status for two items: ‘some cigarette brands have better looking non-smokers, with no significant change in the responses of packs than others’ (interaction: p<0.001) and ‘some cigarette smokers.

Table 4 Among students who have seen cigarette pack in previous 6 months, unadjusted proportions (95% CIs) agreeing*, disagreeing† or who don’t know for statements regarding differences in ease of smoking, ease of quitting, addictiveness, more harmful and have better looking packs between cigarette brands by year of survey (weighted data) χ2 statistic and p value‡

2011 2013 Some cigarette brands are: % (95% CI) % (95%CI) Interaction: year and (Weighted N) (3738) (3856) Year smoking status

Easier to smoke than others 2 2 Agree 28% (26% to 31%) 26% (23% to 28%) χ (2)=14.83, p<0.001 χ (6)=2.33, p=0.89 Disagree 22% (20% to 24%) 19% (17% to 20%) Don’t know 49% (47% to 52%) 56% (53% to 59%) More addictive than others 2 2 Agree 33% (32% to 35%) 34% (32% to 36%) χ (2)=6.68, p=0.04 χ (6)=19.66, p=0.003 Disagree 20% (19% to 22%) 18% (16% to 19%) Don’t know 46% (44% to 49%) 49% (46% to 51%) Easier to quit than others 2 Agree 18% (16% to 19%) 16% (14% to 17%) χ (2)=1.47, p=0.48 NA Disagree 32% (30% to 34%) 31% (29% to 33%) Don’t know 51% (48% to 53%) 54% (51% to 56%) Contain more harmful substances 2 2 Agree 37% (35% to 39%) 38% (36% to 41%) χ (2)=10.63, p=0.005 χ (6)=6.64, p=0.36 Disagree 20% (18% to 22%) 17% (15% to 18%) Don’t know 43% (41% to 45%) 45% (43% to 47%) Have better looking packs than others 2 2 Agree 43% (40% to 44%) 25% (23% to 28%) χ (2)=117.41, p<0.001 χ (6)=28.51, p<0.001 Disagree 25% (24% to 27%) 36% (34% to 38%) Don’t know 32% (30% to 35%) 39% (36% to 41%) Ninety-four students had missing data on the smoking status variable and were excluded from analyses. Students with missing data on a statement excluded for that analysis. The number of students with missing data for the different variables ranged from 129 to 161. *Agree or strongly agree combined. †Disagree or strongly disagree combined. ‡Analyses included smoking status, sex, age, school denomination and state as covariates. NA, not applicable: as there was no main effect of year, the interaction between year and smoking status was not investigated. ii46 White V, et al. Tob Control 2015;24:ii42–ii49. doi:10.1136/tobaccocontrol-2014-052084 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Figure 1 Adjusted proportions of students agreeing, disagreeing or who ‘don’t know’ if ‘some brands are more addictive than others’ (above) or ‘some brands have better looking packs than others’ (below) in 2011 and 2013 by smoking status (proportions adjusted for age, sex, education sector and state).

Sensitivity analyses experimental and qualitative research showing that removing The pattern of results reported above was replicated in both sets design elements from cigarette packs reduces their appeal to of sensitivity analyses. adolescents.4111229We found that among students who had seen a cigarette pack in the previous 6 months, fewer agreed DISCUSSION that some brands have better looking packs than other brands in Ours is the first ‘real-world’ study to investigate the impact of 2013 than in 2011. We also found reductions in positive percep- standardised packaging of tobacco products on the perceptions tions of brand character for all four brands assessed. of the image of cigarette packs and brands among adolescents. Encouragingly, reductions were found among all smoking status Seven to 12 months after the introduction of standardised pack- groups. aging in Australia, the appeal of cigarette packs and brands to Among students who had seen a cigarette pack in the previous adolescents who had seen packs in the previous 6 months had 6 months, there was a significant increase in negative pack decreased significantly, albeit modestly. Additionally, among image ratings between 2011 and 2013 which was consistent students who had seen cigarette packs in the past 6 months, across smoking groups. We also found a decrease in positive more were unsure as to whether brands differed in terms of pack image ratings. Although the absolute change was modest, their ease of being smoked in 2013 than 2011 and there was a this decrease is encouraging as our previous study found that large increase in the proportion disagreeing that some brands although positive image ratings declined when GHWs covering have better looking packs than others. The study suggests that 30% of the front of fully branded packs were introduced in 7–12 months after its introduction, the new controls on the cig- 2006, they had rebounded to pre-2006 GHW levels by 2011.24 arette packaging in Australia were starting to reduce the appeal The current study suggests that the introduction of the new cig- of cigarette packs to adolescents and were beginning to reduce arette packaging has again disrupted the positive image asso- the pack’s ability to communicate messages regarding differ- ciated with cigarette packs. ences in ease of smoking between brands among adolescents. One aim of plain packaging was to reduce cigarette packa- Although the reductions found in our study were relatively ging’s ability to mislead consumers regarding the harmful effects modest, the pattern of results are consistent with those from of smoking. Experimental studies have shown that adolescents

White V, et al. Tob Control 2015;24:ii42–ii49. doi:10.1136/tobaccocontrol-2014-052084 ii47 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper perceive cigarettes from plain packs as ‘tasting cheap’ compared smoking are corrected. Additionally, future studies need to with cigarettes from fully branded packs.10 12 In these studies, determine whether the new packaging has reduced smoking ratings were elicited after showing specific fully branded or uptake among adolescents, as is the long-term goal of the policy. plain pack images to participants. In our study, adolescents were asked whether they agreed or disagreed with a number of general statements regarding whether brands differed on a What this paper adds number of factors including ease of smoking, addictiveness, harmfulness and ease of quitting. Responses to these questions ▸ were mixed. Cigarette packs are powerful communication tools with The proportion of students disagreeing with the statements broad reach including to children and adolescents. ▸ ‘some brands are easier to smoke than others’ and ‘some brands Experimental and qualitative studies suggest that packaging are more addictive than others’ decreased, while the proportion cigarettes in plain or standardised packs reduces perceived reporting ‘don’t know’ regarding the ease of smoking some attractiveness of packs and perceived quality of cigarettes brands increased. The pattern of results found suggests that 7– and corrects misperceptions about harms among ’ 12 months after their introduction, the new packs were creating adolescents. However, no study has examined adolescents uncertainty regarding whether there were differences between responses to plain packaging of cigarettes in a community cigarette brands’ addictive qualities and their ease of being where plain packaging is the norm. ▸ smoked. However, for the statement ‘some cigarette brands Seven to 12 months after the introduction of plain contain more harmful substances than others’ slightly more stu- packaging of cigarettes with large front-of-pack graphic dents agreed with this statement in 2013 than 2011. Further health warnings in Australia, this study suggests that the research is needed to determine whether, with continued expos- appeal of cigarette packs and brands to Australian fi ure to the new packs, adolescents develop stronger disagreement adolescents had decreased signi cantly. regarding differences between brands in the harmfulness, addic- tiveness and ease of smoking. This is particularly important since plain packaging did not limit the use of variant names (eg, Acknowledgements The authors would like to acknowledge the work of Ms ‘Distinct’ or ‘Fine’), which are independently associated with Emily Bariola who co-ordinated the national study in 2011. Dr Kerri Coomber misperceptions of harm.10 11 With the implementation of plain provided advice on statistical analysis. They would like to thank and acknowledge the education authorities in participating states and the principals and contact packaging in Australia, there has been greater use of longer and teachers in participating schools for allowing us to conduct the survey in schools and more evocative variant names on tobacco products.30 for assisting with survey arrangements. We thank the students for participating in Several study limitations need to be noted. First, because the surveys. plain packaging was introduced at the same time as the intro- Contributors VW and MW designed study. VW analysed data and drafted duction of larger GHWs on the front of cigarette packs, we manuscript. TW coordinated data collection in 2013 and oversaw data processing. fi cannot determine whether the changes we found in adolescents’ TW also contributed to data analysis. The nal version of the paper has been approved by all authors. perceptions of cigarette packs are the result of plain packaging alone. For this reason, we must conclude that changes we have Funding Data used in this study were gathered from surveys funded fully or in part by the Australian Government Department of Health and Ageing. Cancer Councils found are the result of the combination of the cigarette pack- and health departments of participating states also contributed funding for the 2011 aging changes and not just the introduction of plain packaging. surveys. Change in consent procedures in one state meant that some Competing interests The authors wish to advise that MW was a member of the schools surveyed only a small number of students and this could Tobacco Working Group of the Australian National Preventive Health Task Force and have introduced some bias. In a study examining the impact of the Expert Advisory Committee on Plain Packaging that advised the Australian active parental consent on students’ responses to surveys about Department of Health on research pertaining to the plain packaging legislation. VW substance use, while prevalence estimates were generally the holds competitive grant funding from the Victorian Cancer Agency and the National Breast Cancer Foundation, VW and MW hold such funding from the Australian same, there was greater similarity in the students surveyed at a National Health and Medical Research Council and MW holds such funding from the particular school.21 Furthermore, sensitivity analyses suggested US National Institutes of Health, Australian National Preventive Health Agency and that the change in consent procedures in one state had minimal BUPA Health Foundation. influence on our results. Ethics approval Both the 2011 and 2013 study had ethical approval from the Despite these limitations, the present study provides new lead author’s institution. In both years, the survey was approved by educational information on the immediate response of adolescents to the authorities in both states. introduction of plain packaging of cigarettes coupled with Provenance and peer review Not commissioned; externally peer reviewed. larger front of pack GHWs. Our data suggest that these changes Open Access This is an Open Access article distributed in accordance with the have reduced the appeal of cigarette packs to adolescents, Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which increased negative perceptions of packs and reduced positive 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 perceptions regarding brand characteristics, although we note properly cited and the use is non-commercial. See: http://creativecommons.org/ that the changes found were modest. While our data do not licenses/by-nc/4.0/ show that the introduction of the new cigarette packs immedi- ately corrected adolescents’ misbeliefs about differences in the ease of smoking, harmfulness and addictiveness of cigarettes REFERENCES fi 1 Hoek J, Gendall P, Gifford H, et al. Tobacco branding, plain packaging, pictorial from different brands, the pattern of ndings does suggest that warnings, and symbolic consumption. Qual Health Res 2012;22:630–9. the new packs may be disrupting these beliefs and increasing 2 Moodie C, Hastings G. Making the pack the hero, tobacco industry response to uncertainty among adolescents about the effects of smoking. marketing restrictions in the UK: findings from a long-term audit. Int J Ment Health Future studies of adolescents’ responses to the new tobacco Addict 2011;9:24–38. 3 Wakefield M, Morley C, Horan J, et al. The cigarette pack as image: new evidence packaging after longer exposure to the new packaging are from tobacco industry documents. Tob Control 2002;11:i73–80. needed to determine whether the effects we have found increase 4 Scheffels J, Saebo G. Perceptions of plain and branded cigarette packaging among or diminish in size and whether misbeliefs about the harms of Norwegian youth and adults: a focus group study. Nicotine Tob Res 2013;15:450–6. ii48 White V, et al. Tob Control 2015;24:ii42–ii49. doi:10.1136/tobaccocontrol-2014-052084 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

5 World Health Organization. World Health Organization Framework Convention on 18 Tobacco Plain Packaging Act 2011. 2011. http://www.comlaw.gov.au/Details/ Tobacco Control New York, United Nations World Health Organization, 2003. http:// C2011A00148 www.who.int.tobacco/framework/WHO_FCTC_english.pdf 19 White V, Williams T, Faulkner A, et al. Do larger graphic health warnings on 6 Scollo M, Lindorff K, Coomber K, et al. Standardised packaging and new enlarged standardised cigarette packs increase adolescents’ cognitive processing of consumer graphic health warnings for tobacco products in Australia—legislative requirements health information and beliefs about smoking related harms? Tob Control 2015;24: and implementation of the Tobacco Plain Packaging Act 2011 and the Competition ii50–7. and Consumer (Tobacco) Information Standard, 2011. Tob Control 2015;24:ii9–16. 20 White VM, Hayman J, Hill DJ. Can population-based tobacco-control policies 7 White CM, Hammond D, Thrasher JF, et al. The potential impact of plain packaging change smoking behaviors of adolescents from all socio-economic groups? Findings of cigarette products among Brazilian young women: an experimental study. BMC from Australia: 1987–2005. Cancer Causes Control 2008;19:631–40. Public Health 2012;12:737. 21 White VM, Hill DJ, Effendi Y. How does active parental consent influence the 8 Moodie C, Mackintosh AM, Hastings G, et al. Young adult smokers’ perceptions of findings of drug-use surveys in schools? Eval Rev 2004;28:246–60. plain packaging: a pilot naturalistic study. Tob Control 2011;20:367–73. 22 White V, Bariola E. Australian secondary school students’ use of tobacco, alcohol, 9 Lund I, Scheffels J. Young smokers and non-smokers perceptions of typical users of and over-the counter and illicit substances in 2011. Melbourne, Australia: Report plain vs. branded cigarette packs: a between-subjects experimental survey. BMC prepared for: Drug Strategy Branch: Australian Government Department of Health Public Health 2013;13:1005. and Ageing, December 2012. http://www.nationaldrugstrategy.gov.au/internet/ 10 Hammond D, Dockrell M, Arnott D, et al. Cigarette pack design and perceptions of drugstrategy/Publishing.nsf/content/BCBF6B2C638E1202CA257ACD0020E35C/$File/ risk among UK adults and youth. Eur J Public Health 2009;19:631–7. National%20Report_FINAL_ASSAD_7.12.pdf 11 Hammond D, Daniel S, White CM. The effect of cigarette branding and plain 23 White V, Webster B, Wakefield M. Do graphic health warning labels have an impact on packaging on female youth in the United Kingdom. J Adolesc Health adolescents’ smoking-related beliefs and behaviours? Addiction 2008;103:1562–71. 2012;52:151–7. 24 White V, Bariola E, Faulkner A, et al. Graphic health warnings on cigarette packs: 12 Germain D, Wakefield MA, Durkin SJ. Adolescents’ perceptions of cigarette how long before the effects on adolescents wear out? Nicotine Tob Res Published brand image: does plain packaging make a difference? J Adolesc Health Online First: 19 Sep 2014. doi:10.1093/ntr/ntu184 2010;46:385–92. 25 Pierce JP, Choi WS, Gilpin EA, et al. Validation of susceptibility as a predictor of 13 Ford A, Moodie C, MacKintosh AM, et al. How adolescents perceive cigarette which adolescents take up smoking in the United States. Health Psychol packaging and possible benefits of plain packaging. Educ Health 2013;31:83–8. 1996;15:355–61. 14 Stead M, Moodie C, Angus K, et al. Is consumer response to plain/standardised 26 Mayhew KP, Flay BR, Mott JA. Stages in the development of adolescent smoking. tobacco packaging consistent with framework convention on tobacco control Drug Alcohol Depend 2000;59(Suppl 1):S61–81. guidelines? A systematic review of quantitative studies. PLoS ONE 2013;8:e75919. 27 Australian Bureau of Statistics. Schools Australia, 2014. http://www.abs.gov.au/ 15 Hammond D. Plain packaging regulations for tobacco products: the impact of AUSSTATS/[email protected]/DetailsPage/4221.02013?OpenDocument. standardizing the color and design of cigarette packs. Salud Pública México 28 StataCorp. Stata Statistical Software: Release 11. College Station, TX: StataCorp LP, 2010;52:226–32. 2009. 16 Hammond D. Standardized packaging of tobacco products: evidence review, 2014. 29 Hammond D, White C, Anderson W, et al. The perceptions of UK youth of http://www.tri.ie/media/3364/standardized-packaging-of-tobacco-products-evidence- branded and standardized, ‘plain’ cigarette packaging. Eur J Public Health review.pdf 2014;24:537–43. 17 Hammond D, Arnott D, Dockrell M, et al. Cigarette pack design and perceptions of 30 Scollo M, Occleston J, Bayly M, et al. Tobacco product developments coinciding risk among adults and youth. 2009 Joint Conference of SRNT and SRNT-Europe; with the implementation of plain packaging in Australia. Tob Control Published Apr 29, 2009, Dublin, Ireland, 2009. Online First: 30 Apr 2014. doi:10.1136/tobaccocontrol-2013-051509.

White V, et al. Tob Control 2015;24:ii42–ii49. doi:10.1136/tobaccocontrol-2014-052084 ii49 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Has the introduction of plain packaging with larger graphic health warnings changed adolescents' perceptions of cigarette packs and brands? Victoria White, Tahlia Williams and Melanie Wakefield

Tob Control 2015 24: ii42-ii49 doi: 10.1136/tobaccocontrol-2014-052084

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii42

These include: References This article cites 19 articles, 7 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii42#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Research paper Do larger graphic health warnings on standardised cigarette packs increase adolescents’ cognitive processing of consumer health information and beliefs about smoking-related harms? Victoria White, Tahlia Williams, Agatha Faulkner, Melanie Wakefield

Centre for Behavioural ABSTRACT required the size of GHW on cigarette packs to Research in Cancer, Cancer Objective To examine the impact of plain packaging increase to 75% of the front-of-pack (with the size Council Victoria, Melbourne, Australia of cigarettes with enhanced graphic health warnings on of the back-of-pack GHW remaining 90%) and a Australian adolescents’ cognitive processing of warnings new set of GHW. With the implementation of the Correspondence to and awareness of different health consequences of Tobacco Plain Packaging Act and this Standard, Associate Professor Victoria smoking. Australia became the first country in the world to White, Centre for Behavioural Methods Cross-sectional school-based surveys have large front-of-pack GHWs on tobacco pro- Research in Cancer, Cancer Council Victoria, 615 St Kilda conducted in 2011 (prior to introduction of standardised ducts with standardised packaging design (see Scollo 4 Road, Melbourne VIC 3004, packaging, n=6338) and 2013 (7–12 months et al for details of new packaging requirements). Australia; vicki.white@ afterwards, n=5915). Students indicated frequency of Owing to the high visibility of cigarette packs, cancervic.org.au attending to, reading, thinking or talking about GHWs have the potential to communicate the warnings. Students viewed a list of diseases or health health effects of smoking to a broad audience Received 6 October 2014 56 Accepted 26 December 2014 effects and were asked to indicate whether each was including to adolescents. Packaging’sinfluence caused by smoking. Two—‘kidney and bladder cancer’ on adolescents’ awareness of health warnings has and ‘damages gums and teeth’—were new while the been investigated since the early 1990s.7 An early remainder had been promoted through previous health study, conducted when health warnings on cigarette warnings and/or television campaigns. The 60% of packs were generally small and text-based, found students seeing a cigarette pack in previous 6 months in that plain packaging only influenced recall of warn- 2011 and 65% in 2013 form the sample for analysis. ings when the cigarette packs used were unfamiliar, Changes in responses over time are examined. with greater recall of warnings found for plain Results Awareness that smoking causes bladder cancer packs of unfamiliar cigarette brands.7 Recent quali- increased between 2011 and 2013 (p=0.002). There tative studies (conducted in an era of larger and was high agreement with statements reflecting health pictorial warnings) have suggested that plain pack- effects featured in previous warnings or advertisements aging will increase the noticeability, believability with little change over time. Exceptions to this were and seriousness of health warnings among adoles- increases in the proportion agreeing that smoking was a cents.58However, as findings from experimental leading cause of death (p<0.001) and causes blindness studies have been mixed, this may not be the case.9 (p<0.001). The frequency of students reading, attending An early experimental study found a higher pro- to, thinking or talking about the health warnings on portion of adolescents recalled two of three text- cigarette packs did not change. based warnings when presented on a plain pack.10 Conclusions Acknowledgement of negative health An eye movements tracking study found that while effects of smoking among Australian adolescents packaging did not influence attention paid to warn- remains high. Apart from increased awareness of bladder ings among adolescents who had never smoked, it cancer, new requirements for packaging and health did influence adolescents who had smoked, with warnings did not increase adolescents’ cognitive greater attention given to warnings on plain processing of warning information. packs.11 However, another experimental study did not find a difference in adolescents’ recall of health warnings regardless of whether they were presented INTRODUCTION on a plain or fully branded pack or whether the fi 12 Open Access Graphic health warnings (GHWs) were rst intro- warning label was large or small. To date, no Scan to access more duced on tobacco products in Australia in 2006, study has examined adolescents’ responses to free content with the warnings required to cover 30% of a cigar- GHWs on plain cigarette packs once this packaging ette pack’s front and 90% of the cigarette pack’s regulation has been introduced into a country. back.1 Coinciding with the full implementation of The present study investigates the impact of Australia’s Tobacco Plain Packaging Act 20112 in Australia’s new cigarette packaging on the fre- December 2012, a new Consumer Information quency of adolescents attending, reading, talking Standard3 was introduced requiring a new set of and thinking about the warnings. These behaviours GHWs. The new GHWs included information are considered indicators of cognitive processing, To cite: White V, about several health effects not covered in previous and greater cognitive processing has been positively Williams T, Faulkner A, et al. warnings (eg, kidney and bladder cancer, and effects associated with quit attempts among adult Tob Control 2015;24: on teeth and gums) and revised images for previ- smokers.13 14 The introduction of GHWs covering – ii50 ii57. ously used health warnings. The new Standard also 30% of the front of branded cigarette packs in ii50 White V, et al. Tob Control 2015;24:ii50–ii57. doi:10.1136/tobaccocontrol-2014-052085 Research paper

2006 increased Australian adolescents’ cognitive processing of Consent procedures warning labels.6 However, recent research has shown that this A ‘passive’ parental consent procedure was used for the 2011 increase diminished after 5 years’ exposure to GHWs, with ado- survey and in one state in 2013. In this procedure, parents are lescents’ cognitive processing of GHWs in 2011 similar to informed of the study and tell the school if they do not want pre-GHW levels.15 This may reflect a process of habituation, their child to take part. In 2013, an ‘active’ parental consent which suggests that attention to a specific object or stimuli will procedure (parents informed the school that their child could reduce with repeated exposure due to its increasing familiar- or could not take part in survey) was used in all government ity.16 17 The changes to both cigarette packaging and GHWs and Catholic schools in the second state. In all surveys, students introduced in Australia in 2012 may interrupt the habituation consented to survey participation. To examine the impact of the process. In the current paper, we use data collected from active parental consent on student participation, we examined surveys of adolescents conducted in the year before the new cig- the proportion of schools where fewer than 40 students com- arette packaging was introduced (2011) and in the 7–12 months pleted the survey. In the state where parental consent procedures after its introduction (2013) to examine whether the introduc- did not change, the proportion of schools where fewer than 40 tion of the new tobacco packaging in Australia increased adoles- students took part in the survey decreased between 2011 (16%) cents’ cognitive processing of warning labels and awareness of and 2013 (6%; p=0.06) while in the other state the proportion health risks associated with smoking among students who had increased (2011: 26%; 2013: 63%, p=0.003). Previous research seen a cigarette pack in the previous 6 months. suggests that although active parental consent reduces participa- tion numbers, overall substance use estimates are similar to METHODS when passive parental consent is used.20 Study overview and design The current study used data from cross-sectional school-based Procedure for all surveys surveys of adolescents in two Australian states conducted in On an agreed day external research staff attended the school to ’ 2011 and 2013. The 2011 data come from the states compo- administer the pencil-and-paper questionnaire to the preselected nent of a national triennial survey of a representative sample of classes of students, during school time. Students without paren- – – Australian students aged 12 17 years in year levels 7 12. The tal consent were excluded from the survey group. After working 2013 survey was separate from the national study, although it through a practice survey with research staff, students gave drew on the procedures and samples of the 2011 survey. Both consent for their participation in the study and were provided surveys had institutional ethics approval and approvals from the with the survey. Students worked independently and completed appropriate school authorities. The sample and procedures used the survey anonymously. for this study have been described elsewhere.18 A brief descrip- tion is provided here. Measures Items used for this investigation were taken from larger surveys School samples in both years. 2011 Survey Schools were randomly selected from the three main Australian education sectors (government, Catholic and independent) to Recency of seeing cigarette packs ensure proportional representation. Principals consented to study In both surveys, students indicated when they last saw a pack of ‘ ’ ‘ participation and when a school declined, it was replaced with the cigarettes (response options: within the last 6 months , more ’ ‘ ’ school geographically closest to the original school within the same than 6 months ago or never ). education sector. We aimed to recruit 117 schools from the two states. To achieve this, we approached 324 schools with 97 schools Perceptions of health consequences of smoking agreeing to study participation (30% response rate). Surveying In both surveys, students were presented with the same list of 18 took place between June and December in 2011. items and asked to indicate whether they agreed or disagreed that they were caused by smoking, using a five-point Likert scale. Fifteen 2013 Survey items—which are the focus of this paper—reflected the GHWs This study aimed to survey students from the 97 secondary appearing on cigarette packs since 2006, warnings that were used in schools participating in the 2011 survey. School principals were post plain packaging implementation or tobacco-related illnesses sent an invitation letter seeking consent for study participation. promoted in tobacco control advertising since 1997 or earlier. The Fifty-eight schools agreed to participate (60%). When a school GHWs used from 2006 to 2012 were a mix of completely new refused, a replacement school selected from the list of replace- messages and re-presentations of text-warning labels used between ment schools drawn for the 2011 survey was approached. 1995 and 2005. Similarly, the GHWs introduced with plain pack- Sixty-three replacement schools were approached and 24 aging in December 2012 were a mix of new and old messages. schools agreed (response rate 38%), giving a total of 82 schools Ta b l e 1 shows the text messages in each of the four sets of seven participating in the 2013 survey. Data collection occurred from GHWs (sets rotating roughly every 12 months) used between 2006 June to November 2013. and 2013. Ta b l e 1 also indicates which warnings used new or similar images or topics post plain packaging, and a visual compari- Student selection son is provided at https://www.cancervic.org.au/plainfacts/browse. Schools were approached regarding surveying one class of asp?ContainerID=packagingexamples. Text-only health warnings students from each of years 7–10 (age 12–15) or two classes of on tobacco products before 2006 included: ‘smoking causes lung students from each of years 11 and 12 (age 16 and 17). cancer’, ‘smoking causes heart disease’, ‘smokingwhenpregnant Researchers worked with each school to ensure selected classes harms your baby’, ‘your smoking can harm others’ and ‘smoking is were representative of all classes (eg, no electives). With an addictive’. Health messages conveyed in tobacco control advertise- average of 21 students per class,19 we aimed to survey approxi- ments before and after 2006 included smoking: clogs arteries; mately 80 students per school. causes blindness; emphysema; lung cancer and heart disease.

White V, et al. Tob Control 2015;24:ii50–ii57. doi:10.1136/tobaccocontrol-2014-052085 ii51 Research paper

Table 1 Health effects depicted on cigarette and tobacco packaging by graphic health warnings in different time periods in Australia Set B* 2006 pre-PP Set A 2006 pre-PP Set 1 post-PP Set 2 post-PP 1 March 2011–28 February 2012 1 November 2011–30 November 2012 1 October 2012–30 November 2013 From 1 August 2013

Smoking is addictive Gangrene Gangrene† Damage to teeth/gums§ Stroke Mouth cancer/throat cancer Mouth cancer‡ Stroke‡ Blindness Artery Blindness† Throat cancer‡ Harm to unborn babies Harm to children Harm to unborn babies‡ Harm to children‡ Heart disease Death Lung cancer‡ Death‡ Lung cancer Emphysema Emphysema† Heart disease‡ Tobacco smoke is toxic Quitting will improve your health Quitting will improve your health‡ Kidney and bladder cancer¶ *Set A and set B: each consisted of seven graphic health warnings with the sets rotated roughly every 12 months. †Same topic, same or similar image. ‡Same topic, new image. §New topic, similar image to a related harm. ¶New topic, new image. PP, plain packaging.

Cognitive processing of warnings Data from students who had seen a cigarette pack in the pre- Those seeing a cigarette pack in the previous 6 months indicated vious 6 months are used. Logistic regression analyses compared how frequently they: read; paid close attention to; thought proportions across the two surveys. Linear regression analyses about and talked about the warning labels using a five-point examined change in the cognitive processing variables between scale: (1) ‘never’; (2) ‘once or twice’; (3) ‘sometimes’; (4) the surveys. In these analyses, each outcome variable was ‘often’ and (5) ‘every time I see them’. In addition, students regressed onto survey period and the control variables age, sex, indicated how frequently they had not had a cigarette because school type, state and smoking status. An interaction term of the warnings. Students who had smoked in the previous between survey period and smoking status was fitted to deter- 12 months were asked how frequently they thought about quit- mine if the effect of survey period was consistent across ting smoking because of the warnings. Responses were made on smoking stage. Students with missing data on variables were a consistent five-point scale. excluded from relevant analyses. All analyses were adjusted for clustering of students within school and SEs obtained that were Student characteristics robust to potential non-independence of students. The statistical 24 Students indicated whether they had ever smoked (no never; package STATA was used for analysis to accommodate the yes, just a few puffs; yes, but less than 10 cigarettes; yes, more complex sample design. than 10 but less than 100 cigarettes; yes, more than 100 cigar- Sensitivity analyses examined whether the change in parental ettes in lifetime); whether they had smoked on each of the pre- consent procedures in one state influenced findings, by repeating vious 7 days and whether they intended to smoke in the next the analyses described above using data only from the state 12 months (7-point scale ranging from certain not to smoke to where there had been no change in parental consent procedures. certain to smoke). We classified students into a four-level Sensitivity analyses produced the same pattern of results as smoking status variable based on their answers to these ques- reported below. tions. The smoking status levels were: non-susceptible never- smokers (NSNS) who had never smoked a cigarette (not even a RESULTS puff) and were certain not to smoke in the next 12 months; sus- Sixty-one per cent of students in 2011 and 65% in 2013 had ceptible never-smokers (SNS) (not even a puff) who were not seen a cigarette pack in the previous 6 months (table 2). certain that they would not smoke in the next 12 months; Characteristics of the sample seeing a cigarette pack are shown experimental smokers (ES) had had at least a puff of a cigarette, for the weighted and unweighted data for the two surveys (table but had not smoked in the previous 7 days and current smokers/ 2). While in the unweighted data, there were some significant established smokers (CS) had smoked in the previous 7 days or differences in sex and age characteristics of students seeing cig- had smoked at least 100 cigarettes in their lifetime. The classifi- arette packs between 2011 and 2013, these differences were not cation used here is similar to that reported by others.21 22 significant in the weighted data set. The proportion of students Students indicated whether their mother and/or father reporting that none of their five closest friends smoked was sig- smoked (yes/no) and indicated the number of smokers among nificantly greater in 2013 than 2011 in the weighted and their five closest friends (0–5). Variables indicating that at least unweighted data. In both survey years, the majority of students one parent smoked and that at least one friend smoked were cal- were NSNS. The proportion of NSNS increased between the culated. Students also indicated their sex and age. Education two surveys for both the weighted and unweighted data. The sector of the school attended (government, Catholic, independ- proportion of CS and ES decreased between 2011 and 2013 for ent) was recorded. both the weighted and unweighted data. Weighted data are reported for the rest of the paper. Statistical analyses To correct for any oversampling or undersampling of students Awareness of health risks of smoking from specific age, sex and education sector groupings within a Table 3 shows the proportion of students who had seen a cigar- state, the sample was weighted to bring the sample distribution ette pack in the previous 6 months agreeing that smoking causes into line with the school enrolment data for the two states.23 χ2 bladder cancer, kidney disease and diseases of the gums and Tests examined the similarity of the sample in the two surveys. teeth, all items reflecting new health warnings introduced with ii52 White V, et al. Tob Control 2015;24:ii50–ii57. doi:10.1136/tobaccocontrol-2014-052085 Research paper

Table 2 Proportion of sample seeing cigarette packs in the previous 6 months and characteristics of these students in 2011 and 2013 surveys Unweighted data Weighted data

2011 2013 p Value 2011 2013 p Value

Total number of students surveyed (N) 6338 5915 6338 5984 Saw cigarette packs in previous 6 months: N (%) 3888 (61) 3852 (65) <0.001 3802 (60) 3886 (65) 0.004 Among students seeing packs in previous 6 months Sex (% males) 43 49 <0.001 45 50 0.26 Age <0.001 0.79 12–15-year olds 62 69 66 68 16–17-year olds 38 31 34 32 Mother smokes (% yes) 27 25 0.04 27 24 0.22 Father smokers (% yes) 33 31 0.14 33 29 0.083 Percentage of students with no friends who smoke 61 71 <0.001 64 72 0.003 Smoking status (%) NSNS 56 65 <0.001 58 66 0.003 SNS 9 9 0.48 9 9 0.87 ES 23 18 <0.001 22 17 0.002 CS 12 8 <0.001 11 7 0.008 CS, smoking in past week/established smokers; ES, experimental smokers; NSNS, non-susceptible never-smokers; SNS, susceptible never-smokers.

the standardised packs. The proportion agreeing that smoking nine health consequences statements (table 4). For seven state- causes bladder cancer increased (p=0.002), with awareness not ments (lung cancer, heart disease, unborn babies, leading cause varying by smoking status (p=0.40). There was a small increase of death, mouth cancer, stroke and diseases of fingers and toes), in agreement that smoking causes kidney disease but this was agreement was higher among NSNS than CS. For two items not statistically significant (p=0.18). Agreement with the state- (emphysema and blindness), agreement was higher among CS ment that smoking causes diseases of teeth and gums was high than NSNS. The two significant interactions between year and in both surveys and did not change. smoking status occurred for items where there was no main Agreement with statements reflecting diseases or conditions effect of year (mouth cancer and emphysema). featured in previous warnings was generally very high and for most items there was no change (table 4). Exceptions were Cognitive processing of warnings increases in the proportion of students agreeing that smoking Among students who had seen a cigarette pack in the previous was a leading cause of death (p<0.001) and causes blindness 6 months, there was no significant change in the cognitive pro- (p<0.001). Smoking status was associated with agreement for cessing items between 2011 and 2013 (table 5). While there was

Table 3 For health messages introduced post plain packaging, unadjusted proportion (95% CI) of students agreeing or strongly agreeing that smoking causes these diseases by year and smoking status among students who had seen a cigarette pack within the previous 6 months (weighted data) χ2 Statistic and p values*

Interaction Years GHW in NSNS SNS ES CS Total of year and circulation % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) Year Smoking status smoking status

2011 (weighted N) (2150) (349) (830) (405) (3733) 2013 (weighted N) (2558) (253) (653) (286) (3851) Diseases of the gums 2 2 2 2011 (%) Message from 1/8/ 88 (86 to 89) 84 (81 to 88) 89 (87 to 91) 86 (82 to 89) 87 (86 to 89) χ(1)=0.42, χ(3)=13.86, χ(3)=3.72, 2013 p=0.52 p=0.003 p=0.29 2013 (%) 87 (85 to 88) 84 (80 to 87) 89 (87 to 91) 85 (81 to 89) 87 (85 to 88) Kidney disease 2 2 2 2011 (%) Message from 1/8/ 78 (76 to 80) 77 (73 to 81) 78 (76 to 81) 75 (71 to 79) 78 (76 to 79) χ(1)=1.77, χ(3)=6.51, χ(3)=2.52, 2013 p=0.18 p=0.09 p=0.47 2013 (%) 80 (78 to 83) 79 (75 to 83) 80 (78 to 83) 77 (73 to 81) 80 (78 to 82) Bladder cancer 2 2 2 2011 (%) Message from 1/8/ 63 (61 to 65) 61 (57 to 66) 64 (60 to 67) 63 (58 to 68) 63 (61 to 65) χ(1)=9.57, χ(3)=2.98, χ(3)=4.13, 2013 p=0.002 p=0.40 p=0.25 2013 (%) 69 (65 to 72) 67 (63 to 71) 69 (66 to 73) 69 (64 to 73) 69 (66 to 71) In total, 104 students had missing data on the smoking status variable and were excluded from analyses. Missing data on the different items over the two surveys ranged from 6 to 20. *Analyses adjusted for covariates of sex, age, school education sector, state and smoking status. CS, smoking in past week/established smokers; ES, experimental smokers; NSNS, non-susceptible never-smokers; SNS, susceptible never-smokers.

White V, et al. Tob Control 2015;24:ii50–ii57. doi:10.1136/tobaccocontrol-2014-052085 ii53 Research paper

Table 4 Unadjusted proportion (95% CIs) of students agreeing or strongly agreeing that smoking causes different diseases promoted in GHWs or tobacco control advertising from 2006 onwards by survey year and smoking status among students who had seen a cigarette pack within the previous 6 months (weighted data) χ2 Statistic and p value*

Interaction Years GHW used prior of year and to plain packaging NSNS SNS ES CS Total Smoking smoking Smoking… introduction % (95%CI) % (95%CI) % (95%CI) % (95%CI) % (95%CI) Year status status

2011 (weighted N) (2150) (349) (830) (405) (3733) 2013 (weighted N) (2558) (253) (653) (286) (2851) Not a plain packaging health warning Is addictive 2007, 2009, 2011 2 2 2 2011 (%) 88 (87 to 90) 90 (87 to 92) 90 (88 to 92) 89 (87 to 92) 89 (88 to 90) χ(1)=1.51, χ(3)=0.50, χ(3)=6.37, 2013 (%) 87 (86 to 89) 88 (86 to 91) 89 (87 to 91) 88 (85 to 92) 88 (86 to 89) p=0.22 p=0.92 p=0.01 Is toxic (from 2007, 2009, 2011 tobacco smoke) 2 2 2 2011 (%) 81 (79 to 83) 77 (73 to 80) 84 (82 to 86) 86 (83 to 89) 82 (80 to 83) χ(1)=0.45, χ(3)=5.24, χ(3)=5.34, 2013 (%) 80 (78 to 83) 76 (72 to 80) 84 (81 to 86) 86 (82 to 89) 81 (79 to 83) p=0.50 p=0.16 p=0.15 Clogs arteries 2006, 2008, 2010, 2012 2 2 2 2011 (%) 81 (79 to 83) 81 (77 to 84) 84 (82 to 87) 84 (80 to 87) 82 (80 to 84) χ(1)=2.23, χ(3)=1.21, χ(3)=2.19, 2013 (%) 79 (77 to 82) 79 (76 to 82) 83 (80 to 86) 82 (78 to 86) 80 (78 to 82) p=0.14 p=0.75 p=0.53 Post-plain packaging Set 1 warning from 1 October 2012 Causes diseases in 2006, 2008, 2010, 2012 toes and fingers (gangrene) 2 2 2 2011 (%) 88 (86 to 89) 89 (86 to 91) 89 (87 to 91) 84 (80 to 88) 88 (86 to 89) χ(1)=0.00, χ(3)=18.03, χ(3)=2.34, 2013 (%) 88 (85 to 91) 89 (86 to 92) 90 (87 to 92) 85 (80 to 89) 88 (86 to 91) p=0.97 p<0.001 p=0.50 Causes mouth cancer 2006, 2008, 2010, 2012 2 2 2 2011 (%) 94 (93 to 95) 93 (91 to 96) 95 (94 to 97) 92 (90 to 94) 94 (93 to 95) χ(1)=3.33, χ(3)=18.29, χ(3)=8.66, 2013 (%) 96 (95 to 96) 95 (93 to 97) 96 (95 to 98) 94 (91 to 96) 95 (95 to 96) p=0.07 p<0.001 p=0.034 Causes blindness 2007, 2009, 2011 2 2 2 2011 (%) 50 (47 to 52) 45 (40 to 50) 49 (46 to 53) 56 (52 to 61) 50 (47 to 52) χ(1)=42.44, χ(3)=8.14, χ(3)=0.85, 2013 (%) 60 (58 to 63) 55 (50 to 60) 60 (57 to 63) 67 (62 to 71) 60 (58 to 63) p<0.001 p=0.043 p=0.84 Harms unborn babies 2007, 2009, 2011 2 2 2 2011 (%) 95 (94 to 96) 92 (89 to 94) 94 (93 to 96) 91 (88 to 94) 94 (93 to 95) χ(1)=0.84, χ(3)=25.81, χ(3)=3.08, 2013 (%) 94 (93 to 95) 91 (88 to 93) 94 (92 to 95) 90 (86 to 94) 93 (92 to 95) p=0.36 p<0.001 p=0.38 Causes lung cancer 2007, 2009, 2011 2 2 2 2011 (%) 98 (98 to 99) 98 (96 to 99) 97 (96 to 98) 94 (92 to 96) 97 (97 to 98) χ(1)=1.37, χ(3)=37.45, χ(3)=0.82, 2013 (%) 98 (97 to 98) 97 (96 to 98) 96 (95 to 97) 93 (90 to 96) 97 (96 to 98) p=0.24 p<0.001 p=0.84 Causes emphysema 2006, 2008, 2010, 2012 2 2 2 2011 (%) 82 (80 to 84) 79 (75 to 83) 89 (87 to 91) 89 (87 to 92) 84 (82 to 86) χ(1)=0.58, χ(3)=14.78, χ(3)=7.97, 2013 (%) 83 (81 to 85) 80 (77 to 84) 90 (88 to 92) 90 (88 to 93) 85 (83 to 86) p=0.45 p=0.002 p=0.047 Post-plain packaging Set 2 warning from 1 August 2013 Is a leading cause of 2006, 2008, 2010, 2012 death 2 2 2 2011 (%) 76 (74 to 78) 70 (66 to 75) 72 (69 to 75) 71 (67 to 76) 74 (72 to 76) χ(1)=18.82, χ(3)=10.35, χ(3)=3.55, 2013 (%) 81 (79 to 83) 76 (72 to 79) 77 (75 to 80) 77 (73 to 81) 80 (78 to 81) p<0.001 p=0.016 p=0.31 Increase the risk of 2006, 2008, 2010, 2012 having a heart attack 2 2 2 2011 (%) 92 (91 to 93) 88 (86 to 91) 92 (90 to 93) 88 (85 to 90) 91 (90 to 92) χ(1)=1.87, χ(3)=28.34, χ(3)=1.93, 2013 (%) 91 (90 to 92) 87 (85 to 90) 91 (89 to 93) 87 (84 to 90) 90 (89 to 91) p=0.17 p<0.001 p=0.59 Doubles risk of stroke 2007, 2009, 2011 2 2 2 2011 (%) 83 (81 to 85) 81 (77 to 84) 83 (80 to 86) 81 (76 to 85) 83 (81 to 84) χ(1)=0.13, χ(3)=8.57, χ(3)=2.93, 2013 (%) 84 (82 to 87) 82 (78 to 85) 84 (81 to 86) 82 (77 to 86) 84 (81 to 86) p=0.72 p=0.036 p=0.40 Items grouped according to whether they were used as a warning with the introduction of plain packaging. In total, 104 students had missing data on the smoking status variable and were excluded from analyses. Missing data on the different items over the two surveys ranged from 5 to 74. *Analyses adjusted for covariates of sex, age, school education sector, state and smoking status. CS, smoking in past week/established smokers; ES, experimental smokers; GHW, graphic health warning; NSNS, non-susceptible never-smokers; PP, plain packaging; SNS, susceptible never-smokers.

ii54 White V, et al. Tob Control 2015;24:ii50–ii57. doi:10.1136/tobaccocontrol-2014-052085 Research paper some difference in these items by smoking status, there was no have also found that this increase dissipated after 5 years of significant interaction between survey year and smoking status exposure, with cognitive processing levels in 2011 similar to (table 5). pre-GHW levels.15 The current study suggests that larger Among ES and CS, there was no significant change in the fre- front-of-pack GHWs on plain packs do not increase cognitive quency of not having a cigarette because of the health warnings processing of warnings among adolescents who have already (p=0.52). Among students who had smoked in the previous been exposed to similarly styled GHWs for about 7 years. 12 months, there was no significant change in the frequency of Our findings are in line with previous studies suggesting that thinking about quitting (p=0.28). There was no interaction familiarity may reduce the effectiveness of plain packaging to between year and smoking status for either item (table 5). increase adolescents’ attention to warning labels.712For instance, an experimental study involving Australian adolescents who had DISCUSSION been exposed to 30% front-of-pack GHWs for several years With around 60% of students surveyed seeing cigarette packs in found no effect of packaging or warning size on warning recall.12 the preceding 6 months, GHWs on tobacco packs have the Of the 14 GHWs used on the new packaging, 9 featured a previ- potential to reach a large proportion of adolescents. Our study ously used message with an updated image and 3 featured mes- found the introduction of the new cigarette packaging in late sages and images used previously. The lack of change in the size 2012 did not have an immediate impact on the cognitive pro- of warning labels on the back-of-packs may also have contributed cessing of GHW among adolescents who had seen a cigarette to a sense of familiarity regarding the warnings. All warnings on pack in the previous 6 months. However, we found a significant the new tobacco packaging promoted relatively long-term health increase in the proportion of these adolescents agreeing that consequences of smoking. It has been suggested that warnings smoking causes bladder cancer, a cancer included in the only promoting the more immediate health or social consequences of completely new health warning introduced with the new cigar- smoking may be more salient and relevant to young people.25 ette packaging. This increase is notable given that the warning However, as our previous work has shown an increase in adoles- only started appearing on significant numbers of packs quite late cents’ cognitive processing of warnings after the introduction of in our survey period. GHWs that promoted long-term health effects in 2006 with We have shown previously that 6 months after the introduc- levels remaining elevated in 2008,15 the use of long-term health tion of GHWs in 2006, adolescents’ cognitive processing of effects per se does not limit adolescents’ cognitive processing of warnings had increased from pre-GHW levels.6 However, we warnings, at least in the short term. Whether the use of

Table 5 For students who had seen a cigarette pack in past 6 months, mean frequency of reading, attending to, thinking about and talking about health warnings and for students with smoking experience, the frequency of not having a cigarette or thinking about quitting in response to the warnings in 2011 and 2013, by smoking status (weighted data) Smoking stage F statistic and p values*

NSNS SNS ES CS Total Interaction year and Mean (SE) Mean (SE) Mean (SE) Mean (SE) Mean (SE) Year Smoking status smoking status

2011 (weighted N) (2150) (349) (830) (405) (3733) F(1,183)= F(3,183)= F(3,183)= 2013 (weighted N) (2558) (353) (653) (286) (3851) Read warnings 2011 2.58 (0.04) 2.50 (0.07) 2.51 (0.04) 2.60 (0.06) 2.56 (0.03) 0.03, p=0.87 2.74, p=0.045 1.07, p=0.36 2013 2.57 (0.05) 2.48 (0.07) 2.50 (0.06) 2.59 (0.07) 2.55 (0.04) Paid close attention 2011 2.77 (0.04) 2.59 (0.07) 2.62 (0.05) 2.37 (0.05) 2.68 (0.03) 0.70, p=0.40 5.96, p<0.001 0.83, p=0.48 2013 2.73 (0.05) 2.55 (0.07) 2.58 (0.06) 2.32 (0.06) 2.66 (0.05) Thought about warnings 2011 2.62 (0.04) 2.40 (0.05) 2.53 (0.04) 2.43 (0.06) 2.56 (0.03) 1.91, p=0.17 4.64, p=0.004 0.41, p=0.75 2013 2.68 (0.04) 2.46 (0.06) 2.59 (0.05) 2.49 (0.07) 2.63 (0.04) Talk about warnings 2011 2.06 (0.03) 1.92 (0.05) 2.05 (0.04) 2.12 (0.05) 2.05 (0.02) 0.34, p=0.56 4.66, p=0.003 1.47, p=0.22 2013 2.03 (0.04) 1.89 (0.06) 2.02 (0.04) 2.09 (0.06) 2.02 (0.03) Not had a cigarette 2011 ––2.60 (0.06) 1.65 (0.06) 2.29 (0.04) 0.042, p=0.52 161.98, p<0.001 1.27, p=0.26 2013 ––2.65 (0.09) 1.69 (0.08) 2.35 (0.08) Among students who had smoked in previous 12 months Thought about quitting 2011 (weighted N) (437) (364) (801) F(1,175)= F(1,175)= F(1,175)= 2013 (weighted N) (403) (285) (688) 2011 ––2.58 (0.09) 2.49 (0.06) 2.53 (0.06) 1.16, p=0.28 0.67, p=0.42 0.94, p=0.33 2013 ––2.47 (0.08) 2.38 (0.07) 2.42 (0.06) Responses measured on a five-point scale from 1 ‘never’ to 5 ‘every time I see them’. In total, 104 students had missing data on the smoking status variable and were excluded from analyses. Missing data on the different items over the two surveys ranged from 37 to 131. *Analyses adjusted for covariates of sex, age, school education sector, state and smoking status. CS, smoking in past week/established smokers; ES, experimental smokers; NSNS, non-susceptible never-smokers; SNS, susceptible never-smokers.

White V, et al. Tob Control 2015;24:ii50–ii57. doi:10.1136/tobaccocontrol-2014-052085 ii55 Research paper immediate health effects or social consequences of smoking in message also included in this new warning. The lack of change GHWs would maintain higher levels of cognitive processing over may be due to the item referring to kidney disease rather than the long term is not known. kidney cancer specifically. However, it may also reflect that the An eye-tracking study has shown that adult smokers spend messages about smoking causing kidney and bladder cancer were less time looking at warnings on plain packs than non-smokers26 conveyed in one combined warning. As the text relating to with a recent eye-tracking study suggesting this is due to bladder cancer was closest to the image used in this GHW,it smokers attempting to avoid warnings.27 While the lack of could be speculated that this aspect of the warning was more change in our cognitive processing indicators may reflect an prominent to adolescents. The item ‘smoking causes diseases of avoidance response stimulated by the larger GHWs, as cognitive the teeth and gums’ did not change, with high levels of agree- processing in 2013 was the same as in 2011 when warnings ment in 2011 and 2013. The lack of change may be due to the covered only 30% of the front-of-pack, our results suggest that similarity of the image used for this GHW and that used for the the new packaging did not increase avoidant responses. GHW ‘smoking causes mouth cancer’ between 2006 and 2012. Field experiments with adult smokers in the UK and France A mouth cancer message additionally featured in a media cam- have found that plain packaging increases reading and thinking paign between 2006 and 2012. It is likely that adolescents under- about warning labels,28 29 while not changing the believability stood the earlier health warning message and image to imply that and noticeability29 or credibility28 of GHWs. Thinking about gums and teeth would be affected by mouth cancer. warnings was the only cognitive processing item in our study to Several limitations should be noted. The study compares increase (although not significantly) between 2011 and 2013. It responses from adolescents participating in two cross-sectional may be that in environments where GHWs on tobacco products surveys and thus does not determine whether the introduction are the norm, plain packaging may first stimulate adolescents to of the new tobacco packaging has changed individuals’ knowl- think a little more about the warning labels they notice. If this is edge or cognitive processing. However, by ensuring that the stu- the case, the non-significant increase we found may indicate that dents taking part in our two cross-sectional surveys were drawn this change was at an early stage when the 2013 survey was con- from the same population and were generally similar on key ducted suggesting that the increase may strengthen with longer demographic characteristics, our design provides information exposure to the new tobacco packaging. However, further regarding whether responses among the target population research is needed to confirm this suggestion. changed over time, which was the aim of the study. We exam- The lack of significant change in adolescents’ cognitive process- ined students’ self-reported behaviours when seeing cigarette ing of warning labels is in contrast to findings from several studies packs rather than assessing actual behaviours. We also examined of Australian adult smokers. Australian population-based studies prompted recall of health consequences of smoking rather than of adult smokers have found the proportion of adult smokers unprompted recall. Unprompted recall generally elicits much having strong cognitive, emotional and avoidant responses to lower estimates of awareness of health risks of smoking and pro- GHWs on cigarette packs increased after the introduction of the vides a more sensitive indication of salient ‘top of mind’ aware- new plain tobacco packaging.30 31 Wa ke field et al31 (this volume) ness. However, as the two surveys used the same questions, found that these effects persisted up to a year after the packaging measurement error associated with survey items should be changes, and Brennan et al32 (this volume) found cognitive and similar across surveys. In addition, it may be that the regular avoidant responses to health warnings on plain packs predicted exposure to messages regarding the many different harms of quitting thoughts, intentions and attempts 1 month later. The dif- smoking through both mass media tobacco control campaigns ferences in these findings and those for the current study may and previous health warnings has created a general belief that reflect the greater involvement adult smokers have with smoking smoking causes many, many diseases which leads students to compared with adolescents. With their higher daily smoking rate, respond positively when presented with a list of diseases. Our adult smokers access and use cigarette packs far more frequently follow-up survey was conducted 7–12 months after the intro- than most adolescents. Many adult smokers want to quit33 and duction of the new tobacco packaging and this may have meant use GHWs for motivation for doing so.34 there was insufficient time for the new packaging to influence There was a very high level of prompted agreement with adolescents. However, an effect on adolescents’ cognitive pro- most of the health consequences in 2011, with levels remaining cessing of warnings was found 6–9 months after the introduc- high in 2013. This likely reflects that most of the health conse- tion of GHWs in 2006.6 Change in consent procedures in one quences assessed had been used in health warnings and/or state meant that some schools surveyed only a small number of tobacco control advertising campaigns since 2006, with some students and this could have introduced some bias. However, warnings used since 1995. There were, however, increases in sensitivity analyses suggested that the change in consent proce- the proportion of students agreeing that ‘smoking is a leading dures in one state had minimal influence on results. cause of death’ and ‘smoking causes blindness’ statements Despite these limitations, the present study provides new infor- reflecting two warnings in use since 2006 and 2007, respect- mation on the immediate impact of adolescents to larger GHWs ively. An increase in the proportion of students agreeing that on standardised cigarette packs. Our data suggest that when the ‘smoking causes blindness’ is encouraging as a previous study of style and content of most warning messages are familiar to the Australian students found no change in agreement levels population, the introduction of plain packaging and larger warn- between 2005 (50%), 2008 (53%) and 2011 (52%).35 Whether ings does not induce adolescents to attend to and process warn- this increase is due to the new tobacco packaging is not clear, as ings on cigarette packs to a greater extent than when GHWs this health warning was used in a number of tobacco control covered 30% of the front of a fully branded pack. However, the mass media campaigns between 2011 and 2013. introduction of larger GHWs and plain packaging has reduced We found an increase in prompted knowledge of one health positive cigarette pack image and positive brand characteristics consequence of smoking that most closely reflected a completely among adolescents (this volume).18 That finding, coupled with new GHW introduced with the new packaging—smoking causes those presented here, suggests that the immediate effect of the bladder cancer. However, we did not find a change in the propor- new tobacco packaging on adolescents has been to reduce pack tion of students agreeing that smoking causes ‘kidney disease’,a appeal and to reduce perceived positive brand characteristics. ii56 White V, et al. Tob Control 2015;24:ii50–ii57. doi:10.1136/tobaccocontrol-2014-052085 Research paper

Continued monitoring of adolescents’ responses to GHWs on the 6 White V, Webster B, Wakefield M. Do graphic health warning labels have an impact on new cigarette packs in Australia would show whether adolescents’ adolescents’ smoking-related beliefs and behaviours? Addiction 2008;103:1562–71. 7 Beede PL, Lawson R. The effect of plain packages on the perception of cigarette cognitive processing of GHWs increases with longer exposure. health warnings. Public Health 1992;106:315–22. 8 Scheffels J, Saebo G. Perceptions of plain and branded cigarette packaging among Norwegian youth and adults: a focus group study. Nicotine Tob Res 2013;15:450–6. 9 Stead M, Moodie C, Angus K, et al. Is consumer response to plain/standardised What this paper adds tobacco packaging consistent with framework convention on tobacco control guidelines? A systematic review of quantitative studies. PLoS ONE 2013;8:e75919. ▸ 10 Goldberg ME, Liefeld J, Madill J, et al. The effect of plain packaging on response to While the introduction of graphic health warning (GHW) in health warnings. Am J Public Health 1999;89:1434–5. Australia in 2006 was associated with an increase in 11 Maynard OM, Munafo MR, Leonards U. Visual attention to health warnings on adolescents’ cognitive processing of warnings, by 2011 plain tobacco packaging in adolescent smokers and non-smokers. Addiction cognitive processing levels had returned to pre-GHW levels. 2013;108:413–19. fi ’ ▸ Along with the introduction of plain packaging for tobacco 12 Germain D, Wake eld MA, Durkin SJ. Adolescents perceptions of cigarette brand image: does plain packaging make a difference? J Adolesc Health 2010;46:385–92. products in December 2012, Australia increased the size of 13 Borland R. Tobacco health warnings and smoking-related cognitions and front-of-pack GHWs to 75%. While qualitative studies suggest behaviours. Addiction 1997;92:1427–35. plain packaging increases the noticeability, believability and 14 Hammond D, Fong GT, McDonald PW, et al. Impact of the graphic Canadian seriousness of health warnings among adolescents, the warning labels on adult smoking behaviour. Tob Control 2003;12:391–5. 15 White V, Bariola E, Faulkner A, et al. Graphic health warnings on cigarette packs: evidence from experimental studies is mixed. how long before the effects on adolescents wear out? Nicotine Tob Res Published ▸ This study which compares data from surveys of Australian Online First: 19 Sep 2014. doi:10.1093/ntr/ntu184 adolescents conducted in 2011 and 2013 showed no change 16 Bouton ME. Learning and behavior: a contemporary synthesis. Sunderland, in adolescents’ cognitive processing of warnings with the Massachusetts: Sinauer Associates, Inc. Publishers, 2007. introduction of new cigarette packaging in late 2012. 17 Kim SW, Wogalter MS. Habituation, dishabituation, and recovery effects in visual warnings. Proc Hum Fact Ergon Soc Annu Meet 2009;53:1612–16. 18 White V, Williams T, Wakefield M. Has the introduction of plain packaging with larger graphic health warnings changed adolescents’ perceptions of cigarette packs Acknowledgements The authors would like to acknowledge the work of Ms Emily and brands? Tob Control 2015;24:ii42–9. Bariola who co-ordinated the national study in 2011. Dr Kerri Coomber provided 19 McKenzie PW, Weldon PR, Rowley G, et al. ACER. Staff in Australia’s schools 2013: advice on statistical analysis. They would like to thank and acknowledge the main report on the survey. In: Education Do, ed. Canberra, ACT: Australian education authorities in participating states and the principals and contact teachers in Government, 2014. participating schools for allowing us to conduct the survey in schools and for assisting 20 White VM, Hill DJ, Effendi Y. How does active parental consent influence the with survey arrangements. We thank the students for participating in the surveys. findings of drug-use surveys in schools? Eval Rev 2004;28:246–60. Contributors VW and MW designed the study. VW undertook analysis and drafted 21 Pierce JP, Choi WS, Gilpin EA, et al. Validation of susceptibility as a predictor of which the manuscript. TW co-ordinated data collection 2013 and AF assisted with data adolescents take up smoking in the United States. Health Psychol 1996;15:355–61. collection in 2011. TW and AF co-ordinated data processing. All authors reviewed 22 Mayhew KP, Flay BR, Mott JA. Stages in the development of adolescent smoking. – and approved the final manuscript. Drug Alcohol Depend 2000;59(Suppl 1):S61 81. 23 Australian Bureau of Statistics. Schools Australia. 2014. http://www.abs.gov.au/ Funding Data used in this study were gathered from surveys funded fully or in part AUSSTATS/[email protected]/DetailsPage/4221.02013?OpenDocument by the Australian Government Department of Health. Cancer Councils and health 24 StataCorp. Stata statistical software: release 8.2. College Station, TX: Stata departments of participating states also contributed funding for the 2011 surveys. Corporation, 2005. Competing interests The authors wish to advise that MW was a member of the 25 Hoek J, Hoek-Sims A, Gendall P. A qualitative exploration of young adult smokers’ Tobacco Working Group of the Australian National Preventive Health Task Force and the responses to novel tobacco warnings. BMC Public Health 2013;13:609. Expert Advisory Committee on Plain Packaging that advised the Australian Department of 26 Munafo MR, Roberts N, Bauld L, et al. Plain packaging increases visual attention to Health on research pertaining to the plain packaging legislation. VW holds competitive health warnings on cigarette packs in non-smokers and weekly smokers but not grant funding from the Victorian Cancer Agency and the National Breast Cancer daily smokers. Addiction 2011;106:1505–10; discussion 11–12. Foundation, VW and MW hold such funding from the Australian National Health and 27 Maynard OM, Attwood A, O’Brien L, et al. Avoidance of cigarette pack health Medical Research Council and MW holds such funding from the US National Institutes of warnings among regular cigarette smokers. Drug Alcohol Depend 2014;136:170–4. Health, Australian National Preventive Health Agency and BUPA Health Foundation. 28 Gallopel-Morvan K, Moodie C, Eker F, et al. Perceptions of plain packaging among young adult roll-your-own smokers in France: a naturalistic approach. Tob Control Ethics approval The study had ethical approval from the institution of the lead Published Online First: 11 June 2014 doi:10.1136/tobaccocontrol-2013-051513 author. State education authorities approved the conduct of the study in both years. 29 Moodie CS, Mackintosh AM. Young adult women smokers’ response to using plain Provenance and peer review Not commissioned; externally peer reviewed. cigarette packaging: a naturalistic approach. BMJ Open 2013;3:e002402. 30 Dunlop SM, Dobbins T, Young JM, et al. Impact of Australia’s introduction of Open Access This is an Open Access article distributed in accordance with the tobacco plain packs on adult smokers’ pack-related perceptions and responses: Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which results from a continuous tracking survey. BMJ Open 2014;4:e005836. permits others to distribute, remix, adapt, build upon this work non-commercially, 31 Wakefield M, Coomber K, Zacher M, et al. Australian adult smokers’ responses to and license their derivative works on different terms, provided the original work is plain packaging with larger graphic health warnings 1 year after implementation: properly cited and the use is non-commercial. See: http://creativecommons.org/ results from a national cross-sectional tracking survey. Tob Control 2015;24:ii17–25. licenses/by-nc/4.0/ 32 Brennan E, Durkin S, Coomber K, et al. Are quitting-related cognitions and behaviours predicted by proximal responses to plain packaging with larger health REFERENCES warnings? Findings from a national cohort study with Australian adult smokers Tob 1 Trade Practices (Consumer Product Information Standards) (Tobacco) Regulations, Control 2015;24:ii33–41. Stat. Statutory Rules 2004 no.264. (29 March 1994, 2004). http://www.comlaw. 33 Australian Institute of Health and Welfare. 2010 National Drug Strategy Household gov.au/Details/F2007C00131 Survey report. Canberra, Australia: Australian Institute of Health and Welfare, 2011. 2 Tobacco Plain Packaging Act 2011. 2011. http://www.comlaw.gov.au/Details/ Drug statistics series No: 25. Cat. no. PHE 145. http://www.aihw.gov.au/ C2011A00148 publication-detail/?id=32212254712 3 Australian Government. Competition and Consumer (Tobacco) Information Standard 34 Borland R, Yong HH, Wilson N, et al. How reactions to cigarette packet health 2011- F2012C00827 Canberra, Australia. 2012. http://www.comlaw.gov.au/Details/ warnings influence quitting: findings from the ITC Four-Country survey. Addiction F2012C00827 2009;104:669–75. 4 Scollo M, Lindorff K, Coomber K, et al. Standardised packaging and new enlarged 35 White V, Johnson K. Tobacco use in the social environment of Victorian secondary graphic health warnings for tobacco products in Australia—legislative requirements school students and their awareness of health effects of smoking in 2011. In: White and implementation of the Tobacco Plain Packaging Act 2011 and the Competition V, Bariola E, eds. Victorian secondary school students’ use of licit and illicit and Consumer (Tobacco) Information Standard, 2011. Tob Control 2015;24:ii9–16. substances in 2001: results for the 2011 Australian Secondary Students’ Alcohol 5 McCool J, Webb L, Cameron LD, et al. Graphic warning labels on plain cigarette and Drug (ASSAD) Survey. Melbourne, Australia: Centre for Behavioural Research in packs: will they make a difference to adolescents? Soc Sci Med 2012;74:1269–73. Cancer, Cancer Council of Victoria, 2012:96–120.

White V, et al. Tob Control 2015;24:ii50–ii57. doi:10.1136/tobaccocontrol-2014-052085 ii57 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper “You’re made to feel like a dirty filthy smoker when you’re not, cigar smoking is another thing all together.” Responses of Australian cigar and cigarillo smokers to plain packaging Caroline L Miller,1,2 Kerry A Ettridge,1 Melanie A Wakefield3

1SAHMRI Population Health ABSTRACT cigar tubes must display text only warnings.2 For Research Group, South Objective To explore experiences of cigar and cigarillo further details see Scollo et al.3 Australian Health and Medical Research Institute, Adelaide, smokers under Australian laws requiring plain packaging In Australia, cigars have been subject to the same South Australia, Australia (PP) and strengthened graphic health warnings (GHWs). marketing restrictions as other types of tobacco with 2School of Population Health, Methods In February/March 2014, we conducted: prohibitions on advertising in mainstream broadcast University of Adelaide, in-depth interviews with 10 regular premium cigar media, retail displays and, in 2012, regulation of Adelaide, South Australia, smokers; two focus groups with occasional premium cigar online ‘point-of-sale’ for Australian-based providers. Australia 3Centre for Behavioural and premium cigarillo smokers (n=14); four focus groups The prevalence of cigar smoking in Australia is low Research in Cancer, Cancer with non-premium cigarillo smokers (n=28); and a and has not risen in recent decades. It is much lower Council Victoria, Melbourne, national online survey of cigar and/or cigarillo smokers than other forms of tobacco smoking. The latest Victoria, Australia (n=268). national Australian data from 2010 found the com- Correspondence to Results Premium cigar smokers had limited exposure to bined percentage of survey participants smoking Dr Caroline L Miller, SAHMRI PP, with many purchasing fully branded cigars in boxes cigars and/or pipes among those who were smokers Population Health Research duty free or online and singles in non-compliant (aged 14+ years) was 8%, or approximately 1% of Group, South Australian Health packaging. Those who were exposed noticed and were all participants.4 and Medical Research Institute, concerned by the warnings, tried to avoid them and felt Despite strong regulation of cigar promotion PO Box 11060, Adelaide, ‘ ’ SA 5061, Australia; more like dirty smokers . Changes in perceived taste, within Australia, Australian legislation does not cover [email protected] harm and value were minimal for experienced premium promotion of cigars and other forms of tobacco in cigar smokers. Occasional premium cigar and premium imported magazines, online marketing from provi- Received 24 September 2014 cigarillo smokers with higher PP exposure (gained by ders outside Australia, product placement in movies, Accepted 5 December 2014 purchasing boxes rather than singles) perceived cigar/ or promotion via social media, all of which expose package appeal and value had declined and noticed the the Australian market to cigar promotion. GHWs. Non-premium cigarillo smokers reported high PP Many international jurisdictions, including the exposure, reduced perceived appeal, quality, taste, USA, have fewer restrictions than Australia on tobacco enjoyment and value, somewhat increased perceived marketing, including cigars. The rise in the popularity harm, greater noticeability of GHWs and concealment of of cigar smoking in the USA and elsewhere has led to packs and more contemplation of quitting. Online survey a focus on the influencing factors, including the participants reported increased noticeability of GHWs increased frequency of cigar content in popular media (33%), decreased appeal of packaging (53%) and and the ways in which cigars have been portrayed.5 reduced consumption of cigars (42%) and cigarillos A number of studies observed increases in cigar mar- (44%) since PP implementation. keting starting in the 1990s, increasing the visibility Conclusions Non-premium cigarillo smokers appear to and normalisation of cigar smoking, with associated – have been most exposed and influenced by PP, with cigar increases in prevalence of use.6 10 Cigar marketing smokers less so, especially regular premium cigar smokers proliferated in print media, social media and other who have maintained access to fully branded products. online platforms. Cigars have been consistently posi- tioned as a luxury good, a status symbol and a sign of affluence and power, and as part of a successful, – sophisticated lifestyle.11 13 Cigars have been pro- INTRODUCTION motedwitharangeofhistoricalandcontemporary fi 611–13 Open Access Australian legislation requiring all tobacco products celebrities and public gures. Analyses of print Scan to access more to be sold in drab dark brown plain packaging media content have also consistently found that the free content from 1 December 2012, also includes specific health effects of cigars have been downplayed and requirements for cigars and cigarillos.1 Like cigar- minimised.12 ettes, the new provisions for cigar and cigarillo In 2011, qualitative research examined the boxes and packs, and bags for packaging of single smoking behaviour, attitudes and perceptions of cigars for sale prohibit logos, brand imagery and Australian cigar smokers to assist development of design. They require standardised package colour plain packaging provisions and GHWs.14 Australian and standardised font colour and size for brand cigar smokers reported viewing cigar smoking as a To cite: Miller CL, and variant names. All cigar packaging, other than luxury and an indulgence, with higher status attribu- Ettridge KA, Wakefield MA. cylindrical tubes in which cigars can be packaged ted to premium cigars. Tob Control 2015;24: for sale, are required to display expanded and This research identified differences between cig- – ii58 ii65. updated graphic health warnings (GHWs), while arillo smokers and lower and higher frequency ii58 Miller CL, et al. Tob Control 2015;24:ii58–ii65. doi:10.1136/tobaccocontrol-2014-052049 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper consumers of premium cigars. More frequent premium cigar all qualitative study participants completed a consent form prior smokers were found to be extremely knowledgeable about the to participation. different types of cigars, regularly smoking different cigar We first conducted in-depth interviews of 40–50 min duration brands. These cigar smokers were also reported to judge the with premium cigar smokers (n=10), defined as those who quality of a cigar on criteria including: the type of tobacco; the smoked a premium cigar (costing $A25+) at least once a week. roll; the age; and the country and region of origin. The brand We then ran two sets of focus groups. The first set consisted of name was taken as a sign of authenticity or legitimacy of the occasional (fortnightly or monthly) premium cigar smokers product. This information was reported to have been included and/or premium cigarillo smokers (2 groups, total n=14). The on the cigar band where more frequent cigar smokers looked second set included non-premium cigarillo smokers (4 groups, for these indications of quality. total n=28). All focus groups were 1 h in duration. All qualita- Less frequent smokers of premium cigars reportedly viewed tive research was conducted in Sydney and Melbourne in cigar smoking as an occasional pleasure, associated with particu- February 2014. lar social occasions or celebrations. They were less knowledge- Finally, in March 2014 we conducted an online survey (5– able than more frequent cigar smokers about premium cigars. 10 min duration) of a general national sample of current cigar These cigar smokers reported taking a stronger interest in the and cigarillo smokers (n=268). This sample was recruited from branding and packaging of cigars than did more frequent cigar an existing national online panel who had expressed their will- smokers. The branding and packaging was reported to have a ingness to be contacted for the purpose of research. This panel strong effect on their perceptions of quality. Similar patterns develops their database from a number of sources including were also reported for smokers of premium and non-premium advertising and ‘word-of-mouth’. A total of 56 589 email invita- cigarillos, with smokers of non-premium cigarillos reporting tions describing the nature of the study (ie, to gather information being more likely to associate appeal and quality with packaging. regarding people’s views and experiences about smoking cigars The current research sought to assess, among different seg- and other tobacco products) were sent out to randomly selected ments of cigar consumers: (1) current cigar and cigarillo members of the online panel, of which 5761 started the survey smoking behaviour and consumers’ exposure to new packaging (response rate of 10%). Only 283 of these participants (ie, 5% of changes; (2) perceived changes relating to the specific objectives the people who started the survey) met the eligibility criteria as of the plain packaging legislation, namely, reduced appeal, assessed by screening questions at the beginning of the survey, increased GHW effectiveness and reduced extent to which con- that is, they were aged 18 years or older and reported that they sumers are misled about harms; and (3) more ‘downstream’ per- currently smoked either cigars and/or cigarillos, with a further 15 ceived changes in smoking behaviours and thoughts since the excluded due to incomplete responses, leaving 268 participants implementation of plain packaging. in the sample.

Research materials METHODS Semistructured discussion guides were used for the qualitative Study design and participants studies, comprised of: general introduction; existing smoking and Owing to the low population prevalence of cigar smoking and purchasing behaviour; relationships with brands (if any); views on diverse demographic composition of segments of cigar smokers plain packaging; reactions to GHWs; and intentions to quit. in Australia, this research employed a mixed methods design Participant self-completion forms including pictorial examples of with qualitative and quantitative components to maximise plain packaging and GHWs were used as stimulus material. capture and engagement of different segments of cigar consu- The online survey used questions adapted from existing mers. The methods used in the qualitative studies were closely tobacco control monitoring surveys where available and new, aligned to pre-plain packaging qualitative research and had survey-specific questions where necessary, with response options proved successful with engaging different segments of cigar informed by the qualitative research. Questions addressed smokers.14 Based on this previous research experience,14 a focus tobacco use, purchasing behaviour, cigar/cigarillo specific group methodology was not considered to be a viable approach smoking behaviour, brand choice and preferences, recall and to recruit and/or engage regular premium cigar smokers; a perceptions of plain packaging and GHWs on cigars/cigarillos; group which is quite distinct from other cigar/cigarillo smokers and self-perceived changes in beliefs and behaviour since the in that they were of higher socioeconomic status and placed implementation of plain packaging. higher value on their time. However, collecting data via struc- tured interviews had been successful with this group. Thus, Analyses structured interviews were employed for the regular premium Two researchers moderated the focus groups and conducted the cigar smokers, and group discussions were employed for the interviews. A thematic analysis model was used to analyse the other cigar and cigarillo smokers. An online survey enabled qualitative data. The two researchers on the project contributed quantitative data collection from a substantial general sample of to this analysis by independently referring to the data sources cigar and cigarillo smokers. (transcripts, notes and the self-completion forms). Each devel- Participants in the qualitative studies were recruited through oped and tested hypotheses based on the interviews and group an accredited recruitment specialist with access to databases of discussions they conducted, and identified emergent themes and individuals who had already indicated a willingness to be con- patterns. Following this, these researchers participated in several tacted for the purpose of research, using Interviewer Quality debriefing sessions to discuss the analysis and results, to reach Control Australia (IQCA) standards. Current smokers were iden- consensus on potential themes using all the data relevant to tified from the database, and were administered a recruitment each and to identify frameworks for further in-depth analysis. screening survey either online or by telephone. Eligible partici- On-going analysis and revision of data sources was conducted to pants were encouraged to pass on the details of the study to ensure coding consistency, to refine the specifics of each theme, others who fit the selection criteria. All studies included partici- to strengthen within-theme coherence, to ensure evidence of pants who were 18 years or older and proficient in English and saturation and to ensure the credibility of the data with the

Miller CL, et al. Tob Control 2015;24:ii58–ii65. doi:10.1136/tobaccocontrol-2014-052049 ii59 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper research objectives. This was an iterative process that continued end of working week, or to have some ‘time out’. Owing to the throughout the writing of the results. cost of the cigars, cigar smoking was also perceived to provide a Data from the online survey were analysed using IBM SPSS sense of exclusivity, prestige and sophistication. Statistics V.22. For continuous variables, means and SDs were calculated, and frequencies are reported for categorical data. We do poker nights, it’s a boys club – meet weekly. The host will Small cell sizes prevented more complex analyses. have a good bottle of scotch or bourbon. It began with a few of the boys who were settling down and getting married – they wanted to keep up the lifestyle with no family around. RESULTS Table 1 summarises the samples for each of the three studies, Smoking cigars is more of a treat for me at the end of a week. Or and table 2 summarises the results for each study. if I’ve had a busy day and I just want to go home and relax.

In-depth interviews of regular premium cigar smokers There was a general perception that the lower frequency of Of the Regular (at least weekly) premium cigar smokers, half cigar smoking compared to cigarette smoking, resulted in a low also smoked cigarettes. Among these dual-product smokers, relative risk of health concerns. each product was claimed to fulfil a different need. Cigar I’ve always believed, and I could be entirely wrong, that cigars smoking was considered to be more a choice and an enjoyment, aren’t as bad for you as cigarettes. while cigarette smoking was considered more of an addiction fi and ful lling a physiological need. I would never have thought lung cancer. I would have related that to inhaling the smoke which I don’t do. Cigarettes have become more of a habit and just satisfying an urge. Cigar smoking is about the smoking… about that very rich The longer-term cigar smokers bought cigars by the box via taste. planned online and duty free purchases; whereas more occasion- For most, cigar smoking was driven by a specific, regular based smokers made opportunistic purchases, usually from a occasion, for example, ‘whisky’ or ‘card’ nights with friends, retailer or tobacconist. All regular premium cigar smokers had a

Table 1 Characteristics of samples Qualitative Quantitative 1. Regular premium 2. Occasional premium cigar/ 3. Non-premium cigarillo 4. Cigar and cigarillo cigar smokers (n=10) premium cigarillo smokers (n=14) smokers (n=28) smokers (n=268) (%)

Gender Female 2 2 8 48 Male 8 12 20 52 Age group 18–24 – 3312 25–34 1 5 8 29 35–44 3 2 5 19 45–54 3 3 9 24 55+ 3 1 3 16 Frequency of smoking cigar/cigarillos Daily 2 1 3 9* At least weekly 8 1 9 14* Fortnightly (qualitative)/less – 4 9 71* than weekly (quantitative) Monthly – 64NA Less than monthly – 23NA Packaging purchased Boxed or mostly boxed 4 3 15 55*† Single or mostly single 4 2 9 60*† Both equally 2 9 4 NA Tobacco smoked (multiple response)‡ Cigarettes (factory made) 5 8 19 79 Roll your own cigarettes – 31062 Cigarillos 2 6 28 44 Cigars 10 12 14 94 1. Regular premium cigar smokers: currently smoking premium cigars (>$25) at least weekly; 2. Occasional premium cigar/premium cigarillo smokers: currently smoking premium cigars fortnightly or monthly and/or currently smoking premium cigarillos (any frequency); 3. Non-premium cigarillo smokers: currently smoking non-premium cigarillos (any frequency); and 4. Cigar and cigarillo smokers: currently smoking cigars and/or cigarillos (any frequency). *For the purpose of this table, frequency of smoking cigars (and not cigarillos) was reported for online survey participants, as frequency of cigarillo consumption was reported separately. †Categories were not mutually exclusive for online survey results. ‡For qualitative interviews and groups, tobacco smoked was assessed for each form of tobacco by yes/no response; for the quantitative study respondents were asked to indicate the frequency with which they smoked each product (ie, daily, weekly or less than weekly), with any frequency of smoking reported for the purpose of this item. NA, not assessed for online survey participants. ii60 Miller CL, et al. Tob Control 2015;24:ii58–ii65. doi:10.1136/tobaccocontrol-2014-052049 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 2 Summary of results across the four participant groups Qualitative Quantitative 2. Occasional premium cigars and 4. General sample of cigar or 1. Regular premium cigars premium cigarillo smokers 3. Non-premium cigarillos cigarillo smokers (interviews) (groups) (groups) (n=268)

Perceived impact of plain packaging Exposure ▸ Half seen; half unaware Mixed, low in young purchasing High exposure ▸ Recall of any cigar GHWs ▸ Purchasing from many singles, higher for boxes (50%; ±6% CI) channels ▸ Last purchase was compliant with plain packaging and/or GHWs (46%; ±6% CI) ▸ Smoked only compliant in past 6 months (48%; ±6% CI) Appeal of Reduced if exposed (for box and Reduced among those few exposed Greatly reduced ▸ 53% lower packaging appeal packaging bag) (±6% CI) (12% higher; 35% same) ▸ 28% lower product appeal (±5% CI) (12% higher; 60% same) Perceived quality Minimal change, relied less on Unsure, packaging is a cue for quality Reduced 16% lower (±4% CI) pack (15% higher; 69% same) Changes in taste, Minimal Minimal for experienced smokers Reduced 19% lower (±5% CI) enjoyment (15% higher; 66% same) Perceived harm Low Lower than cigarettes Somewhat increased 19% higher (±5% CI) (15% lower; 66% same) Perceived value Unchanged (less price sensitive) Reduced, pack part of value Reduced, noted tax increased 41% lower (±6% CI) too (18% higher; 41% same) Notice Warnings High if exposed, especially High if exposed, especially graphics High, especially graphics 33% more often (±6% CI) graphics (16% less often, 43% same) Reported ▸ Decant from box ▸ Decant from box Conceal packs, less brand Consumption changes since 2 years behavioural ▸ Retail purchasing made them ▸ Less brand trialling trialling, contemplating quitting ago: changes feel closer to “dirty smokers” ▸ Cigars: 42% less; 13% more; 45% same ▸ Cigarillos: 44% less; 15% more; 42% same Deliberately concealed or decanted: ▸ 11% more often (±4% CI) (21% less often; 56% same) 1. Regular Premium Cigar smokers: currently smoking premium cigars (>$25) at least weekly; 2. Occasional premium cigar/premium cigarillo smokers: currently smoking premium cigars fortnightly or monthly and/or currently smoking premium cigarillos (any frequency); 3. Non-premium cigarillo smokers: currently smoking non-premium cigarillos (any frequency); and 4. Cigar and cigarillo smokers: currently smoking cigars and/or cigarillos (any frequency). GHWs, graphic health warnings. preferred brand; however, trying new brands was part of the smokers and counteracted the feelings of reward, prestige and experience. sophistication that were often experienced during the purchase of a cigar. When purchasing from places other than tobacco- I change brands for something different. I might buy a more nists, there was some concern that they would be subject to the expensive brand as a treat. same social stigma that was directed towards smokers. While brand name and quality packaging contributed to the cigar smoking experience, they reportedly did not shape per- ceived quality among these experienced premium cigar smokers. Yo u ’re made to feel like a dirty filthy smoker when you’re not, ’ Half of premium cigar smokers reported that they had been cigar smoking is another thing all together. It s not like cigarettes exposed to cigar plain packaging, with the others unaware that and it puts it in the same category. It makes a cigar smoker feel like they are doing something wrong, that it’s a dirty habit like the legislation applied to cigars. Those exposed indicated that smoking. plain packaging and GHWs on the boxes had reduced the appeal of the cigar box. Perceptions of the actual product and experience were not reported to have changed, as they did not Those exposed to plain packaging and GHWs via cigar boxes attribute the quality of the product to the packaging. Consistent no longer left the box where it could be seen by others and with this, there was also no reported perceived change in value those with a humidor decanted individual cigars to the humidor for money. whereas some had previously placed the whole box in the All the health warnings on it, it’s designed to entice you not to humidor. smoke. When you open the box, it’s like a box of goodies! … At the end of the day, that’s what you want! I keep them in the box in the drawer...I remember in the old The GHWs on bags (required by health warning legislation days you’d get the big cigars with the nice flash boxes, they’dbe for single cigar purchases) had reportedly detracted from the all colourful with no warnings on it, and then you’d keep the experience of purchasing singles among these premium cigar box out.

Miller CL, et al. Tob Control 2015;24:ii58–ii65. doi:10.1136/tobaccocontrol-2014-052049 ii61 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Focus group discussions with occasional premium cigar Exposure to plain packaging and GHWs was greater among smokers and premium cigarillo smokers the more experienced occasional cigar/cigarillo smokers who This group of occasional (less than fortnightly/monthly) purchased cigars or cigarillos by the box, or who were purchas- premium cigar and premium cigarillo smokers was comprised of ing multiple singles at a time. younger people relatively new to smoking premium cigars and Instead of buying a nice quality, beautiful cigar, you get some cigarillos, as well as older smokers who were quite similar in health warning that reminds you that if you smoke this you’re reported smoking situations to regular premium cigar smokers. gonna die. I guess it does take away from the enjoyment a little bit. Among younger cigar/cigarillo smokers, smoking behaviour was reportedly largely opportunistic. These participants reported that they had made changes to how they stored their cigars since the implementation of plain ’ If guys have them at parties and they offer you one. (I ll have one) packaging, that is, those with a humidor decanted individual (At the bar) Normally they have four options – I always get the cigars to the humidor rather than the whole box. fl ’ chocolate avoured one. They re the best! I always move the cigars out of the box into my humidor. I used ’ Smoking cigars/cigarillos occurred in a more public environ- to keep the boxes cause they were quite cool. ment as a sign of symbol or status and the product and pack- Now you don’t want to look at the boxes – they’re just aging of cigarillos was considered to establish the smokers as disgusting! different to an average ‘cigarette’ smoker. Older occasional premium cigar/cigarillo smokers reported behaviour very similar Lack of exposure to plain packaging and relatively limited to regular premium cigar smokers, that is, smoking cigars at a experience and knowledge of cigars and cigarillos meant that regular social activity, an impending occasion or by themselves younger occasional smokers of these products were limited in at home, and they usually had a preferred brand. They also asso- their capacity to discuss their perceptions of markers of quality ciated a sense of exclusivity, prestige and sophistication with of cigars. However, the discussion in the groups indicated that smoking cigars. packaging was often a key feature on which they rely to make decisions about the quality of the product they were consuming. (Cigars have) a refined image. Bit of a boss look. Prestige, … wealthy. Packaging if it looks nice, you think the cigar will be nicer. I want to pay for good packaging. I want to get excited about it! I think there’s an element of perceived class with cigars and cigar- ’ illos. Whereas you re almost a junkie if you smoke cigarettes! Now when you go to the tobacconist and it’s all in that pack- aging, you don’t know what to pick and it’s difficult to go with a It’s cooler than a cigarette. particular brand based on its prestige and the way it looks. Many of the smokers within this sample had limited knowl- I hate the packaging. I can’t stand it! edge about how to determine the quality of the cigars compared with regular premium cigar smokers. However, the experience The words have no impact on me – it’s the images. Because the they did have gave them some familiarity with brands and the cigar smoke’s in your mouth most of the time. other factors that were perceived to indicate the quality of the Plain packaging appeared to have less impact on more experi- product, making them less subject to packaging as a way to enced smokers’ perceptions of quality and product than others, determine quality. Younger, less experienced, smokers were par- as they reported relying on brand name, reputation, acquired ticularly guided by visual cues such as cigar size, packaging and knowledge and recommendations to determine quality. They price. did not perceive changes in brands since plain packaging in I tend to look at the price. terms of quality, taste, product or value for money.

If you take one out of those tubes, you think, oh, this is going to You used to be able to go in and browse. The ability to do that … ’ be good. has decreased enormously the way they re packaged, displayed. You have to go in already knowing what you’re going to buy. I’ve (packaging)…it establishes the brand. given up browsing. Both younger and older occasional cigar/cigarillo smokers Once you get rid of the box and start smoking it, the romance were cognisant that cigar smoking could cause harm to their comes back! health. However, all held the perception that the comparatively low frequency of cigar smoking compared to other tobacco pro- ducts resulted in low relative risk. Focus group discussions with non-premium cigarillo smokers Among this group of non-premium cigarillo smokers, consump- Cigars are much better ’cause they only give you mouth cancer – as opposed to ‘everything else’ cancer! (like cigarettes) tion of non-premium cigarillos varied from daily to monthly (or less) and many felt cigarillo consumption regulated their cigar- Because you don’t have them too often, no-one’s going to be like ette consumption. ‘you should slow down on that.’ I smoke them because when I do, I don’t smoke nearly as many Exposure to cigar plain packaging was relatively low consist- as I do when I’m smoking cigarettes. ent with the popularity of single cigar/cigarillo only purchases Smoking cigarillos was part of their routine to mark a point among this group, particularly among the younger smokers. in time, usually the afternoon or evening for relaxation. They You can get them online…They come in silver tubes… I’ve never often smoked alone; however, some enjoyed cigarillos in a seen the green (sic) packaging. social setting.

They just give it to you how it is. (when purchasing from bars) After dinner, I’m ready to smoke them after a good meal. ii62 Miller CL, et al. Tob Control 2015;24:ii58–ii65. doi:10.1136/tobaccocontrol-2014-052049 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

For those non-premium cigarillo smokers who smoked in Cigars are supposed to be a classy experience … you don’t want social settings, cigarillos were reportedly smoked as part of their to be reminded… and they detract from the overall experience… personal image. Before it was pleasurable, now we have to put up with these ter- rible packs. You feel like you’re more sophisticated than the average smoker.

Non-premium cigarillo smokers perceived cigarillos to be more natural and less harmful than cigarettes. Online survey of cigar and cigarillo smokers Among this general sample of cigar and cigarillo smokers, Cigarillos and cigars are obviously a more natural product nearly all smoked cigars (table 1), with just under half also whereas cigarettes are way too much junk these days. smoking cigarillos, and most smoking cigarettes (and a very small proportion smoking only cigarillos). This was largely a ’ ’ It s this little pleasure and doesn t do that much harm. sample of less frequent cigar smokers and only a small propor- All non-premium cigarillo smokers were highly brand loyal tion smoked premium cigar/cigarillos (11%). Participants most and only familiar with one or two brands, with limited knowl- commonly reported that they purchased their last cigar or cigar- edge of the factors that differentiate brands beyond the taste. illo from (43%), with fewer purchasing from other They reported little interest in trying new or different kinds of sources (9% bars; 7% duty free/overseas; 3% internet). cigarillos due to the lack of visual cues to determine possible Half of the participants (50%) reported smoking cigars and quality and therefore, the risk of disappointment. cigarillos for a special occasion; while approximately one-third reported smoking with a friend (31%), a group of friends I just smoke the one brand, it’s a popular brand … and I stick to (33%) and at home (33%); and fewer reported smoking cigars it not knowing anything about others. and cigarillos alone (22%), at a pub/bar (16%) or at a cigar bar/ lounge (15%). ’ If you want to try something different and you re only getting 10 Cigar smokers (n=251) most commonly reported that they ’ … ’ cigars for $10 or $20 then it s a risk You don t want to pay all liked smoking cigars for relaxation (47%), taste (46%) and cele- that … and hate it. bratory purposes (37%). Benefits of smoking cigars were per- These smokers reported a high level of exposure to plain pack- ceived to be: not as harmful as cigarettes (30%); a good thing aging both of cigarillos and cigarettes. Some non-premium cigar- to do with friends (26%); a way of having time out for them- illo smokers who had been exposed to plain packaging were selves (24%); a sense of sophistication (17%), exclusivity (16%), convinced that the product was not equivalent to that purchased prestige (16%), and acceptance (6%), and few reported it was a before the implementation of the legislation, describing per- hobby/interest (9%). ceived changes in flavour, taste, size and ‘tightness’ of the roll. Cigar smokers reported strong brand loyalty with 1 (32%) or a few (38%) preferred brands, even if they sometimes tried I don’t smoke them in the new packs because they taste awful … other brands. Furthermore, for most (81%) with a brand prefer- they are completely different. ence, it was the same as it was 2 years ago. Half (or more) of the sample perceived different cigar and cigarillo brands to Captain Black have shrunk by a third and they taste disgusting, differ in taste, prestige and quality. Many of these participants very artificial … before they had a distinct cherry taste. used the taste (68%) and smell (61%) of the cigar as indicators Many of the non-premium cigarillo smokers who had seen of quality. Other indicators were: country of origin (51%), price the GHWs repeatedly acknowledged that the images and warn- (41%), brand name (35%), appearance of cigar (33%) and pack- ings had prompted them to think about the product’s harm. aging (15%). Self-reported changes in the behaviour of non-premium cigarillo Approximately one-third (30%) of the sample believed that smokers included keeping cigarillo packaging out of sight from cigars and/or cigarillos were less harmful than cigarettes, 47% others, turning over the pack, averting their eye from the believed they caused the same amount of harm and small pro- graphic images and warnings, decanting the cigarillos into an portions believed they were more harmful (13%) or did not unbranded tin, changing brands and intentionally buying from know (10%). Levels of unprompted knowledge regarding ill- non-compliant retailers. Plain packaging and GHWs were nesses associated with cigars and cigarillos (that were featured reported to have reduced perceived sophistication associated on the GHWs) was low among this sample, with 38% saying with non-premium cigarillos and increased participants’ ten- lung cancer was caused by cigar/cigarillo smoking, 11% throat dency to consider the health impacts of smoking cigarillos and cancer, 10% mouth cancer, and 2% mentioning gum disease or to re-evaluate and/or change their smoking behaviour. damaged teeth. Exposure to cigar and/or cigarillo plain packaging was It’s psychological, it just puts me off … definitely with normal reported consistently by approximately half of participants cigarettes. (table 2), in terms of purchasing and smoking cigar/cigarillos The tongue cancer ones worried me a bit, that was the one I saw that had come in compliant packaging and recalling any one of fi first and I worried a bit about what it could do to me. It was at a the cigar/cigarillo speci c GHWs. time when I was thinking about smoking them more regularly While participants most commonly reported that the product and so you could say it made me cut down a bit. they currently smoked was ‘about the same’ as they smoked 2 years ago in terms of ‘taste’, ‘appeal’, ‘satisfaction’, ‘quality’ Cigarillos used to be in the beautiful boxes, now it’s all generic and ‘harmfulness’, over half of the sample reported that the … greeny (sic) colour it really puts you off. ‘appeal of packaging’ was lower and over a third reported that ‘value for money’ was also lower (table 2). Before they used to look like a box of chocolates and now … Participants were most likely to report that the frequency You just want to keep it hidden in your pocket…. You don’t want with which they experienced a range of smoking thoughts and rotten teeth looking at you. behaviours had not changed since 2 years ago. However, they

Miller CL, et al. Tob Control 2015;24:ii58–ii65. doi:10.1136/tobaccocontrol-2014-052049 ii63 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper also reported that they were slightly more likely (33% as measures restricting marketing from sources originating in opposed to less likely (16%)) to notice the health warnings. Australia. This most likely reflects the historical influence of many decades of cigar marketing and the contemporary, global DISCUSSION reach of social media and online marketing. Unlike the USA and The low population prevalence of cigar smoking in Australia elsewhere,8 where cigar marketing is far less regulated, Australia posed challenges for the research. A mixed qualitative and quan- has not seen the rise in prevalence of use in young people or titative research methodology enabled quantitative data collec- overall. Views that cigar smoking was associated with prestige, tion from a substantial sample of cigar and cigarillo smokers, status and exclusivity were common, as were reports that cigars complemented by the in-depth insight into thoughts, percep- and cigarillos were for ‘relaxation’ and ‘time out for oneself’. tions and behaviours of cigar and cigarillo smokers from the Cigars and cigarillos were consistently viewed as less harmful qualitative interviews and focus groups. and distinct from cigarettes, all of which is consistent with the – The primary limitations of this study are the representative- way cigars have been positioned.61113 There were several indi- ness of the samples and the accuracy of self-report measures, cations that plain packaging and new GHWs were challenging most notably recall. Samples in all studies are not necessarily views that cigars were less harmful. There were also strong indi- representative of the population. Participants were asked to cations that plain packaging and GHWs were substantially redu- compare current thoughts, behaviours and attitudes to those cing the perceived distinctions in prestige and desirability they had experienced 2 years ago, which poses a risk of inaccur- between cigars and cigarettes for cigar smokers. This is a key acy. However, in a postintervention only study design, this is the finding highlighting the influence of plain packaging and GHW, most appropriate way to gauge reported changes. and suggests a need to ensure tobacco control measures aimed Other factors to consider in the interpretation of the results at cigarettes include cigars. are social desirability and political sensitivities. Overall, care was Self-reported changes in perceptions and behaviours since the taken in the ordering and framing of questions and discussion implementation of plain packaging differed between groups and prompts to minimise socially desirable responses. Plain packaging studies, although there were many positive indicators that legislation received considerable industry-driven and general exposure to GHWs and plain packaging may have increased the media coverage during its inception. Participants may not have frequency of self-reported quitting thoughts and behaviours. For supported plain packaging and GHWs. Support for regulatory example, online survey participants reported changes in measures to curb tobacco consumption often receive less support smoking behaviour such that they currently smoked less often – from smokers than non-smokers,15 17 and this was observed for than they had 2 years ago. plain packaging in a 2010 study conducted in Western Other self-reported behaviour changes were documented in Australia.16 However, during the transition to plain packaging, only some studies. For example, perceived increases in the fre- cigarette smokers who were smoking from plain packs had quency of decanting cigars was a prominent finding in the quali- higher approval of the legislation than those still smoking from tative studies, but participants in the online survey did not fully branded packs;18 and support for plain packaging increased report that they were decanting their cigars and/or cigarillos significantly from preimplementation to postimplementation more often than they did 2 years ago. This may be due to data among a population-based sample of Australian smokers.19 This collection methods, for example, group discussions may have pattern of differences in approval and increases in support post- prompted more accurate recall, or different behaviours may be implementation are similar to those observed when smoke-free observed among the different samples. – laws and display bans have been implemented.20 22 Where pos- Overall, the studies provide useful insight into the smoking sible, questions were not framed in the context of plain pack- thoughts, attitudes and behaviours of the samples of cigar and aging, however, it was necessary in some instances to ask cigarillo smokers, as well as evidence that when exposure participants to recall perceived changes since its implementation. occurred, plain packaging influenced these smokers in ways that Some participants may have been unaware of or reluctant to were consistent with the specific objectives of the legislation. make attributions about the influence of plain packaging and GHWs on their perceptions, thoughts and behaviours. A major strength of this research is the in-depth insight it pro- vides into the attitudes, perceptions and behaviours of the What this paper adds hard-to-reach group of cigar smokers. Furthermore, the online survey yielded a substantial sample of cigar/cigarillo smokers ▸ Perceptions of cigars and cigarillos were often consistent and there were some similarities observed across the study with the ways in which they have been marketed. samples. A key consistent finding was the variable exposure to ▸ There was incomplete exposure to Australian plain new packaging requirements. Furthermore, there were reports packaging and graphic health warnings (GHWs) on cigars of products in non-compliant packaging in most of the samples, and cigarillos; with reported exposure seemingly highest although the extent of non-compliance was difficult to quantify. among non-premium cigarillo smokers. This makes it difficult to determine the full effects of plain pack- ▸ When exposure occurred, there were many positive indicators aging as even for those reporting exposure, there may not have that GHWs and plain packaging meets tobacco control been consistent exposure to plain packaging and GHWs. When objectives; with premium cigar smokers also indicating some exposure had occurred, plain packaging and GHWs were attrib- fear of being equated with cigarette smokers. uted to have impact, most notably on non-premium cigarillo smokers who reported the highest exposure, but in all samples, Acknowledgements The authors would like to acknowledge Victoria Parr and the new packaging reduced perceived packaging appeal and Margie Lane for qualitative data collection and the provision of qualitative data analysis. increased noticeability of warnings. fi Contributors CLM designed the study. KAE cleaned data files, undertook analysis Another notable nding was that perceptions of cigars and and reported results. MAW provided advice on study design and interpretation of cigarillos were often consistent with the ways in which they results. CLM drafted the manuscript with contributions from all authors. All authors have been marketed internationally, despite there being strong approved the final manuscript. ii64 Miller CL, et al. Tob Control 2015;24:ii58–ii65. doi:10.1136/tobaccocontrol-2014-052049 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Funding This study was funded under a contract with the Australian Government 7 Delnevo CD, Hrywna M. Clove cigar sales following the US flavoured cigarette ban. Department of Health. Tob Control Published Online First: 20 March 2014. 8 Cullen J, Mowery P, Delnevo C, et al. Seven-year patterns in US cigar use Competing interests The authors wish to advise that CM and MW were epidemiology among young adults aged 18–25 years: a focus on race/ethnicity and members of the Expert Advisory Committee on Plain Packaging that advised the brand. Am J Public Health 2011;101:1955–62. Australian Department of Health on research pertaining to the plain packaging 9 McNichol T. Cigar asphyxionado. The New York Times, 1997; 29 June. legislation. MW holds competitive grant funding from the Australian National Health 10 Substance Abuse and Mental Health Services Administration (US). Overview of and Medical Research Council, US National Institutes of Health, Australian National findings from the 2003 National Survey on Drug Use and Health. Rockville, MD: US Preventive Health Agency and BUPA Health Foundation. CM and MW hold such Department of Health and Human Services, Substance Abuse and Mental Health grant funding from Cancer Council South Australia. Services Administration, Office of Applied Studies, 2004. Ethics approval The University of South Australia Human Research Ethics 11 Wenger L, Malone R, Bero L. The cigar revival and the popular press: a content Committee provided approval for this research. analysis, 1987–1997. Am J Public Health 2001;91:288–91. Provenance and peer review Not commissioned; externally peer reviewed. 12 Wenger L, Malone R, George A, et al. Cigar magazines: using tobacco to sell a lifestyle. Tob Control 2001;10:279–84. Open Access This is an Open Access article distributed in accordance with the 13 Delnevo CD. Smokers’ choice: What explains the steady growth of cigar use in the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which U.S.? Public Health Report 2006;121:116–19. permits others to distribute, remix, adapt, build upon this work non-commercially, 14 Parr V, Ell P. Market testing of new health warnings and information messages for and license their derivative works on different terms, provided the original work is tobacco product packaging: Premium cigars, cigarillos/little cigars and roll your own. properly cited and the use is non-commercial. See: http://creativecommons.org/ Sydney, Australia. Prepared for: Australian Department of Health and Ageing, 2011. licenses/by-nc/4.0/ http://www.health.gov.au/internet/main/ (accessed Sep 2014). 15 Farley SM, Coady MH, Mandel-Ricci J, et al. Public opinions on tax and retail-based REFERENCES tobacco control strategies. Tob Control Published Online First: 23 December 2013. 1 Tobacco Plain Packaging Act. 2011. http://www.comlaw.gov.au/details/ 16 Rosenberg M, Pettigrew S, Wood L, et al. Public support for tobacco control policy c2011a00148 (accessed Sept 2014). extensions in Western Australia: a cross-sectional study. BMJ Open 2012;2: 2 Competition and Consumer (Tobacco) Information Standard. 2011. http://www. e000784. (accessed Sept 2014). comlaw.gov.au/Details/F2011L02766 (accessed Sept 2014). 17 Walsh RA, Paul CL, Tzelepis F, et al. Is government action out-of-step with public 3 Scollo M, Lindorff K, Coomber K, et al. Standardised packaging and new enlarged opinion on tobacco control? Results of a New South Wales population survey. graphic health warnings for tobacco products in Australia—legislative requirements Aust N Z J Public Health 2008;32:482–8. and implementation of the Tobacco Plain Packaging Act 2011 and the Competition 18 Wakefield M, Hayes L, Durkin S, et al. Introduction effects of the Australian plain and Consumer (Tobacco) Information Standard, 2011. Tob Control packaging policy on adult smokers: a cross-sectional study. BMJ Open 2013;3: 2015;24:ii9–16. e003175. http://bmjopen.bmj.com/content/3/7/e003175.full (accessed Sept 2014). 4 Australian Institute of Health and Welfare. 2010 National Drug Strategy Household 19 Swift E, Borland R, Cummings MK, et al. Australian smokers’ support for Survey report. Drug statistics series no. 25. Cat. no. PHE 145. Canberra: AIHW, standardized packs before and after implementation: findings from the ITC Four 2011. http://www.aihw.gov.au/WorkArea/DownloadAsset.aspx?id=10737421314 Country survey. Tob Control Published Online First: 12 November 2014. (accessed Sept 2014). 20 Cooper J, Borland R, Yong HH, et al. Compliance and support for bans of smoking 5 Jamner M. Cigar smoking among college students: prevalence and correlates. in licensed venues in Australia: findings from the International Tobacco Control Four Prev Med 1999;29:187–94. Country Survey. Aust N Z J Public Health 2010;20:137–43. 6 Slade J. Marketing and Promotion of cigars (Chapter 7). In: Burns D, Cummings 21 Hyland A, Higbee C, Borland R, et al. Attitudes and beliefs about secondhand KM, Hoffmann D, eds. Cigars. Health Effects and Trends, Monograph No. 9. USA: smoke and smoke-free policies in four countries: findings from the International Department of Health and Human Services, National Institutes of Health, NIH Tobacco Control Four Country Survey. Nicotine Tob Res 2009;11:642–9. publication, 1998. http://cancercontrol.cancer.gov/brp/tcrb/monographs/9/m9_ 22 McNeill A, Lewis S, Quinn C, et al. Evaluation of the removal of point-of-sale complete.pdf (accessed Sept 2014). tobacco displays in Ireland. Tob Control 2011;20:137–43.

Miller CL, et al. Tob Control 2015;24:ii58–ii65. doi:10.1136/tobaccocontrol-2014-052049 ii65 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

''You're made to feel like a dirty filthy smoker when you're not, cigar smoking is another thing all together.'' Responses of Australian cigar and cigarillo smokers to plain packaging Caroline L Miller, Kerry A Ettridge and Melanie A Wakefield

Tob Control 2015 24: ii58-ii65 doi: 10.1136/tobaccocontrol-2014-052049

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii58

These include: References This article cites 11 articles, 5 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii58#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper Changes in use of types of tobacco products by pack sizes and price segments, prices paid and consumption following the introduction of plain packaging in Australia Michelle Scollo, Meghan Zacher, Kerri Coomber, Megan Bayly, Melanie Wakefield

▸ Additional material is ABSTRACT less than 10% of the market in Australia.5 While a published online only. To view Objectives To describe changes among smokers in use small number of slims and king size brands are please visit the journal online 1 (http://dx.doi.org/10.1136/ of various types of tobacco products, reported prices available, most cigarettes sold are regular size. tobaccocontrol-2014-052071). paid and cigarette consumption following the Cigarette brands in Australia until 2006 were typic- standardisation of tobacco packaging in Australia. ally sold in four different ‘tar’ levels—16, 12, Centre for Behavioural Research in Cancer, Cancer Methods National cross-sectional telephone surveys of 8 and 4 mg, with 2 mg also available in some 6 Council Victoria, Melbourne, adult smokers were conducted from April 2012 brands. In 2006 the three major manufacturers Australia (6 months before transition to plain packaging (PP)) to stopped using such labelling,7 however, colour March 2014 (15 months afterwards). Multivariable coding and terms such as Rich, Smooth and Ultra Correspondence to Dr Michelle Scollo, Centre for logistic regression assessed changes in products, brands were incorporated into pack design and in variant Behavioural Research in and pack types/sizes; multivariable linear regression names providing consumers with cues about con- Cancer, Cancer Council examined changes in inflation-adjusted prices paid and cordance with the old tar-levels.8 Victoria, 615 St Kilda Rd, reported cigarette consumption between the pre-PP and As in most countries, cigarettes in Australia have Melbourne, VIC 3004, three subsequent periods—the transition phase, PP year been categorised into three price segments— Australia; 59 [email protected] 1 and PP post-tax (post a 12.5% tax increase in premium, mainstream and value. The recom- December 2013). mended retail price per typical pack of 25s in Received 29 September 2014 Results The proportion of current smokers using roll- early-2014 was over $20 for premium brands and Accepted 13 January 2015 your-own (RYO) products fluctuated over the study under $17.70 for value brands.5 While value period. Proportions using value brands of factory-made brands have traditionally been sold in Australia in (FM) cigarettes increased from pre-PP (21.4%) to PP packs of 35, 40 and 50 cigarettes10 many value year 1 (25.5%; p=0.002) and PP post-tax (27.8%; brands have recently become available in packs of p<0.001). Inflation-adjusted prices paid increased in the 20 which retail well under the $15 mark,11 and PP year 1 and PP post-tax phases; the largest increases some commentators have pointed to the emergence were among premium FM brands, the smallest among of a super-value segment.12 A full list of products value brands. Consumption did not change in PP year 1 available for sale in Australia between 2012 and among daily, regular or current smokers or among 2014 can be viewed here (see online supplementary smokers of brands in any market segment. Consumption table S1). among regular smokers declined significantly in PP post- Prices charged at the retail level in small mixed tax (mean=14.0, SE=0.33) compared to PP year 1 businesses and petrol stations are generally in line (mean=14.8, SE=0.17; p=0.037). with recommended retail prices regularly published Conclusions Introduction of PP was associated with by Australia’s association of retail tobacconists,13 an increase in use of value brands, likely due to however, larger retailers such as supermarkets and increased numbers available and smaller increases in large tobacconists often negotiate lower wholesale prices for value relative to premium brands. Reported prices for greater volumes and sell tobacco pro- consumption declined following the December 2013 tax ducts at prices discounted well below the recom- increase. mended levels.14 A recent check of prices in supermarkets (May 2014) revealed prices approxi- mately 2–7% cheaper for premium brands, 5–15% INTRODUCTION cheaper for mainstream brands and 3–16% cheaper 15 Open Access The tobacco market in Australia includes factory- for value brands. Thus, the price paid by a con- Scan to access more made (FM) cigarettes, cigars and smoking tobacco sumer for a pack of cigarettes may vary consider- free content (some brands for use in pipes and some brands for ably not just based on price segment of the brand use in roll-your-own (RYO) cigarettes).1 It differs they choose (value, mainstream or premium), but from some other markets internationally in that also depending on whether they purchase it from a smokeless tobacco for oral use, while legal to convenience or a discount outlet, as a single pack, a import, cannot be sold.2 Similarly e-cigarettes may multipack (two or three packs bundled together) or be sold, but nicotine capsules for use in such carton, and what pack size they purchase. devices can currently only be imported under par- This paper aims to assess the impact on product 3 To cite: Scollo M, ticular arrangements. Use of RYO is less common choice, prices paid and consumption following Zacher M, Coomber K, et al. than in neighbouring New Zealand and in the UK changes in the packaging of tobacco products intro- Tob Control 2015;24: or Canada.4 FM cigarettes dominate and almost all duced by the Australian Government at the end of – ii66 ii75. cigarettes are filtered. Menthol cigarettes make up 2012, the details of which are set out in a ii66 Scollo M, et al. Tob Control 2015;24:ii66–ii75. doi:10.1136/tobaccocontrol-2014-052071 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper companion paper in this supplement.16 Choice of product type, recommended retail prices increased more steeply among purchase channel and pack size and prices paid may affect con- premium than among value brands over the period of introduc- – sumption patterns17 18 as well as quitting behaviours19 21 so it tion of PP. A further aim of this study was to assess whether a was important to track all of these behaviours. Shifts in purchas- similar pattern of increases was observed for prices paid by con- ing decisions and reductions in consumption could occur after sumers taking into account not just recommended prices but the introduction of plain packaging (PP) with larger and prices charged, and changes in use of brands within segments. refreshed graphic health warnings. Smokers could attempt to After a 25% increase in tobacco excise/customs duty in ameliorate the effects of the new, concern-provoking packaging Australia in April 2010 (on top of the earlier 20–25% increase in by smoking fewer cigarettes, rather than trying to quit. On the duty applicable to large packs in November 1999 and continuing other hand, the tobacco industry has argued that PP would increases with twice-yearly indexation since that time), value result in a fall in overall prices in the market as companies brands in traditionally large pack sizes in Australia had been would be forced to compete on price alone and that this would becoming increasingly less affordable for price-sensitive lead to a shift to value brands and illicit tobacco, and an overall smokers.10 While still cheaper per stick than many other brands, – increase in consumption.22 24 the up-front cost of a pack of Horizon 50s (the leading value This paper aims to test predictions by the tobacco industry brand in 2011) had increased from $15.42 (AUD) in February that the legislation would result in 2001 to $29.80 in February 2012, just under $30, no doubt a 1. a decline in the use of premium and mainstream brands and critical ‘left-digit price point’.37 38 Tobacco companies needed to an increase in use of value brands find a way to provide value smokers with products with lower 2. a decline in prices paid for tobacco products, particularly for up-front costs. In the lead up to PP, several new value brands premium brands and were introduced in small pack sizes early in 201235 and several 3. a consequent increase in the consumption of tobacco brands of FM cigarettes already sold in 20s were expanded to products. include packs of 21, 22 and 23.35 Some changes in pack size These predictions were vigorously promulgated in response to offerings effectively provided ‘free’ cigarettes, with the recom- – the Government’s calls for comment on draft legislation.22 24 mended retail prices for packs of 21s and 22s, for example, being Claims that sales of tobacco products have increased since listed at the same or lower price as packs of 20s.39 In addition, implementation have continued to be made in the media,25 26 several brands appeared in a pack size of 26, and packs of 35s in submissions to UK government inquiries considering propo- were replaced with 30s. Several brands traditionally offered in – sals to implement standardised packaging27 30 and in New packs of 50 cigarettes introduced 40s or priced existing 40s pro- Zealand.31 Another key prediction—that use of illicit tobacco ducts more competitively than 50s.36 Online supplementary would increase—is tested elsewhere (this volume).32 table S2 shows the variant–pack size combinations available for When attempting to assess the impact of PP legislation on John Player Special (JPS) which manufacturer Imperial Tobacco consumers’ choice of tobacco product type, prices paid and reports was the fastest growing brand over 2013.40 41 As indi- number of cigarettes consumed, account must also be taken of cated in table 1 in Scollo et al36 in this volume, between the effects of other government policies and industry strategies November 2011 and November 2013, the number of brands that attempt to mitigate the effects of such policies.33 34 The available in new pack sizes increased from 0 to 2 for 21s, from 0 introduction of PP coincided with numerous instances of to 1 for 22s, from 0 to 2 for 23s, from 0 to 3 for 26s. The expan- tobacco companies expanding variant names to reinforce the sion of pack sizes observed across the market was likely intended colour system of naming,35 with Pall Mall Slims variant names to retain price sensitive smokers including those wanting to buy Red, Blue, Amber and Green, for instance, changing to Original value cigarettes, but likely to baulk at the very high up-front pur- Red, Rich Blue, Smooth Amber and Menthol Green.13 35 chase cost of a pack of 50s. A further aim of this study therefore Manufacturers also took the opportunity to reformulate was to document any changes in the distribution of use of differ- product and pack size offerings and adjust recommended retail ent pack sizes in Australia. prices. As detailed elsewhere35 product reformulations included During the course of the current study, the Australian the extension of several brands to include king size cigarettes Government announced that it would introduce four 12.5% with extralong filters (eg, Pall Mall Extra Kings 20s) and the increases in excise/customs duty, the first in December 2013,42 insertion of padding in the packaging of slim cigarettes in order and the other three scheduled to occur annually from to comply with required larger pack size dimensions without September 2014.16 43 44 To prevent increases in the affordability increasing the diameter of the cigarettes (for Vogue Slims 20s). of tobacco products in the longer term, the legislation enacting Several new and novel menthol products—featuring either or these increases also amended the basis for 6-monthly indexation both loading with extra menthol (eg, Holiday Cool Chill, Cool so that from March 2014, indexation in line with the consumer Frost and Cool Blast), or squeezable capsules in the filter to price index (CPI) would cease and tobacco excise and customs convert the cigarette to menthol (eg, Peter Jackson Hybrid)— duty would be indexed instead in line with changes in Average were introduced onto the market before PP.35 This paper also Weekly Earnings.43 44 Reducing affordability through tax assesses whether use of these products increased given that increases is a highly effective approach available to governments – opportunities to promote product characteristics through to reduce the use of tobacco products.45 47 Tw o final aims of package design were curtailed after the introduction of PP this study were to assess changes in reported prices paid and legislation. changes in reported numbers of cigarettes consumed in the Australia is unusual internationally in that excise/customs duty period up to and after the large increase in excise duty on 1 since 1986 has been automatically inflation-adjusted 6-monthly December 2013. resulting in twice yearly increases in retail prices since that time.10 Increases in excise/customs duty most commonly have METHODS been ‘over-shifted’ to consumers (ie, increases have been higher Design than that required merely to pass on the effects of indexation),10 National cross-sectional telephone surveys were conducted and Scollo et al36 have reported in this volume that between 9 April 2012 and 30 March 2014, with approximately

Scollo M, et al. Tob Control 2015;24:ii66–ii75. doi:10.1136/tobaccocontrol-2014-052071 ii67 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 1 Type of cigarette smoked, market segment of current cigarette smokers’ current brands of FM and RYO cigarettes, and features of FM cigarettes by phase: percentages and results of logistic regression models

Differences between PP phases—unadjusted models Differences between PP phases—adjusted models Time period Per cent OR 95% CI Per cent OR 95% CI

Type of cigarette smoked (n=8610)† Any FM Pre-PP 79.6 1.00 79.5 1.00 Transition 83.1 1.26 1.00 to 1.59 83.2 1.28* 1.02 to 1.62 PP year 1 80.8 1.08 0.93 to 1.24 81.0 1.10 0.96 to 1.28 PP post-tax 78.8 0.95 0.79 to 1.15 79.2 0.98 0.81 to 1.19 Any RYO Pre-PP 32.8 1.00 32.9 1.00 Transition 27.0 0.76** 0.62 to 0.92 26.7 0.73** 0.60 to 0.90 PP year 1 33.6 1.04 0.92 to 1.17 33.4 1.02 0.90 to 1.16 PP post-tax 32.0 0.96 0.82 to 1.13 31.4 0.93 0.79 to 1.10 Both FM and RYO Pre-PP 13.9 1.00 13.8 1.00 Transition 11.7 0.82 0.62 to 1.09 11.5 0.81 0.60 to 1.08 PP year 1 15.9 1.17 0.99 to 1.38 15.9 1.19* 1.00 to 1.41 PP post-tax 12.9 0.92 0.73 to 1.16 12.8 0.92 0.72 to 1.17 Market segment (n=7076)‡ RYO tobacco Pre-PP 22.7 1.00 22.9 1.00 Transition 17.8 0.74* 0.57 to 0.95 17.6 0.71* 0.55 to 0.93 PP year 1 21.4 0.93 0.80 to 1.08 21.3 0.91 0.78 to 1.06 PP post-tax 21.3 0.92 0.75 to 1.13 21.7 0.93 0.75 to 1.15 Value FM cigarettes Pre-PP 21.1 1.00 21.4 1.00 Transition 22.2 1.07 0.84 to 1.35 22.1 1.04 0.82 to 1.33 PP year 1 25.8 1.30** 1.12 to 1.50 25.5 1.27** 1.10 to 1.48 PP post-tax 27.4 1.41** 1.16 to 1.71 27.8 1.44*** 1.18 to 1.76 Mainstream FM cigarettes Pre-PP 39.2 1.00 39.1 1.00 Transition 42.7 1.16 0.94 to 1.42 42.9 1.17 0.96 to 1.44 PP year 1 37.4 0.92 0.81 to 1.05 37.4 0.93 0.82 to 1.06 PP post-tax 36.2 0.88 0.74 to 1.05 36.1 0.88 0.73 to 1.05 Premium FM cigarettes Pre-PP 16.9 1.00 16.5 1.00 Transition 17.3 1.03 0.79 to 1.33 17.7 1.09 0.83 to 1.43 PP year 1 15.4 0.90 0.76 to 1.05 15.7 0.93 0.79 to 1.11 PP post-tax 15.1 0.88 0.70 to 1.09 14.5 0.85 0.68 to 1.07 FM cigarette features (n=4824)§ Menthol Pre-PP 11.0 1.00 11.1 1.00 Transition 10.9 0.99 0.69 to 1.42 11.3 1.02 0.70 to 1.47 PP year 1 10.9 0.98 0.78 to 1.24 11.0 0.99 0.78 to 1.25 PP post-tax 11.3 1.03 0.76 to 1.39 10.9 0.98 0.71 to 1.34 Cigarette length—king size Pre-PP 2.2 1.00 2.2 1.00 Transition 1.1 0.52 0.21 to 1.24 1.2 0.52 0.21 to 1.25 PP year 1 2.0 0.91 0.56 to 1.47 2.0 0.90 0.55 to 1.45 PP post-tax 2.4 1.10 0.58 to 2.10 2.4 1.09 0.57 to 2.09 Cigarette diameter—slims Pre-PP 0.8 1.00 0.7 1.00 Transition 1.4 1.77 0.62 to 5.06 1.4 1.86 0.65 to 5.31 PP year 1 1.8 2.30* 1.12 to 4.72 1.8 2.39* 1.16 to 4.94 PP post-tax 1.3 1.65 0.68 to 4.00 1.3 1.65 0.68 to 4.04 *p<0.05, **p<0.01, ***p<0.001. †Of n=8811 smokers, we excluded those who did not have SES information (n=201). Analysed n=8610. ‡Of n=8679 cigarette smokers, we excluded those who: did not provide their current brand name (n=283); currently smoked unbranded tobacco or gave a brand of cigars/cigarillos or e-cigarettes (n=60); did not recall where they purchased their pack or received it as a gift (n=75); did not provide a valid pack size (n=231); did not report a price (n=413); purchased tobacco duty free or overseas (n=70); skipped relevant questions due to back-coding of verbatim responses (n=77); gave responses that were deemed unlikely to be correct (such as a non-existent pack size or unrealistically low or high price) (n=243); or did not have SES information (n=151). Analysed n=7076. §Of n=4958 FM only cigarette smokers included in the market segment analyses, we excluded those who: did not provide their current variant name (n=42); or did not have SES information (n=92). Analysed n=4824. FM, factory made; PP, plain packaging; RYO, roll-your-own; SES, socioeconomic status. ii68 Scollo M, et al. Tob Control 2015;24:ii66–ii75. doi:10.1136/tobaccocontrol-2014-052071 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

100 interviews completed each week. As detailed in Wakefield not be determined) in odds of smoking FM cigarettes, RYO et al48 (this volume), a dual frame design using random digit cigarettes, cigars and pipes (alone or in combination with other dialling to landline and mobile phones was used, with an types of tobacco). average adjusted response rate per 4-week period of 57%. We then analysed features of current FM cigarette and RYO tobacco products among 7076 smokers of FM or RYO cigarettes Participants for whom SES could be determined and who provided full Eligible participants were current smokers and recent (past information on their current brand, place of purchase, pack size year) quitters of tobacco products, aged 18–69 years. For the and price (see notes to table 1). Logistic regression models present study, we included all current smokers (n=8811) and examined change between the pre-PP and later time periods in limited the sample as appropriate for various analyses. prevalence of brand extensions/variants classified as menthol, slims and king size cigarette types. We used logistic regression to Measures assess change between the pre-PP and later time periods in Tobacco products used prevalence of use of brands of each market segment (RYO All current smokers were asked, “Do you currently smoke FM tobacco; value; mainstream; or premium FM cigarettes) and use cigarettes only, RYO cigarettes only, both, or neither of these?” of different pack types and sizes (see notes to table 2). Linear They were then asked whether they currently smoked any other regression was used to assess whether real prices paid changed form of tobacco and to list which types (ie, cigars, pipes, etc). from the pre-PP period. We also tested interactions between the pre-PP and PP year 1 and the pre-PP and PP post-tax phases and Market segment of current brands market segment and pack size to determine whether prices Those who smoked cigarettes were asked to provide their changed more among some brands and pack sizes than others. current brand and variant names (the brand of their last cigar- Logistic regression was used to assess the proportion of FM ette smoked). We coded current brands of FM cigarettes into and/or RYO smokers who had tried a different brand because three market segments (value, mainstream and premium) using they were cheaper or thought about money spent on smoking definitions from Retail World9 trade magazine where they during past month. To assess changes in daily cigarette con- existed and using stick prices relative to those for listed prices sumption for daily, weekly and monthly smokers, we used linear where they did not. Brands of RYO tobacco were coded as a regression. fourth market segment. All regression models controlled for sex, age group (18–29, The brand variants nominated were also classified according 30–49 and 50–69 years), area SES (computed using an index of to whether or not they were king size, regular size or slims and relative disadvantage)51 and education (less than high school, whether or not the name of the brand variant suggested that the high school and/or some tertiary and tertiary or higher). For product was mentholated. analyses examining daily cigarette consumption, we also con- trolled for past 3-month exposure to antismoking campaigns Pack types and sizes of tobacco products purchased that aired on television during the survey period, as measured and reported prices paid by Target Audience Rating Points (TARPs) for adults aged 18 Cigarette smokers were then asked the pack size of their current and above. TARPs represent an average potential exposure to brand (in number of cigarettes or grams) and questions to ascer- advertising, with 100 TARPs equivalent to an average of one tain pack/pouch price, including pack type for FM cigarettes potential advertisement exposure per month for all adults (single pack, multipack containing ≤100 cigarettes, or carton within a media market. containing >100 cigarettes). We configured these responses into Data were weighted as described in Wakefield et al48 (this cents per cigarette (or per 0.8 g of RYO tobacco (the weight at volume). All analyses were conducted in Stata V.12.1, adjusting which duty on a RYO cigarette is equivalent to that of a FM for the effects of sample weighting on parameter estimates and cigarette)).49 Reported prices were then adjusted for inflation to SEs. In addition, an unconditional approach (ie, the ‘subpopula- March 2014 prices using the quarterly CPI.50 Respondents were tion’ command in Stata V.12.1) was used to limit the sample as also asked if they had tried any other brand in the previous appropriate for each set of analyses, ensuring correct estimation month because it was cheaper than usual, and how often in the of the SEs. Both unadjusted and adjusted results are provided in past month they had thought about much they spend on tables; all results reported in text below are adjusted for smoking (many or several times vs never or once or twice). covariates.

Reported consumption RESULTS Finally, daily, weekly and monthly cigarette smokers were asked Participants how many cigarettes they smoked per day, week or month Of the 8811 current smokers, slightly more than one quarter of (respectively). For analysis, these were configured into cigarettes participants (29%) were aged 18–29 years, 45% were aged 30– per day. 49 years and 26% were aged 50–69 years; 55% were men. Additionally, 39% of smokers resided in low-SES, 42% mid-SES Analysis and 18% in high-SES areas. Overall, 31% had not completed We used logistic and linear regression to assess changes between high school, 56% had completed high school and/or had some the pre-PP period (April to September 2012; n=2223) and tertiary education and 13% had completed tertiary or higher three subsequent time periods: the transition phase during education. which plain packages were being introduced into the Australian market (October and November 2012; n=776); PP year 1 Types of products used (December 2012 to November 2013; n=4431); and PP post-tax Types of tobacco smoked—cigarettes versus cigars and pipes (December 2013 to March 2014; n=1381). First, we used logis- Use of cigars or pipes was uncommon (cigars=2.9%; tic regression to assess changes among all 8811 current smokers pipes=0.8%) with no difference between the pre-PP phase and (excluding n=201 for whom socioeconomic status (SES) could any of the subsequent phases (results not in tables).

Scollo M, et al. Tob Control 2015;24:ii66–ii75. doi:10.1136/tobaccocontrol-2014-052071 ii69 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 2 FM pack configurations, and size of FM cigarette packs and RYO pouches of current smokers’ brands by phase: percentages and results of logistic regression models Differences between PP phases—unadjusted models Differences between PP phases—adjusted models Time period % OR 95% CI % OR 95% CI

FM pack configuration (n=5559) Single pack Pre-PP 84.3 1.00 84.2 1.00 Transition 84.0 0.98 0.73 to 1.31 83.8 0.97 0.73 to 1.31 PP year 1 79.9 0.74** 0.61 to 0.89 80.0 0.75** 0.62 to 0.90 PP post-tax 78.3 0.67** 0.53 to 0.85 78.2 0.67** 0.52 to 0.86 Multipack Pre-PP 2.9 1.00 2.9 1.00 Transition 4.9 1.69 0.96 to 2.98 4.8 1.71 0.96 to 3.02 PP year 1 6.2 2.20*** 1.50 to 3.23 6.4 2.29*** 1.55 to 3.39 PP post-tax 9.1 3.33*** 2.15 to 5.14 9.4 3.49*** 2.24 to 5.43 Carton Pre-PP 12.7 1.00 12.9 1.00 Transition 11.1 0.86 0.62 to 1.19 11.3 0.85 0.61 to 1.19 PP year 1 13.9 1.10 0.90 to 1.35 13.6 1.06 0.87 to 1.31 PP post-tax 12.5 0.98 0.74 to 1.29 12.4 0.95 0.71 to 1.27 Pack/pouch size FM cigarettes (n=5559) 20 cigarettes Pre-PP 16.9 1.00 16.8 1.00 Transition 17.5 1.04 0.78 to 1.40 17.4 1.05 0.78 to 1.41 PP year 1 15.7 0.92 0.76 to 1.11 15.8 0.93 0.77 to 1.13 PP post-tax 19.3 1.18 0.92 to 1.51 19.2 1.19 0.92 to 1.52 21, 22 or 23 cigarettes Pre-PP 1.6 1.00 1.6 1.00 Transition 1.4 0.87 0.36 to 2.13 1.4 0.84 0.34 to 2.06 PP year 1 4.0 2.52*** 1.52 to 4.18 3.9 2.45** 1.48 to 4.08 PP post-tax 3.0 1.86 0.98 to 3.52 3.0 1.86 0.98 to 3.53 25 cigarettes Pre-PP 43.1 1.00 42.7 1.00 Transition 46.9 1.17 0.93 to 1.46 47.0 1.20 0.95 to 1.51 PP year 1 39.4 0.86* 0.74 to 0.99 39.7 0.88 0.76 to 1.01 PP post-tax 33.9 0.68*** 0.56 to 0.82 33.7 0.67*** 0.55 to 0.82 26 cigarettes Pre-PP 0.3 1.00 0.4 1.00 Transition 1.0 2.92 0.62 to 13.85 1.0 2.82 0.60 to 13.25 PP year 1 2.5 7.35** 2.13 to 25.41 2.5 7.20** 2.08 to 24.91 PP post-tax 3.2 9.47** 2.62 to 34.20 3.3 9.62** 2.65 to 34.92 30 or 35 cigarettes Pre-PP 17.4 1.00 17.5 1.00 Transition 14.5 0.80 0.59 to 1.09 14.6 0.80 0.59 to 1.09 PP year 1 16.5 0.94 0.78 to 1.13 16.4 0.93 0.77 to 1.12 PP post-tax 15.9 0.90 0.69 to 1.16 15.9 0.89 0.69 to 1.15 40 cigarettes Pre-PP 13.6 1.00 13.7 1.00 Transition 13.1 0.96 0.70 to 1.32 13.2 0.95 0.69 to 1.32 PP year 1 16.1 1.22 1.00 to 1.49 15.9 1.20 0.98 to 1.47 PP post-tax 18.6 1.46** 1.12 to 1.90 18.8 1.48** 1.13 to 1.93 50 cigarettes Pre-PP 7.1 1.00 7.2 1.00 Transition 5.6 0.78 0.51 to 1.19 5.7 0.77 0.50 to 1.19 PP year 1 5.8 0.81 0.62 to 1.06 5.8 0.78 0.59 to 1.03 PP post-tax 6.1 0.85 0.59 to 1.23 6.2 0.84 0.58 to 1.23 RYO tobacco (n=1517) 25 g Pre-PP 2.6 1.00 2.6 1.00 Transition 5.4 2.11 0.66 to 6.76 5.4 2.12 0.66 to 6.82 Continued ii70 Scollo M, et al. Tob Control 2015;24:ii66–ii75. doi:10.1136/tobaccocontrol-2014-052071 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 2 Continued Differences between PP phases—unadjusted models Differences between PP phases—adjusted models Time period % OR 95% CI % OR 95% CI

PP year 1 11.6 4.88*** 2.34 to 10.18 11.6 4.93*** 2.36 to 10.32 PP post-tax 23.3 11.30*** 5.18 to 24.67 23.2 11.36*** 5.25 to 24.56 30 g Pre-PP 41.0 1.00 40.7 1.00 Transition 39.0 0.92 0.57 to 1.48 39.1 0.93 0.58 to 1.52 PP year 1 36.6 0.83 0.63 to 1.10 37.1 0.86 0.65 to 1.13 PP post-tax 26.6 0.52** 0.35 to 0.78 25.7 0.50** 0.33 to 0.75 35 or 40 g Pre-PP 1.3 1.00 1.3 1.00 Transition 2.2 1.77 0.29 to 10.63 2.2 1.77 0.30 to 10.34 PP year 1 0.6 0.45 0.11 to 1.90 0.6 0.44 0.10 to 1.93 PP post-tax 0.2 0.19 0.02 to 1.83 0.2 0.18 0.02 to 1.91 50 or 55 g Pre-PP 55.2 1.00 55.5 1.00 Transition 53.4 0.93 0.59 to 1.49 53.3 0.91 0.57 to 1.47 PP year 1 51.2 0.85 0.65 to 1.12 50.7 0.82 0.62 to 1.08 PP post-tax 49.9 0.81 0.56 to 1.17 50.9 0.83 0.57 to 1.20 Of n=8679 cigarette smokers, we excluded those who: did not provide their current brand name (n=283); currently smoked unbranded tobacco or gave a brand of cigars/cigarillos or e-cigarettes (n=60); did not recall where they purchased their pack or received it as a gift (n=75); did not provide a valid pack size (n=231); did not report a price (n=413); purchased tobacco duty free or overseas (n=70); skipped relevant questions due to back-coding of verbatim responses (n=77); gave responses that were deemed unlikely to be correct (such as a non-existent pack size or unrealistically low or high price) (n=243); or did not have socioeconomic status information (n=151). Analysed n=7076. *p<0.05, **p<0.01, ***p<0.001. FM, factory made; PP, plain packaging; RYO, roll-your-own.

Types of cigarettes used—FM and/or RYO cigarettes study period, from 2.9% pre-PP to 4.8% in the transition The proportion of all cigarette smokers reporting any use of phase, 6.4% in PP year 1 and 9.4% in the PP post-tax phase FM cigarettes significantly increased from pre-PP to the transi- (table 2). Odds of having purchased one’s current pack as a tion phase, but returned to pre-PP levels during PP year 1 and single pack significantly declined between the pre-PP (84.2%) PP post-tax phases (table 1). The proportions of current cigar- and the PP year 1 and PP post-tax periods. Carton purchases ette smokers reporting any use of RYO cigarettes was lower in did not differ between the pre-PP and subsequent phases; the transition phase than pre-PP, but was similar to pre-PP 13.1% of packs were purchased this way overall. during the PP year 1 and PP post-tax phases. Dual use of FM Among smokers of FM cigarettes, use of packs containing and RYO cigarettes was slightly higher in PP year 1, but 20 cigarettes did not change over the course of the study; returned to PP levels during the PP post-tax phase. neither did packs of 30s/35s or 50s (table 2). Use of packs con- taining between 21 and 23 cigarettes increased from pre-PP to Market segments PP year 1, but was not significantly higher in the PP post-tax Ta b l e 1 shows the proportion of current cigarette smokers whose than pre-PP phase. Packs containing 25 cigarettes significantly current brand was RYO tobacco did not change between the pre-PP declined in popularity from the pre-PP to PP year 1 and PP andPPyear1orPPpost-taxphases,despiteasignificantly lower post-tax periods, while packs of 26 cigarettes significantly proportion in the transition period. The proportion of current increased. Packs containing 40 cigarettes (mostly value brands, smokers using value FM cigarettes increased significantly between but also Longbeach, considered a mainstream brand) became the pre-PP and PP year 1 and PP post-tax periods. Proportions more common in the PP post-tax period. using mainstream and premium brands of FM cigarettes did not Among current smokers using RYO cigarettes, the proportion change significantly between the pre-PP and later periods. using pouches containing 25 g of tobacco increased substantially from pre-PP to the two phases after implementation of PP Use of various types of FM cigarettes (table 2). Correspondingly, use of 30 g pouches was significantly The percentage of FM smokers using menthol cigarettes lower in the PP post-tax phase compared with pre-PP. Use of remained at just under 12% over the course of the study and larger pouch sizes did not change between periods. did not change between the pre-PP and the later three phases (table 1). Overall, 2.1% of FM smokers were currently smoking a king size brand; the proportion did not change over the study FM brand–variant–pack size combinations period. The percentage of FM cigarette smokers currently Prior to the introduction of PP, 95% of current FM smokers smoking a slims brand showed a significant increase PP year 1 smoked one of 17 brands, 83 brand–variant combinations and compared to pre-PP, however, proportions declined slightly and 150 brand–variant–pack-size combinations (results not depicted were not significantly higher during PP post-tax (table 1). in tables). During PP year 1, while the number of brands and brand-variants stayed constant, brand–variant–pack-size combi- Pack types and sizes of tobacco products purchased nations had increased to 165. The reported number of brand– Table 2 shows that the odds of having purchased current packs variant–pack size combinations increased only slightly from of FM cigarettes in multipacks significantly increased over the pre-PP to PP year 1 for premium brands (33–35 (a 6%

Scollo M, et al. Tob Control 2015;24:ii66–ii75. doi:10.1136/tobaccocontrol-2014-052071 ii71 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 3 Mean real per stick price ($March 2014) of current FM and RYO cigarettes by phase: means and results of linear regression models Differences between PP phases—unadjusted models Differences between PP phases—adjusted models Mean cigarette Mean cigarette Sample Time period price in cents (SE) B-weight 95% CI price in cents (SE) B-weight 95% CI

All RYO cigarettes Pre-PP 57.2 (0.32) Ref 57.2 (0.32) Ref (per 0.8 g) (n=1517) Transition 56.6 (0.58) −0.61 −1.91 to 0.70 56.7 (0.57) −0.50 −1.78 to 0.78 PP year 1 59.1 (0.27) 1.82*** 0.99 to 2.64 59.1 (0.27) 1.91*** 1.09 to 2.74 PP post-tax 65.2 (0.59) 7.99*** 6.67 to 9.31 65.1 (0.58) 7.92*** 6.62 to 9.23 All FM cigarettes Pre-PP 65.2 (0.27) Ref 65.1 (0.25) Ref (n=5559) Transition 65.9 (0.45) 0.70 −0.32 to 1.72 65.9 (0.44) 0.82 −0.16 to 1.80 PP year 1 66.8 (0.21) 1.59*** 0.92 to 2.25 66.9 (0.21) 1.80*** 1.18 to 2.42 PP post-tax 72.6 (0.47) 7.41*** 6.34 to 8.47 72.6 (0.45) 7.46*** 6.46 to 8.45 By FM market segment Value FM cigarettes Pre-PP 55.9 (0.30) Ref 56.0 (0.30) Ref (n=1731) Transition 56.6 (0.55) 0.66 −0.56 to 1.89 56.5 (0.55) 0.59 −0.63 to 1.82 PP year 1 57.3 (0.22) 1.43*** 0.69 to 2.17 57.3 (0.22) 1.38*** 0.65 to 2.11 PP post-tax 62.2 (0.45) 6.32*** 5.27 to 7.38 62.2 (0.45) 6.23*** 5.17 to 7.29 Mainstream FM Pre-PP 66.9 (0.31) Ref 66.9 (0.30) Ref cigarettes (n=2700) Transition 67.5 (0.47) 0.60 −0.50 to 1.70 67.4 (0.47) 0.50 −0.58 to 1.58 PP year 1 69.4 (0.23) 2.52*** 1.77 to 3.28 69.4 (0.23) 2.58*** 1.85 to 3.30 PP post-tax 75.7 (0.59) 8.79*** 7.48 to 10.10 75.7 (0.57) 8.87*** 7.63 to 10.11 Premium FM cigarettes Pre-PP 73.1 (0.40) Ref 72.9 (0.38) Ref (n=1126) Transition 74.1 (0.78) 1.07 −0.65 to 2.80 74.2 (0.76) 1.35 −0.31 to 3.02 PP year 1 76.4 (0.35) 3.37*** 2.33 to 4.40 76.5 (0.35) 3.64*** 2.62 to 4.67 PP post-tax 84.3 (0.66) 11.18*** 9.67 to 12.69 84.4 (0.62) 11.57*** 10.14 to 12.99 Of n=8679 cigarette smokers, we excluded those who: did not provide their current brand name (n=283); currently smoked unbranded tobacco or gave a brand of cigars/cigarillos or e-cigarettes (n=60); did not recall where they purchased their pack or received it as a gift (n=75); did not provide a valid pack size (n=231); did not report a price (n=413); purchased tobacco duty free or overseas (n=70); skipped relevant questions due to back-coding of verbatim responses (n=77); gave responses that were deemed unlikely to be correct (such as a non-existent pack size or unrealistically low or high price) (n=243); or did not have socioeconomic status information (n=151). Analysed n=7076. ***p<0.001. FM, factory made; PP, plain packaging; RYO, roll-your-own.

increase)) and for mainstream brands (36–38 (6%)), but differed by market segment (interaction F=4.94, p=0.002), with increased by 17 (19%) for value brands (results not in tables). premium brands of FM cigarettes increasing the most (5.0%), and value brands increasing the least (2.3%) (mainstream FM: 3.8%; Prices paid RYO: 3.4%). The interaction between market segment and the The proportion of respondents who reported having tried a dif- pre-PP and PP post-tax phases was also significant (F=11.88, ferent brand because it was cheaper, or who had thought in the p<0.001). Again, prices of premium FM cigarettes increased the past month about money spent on smoking, were examined most (by 15.7%), while prices of value cigarettes increased the among 8672 FM and/or RYO cigarette smokers (analysis for least (by 10.7%) (mainstream FM: 13.2%; all RYO: 13.8%) so trying a different brand excluded n=55 who refused, did not that the differential between the average stick price of premium know or responded not applicable and n=197 for whom SES compared to value brands widened considerably over the course could not be determined; analysed n=8420). Analysis for of the study, from 30.2% pre-PP to 31.3% during transition, ‘thought about money spent’ excluded n=88 who refused and 33.5% in PP year 1 and 35.7% PP post-tax (see figure 1). n=195 for whom SES could not be determined; analysed The magnitude of price increases between the pre-PP and PP n=8388. There was no change in the proportion of cigarette year 1 phases also differed by pack size among FM cigarettes smokers trying a different brand from pre-PP (14.6%) to the (interaction F=3.20, p=0.004; data not shown in tables), with transition (14.9%) or PP year 1 phases (14.7%), but significantly inflation-adjusted prices increasing the most among packs con- more smokers did so during the PP post-tax phase (17.9%) taining 21, 22 or 23 cigarettes (9.9%) or 26 cigarettes (7.8%). compared to pre-PP (adjusted (Adj)OR=1.29, 95% CI 1.03 to Inflation-adjusted prices increased the least among packs of 40 1.60, p=0.025). Similarly, compared to pre-PP (66.9%) a sig- cigarettes (by 0.7%) (20s: 4.0%; 25s: 3.9%; 30s and 35s: 4.2%; nificantly higher percentage of respondents thought about 50s: 5.1%). The interaction between pre-PP and the PP post-tax money spent on smoking several or many times in the previous period was also significant (F=6.48, p<0.001): packs of 21–23 month during the PP post-tax phase (71.0%) (AdjOR=1.22, cigarettes increased by 22.7%, while packs of 40s increased by 95% CI 1.03 to 1.45, p=0.021) (results not in tables). only 6.9% (20s: 14.9%; 25s: 14.4%; 26s: 12.7%; 30s and 35s: Table 3 shows that on average, between the pre-PP and PP 12.7%; 50s: 13.2%). year 1 phases, inflation-adjusted price paid increased by 1.9 Among current RYO smokers, the interaction between pouch cents per stick for RYO cigarettes (assuming rolled with 0.8 g of size and pre-PP and PP year 1 was not significant (F=2.27, tobacco) and by 1.8 cents per stick for FM cigarettes. The p=0.079); however, the interaction from pre-PP to PP post-tax increases in the PP post-tax phase were even greater. After was (F=4.39, p=0.004). Pouches containing 25 g of tobacco adjusting for inflation, RYO cigarettes cost an average 7.9 cents increased in price by 7.8%, while those containing 30 g more per 0.8 g after the tax increase, while FM cigarettes cost increased by 11.7% and those with 50 or 55 g increased by 7.5 cents more per stick. 16.5%. Inflation-adjusted prices paid for pouches containing 35 We then tested interactions between phase and market segment. and 40 g appeared to increase by a smaller amount (3.7%), but The magnitude of the price increase between pre-PP and PP year 1 the sample size was very small (n=14). ii72 Scollo M, et al. Tob Control 2015;24:ii66–ii75. doi:10.1136/tobaccocontrol-2014-052071 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

introduction of PP. Contrary to predictions by the tobacco industry, however, prices paid for cigarettes were higher in all segments of the market, with larger increases rather than the industry-predicted falls in prices among premium brands. Reported consumption did not change following implementa- tion despite the increased availability and use of value brands. The shift to value brands in 2013 appears to have been driven largely by the increased availability and favourable introductory pricing of value brands in smaller pack sizes. Traditionally value smokers in Australia favoured brands such as Brandon and Stradbroke (introduced in the early-1990s in packs of 40) and Holiday and Horizon (introduced in the mid-1990s in packs of 50), however, in more recent times value brands such as Choice, Pall Mall and JPS have been promoted in packs of 20 and 25 as Figure 1 Average reported price paid per single cigarette stick or well as the more traditional 40s value format.10 In the lead-up to 0.8 g or roll-your-own tobacco ($March 2014)—pre-PP, transition, PP PP, two new value brands in small pack sizes were introduced year 1 and PP post tax, by market segment and gap between value onto the market in Australia—Philip Morris’ Bond Street 20s and and premium brands, (cents per stick). Average using adjusted models. 26s and British American Tobacco’s Just Smokes 25s. Imperial PP, plain packaging. Tobacco also introduced several new extensions to its JPS range such as JPS Nano 23s, JPS 26s, JPS Ice 20s (menthol) and JPS Duo 20s (menthol hybrid). All of these packs featured striking Reported cigarette consumption pack design (such as grip patterns similar to that on some mobile Of the 8408 daily, weekly and monthly cigarette smokers in the phones on the Nano and fluorescent blue on the JPS Ice total sample, we excluded seven respondents who smoked menthols), and substantially lower up-front purchase prices.35 unbranded cigarettes only (ie, they did not smoke any commer- JPS 40s was also added to the range for smokers who still pre- cial cigarettes), 52 who did not know or refused to give their ferred large pack sizes. Products such as these might be thought level of consumption, and 185 for whom SES could not be to constitute a fourth new price segment in Australia, being determined. In total, we analysed consumption among 7218 termed in the UK and New Zealand as ‘ultra-value’52 or daily and 946 weekly or monthly smokers. ‘budget’.53 ‘Supervalue’ products have directly competed not just with value packs of 40s and 50s with much higher upfront pur- Changes in consumption following the introduction of PP chase costs, but also with mainstream brands such as Winfield 25s fi fi We rst examined daily cigarette consumption within the rst and Peter Jackson 30s which cost considerably more per stick and three phases of PP only. Among daily cigarette smokers, there per pack. Similar to industry reports to shareholders,40 41 data was no change in consumption between pre-PP and the transi- from this tracking survey suggests that JPS has been particularly tion phase or PP year 1 period, with an average daily consump- successful in attracting market share for Imperial Tobacco.54 fi tion over the rst three phases of 15.7 (SE=0.13) cigarettes per Our study has several limitations. Although the sample size day. Nor was any change detected when mean daily consump- was large, it was difficult to detect small changes in use of low- tion was analysed among regular smokers (all those who frequency products such as slims and unusual new pack sizes smoked at least once per week) (mean=14.6, SE=0.13). Mean such as 22s. Since this is a cross-sectional study we were unable daily consumption also did not change from the pre-PP to sub- to ascertain what component of the higher proportion of — sequent two phases among current smokers those who smokers smoking value brands in later phases compared to the smoked at least monthly, including daily and weekly smokers pre-PP period were new smokers entering the market or existing (mean=14.1, SE=0.13). Furthermore, consumption did not smokers who had shifted from premium to value brands, or change from pre-PP to the subsequent two phases among whether there may have been differentially more quitting current smokers of brands of any market segment (averages among premium compared to value brand smokers. fi — over the rst three phases RYO: mean=14.7, SE=0.28; value: Longitudinal studies will be required to establish the extent of mean=16.8, SE=0.27; mainstream: mean=14.1, SE=0.20; shifts in brands and brand styles among existing smokers. premium: mean=10.6, SE=0.29). Several pieces of evidence suggest that the shift to value brands observed in this study and reported in industry share- Changes in consumption following the 1 December 2013 increase holder communications12 55 were predominantly due to the tax in excise/customs duty increases and pricing strategy purposefully adopted by tobacco Daily cigarette consumption was then examined using within companies rather than to a reduced commitment to and valuing the PP year 1 and PP post-tax phases only. Mean reported daily of established brands by consumers resulting from PP.First, sales consumption among daily smokers was significantly lower PP of super-value brands have been increasing exponentially in post-tax (mean=15.1, SE=0.34; Adj B-weight=−0.80, 95% CI Australia since the major increase in excise/customs duty in – −1.56 to −0.05, p=0.037) compared to PP year 1 (mean=15.9, Australia in April 2010, well before the introduction of PP.956 SE=0.17). Similarly, there was a significant decline in mean cig- 60 Second, the increase in the number of brand–variant–pack arette consumption from PP year 1 (mean=14.8, SE=0.17) to size combinations available in the value segment predated the PP post-tax (mean=14.0, SE=0.33) among regular smokers introduction of PP and indeed appeared to rely on packaging to (Adj B-weight=−0.79, 95% CI −0.53 to −0.05, p=0.037). emphasise product characteristics.35 Third, although excise/ customs duty in Australia applies equally to every cigarette DISCUSSION regardless of price segment, average recommended retail prices This study found clear evidence of an increase in the proportion increased more for premium than for value brands after the of current smokers using value brands following the August 2012 and February and August 2013 indexations of duty

Scollo M, et al. Tob Control 2015;24:ii66–ii75. doi:10.1136/tobaccocontrol-2014-052071 ii73 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper were accommodated.36 Fourth, the percentages of smokers who contemplating tobacco PP legislation should specifically prohibit reported that they had tried a different brand because it was slim cigarettes as the Australian example demonstrates that man- cheaper or thought about how much they were spending on dated increases in packaging dimensions is not a sufficient cigarettes in the PP year 1 phase were no higher than in the means of outlawing these products.35 pre-PP phase but were significantly higher after the December Industry claims that PP would lead to a collapse in prices and 2013 tax increase a year after introduction of PP. Finally and increased consumption of tobacco products have not been most importantly it should be noted that similar shifts, particu- borne out. Scheduled large increases in excise/customs duty,43 44 larly by low-SES smokers, from premium to discount brands are likely to further reduce consumption but will no doubt also have been observed in many other countries following large tax further accelerate the shift to value brands in Australia. The – increases.53 61 63 A particularly marked increase in brand findings of this study suggest regulation of price-based promo- switching from premium to discount cigarettes was observed tion such as multibuys and novel pack sizes as a new frontier in after 2009 corresponding to the $0.61 increase in the federal tobacco control to be tackled by governments in Australia. excise tax on cigarettes in the USA.61 62 Notable shifts to use of fi ultra-value brands have also recently been documented after Contributors MS and MW designed this study. MZ cleaned data les, undertook 52 53 analysis and reported results. KC undertook additional analysis and MB provided increases in taxes on tobacco in the UK and New Zealand additional advice on analysis. MS drafted the manuscript with contributions from which were comparable in magnitude to recent increases in all authors. All authors approved the final manuscript. 43 44 taxes on tobacco products in Australia. Funding The National Plain Packaging survey was funded under a contract with Variation in pack size remains a powerful form of promotion the Australian Government Department of Health and Ageing. in Australia that reduces the effectiveness of tax policy in dis- Competing interests The authors wish to advise that MS was a technical writer couraging consumption of tobacco products. The strategy of for and MW a member of the Tobacco Working Group of the Australian National providing extra cigarettes per pack was used to full effect by Preventive Health Task Force and MW was a member of the Expert Advisory tobacco companies in the lead up to introduction of PP. Committee on Plain Packaging that advised the Australian Department of Health on research pertaining to the plain packaging legislation. MW holds competitive grant Governments contemplating PP legislation elsewhere would be funding from the Australian National Health and Medical Research Council and MW well advised to include provision in such legislation to standard- holds competitive grant funding from the US National Institutes of Health, Australian ise the number of cigarettes per pack. National Preventive Health Agency and BUPA Health Foundation. The decline in reported consumption by (remaining) regular Ethics approval The survey was approved by the Cancer Council Victoria Human smokers following the 12.5% increase in excise/customs duty in Ethics Committee (HREC 0018). December 2013 was as expected based on internationally Provenance and peer review Not commissioned; externally peer reviewed. − 45 64 accepted estimates of price elasticity of about 0.4%. This is Open Access This is an Open Access article distributed in accordance with the despite December 1 perhaps not being optimal timing for a duty Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which increase with end-of-year festivities and summer holidays posing permits others to distribute, remix, adapt, build upon this work non-commercially, many risks of relapse for smokers who may have quit. Changes in and license their derivative works on different terms, provided the original work is consumption following increases in 2014, 2015 and 201643 44 properly cited and the use is non-commercial. See: http://creativecommons.org/ licenses/by-nc/4.0/ should be monitored to ascertain whether at the more conducive timing of September 1, price elasticity of demand in the PP envir- REFERENCES onment might be even greater than previously observed. 1 Freeman B. Chapter 10. The tobacco industry in Australian society. In: Scollo M, Use of king-sized cigarettes remained low and there was no Winstanley M. eds. Tobacco in Australia: facts and issues. Melbourne: Cancer Council increase in the use of menthol cigarettes. It is possible that PP Victoria, 2013. http://www.tobaccoinaustralia.org.au/chapter-10-tobacco-industry legislation has prevented the development of a market for such 2 Gartner C. Chapter 12, Attachment 12.3 Smokeless tobacco and harm reduction. fi In: Scollo M, Winstanley M. eds. Tobacco in Australia: facts and issues. Melbourne: products in Australia, however, the small but signi cant increase Cancer Council Victoria, 2013. http://www.tobaccoinaustralia.org.au/chapter-12- in use of slims following the introduction of PP suggests that tobacco-products use of such novel cigarettes should continue to be monitored 3 Therapeutic Goods Administration. Electronic cigarettes. Canberra: Department of particularly among young people. Other countries health, 2014. http://www.tga.gov.au/consumers/ecigarettes.htm#supplying (accessed Sep 2014). 4 Young D, Yong HH, Borland R, et al. Trends in roll-your-own smoking: findings from the ITC Four-Country Survey (2002–2008). J Environ Public Health What this paper adds 2012;2012:406283. 5 Euromonitor International. Tobacco-Australia 2002 to 2013, Available for purchase. London: Euromonitor International, 2014. http://www.euromonitor.com/tobacco What is already known on this topic (accessed Aug 2013). ▸ 6 King B. Chapter 12. The construction and labelling of Australian cigarettes. In: The tobacco industry vigorously opposed the introduction of Scollo M, Winstanley M. eds. Tobacco in Australia: facts and issues. Melbourne: plain packaging in Australia and warned of many unintended Cancer Council Victoria, 2013. http://www.tobaccoinaustralia.org.au/ consequences, in particular a shift to use of low-cost cigarettes chapter-12-tobacco-products and a consequent increase in tobacco consumption. 7 Liberman J. Chapter 16. Tobacco litigation in Australia, 16.2 Other litigation involving the tobacco industry. In: Scollo M, Winstanley M. eds. Tobacco in What this paper adds Australia: facts and issues. Melbourne: Cancer Council Victoria, 2013. http://www. ▸ Use of value brands did increase following the introduction tobaccoinaustralia.org.au/chapter-16-litigation/16-2-other-litigation-involving-the- tobacco-indust of plain packaging (PP), however, this appears to have been 8 King B, Borland R. What was ‘light’ and ‘mild’ is now ‘smooth’ and ‘fine’:new driven by the introduction of smaller pack sizes in value labelling of Australian cigarettes. Tob Control 2005;14:214–15. brands in the months leading up to the introduction of PP, 9 Anon. Retail World Annual Report 2013. Market sizes and shares. Retail World combined with larger increases in prices of premium brands 2013;December:30. 10 Scollo M. Chapter 13. The pricing and taxation of tobacco products in Australia. relative to value brands. In: Scollo M, Winstanley M. eds. Tobacco in Australia: facts and issues. Melbourne, ▸ Among smokers, reported number of cigarettes smoked did Australia: Cancer Council Victoria, 2013. http://www.webcitation.org/6QsaCaM3R not change after PP introduction, but did reduce among 11 Scollo M, Bayly M, Wakefield M. The advertised price of cigarette packs in retail regular smokers post a 12.5% tax increase. outlets across Australia before and after the implementation of plain packaging: a repeated measures observational study. Tob Control 2015;24:ii82–9. ii74 Scollo M, et al. Tob Control 2015;24:ii66–ii75. doi:10.1136/tobaccocontrol-2014-052071 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

12 Calantzopoulos A. Philip Morris presentation. Morgan Stanley Global Consumer & Retail 36 Scollo M, Bayly M, Wakefield M. Did the recommended retail price of tobacco products Conference; New York; 19 November 2014. Morgan Stanley Global, 2014. http:// fall in Australia following the implementation of plain packaging? Tob Control 2015;24: investors.pmi.com/phoenix.zhtml?c=146476&p=irol-presentations (accessed Nov 2014). ii26–32. 13 NSW Retail Tobacco Traders’ Association. Price lists–cigarettes. Aust Retail 37 MacKillop J, Amlung MT, Blackburn A, et al. Left-digit price effects on smoking Tobacconist 2014;90:13–20. cessation motivation. Tob Control 2014;23:501–6. 14 Burton S, Williams K, Fry R, et al. Marketing cigarettes when all else is unavailable: 38 Scollo M, Hayes L, Wakefield M. What price quitting? The price of cigarettes at which evidence of discounting in price-sensitive neighbourhoods. Tob Control 2014;23:e24–9. smokers say they would seriously consider trying to quit. BMC Public Health 2013;13:650. 15 Scollo M, Bayly M. Section 13.3 The price of tobacco products in Australia in 39 NSW Retail Tobacco Traders’ Association. Price lists–cigarettes. Aust Retail Chapter 13. The pricing and taxation of tobacco products in Australia (December Tobacconist 2013;87:1–2. 2014 update). Melbourne, Australia: Cancer Council Victoria, 2014. http://www. 40 Cooper A. Preliminary results 2013, Transcript of presentation to shareholders. tobaccoinaustralia.org.au/chapter-13-taxation/13-3-the-price-of-tobacco-products-in- London: Imperial Tobacco Ltd, 2013. http://www.imperial-tobacco.com/index.asp? australia (accessed Nov 2014). page=56&newsid=1844 (accessed Aug 2014). 16 Scollo M, Lindorff K, Coomber K, et al. Standardised packaging and new enlarged 41 Cooper A. Preliminary Results 2014 Imperial Tobacco Group PLC. Imperial Tobacco, graphic health warnings for tobacco products in Australia—legislative requirements 2014. http://www.imperial-tobacco.com/index.asp?page=78&newsid=2094 and implementation of the Tobacco Plain Packaging Act 2011 and the Competition (accessed Nov 2014). and Consumer (Tobacco) Information Standard, 2011. Tob Control 2015;24:ii9–16. 42 Government of Australia. Explanatory Memorandum to Excise Tariff Amendment 17 Hyland A, Higbee C, Bauer J, et al. Cigarette purchasing behaviors when prices are (Tobacco) Bill 2014. Australia, 2014. http://parlinfo.aph.gov.au/parlInfo/search/ high. J Public Health Manag Pract 2004;10:497–500. display/display.w3p;query=Id%3A%22legislation%2Fems%2Fr5172_ems_ 18 Hyland A, Bauer JE, Li Q, et al. Higher cigarette prices influence cigarette purchase 78ed6212-03a3-4669-9947-c3fc17c8b5e0%22 patterns. Tob Control 2005;14:86–92. 43 Excise Tariff Amendment (Tobacco) Act 2014, Stat. No 9 (2014). http://www. 19 Hyland A, Higbee C, Li Q, et al. Access to low-taxed cigarettes deters smoking comlaw.gov.au/Details/C2014A00009 cessation attempts. Am J Public Health 2005;95:994–5. 44 Customs Tariff Amendment (Tobacco) Act 2014, Stat. No 8 (2014). http://www. 20 Hyland A, Laux FL, Higbee C, et al. Cigarette purchase patterns in four countries comlaw.gov.au/Details/C2014A00008 and the relationship with cessation: findings from the International Tobacco Control 45 The World Bank. Curbing the epidemic: governments and the economics of tobacco (ITC) Four Country Survey. Tob Control 2006;15:iii59–64. control. Tob Control 1999;8:196–201. 21 Licht AS, Hyland AJ, O’Connor RJ, et al. Socio-economic variation in price 46 World Health Organization. Appendix IX. Tobacco taxes and prices. WHO report on minimizing behaviors: findings from the International Tobacco Control (ITC) Four the global tobacco epidemic, 2013 Enforcing bans on tobacco advertising, Country Survey. Int J Environ Res Public Health 2011;8:234–52. promotion and sponsorship. Geneva: WHO, 2013. http://www.who.int/tobacco/ 22 British American Tobacco Australia. Submission on the Tobacco Plain Packaging Bill global_report/2013/en/index.html 2011 . Canberra: Department of Health and Ageing Consultation website, 2011. 47 Chaloupka FJ, Yurekli A, Fong GT. Tobacco taxes as a tobacco control strategy. http://content.webarchive.nla.gov.au/gov/wayback/20130904173432/http://www. Tob Control 2012;21:172–80. yourhealth.gov.au/internet/yourhealth/publishing.nsf/Content/ 48 Wakefield M, Coomber K, Zacher M, et al. Australian adult smokers’ responses 2EE2F6F3EC74F628CA2579540005F68B/$File/British%20American%20Tobacco% to plain packaging with larger graphic health warnings 1 year after implementation: 20Australia%20-%20Public%20Submission.pdf (accessed Oct 2014). results from a national cross-sectional tracking survey. Tob Control 23 Philip Morris Limited. Commoditising tobacco products through plain packaging will 2015;24:ii17–25. harm public health, violate treaties and does not meet the test of ‘evidence-based 49 Australian Taxation Office. Excise duty on tobacco products. Canberra: ATO, 2014. policy’. Melbourne, 2011. http://webarchive.nla.gov.au/gov/20130329051831/http:// http://law.ato.gov.au/atolaw/view.htm?docid=PAC/BL030002/1 (accessed Mar 2014). www.yourhealth.gov.au/internet/yourhealth/publishing.nsf/Content/ 50 Australian Bureau of Statistics. 6401.0 Consumer Price Index, Australia Table 11. Phillip-Morris-Limited~Phillip-Morris-Limited-2 (accessed Oct 2014). CPI: group, sub-group and expenditure class, inex numbers by capital city [database 24 Imperial Tobacco Australia Limited. Submission to the department of health and on the Internet]. ABS, 2014. (cited March 2014). http://www.abs.gov.au/AUSSTATS/ ageing regarding the Tobacco Plain Packaging Bill 2012 (exposure draft) and [email protected]/DetailsPage/6401.0Jun%202012?OpenDocument consultation Paper. Sydney, 2011. http://webarchive.nla.gov.au/gov/ 51 Australian Bureau of Statistics. Technical paper: socio-economic indexes for areas 20130329044623/http://www.yourhealth.gov.au/internet/yourhealth/publishing.nsf/ (SEIFA) 2011. Canberra: Australian Capital Territory: ABS, 2013. http://www.abs. Content/ppit~ncr~loe (accessed 10 Jan 2012). gov.au/websitedbs/censushome.nsf/home/seifa2011?opendocument&navpos=260 25 Davidson S. Cherry-picking stats an uncertain science. The Australian. 2014. http:// (accessed Apr 2014). www.theaustralian.com.au/ 52 Gilmore AB, Tavakoly B, Hiscock R, et al. Smoking patterns in Great Britain: the rise of 26 Sloane J. Claims plain packaging works go up in smoke. The Australian. 2014. cheap cigarette brands and roll your own (RYO) tobacco. J Public Health (Oxf) http://www.theaustralian.com.au/ Published Online First: 11 Aug 2014. doi:10.1093/pubmed/fdu048 27 British American Tobacco (BAT). UK Standardised Packaging Consultation: Response 53 Action on Smoking and Health New Zealand. 2013 tobacco returns. Auckland: ASH of British American Tobacco UK Limited. 2012. http://www.bat.com/group/sites/ NZ, 2014. http://www.ash.org.nz/ (accessed Sep 2014). uk__3mnfen.nsf/vwPagesWebLive/DO8WZC5E/$FILE/medMD8WZC6J.pdf? 54 Imperial Tobacco. Half-year results 2013, Imperial Tobacco Group plc: report to openelement (accessed May 2014). investors, slide 51. London: Imperial Tobacco, 2013. http://www.imperial-tobacco. 28 British American Tobacco (BAT). UK Standardised Appendix A, Report of Dr. Jonathan com/index.asp?page=150 (accessed 31 Jul 2013). Klick. Appendix A. 2012. http://www.bat.com/group/sites/uk__3mnfen.nsf/vwPages 55 Calantzopolous A. Presentation from CEO of Philip Morris International. CAGNY WebLive/DO8WZC5E/$FILE/medMD8WZC6J.pdf?openelement (accessed May 2014). Conference; Boca Raton Florida; Tuesday, 18 February 2014. http://investors. 29 British American Tobacco (BAT). UK Standardised, Report of Mr. Stephen Gibson. philipmorrisinternational.com/phoenix.zhtml?c=146476&p=irol-eventDetails& Appendix B. 2012. http://www.bat.com/group/sites/uk__3mnfen.nsf/vwPages EventId=5082097 (accessed Jul 2014). WebLive/DO8WZC5E/$FILE/medMD8WZC6J.pdf?openelement (accessed May 2014). 56 Anon. Retail World Annual Report 2008. Market sizes and shares. Retail World 30 British American Tobacco Australia. Transcript of interview of representatives of 2008;December:30. British American Tobacco Australia and Imperial Tobacco Australia with Sir Cyril 57 Anon. Retail World Annual Report 2009. Market sizes and shares. Retail World Chantler. London: Kings College, 2014. http://www.kcl.ac.uk/health/Packaging- 2009;December:30. review/packaging-review-docs/meetingsandbriefings/British-American-Tobacco- 58 Anon. Retail World Annual Report 2010. Market sizes and shares. Retail World Australian-and-Imperial-Tobacco-Australia-12-March-2014.pdf (accessed Aug 2014). 2010;December:30. 31 British American Tobacco (New Zealand) Ltd. Proposal to introduce plain packaging 59 Anon. Retail World Annual Report 2011. Market sizes and shares. Retail World of tobacco products in New Zealand: submission by British American Tobacco (New 2011;December:30. Zealand) Limited. Wellington: Parliament of New Zealand, 2014. http://www. 60 Anon. Retail World Annual Report 2012. Market sizes and shares. Retail World parliament.nz/en-nz/pb/sc/documents/evidence?Custom=00dbhoh_bill12969_ 2012;December:30. 1&Criteria.PageNumber=2 (accessed May 2014). 61 Cornelius ME, Driezen P, Fong GT, et al. Trends in the use of premium and discount 32 Scollo M, Zacher M, Coomber K, et al. Use of illicit tobacco following cigarette brands: findings from the ITC US Surveys (2002–2011). Tob Control implementation of plain packaging of tobacco products in Australia: results from 2014;23(Suppl 1):i48–53. a national cross-sectional survey. Tob Control 2015;24:ii76–81. 62 Cornelius ME, Driezen P, Hyland A, et al. Trends in cigarette pricing and purchasing 33 WakefieldM,ZacherM,ScolloM,et al. Brand placement on price boards after tobacco patterns in a sample of US smokers: findings from the ITC US surveys (2002–2011). display bans: a point-of-sale audit in Melbourne, Australia. Tob Control 2012;21:589–92. Tob Control Published online first 10 Jun 2014; doi:10.1136/tobaccocontrol-2013- 34 Zacher M, Germain D, Durkin S, et al. A store cohort study of compliance with a point- 051376 of-sale cigarette display ban in Melbourne, Australia. Nicotine Tob Res 2013;15:444–9. 63 Gilmore AB, Reed H. The truth about cigarette price increases in Britain. 35 Scollo M, Occleston J, Bayly M, et al. Tobacco product developments coinciding Tob Control 2014;23:e15–16. with the implementation of plain packaging in Australia. Tob Control Published 64 Gallet C, List J. Cigarette demand: a meta-analysis of elasticities. Health Econ Online First: 30 Apr 2014; doi:10.1136/tobaccocontrol-2013-051509 2003;12:821–35.

Scollo M, et al. Tob Control 2015;24:ii66–ii75. doi:10.1136/tobaccocontrol-2014-052071 ii75 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Changes in use of types of tobacco products by pack sizes and price segments, prices paid and consumption following the introduction of plain packaging in Australia Michelle Scollo, Meghan Zacher, Kerri Coomber, Megan Bayly and Melanie Wakefield

Tob Control 2015 24: ii66-ii75 doi: 10.1136/tobaccocontrol-2014-052071

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii66

These include: Supplementary Supplementary material can be found at: Material http://tobaccocontrol.bmj.com/content/suppl/2015/02/09/tobaccocon trol-2014-052071.DC1.html References This article cites 19 articles, 11 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii66#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper Use of illicit tobacco following introduction of standardised packaging of tobacco products in Australia: results from a national cross-sectional survey Michelle Scollo, Meghan Zacher, Kerri Coomber, Melanie Wakefield

– Centre for Behavioural ABSTRACT universe of all packs used by smokers.14 17 Further, Research in Cancer, Cancer Objectives To assess whether following such studies cannot determine the proportion of Council Victoria, Melbourne, ’ Victoria, Australia standardisation of tobacco packaging in Australia, packs non-compliant with Australia s packaging smokers were, as predicted by the tobacco industry, legislation that were legitimately purchased overseas – Correspondence to more likely to use illicit tobacco. either duty-free or with duty paid on entry.16 18 Dr Michelle Scollo, Centre for Methods National cross-sectional telephone surveys While consumer surveys provide alternative esti- Behavioural Research in conducted continuously from April 2012 (6 months mates of use of illicit tobacco not subject to these Cancer, Cancer Council Victoria, 615 St Kilda Rd, before implementation of plain packaging (PP)) to March problems, respondents may not know the tax-paid Melbourne, VIC 3004, 2014 (15 months after) using responses from current status of the tobacco products they purchase, may Australia; cigarette smokers (n=8679). Changes between pre-PP, not admit to knowingly buying illicit tobacco or [email protected] the transition to PP and PP phase were examined using conversely may be more ready to believe than is Received 29 September 2014 logistic regression models. warranted that cigarettes sold at low prices are Accepted 13 January 2015 Results Among those whose factory-made cigarettes counterfeit. However, directly observing smokers’ were purchased in Australia, compared with pre-PP, packs19 or asking smokers objective concrete facts there were no significant increases in the PP phase in about their packs is likely to elicit responses that use of: ‘cheap whites’ (<0.1%; OR=0.24, 95% CI 0.04 are both honest and less subject to respondent to 1.56, p=0.134); international brands purchased for error. 20% or more below the recommended retail price In their study across 18 European countries, (0.2%; OR=3.49, 95% CI 0.66 to 18.35, p=0.140); or Joossens et al19 asked cigarette smokers to show packs purchased from informal sellers (<0.1%; field interviewers their latest purchased pack of OR=0.24, 95% CI 0.04 to 1.47, p=0.124). The cigarettes or hand-rolled tobacco. Packs were prevalence of any use of unbranded illicit tobacco defined as illicit if they had at least one of the fol- remained at about 3% (adjusted OR=0.79, 95% CI lowing tax evasion indicators: (1) it was bought 0.58 to 1.08, p=0.141). from illicit sources, as reported by smokers, (2) it Conclusions While unable to quantify the total extent had an inappropriate tax stamp, (3) it had an of use of illicit manufactured cigarettes, in this large inappropriate health warning or (4) its price was national survey we found no evidence in Australia of substantially below the known price in their increased use of two categories of manufactured market. Adaptation of such a study in Australia cigarettes likely to be contraband, no increase in needed to take account of that fact that tax stamps purchase from informal sellers and no increased use of are not required in Australia and that only very unbranded illicit ‘chop-chop’ tobacco. well-funded studies could include fieldworkers able to examine packs. While an audit of Australian retailers’ willingness to sell illicit tobacco conducted INTRODUCTION over 2012 and 2013 found some clear instances of Opponents of plain packaging (PP) in Australia products without health warnings,20 the price was claimed that the legislation would have major unin- somewhat volatile over this period with instances – tended consequences,1 3 in particular that the stan- of retailers trying to run down supplies of fully dardised appearance of the packs would make branded tobacco products, and a few cases of retai- them easier to counterfeit4 and that this, combined lers selling stock in the old packaging at reduced with the reduced valuing of brands, would lead to rates afterwards.20 Further, tobacco companies an increase in use. The increased use of contraband claimed that prices of tax-paid products would fall sharply as a result of PP. Price was therefore not a Open Access cigarettes as well as unbranded illicit tobacco, it Scan to access more was argued, would disadvantage retailers and stable or reliable indicator of illicit status over the free content advantage ‘criminal gangs’ purported to be widely period immediately prior to and after the introduc- involved in its distribution.56Many of the same tion of PP. 20 arguments have been proffered by those opposed The World Customs Organization has identified to the introduction of standardised packaging in the problem of cigarettes produced by operators in the UK,78Ireland9 and New Zealand.10 11 one country being smuggled to the illicit market in Studies funded by Australian tobacco companies, other countries—so-called ‘cheap whites’.21 Its which rely heavily on surveys of discarded cigarette 2102 annual report listed a number of ‘brand’ To cite: Scollo M, packs, have reported increases in the size of the names of cigarettes falling into this category. The Zacher M, Coomber K, et al. illicit tobacco market following the introduction of report also listed several major international brands Tob Control 2015;24: PP legislation.12 13 Packs discarded in public places that appeared frequently among cigarettes identi- – ii76 ii81. are likely to provide a poor representation of the fied by customs officials and intercepted because it ii76 Scollo M, et al. Tob Control 2015;24:ii76–ii81. doi:10.1136/tobaccocontrol-2014-052072 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper was believed that the required taxes had not been paid. Any additional brands (Nanjing, Su Yan and Zig Zag)as‘cheap reported purchase of ‘cheap whites’ in Australia would represent whites’, as no registered manufacturer could be located. clear cases of use of illicit tobacco. International brands such as Marlboro, Dunhill and Benson & Hedges are regarded as ‘International brands purchased at suspiciously low prices’ premium products in Australia and are rarely discounted.22 Any FM brands named in World Customs reports21 as frequently instances of purchase of such products at highly discounted smuggled included Benson & Hedges, Brooks, Camel, Craven A, prices would also therefore be highly likely to be illicit. Dunhill, Furongwang, Golden Eagle, LM, Lucky Strike, A study our group conducted in the Australian state of Marlboro, Mild Seven, Philip Morris, Rothmans, Shuangxi and Victoria between 2011 and 2013—1 year before and 1 year 555. We considered current packs of these brands suspicious after the introduction of PP—detected no changes among where the pack contained 20 cigarettes (the standard inter- smokers in use of unbranded tobacco, and low levels of use in national pack size, and the one most frequently smuggled) and November 2013 of cigarettes deemed likely to be contraband.23 where the reported price paid was 20% or more below the The current study aims to extend this analysis using a large recommended retail price (RRP), as reported in the Australian national data set and including baseline estimates of prevalence Retail Tobacconist.27 Where cigarettes were purchased in cartons of smokers reporting any use of several categories of cigarettes from large retail outlets, this cut-off point was relaxed slightly to which are highly likely to be contraband. We aimed to assess the accommodate prices just a few cents below the cut-off and to significance of changes since the period prior to PP in the per- take account of typically lower average prices for carton sales. We centage of current cigarette smokers whose current pack of also considered packs of 25 cigarettes suspicious if the reported cigarettes: price was 20% or more below the RRP of a pack of 20 cigarettes ▸ Carried a ‘brand’ name corresponding with known ‘cheap of that brand. whites’; ▸ Was a brand frequently observed in international intercep- Purchase from informal sellers tions of contraband tobacco and was purchased in Australia Current smokers were also asked “Where did you get the cigar- at a suspiciously cheap price; ettes you are currently smoking?” Those who indicated that rela- ▸ Was purchased from an informal seller. tives or friends had purchased their cigarettes were further asked We also aimed to assess any changes in the proportion of smokers from where this person had purchased them. Responses coded as ▸ Who had any self-reported recent use of illicit unbranded a ‘purchase from an informal seller’ were ‘from someone selling tobacco. independently and/or illegally (not at a store, shop or other main- stream establishment, but perhaps at local markets, delivery METHODS service, door-to-door, in a pub or just in the street)’. Design and participants We used data from national cross-sectional telephone surveys Unbranded illicit ‘chop-chop’ tobacco which were conducted continuously between 9 April 2012 and Use of unbranded illicit tobacco known in Australia as ‘chop- 30 March 2014, with about 100 respondents per week, chop’ could be reported in several ways throughout the survey. described more fully in Wakefield et al24 (this volume). After the initial question inquiring about tobacco products used, This study includes n=8679 adult smokers who reported smokers were asked whether they currently smoked any other smoking at least once per month. Overall, 55% were males; form of tobacco and, if so, which types; some respondents men- slightly more than one-quarter (29%) were aged 18–29 years, tioned unbranded tobacco here. Respondents could also volun- 46% were 30–49 years, and 26% were 50–69 years. teer ‘unbranded tobacco’ as their ‘current brand’, usual brand Additionally, 39% of smokers lived in low socioeconomic status (the brand smoked more than any other) or as ‘another brand’ (SES) areas (as determined by a postcode-based measure of rela- they frequently smoked. The few respondents who said they did tive social disadvantage),25 42% in mid-SES and 18% in not know the brand name or refused to give it were directly high-SES areas. About one in three had not completed high asked whether it was unbranded tobacco. Finally, all respon- school (32%), about half had completed high school and/or had dents who did not report the use of unbranded tobacco in some tertiary education (56%), and 13% had completed tertiary another way were directly asked if, in the past month, they had or higher education. The sample was limited to eligible respon- bought loose unbranded tobacco or cigarettes (“bought loose dents for each of the various analyses. unbranded tobacco in a plastic bag, that is also known as chop- chop” or “bought unbranded or chop-chop tobacco that has Measures already been rolled into cigarettes”). From these questions, we All respondents were asked “Do you currently smoke factory- created a measure indicating whether respondents reported any made cigarettes only, roll-your-own cigarettes only, both, or use of unbranded tobacco. neither of these?” Those who smoked factory-made (FM) or roll-your-own (RYO) cigarettes were asked to provide the name Analysis of the brand of the last cigarette they smoked as well as the We included current smokers of cigarettes (FM, RYO or variant name, pack size and pack number (single pack, multibuy unbranded) in our sample (n=8679). Data were weighted to or carton). account for telephony status (landline or mobile phone), gender, age by education and state of residence.24 All analyses were con- ‘Cheap whites’ ducted in Stata V.12.1, adjusting for the effects of sample Among FM cigarettes, ‘cheap whites’ included brands listed as weighting on parameter estimates and SEs. We undertook logis- produced specifically for the illicit market in the World Customs tic regression analyses to assess changes in outcomes between Organization’s Illicit Trade report 201221 and in reports by the pre-PP (April to September 2012, n=2193), the transition international consulting company KPMG LLP12 26 including phase (October and November 2012, n=765) and the PP phase Dunstan, Jin Ling, Manchester, Modern, Regal, Sunlite, Timeless (December 2012–March 2014, n=5721). We also tested Time, Win, YunYan and . We coded three whether there were linear changes (eg, gradual increases) over

Scollo M, et al. Tob Control 2015;24:ii76–ii81. doi:10.1136/tobaccocontrol-2014-052072 ii77 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper the months during the PP phase. In analyses of use of was no difference in odds from the pre-PP to the PP phase unbranded tobacco, we controlled for sex, age group, SES and (table 1). education. In analyses to see whether prevalence of three cat- egories of packs likely to be contraband changed over time, we Unbranded ‘chop-chop’ tobacco excluded packs reported as purchased duty-free or from over- Table 2 shows the unadjusted and adjusted proportions of four seas. Owing to extremely low prevalence, these analyses did not categories of use of unbranded illicit tobacco and combined as adjust for covariates. ‘any reported unbranded illicit tobacco use’. There was no To exclude the possibility that the pattern of changes (or lack change between the pre-PP and the latter two phases for thereof) was attributable to the 12.5% excise/customs duty regular, current or other unbranded tobacco use; the reported increase that was effective from 1 December 2013 rather than use in these three categories was less than 1% in all phases. The to PP, we conducted sensitivity testing excluding the months prevalence of purchasing unbranded tobacco in the past month after the tax increase (December 2013 through March 2014). declined significantly from the pre-PP to the PP phase. Lastly, any reported use of unbranded illicit tobacco was between 3% and 4% in each phase and did not change significantly between RESULTS the pre-PP and the PP phase or over the months of PP. ‘Cheap whites’ Across the entire survey period, under 0.5% (unadjusted) of Sensitivity testing brands of last FM cigarette smoked were classified as ‘cheap Sensitivity tests excluding the months after the 12.5% increase whites’, and prevalence did not change overall between the in tobacco excise and customs duties in December 2013 (PP pre-PP and the PP phase (table 1). Over the months within PP, post-tax) produced a slightly different pattern of findings. The the prevalence of these brands declined significantly. significant linear decrease in the prevalence of ‘cheap whites’ in the months of PP was no longer significant (unadjusted ‘International brands purchased at suspiciously low prices’ OR=1.01, p=0.789). No respondents at all purchased their The prevalence of international brands of FM cigarettes pur- pack of FM cigarettes from an informal seller in this restricted chased in Australia for 20% or more below the RRP was low in sample (December 2012–November 2013). The non-significant all phases (under 0.5%, unadjusted) and did not increase linear decline (p=0.054) in current use of unbranded tobacco between pre-PP and PP,nor linearly during PP. in the months during PP became significant when excluding the post-tax period (adjusted OR=0.87, p=0.042). The non- significant linear decline in use of illicit unbranded tobacco in Packs purchased from informal sellers the months during PP became significant when we excluded the ’ The prevalence of having purchased one s current pack of FM post-tax period (adjusted OR=0.94, p=0.042). These results cigarettes from an informal seller did increase over the months suggest that the use of all three forms of illicit tobacco declined of PP, but overall the prevalence remained negligible and there after the introduction of PP, but that the use of contraband cigarettes and unbranded tobacco may have increased slightly after the 1 December 2013 increase in excise/customs duty. Table 1 Prevalence of use of three categories of cigarettes likely to be contraband among the most recently purchased pack of FM DISCUSSION cigarettes purchased in Australia—percentages and results of The results of this large national study confirm those of our logistic regression models smaller state-based study conducted in Victoria which found no Differences between PP phases increase in use of unbranded illicit tobacco and low levels of use — †‡ Linear trend during 23 unadjusted models — of cigarettes likely to be contraband. This study extends these PP unadjusted model fi Time period Per cent OR 95% CI OR ndings by showing no increase in use, before versus after the introduction of PP, of two key subsets of cigarettes likely to be ‘Cheap whites’ contraband—those that are so-called ‘cheap whites’ produced Pre-PP 0.2 1.00 specifically for the illicit market, and those brands frequently Transition 0.2 1.15 0.12 to 11.39 identified in customs interceptions internationally and purchased PP <0.1 0.24 0.04 to 1.56 0.92* by survey respondents for a suspiciously low price. We also Suspiciously priced international brand found no increase in use of cigarettes purchased from informal Pre-PP <0.1 1.00 sources. The prevalence of any level of use of unbranded Transition 0.1 1.80 0.16 to 20.85 tobacco did not change and purchasing unbranded tobacco in PP 0.2 3.49 0.66 to 18.35 0.97 the past month declined following the implementation of PP. Purchased from informal seller§ This study used a considerably larger national sample and was Pre-PP 0.1 1.00 also stronger than the Victorian one in that it asked respondents Transition 0.2 1.83 0.19 to 17.84 to report on the brand of cigarettes they were currently smoking PP <0.1 0.24 0.04 to 1.47 1.63* rather than only asking respondents about their ‘regular’ brand ‘ ’ *p<0.05. and their usual price and place of purchase. Respondent †Of n=8679 cigarette smokers, n=6658 said the last cigarette they smoked was a reporting of the brand they are currently smoking is likely to be brand of FM cigarettes. We excluded those who did not provide their brand name of the last cigarette they smoked (n=193), or who reported a brand of e-cigarettes, less subject to error and approximation. Like the Victorian cigars or cigarillos (n=11), or who purchased their current pack duty-free or from study, this survey recruited participants via random digit dialling overseas (n=60). Analysed n=6395. to landline and mobile phones, and is more likely than ‡Models did not control for respondent characteristics due to low cell sizes. §In addition, those who did not provide a valid pack size (n=126) were not asked landline-only or online surveys to be representative of smoking where they purchased their cigarettes from and were excluded from the ‘informal patterns in the Australian smoker population. seller’ analyses; analysed n=6268. Our study has several limitations. It excluded those people FM, factory-made; PP, plain packaging. who do not speak English—just over 2.6% of the Australian ii78 Scollo M, et al. Tob Control 2015;24:ii76–ii81. doi:10.1136/tobaccocontrol-2014-052072 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 2 Prevalence of categories of unbranded illicit tobacco use among cigarette smokers—percentages and results of logistic regression model Differences between PP phases—unadjusted Differences between PP phases—adjusted Linear trend during models† models†‡ PP—adjusted models Time period Per cent OR 95% CI Per cent OR 95% CI OR

Regular unbranded use Pre-PP 0.1 1.00 0.1 1.00 Transition 0.4 2.50 0.43 to 14.58 0.4 2.45 0.44 to 13.63 PP 0.2 1.48 0.47 to 4.67 0.2 1.48 0.46 to 4.73 1.03 Current unbranded use Pre-PP 0.2 1.00 0.2 1.00 Transition 0.1 0.42 0.05 to 3.61 0.1 0.41 0.05 to 3.58 PP 0.4 2.42 0.90 to 6.54 0.4 2.40 0.89 to 6.50 0.90 Nominated as ‘another type of tobacco smoked’ (other unbranded use) Pre-PP 0.2 1.00 0.2 1.00 Transition 0.1 0.68 0.07 to 6.14 0.1 0.62 0.07 to 5.79 PP 0.3 1.43 0.44 to 4.63 0.3 1.37 0.42 to 4.45 0.94 Purchased unbranded tobacco in the past month Pre-PP 3.2 1.00 3.2 1.00 Transition 3.0 0.94 0.55 to 1.60 2.9 0.91 0.53 to 1.57 PP 2.1 0.64* 0.45 to 0.91 2.1 0.63* 0.44 to 0.90 0.97 Any unbranded tobacco use§ Pre-PP 3.7 1.00 3.7 1.00 Transition 3.5 0.96 0.59 to 1.57 3.5 0.93 0.57 to 1.53 PP 3.0 0.80 0.58 to 1.09 3.0 0.79 0.58 to 1.08 0.96 *p<0.05. †Of n=8679 cigarette smokers, we excluded those who did not have SES information (n=199). Analysed n=8480. ‡Model controlled for sex, age group, SES and education. §Any of the four categories of unbranded use. PP, plain packaging; SES, socioeconomic status.

population28—so it may have slightly under-represented those identified counterfeited cigarettes as a significant problem. In recent immigrants who still prefer brands from their country of fact, the latest industry-funded study by KPMG LLP states that to birth and who continue to buy these either on return visits or date in empty pack studies ‘there has been no evidence of coun- who receive them through the mail or by some other means. It terfeit plain packaging cigarettes’ (ref. 13, p.40). did not include packs purchased over the internet or overseas, Some respondents may have been reluctant to report on the some of which may also have been beyond duty-free allowances use of unbranded tobacco; however, there is no reason to and on which duty was not paid. However, as we have reported believe any such reluctance would be more common after, com- elsewhere,29 the numbers who purchased via these means pared with prior to, December 2012. Use of unbranded tobacco remained very low throughout the survey—about 0.1% for could be detected at multiple question points throughout the internet purchases and 1.8% pre-PP declining to 0.6% during survey, so it is unlikely that the total estimates of prevalence PP for purchases from overseas and duty-free. Another limita- have excluded any current or recent use. Survey items used to tion results from exclusions resulting from the sequencing of determine whether packs of FM cigarettes were likely to be questions in the survey. By excluding those who did not provide contraband were objective and easy to understand, and in the a brand name, we may have underestimated those who pur- context of the overall survey were unlikely to have been alarm- chased from informal sources. Note, however, that the numbers ing to respondents. are small and that this limitation applies equally to phases While small retailers interviewed in industry-funded studies before and after introduction of PP. may have believed that rates of use of illicit tobacco have Since we limited the analysis to categories of contraband cigar- increased following PP implementation,31 we found no evi- ettes for which we could obtain reliable objective data, our study dence of such increases. We also found much lower levels of was unable to assess changes in use of any top-selling Australian use of illicit unbranded tobacco than those reported in less rep- cigarettes that had somehow got into the illicit market. Also since resentative industry-funded internet surveys12 26 32 where the we were unable to apply stable price-based criteria to assess likely sampling frame comprises people who opt in and earn incen- illicit status, we would have failed to detect cigarettes that have tives for survey participation.16 17 While this study was unable been counterfeited to appear like common Australian brands. to assess the absolute level of use of illicit tobacco, three Price-based criteria may need to be included in future studies important objective measures of use of contraband cigarettes attempting to quantify the use of illicit tobacco. However, at this did not increase following the introduction of PP, and there time, this does not appear to be a consequential omission. was no increase in use of unbranded tobacco. Results are in Neither the Australian Customs and Border Protection Agency30 line with those reported in Australia’s National Drug Strategy nor the industry-funded discarded pack studies12 13 have so far Household Survey which found fewer than 3.6% of smokers

Scollo M, et al. Tob Control 2015;24:ii76–ii81. doi:10.1136/tobaccocontrol-2014-052072 ii79 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper reporting any current use of illicit unbranded tobacco in 2013, 3 Imperial Tobacco Australia Limited. Submission to the Department of Health and significantly fewer than in 2010 or 2007. Under 1% reported Ageing regarding the Tobacco Plain Packaging Bill 2012 (exposure draft) and ‘ ’ 33 consultation paper. Sydney, 2011. http://webarchive.nla.gov.au/gov/ use of illicit unbranded tobacco half the time or more . 20130329044623/http://www.yourhealth.gov.au/internet/yourhealth/publishing.nsf/ While various indicators of use of illicit tobacco should con- Content/ppit~ncr~loe (accessed 10 Jan 2012). tinue to be monitored through population surveys, retail audits 4 Crowe D. Campaign questions expensive plain packaging experiment, May 17. and analysis of customs interceptions, our study suggests that 2011. http://www.bata.com.au/group/sites/bat_7wykg8.nsf/vwPagesWebLive/ industry estimates of the extent of use of illicit tobacco are DO7WZEX6/$FILE/medMD8GX5XH.pdf?openelement (accessed Jul 2014). 5 British American Tobacco Australia. Sydney number one hot spot for illegal tobacco. exaggerated. Sydney: BATA, 2011. http://www.bata.com.au/group/sites/BAT_7WYKG8.nsf/ vwPagesWebLive/DO7WZEX6?opendocument&SKN=1 (accessed May 2014). 6 British American Tobacco Australia. Booming illegal tobacco costs government billions. Sydney: BATA, 2013. http://www.bata.com.au/group/sites/BAT_7WYKG8. What this paper adds nsf/vwPagesWebLive/DO9879X3?opendocument&SKN=1 (accessed May 2014). 7 British American Tobacco (BAT). UK Standardised Packaging Consultation: Response of British American Tobacco UK Limited. 2012. http://www.bat.com/group/sites/ What is already known on this subject uk__3mnfen.nsf/vwPagesWebLive/DO8WZC5E/$FILE/medMD8WZC6J.pdf? ▸ The tobacco industry vigorously opposed the introduction of openelement (accessed May 2014). plain packaging in Australia and warned of an increase in 8 Japan Tobacco International ( JTI). Response to the Department of Health’s use of illicit tobacco. consultation on the standardised packaging of tobacco products. 2012. http://www. ▸ jti.com/how-we-do-business/key-regulatory-submissions/ (accessed May 2014). Research conducted prior to and after the introduction of plain 9 Oireachtas Joint Committee on Health and Children. Public Health (Standardised packaging found no evidence of an increase in the availability Packaging of Tobacco) Bill presentations and submissions. Dublin: Houses of the of illicit tobacco from a representative sample of small retail Oireachtas, 2014. http://www.oireachtas.ie/parliament/oireachtasbusiness/committees_ outlets and no evidence of increased use of unbranded tobacco list/health-and-children/submissionsandpresentations/ (accessed May 2014). among a representative sample of smokers. 10 British American Tobacco (New Zealand) Ltd. Proposal to introduce plain packaging of tobacco products in New Zealand: submission by British American Tobacco (New What this paper adds Zealand) Limited. Wellington: Parliament of New Zealand, 2014. http://www. ▸ parliament.nz/en-nz/pb/sc/documents/evidence?Custom=00dbhoh_bill12969_ In a larger national sample, this study found no evidence of 1&Criteria.PageNumber=2 (accessed May 2014). increases in two important categories of contraband cigarettes 11 Imperial Tobacco New Zealand Limited. Submission to Health Select Committe by and no increase in purchases from informal sellers. The Imperial Tobacco New Zealand Limited. Wellington: Parliament of New Zealand, reported use of illicit unbranded tobacco also did not increase. 2014. http://www.parliament.nz/en-nz/pb/sc/documents/evidence? Custom=00dbhoh_bill12969_1&Criteria.PageNumber=4 (accessed Sep 2014). 12 KPMG LLP. Illicit tobacco in Australia: 2013 full-year report. Sydney, 2014. http:// www.bata.com.au/group/sites/BAT_7WYKG8.nsf/vwPagesWebLive/DO9FC38M? Acknowledgements Thank you to Dr Emily Brennan who provided several opendocument&SKN=1 (accessed Apr 2014). suggestions for analysis and to Ms Megan Bayly for data checking and review. 13 KPMG LLP. Illicit tobacco in Australia: 2014 half year report. Sydney, 2014. http:// fi www.bata.com.au/group/sites/BAT_7WYKG8.nsf/vwPagesWebLive/DO9FC38M? Contributors MS and MW designed this study. MZ cleaned the data les, opendocument&SKN=1 (accessed 25 Nov 2014). undertook the analysis and reported the results. MS drafted the manuscript with 14 Ciecierski C. Working Paper—The Market for Legal and Illegal Cigarettes in Poland: contributions from all authors. KC undertook further analysis. All authors approved fi A Closer Look at Demand and Supply-Side Characteristics. 2007. http://web.idrc.ca/ the nal manuscript. fr/ev-119595-201-1-DO_TOPIC.html (accessed Nov 2013). Funding The National Plain Packaging survey was funded under a contract with 15 Stoklosa M, Ross H. Contrasting academic and tobacco industry estimates of illicit the Australian Government Department of Health and Ageing. cigarette trade: evidence from Warsaw, Poland. Tob Control 2014;23:e30–4. 16 Quit Victoria. Critique of, illicit tobacco in Australia: full year report 2013 by KPMG Competing interests The authors wish to advise that MS was a technical writer LLP. Melbourne, Australia: Cancer Council Victoria, 2014. http://www.cancervic.org. for and MW a member of the Tobacco Working Group of the Australian National au/plainfacts/browse.asp?ContainerID=illicittobacco (accessed Aug 2014). Preventive Health Task Force and MW was a member of the Expert Advisory 17 Quit Victoria. Critique of, illicit tobacco in Australia: half year report 2014 by KPMG Committee on Plain Packaging that advised the Australian Department of Health on LLP. Melbourne, Australia: Cancer Council Victoria, 2014. http://www.cancervic.org. research pertaining to the plain packaging legislation. MW hold competitive grant au/plainfacts/browse.asp?ContainerID=illicittobacco (accessed Nov 2014). funding from the Australian National Health and Medical Research Council and MW 18 Gilmore AB, Rowell A, Gallus S, et al. Towards a greater understanding of the illicit holds competitive grant funding from the US National Institutes of Health, Australian tobacco trade in Europe: a review of the PMI funded ‘Project Star’ report. National Preventive Health Agency and BUPA Health Foundation. Tob Control 2014;23:e51–61. Ethics approval The survey was approved by the Cancer Council Victoria Human 19 Joossens L, Lugo A, La Vecchia C, et al. Illicit cigarettes and hand-rolled tobacco in Ethics Committee (HREC 0018). 18 European countries: a cross-sectional survey. Tob Control 2014;23:e17–23. fi Provenance and peer review Not commissioned; externally peer reviewed. 20 Scollo M, Bayly M, Wake eld M. Availability of illicit tobacco in small retail outlets before and after the implementation of Australian plain packaging legislation. Tob Open Access This is an Open Access article distributed in accordance with the Control Published Online First: 10 Apr 2014 doi:10.1136/tobaccocontrol-2013-051353 Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which 21 World Customs Organization. Illicit trade report, 2012. Brussels: WCO, 2013. http:// permits others to distribute, remix, adapt, build upon this work non-commercially, www.wcoomd.org/en/media/newsroom/2013/june/wco-publishes-its-first-illicit-trade- and license their derivative works on different terms, provided the original work is report.aspx (accessed Jul 2013). properly cited and the use is non-commercial. See: http://creativecommons.org/ 22 Scollo M, Bayly M. Section 13.3 The price of tobacco products in Australia in licenses/by-nc/4.0/ Chapter 13. The pricing and taxation of tobacco products in Australia (December 2014 update). Melbourne, Australia: Cancer Council Victoria, 2014. http://www. tobaccoinaustralia.org.au/chapter-13-taxation/13-3-the-price-of-tobacco-products- REFERENCES inaustralia (accessed Nov 2014). 1 British American Tobacco Australia. Submission on the Tobacco Plain Packaging Bill 23 Scollo M, Zacher M, Durkin S, et al. Early evidence about the predicted unintended 2011. Canberra: Department of Health and Ageing Consultation website, 2011. consequences of standardised packaging of tobacco products in Australia: http://content.webarchive.nla.gov.au/gov/wayback/20130904173432/http://www. a cross-sectional study of the place of purchase, regular brands and use of illicit yourhealth.gov.au/internet/yourhealth/publishing.nsf/Content/ tobacco. BMJ Open 2014;4:e005873. 2EE2F6F3EC74F628CA2579540005F68B/$File/British%20American%20Tobacco% 24 Wakefield M, Coomber K, Zacher M, et al. Australian adult smokers’ responses to 20Australia%20-%20Public%20Submission.pdf (accessed Oct 2014). plain packaging with larger graphic health warnings 1 year after implementation: 2 Philip Morris Limited. Commoditising tobacco products through plain packaging will results from a national cross-sectional tracking survey. Tob Control 2015;24:ii17–25. harm public health, violate treaties and does not meet the test of ‘evidence-based 25 Australian Bureau of Statistics. Technical paper: socio-economic indexes for areas policy’. Melbourne, 2011. http://webarchive.nla.gov.au/gov/20130329051831/http:// (SEIFA) 2011. Canberra, Australian Capital Territory: ABS, 2013. http://www.abs. www.yourhealth.gov.au/internet/yourhealth/publishing.nsf/Content/ gov.au/websitedbs/censushome.nsf/home/seifa2011?opendocument&navpos=260 Phillip-Morris-Limited~Phillip-Morris-Limited-2 (accessed Oct 2014). (accessed Apr 2014). ii80 Scollo M, et al. Tob Control 2015;24:ii76–ii81. doi:10.1136/tobaccocontrol-2014-052072 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

26 KPMG LLP. Illicit tobacco in Australia: 2013 half year report. Sydney: 2013. http:// 31 Roy Morgan Research. Impact of plain packaging on small retailers—wave 2. www.bata.com.au/group/sites/BAT_7WYKG8.nsf/vwPagesWebLive/DO9879X3? Melbourne: Australian Association of Convenience Stores, 2013. http://www.aacs. opendocument&SKN=1 (accessed Nov 2013). org.au/wp-content/uploads/2013/10/Impact-of-Plain-Packaging-on-Small-Retailers- 27 NSW Retail Tobacco Traders’ Association. Price lists–cigarettes. Aust Retail W2-Final-Report.pdf (accessed May 2014). Tobacconist 2013;87:1–2. 32 Deloitte. Illicit trade of tobacco in Australia: update for 2012: a report 28 Australian Bureau of Statistics. 2071.0—Reflecting a Nation: Stories from the 2011 prepared for British American Tobacco Australia Limited, Philip Morris Limited Census, 2012–2013. 2013. http://www.abs.gov.au/ausstats/[email protected]/Lookup/2071. and Imperial Tobacco Australia Limited. Sydney: British American Tobacco 0main+features902012-2013 (accessed Jul 2014). Australia, 2012. http://www.bata.com.au/group/sites/BAT_7WYKG8.nsf/ 29 Scollo M, Zacher M, Coomber K, et al. Did consumers shift from small mixed vwPagesWebLive/DO9879X3?opendocument&SKN=1 (accessed Aug 2014). businesses to discount outlets following the introduction of plain packaging in 33 Australian Institute of Health and Welfare. National Drug Strategy Household Australia? A national cross-sectional survey. Tob Control 2015;24:ii98–9. Survey detailed report: 2013—supplementary tables. Canberra: AIHW, 2014. 30 Australian Customs and Border Protection Service. Annual report 2013–14. http://www.aihw.gov.au/publication-detail/?id=60129549469&tab=3 (accessed Canberra: ACBPS, 2014. http://www.customs.gov.au/site/page4283.asp Nov 2014).

Scollo M, et al. Tob Control 2015;24:ii76–ii81. doi:10.1136/tobaccocontrol-2014-052072 ii81 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Use of illicit tobacco following introduction of standardised packaging of tobacco products in Australia: results from a national cross-sectional survey Michelle Scollo, Meghan Zacher, Kerri Coomber and Melanie Wakefield

Tob Control 2015 24: ii76-ii81 doi: 10.1136/tobaccocontrol-2014-052072

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii76

These include: References This article cites 5 articles, 3 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii76#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper The advertised price of cigarette packs in retail outlets across Australia before and after the implementation of plain packaging: a repeated measures observational study Michelle Scollo, Megan Bayly, Melanie Wakefield

▸ Additional material is ABSTRACT audits of retail outlets7 and surveys of consumers8 published online only. To view Objective This study monitored the advertised price of find no evidence of this anticipated trend having please visit the journal online (http://dx.doi.org/10.1136/ the most prominently promoted and the cheapest single occurred. Also in contrast to industry predictions, tobaccocontrol-2014-051950) packs of cigarettes in Australian retail outlets before and the volume of tobacco products on which excise after the implementation of plain packaging. and customs duty has been paid in the calendar Centre for Behavioural Research in Cancer, Cancer Methods A panel of 421 outlets in four large year 2013 was 3.4% lower than in 2012 which Council Victoria, Melbourne, Australian cities was visited monthly from May 2012 to included 3 months during which plain packs were Victoria, Australia August 2013 and the brand, pack size and price of the already on the market in Australia.9 The reported most-prominently listed and lowest-priced single number of cigarettes smoked each day by current Correspondence to cigarette pack were recorded from each store’s tobacco smokers was 15% lower in 2013 than it was in Dr Michelle Scollo, Centre for 10 Behavioural Research in price board. Changes in the inflation-adjusted stick price 2010. The current study is one of several by our – Cancer, Cancer Council were examined using linear mixed models, controlling for group in this supplement11 13 aiming to investigate Victoria, 615 St Kilda Rd, fixed effects of city, store type, area socioeconomic what happened to prices of tobacco products after Melbourne, Vic 3004, status and random effects of time. The adjusted stick the introduction of plain packaging in Australia. Australia; [email protected] price was also examined over time by tobacco Cigarette prices can be conceptualised and mea- manufacturer and pack size. sured in several different ways.14 Recommended Received 6 August 2014 Results The inflation-adjusted stick price of the most- retail prices (RRPs) published in trade magazines Accepted 23 October 2014 prominently advertised single packs was significantly provide a comprehensive picture of recommended higher than in May–July 2012 from August–October prices of almost all products sold by all major man- 2012 for mainstream and premium brands and from ufacturers and most importers. However, RRPs February–April 2013 for value brands. Adjusted average may not reflect tobacco prices promoted in real- stick prices of lowest-priced packs in August 2013 were world settings, can be slow to document changes in $0.02 (95% CI $0.02 to $0.03, p<0.001) higher than product offerings and do not include some of the in May–July 2012 ($Aug13). A large real increase in brands imported by smaller distributors. Actual stick price was seen in February–April 2013 across all prices paid for tobacco products can be estimated major manufacturers, market segments and pack size from self-report surveys, but they reflect the categories. outcome of consumer purchasing choices—brand, Discussion The price of cigarettes most prominently pack size, place of purchase; they do not necessarily promoted on price boards did not decrease in the directly reveal industry pricing and other marketing months following implementation of Australia’s plain strategies. Observational studies of tobacco retai- packaging legislation. Retail prices continued to increase lers, on the other hand, provide a means of moni- above the level resulting from automatic indexation of toring real-world advertised tobacco prices. excise/customs duty even at the lowest-priced end of the Scollo et al15 monitored changes in tobacco Australian market. prices during Australia’s National Tobacco Campaign from 1997 to 2000, a period which coincided with changes to tobacco taxation. RRPs, INTRODUCTION the self-reported price paid and an observational Australia’s plain packaging legislation was met with retail monitoring study showed similar magnitudes fierce opposition from the tobacco industry.1 of increase in price for all monitored cigarette Among many criticisms, tobacco companies brands, although RRPs were consistently higher claimed that the removal of design features from than the self-reported price paid and advertised Open Access Scan to access more tobacco packaging would force brands to compete retail prices (reflecting price discounting strategies). free content only on the basis of cost and that this ‘commodit- A more recent observational study of small mixed- isation’ of tobacco products would lead to dramatic business retailers in close proximity to secondary drops in prices, resulting in an increase in con- schools in Melbourne, Australia showed that the – sumption.2 4 It was also claimed that standardised advertised price of popular cigarette brands was packaging would lead to an influx of counterfeit lower than the RRP in more than 10% of stores, and other contraband tobacco, also putting down- and that discounting was more common in areas of – ward pressure on prices.2 4 While industry reports low socioeconomic status (SES).16 Another audit To cite: Scollo M, Bayly M, based on studies of discarded packs claim that use of more than 1500 tobacco retailers across the state Wakefield M. Tob Control of illicit tobacco has increased since the introduc- of New South Wales found that the advertised – 2015;24:ii82 ii89. tion of plain packaging legislation,56independent price of the most popular cigarette pack in ii82 Scollo M, et al. Tob Control 2015;24:ii82–ii89. doi:10.1136/tobaccocontrol-2014-051950 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Australia—Winfield 25s—varied by store type, area SES and Store sample and selection with the proportion of children residing in the area, but did not In May 2012, stores were selected from 16 sampling areas differ between regional and metropolitan areas.17 A similar within Australia’s two largest cities—Melbourne and Sydney— study in south-east Queensland found lower prices for leading and eight areas in each of the cities of Adelaide and Perth. mainstream and value brands in low-SES suburbs.18 Sampling areas were selected by stratifying all postcodes within Opponents of the plain packaging legislation had argued that the metropolitan boundaries of each city into four SES categor- premium brands would be most severely affected by plain pack- ies using the Socioeconomic Indices for Areas 2011 Index of aging as it is these products that rely most heavily on package Relative Disadvantage,23 then randomly selecting equal numbers design to communicate prestige and quality.19 Examination of of postcodes from each SES category in each city. Each of these retailer association recommended price lists, however, revealed postcodes formed the primary postcode of that sampling area. that RRPs of all tobacco products produced by the major manu- For efficiency during store selection, a nearby secondary post- facturers/importers increased above inflation in the 12 months code was selected for each area, matched by SES category. after plain packaging was introduced in Australia, despite there Therefore, each sampling area was comprised of two postcodes. being no increases in excise/customs duty beyond those required Maps of each sampling area identifying potentially useful by automatic 6-monthly indexation.11 This raises the question shopping strips and centres were created. Fieldworkers were of what happened to the real-world prices in retail outlets fol- instructed to visit the largest shopping strip in the primary post- lowing the introduction of the legislation. code and, using a rigorous set of walking rules, sample all The price board—a and prices on potentially eligible stores until either their store quotas were display at the point of sale—is one of the few ways in which met or they had sampled the entire shopping strip. In the latter companies can inform Australian smokers of the prices of situation, the second largest shopping area in the primary post- tobacco products. Restrictions on the size and contents of price code was then sampled, and so on. If all shopping strips were boards differ across the states and territories in Australia with exhausted in the primary postcode, the sampling procedure was Western Australia allowing very limited information at the point continued in the secondary postcode. Potentially eligible stores of sale, and Queensland banning price boards altogether.20 were: Previous research following the implementation of bans of cigar- ▸ Supermarkets: large grocery chains; ette displays has demonstrated the strategic use of boards to ▸ Convenience stores: small independent supermarkets and promote particular products, with space at the top more likely branded chain convenience stores with a broad range of pro- to be occupied by premium brands which provide high margins ducts and extended opening hours; to manufacturers and retailers.21 If prices were ‘softening’ ▸ Small mixed-businesses: milk bars (Melbourne), delis among premium brands (declining or increasing at lower rates (Adelaide and Perth), corner stores and general stores; than prices across the market as a whole), this ought to be ▸ Petrol stations: any store that sold petrol, including those apparent among the most prominently promoted premium branded as convenience stores or supermarket chains; brand. The lowest-priced brand on the board represents the ▸ Newsagents/lottery outlets: stores that sold newspapers, other extreme—the pack with the lowest purchase price of magazines and stationery as their primary business and/or which consumers are likely to be aware. lottery tickets and services. Cigarette packs in Australia have typically contained multiples Quotas were set at nine stores per area, including one of each of five sticks, ranging from 20 to 50 cigarettes per single pack.14 store type in every area wherever possible. We visited a total of However, in the lead-in period to plain packaging, a variety of 796 stores, of which a total of 375 (47.1%) did not meet the new pack sizes—21s, 22s, 23s, 26s—were introduced predomin- eligibility criteria in that they did not sell tobacco (n=129, antly among value brands. These packs with ‘bonus’ cigarettes 16.2%), had no visible price board (n=225, 28.3%), or dis- tended to have the same or lower RRPs than packs from the played fewer than three single cigarette pack prices (n=21, same brand in the equivalent 20s or 25s size.11 22 The introduc- 2.6%), leaving 421 stores in the sample. tory prices for these products were therefore lower per stick than the price per stick of other value brands. It might be expected that the lower prices of these new packs with bonus Data collection procedure sticks would have exerted downward pressure on the prices of This study was part of a larger exercise that aimed to collect other value brands of similar pack size. information about transaction times24 25 and willingness to sell The current study aimed to monitor before and after the illicit tobacco,7 which necessitated unobtrusive data collection. implementation of plain packaging the real-world advertised Each month, fieldworkers entered the stores under the guise of price of the most prominently promoted and the lowest priced a regular customer, and discretely observed the contents of the single packs of cigarettes in retail outlets in four capital cities price board. The brand name, pack size and price of the most- across Australia where price boards are still permitted. prominent and lowest-priced single cigarette packs were noted. The most-prominent single pack was defined as that listed highest on the price board. A national fieldwork agency experi- METHOD enced in store research undertook data collection. Fieldworkers Design were trained to use effective observational techniques developed This study was a repeated-measures observational design. A in pilot studies and using mocked-up price board examples. team of trained fieldworkers visited a panel of tobacco retailers monthly from May 2012—7 months prior to the 1 December 2012 implementation of plain packaging—to August 2013, Statistical analysis 9 months after implementation. Stores that closed during the Data management and analyses used Stata V.12.1.26 To account study period were not replaced. Fieldworkers rotated across for variability in the size of the observed packs, stick price was areas within cities to avoid visiting a given store in consecutive calculated by dividing the advertised pack price by the number months. of sticks in each pack.

Scollo M, et al. Tob Control 2015;24:ii82–ii89. doi:10.1136/tobaccocontrol-2014-051950 ii83 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

The study period was aggregated into six phases which took model. A systematic examination of various model specifications into account both the stage of implementation of plain pack- showed that an unstructured covariance matrix for the random aging and the dates on which tobacco excise/customs duty was effect provided the best-fitting model. Finally, remaining within- indexed (on 1 February and 1 August each year, based on store residuals were modelled independently by store type to changes in the Consumer Price Index (CPI) over the previous further improve model fit. December to June or June to December 6-month period). This These specifications were the best-fitting models for the most- ‘phase’ variable was created by combining all months prior to prominent and lowest-priced packs, and were used for all ana- the 1 August 2012 CPI adjustment (May–July 2012), then lyses. Wald χ2 postestimation tests were performed to examine August–October 2012 (when only small numbers of plain packs main fixed effects of categorical covariates. The most-prominent had started to appear in shops), November 2012–January 2013 observations were coded by market segment (value, mainstream (early after implementation of plain packaging but prior to and premium) using definitions from Retail World27 trade maga- indexation on 1 February 2013), February–April 2013 and zine. The stick prices of any brands not listed in Retail World May–July 2013 (before the 1 August CPI-related increase in were compared to those listed and coded accordingly. The final excise/customs duty). August 2013 represented the final phase. models were then repeated including only brands from each The phases are referred to as ‘plain packaging (PP)/CPI phase’ tobacco manufacturer (British American Tobacco Australia throughout. Prices per stick were adjusted to take into account (BATA), Imperial Tobacco and Philip Morris separately; brands changes in the CPI. These were calculated using changes in from all other small manufacturers combined). Similarly, separ- Index figures between December 2011 and June 2012, June ate models were fitted for different pack sizes—packs of 20 2012 and December 2012, and December 2012 and June 2013, sticks; 21s, 22s and 23s combined; and 25s, 26s and 30s com- so that in effect the analysis examined changes over time above bined. From all models, average stick prices were calculated the legislated CPI-linked increases in excise/customs duty that adjusting for fixed and random effects for each PP/CPI phase. went into effect in Australia on 1 August 2012, 1 February 2013, and 1 August 2013. RESULTS Linear mixed models using maximum likelihood estimation Store characteristics were performed to examine the effect of time on Table 1 shows that supermarkets and petrol stations were the inflation-adjusted stick price, controlling for fixed effects of most common store types, while small mixed businesses were store type, city and SES category. As the price observations were least common. More than two-thirds of the stores were located repeated every month in the same panel of stores, random inter- in Melbourne and Sydney and there were more stores in low cepts for stores were included in all models to adjust for correla- and least disadvantaged (higher SES) areas than areas of most tions between observations within each venue. A random effect and moderate disadvantage (lower SES). Importantly, only mar- for time was also included in order to control for the possibility ginal differences between the most-prominent and lowest-priced that prices changed more in some stores than others over the observations and the original sample distribution were found; study period. Likelihood ratio tests confirmed that these inclu- that is, no one store type, city or SES category tended to sions significantly improved model fit. The effect of time was become ineligible over time more than any others. explored in two ways: as a linear function of month, and as a In the 6736 store visits over the 16 months, the store was stepwise function of phase. Comparison of Akaike information closed or no longer sold tobacco on 114 (1.7%) occasions. criterion values showed that the latter was the better fitting Prices were no longer displayed in a further 624 store visits

Table 1 Characteristics of the store sample, and distribution of price observations by store type, city and area SES Total most-prominent Total lowest-priced Eligible store sample pack observations pack observations n (%) n (%) n (%)

Total n=421 n=5473 n=5859 Store type Supermarket 118 (28.0%) 1268 (23.2%) 1684 (28.7%) Convenience store 64 (15.2%) 907 (16.6%) 890 (15.2%) Small mixed business* 42 (10.0%) 561 (10.3%) 546 (9.3%) Petrol station 124 (29.5%) 1763 (32.2%) 1785 (30.5%) Newsagent/lottery 73 (17.3%) 974 (17.8%) 954 (16.3%) City Melbourne 146 (34.7%) 1879 (34.3%) 2047 (34.9%) Sydney 146 (34.7%) 1803 (32.9%) 1946 (33.2%) Adelaide 64 (15.2%) 894 (16.3%) 912 (15.6%) Perth 65 (15.4%) 897 (16.4%) 954 (16.3%) SES category Most disadvantage 93 (22.1%) 1219 (22.3%) 1277 (21.8%) Moderate disadvantage 78 (18.5%) 986 (18.0%) 1074 (18.3%) Low disadvantage 126 (29.9%) 1656 (30.3%) 1782 (30.4%) Least disadvantage 124 (29.5%) 1612 (29.5%) 1726 (29.5%) *Includes corner stores, milk bars, delicatessens, general stores, etc. SES, socioeconomic status. ii84 Scollo M, et al. Tob Control 2015;24:ii82–ii89. doi:10.1136/tobaccocontrol-2014-051950 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

(9.3%). The incidence of these factors increased gradually over to $0.05, p<0.001). Among premium brands, stick prices in all the study period. In the remaining 5998 store visits, price phases were significantly higher than in May–July 2012 (all boards were alphabetised by brand on only 8.5% of occasions, p<0.001), with a substantial increase in February–April 2013 and this did not differ significantly by month (χ2=16.15, relative to the first phase ($0.04; 95% CI $0.03 to $0.04, p=0.37). Alphabetised boards were more common in supermar- p<0.001). kets (13.1%) and petrol stations (10.2%), while each of the other Given the differences in stick prices between market seg- store types had alphabetised boards on <5% of store visits. ments, and the high proportion of premium brands among the Eligible observations of the most-prominent single pack were most-prominent single packs, changes in the inflation-adjusted made on 5473 (91.2%) occasions, and eligible observations of stick price over the study phases were examined among just the lowest-priced pack were made in 5859 (97.7%) store visits. premium brands for each of the three major Australian tobacco Nearly all exclusions of the most-prominent pack price were manufacturers (BATA: n=2393, Imperial Tobacco: n=350, and because a multibuy product (two or three packs bundled Philip Morris: n=789). Figure 2 and table A1 (see online sup- together) was inadvertently recorded as it was listed by the total plementary appendix) show significant changes in the number of cigarettes on offer, for example, ‘Winfield 50s’, and inflation-adjusted stick price among premium brands from all therefore appeared to be a single cigarette pack. These were manufacturers over the study period (all p<0.001). excluded given the focus of this study on single packs as Among BATA and Imperial Tobacco premium brands, only opposed to cartons or multibuys. minor changes were observed in the first three phases before In the majority of months, five or fewer lowest-priced records large and significant increases were seen in February–April 2013 were deemed ineligible (eg, where the fieldworker mistakenly relative to the first phase (BATA: $0.03, 95% CI $0.03 to recorded a non-existent brand/pack size such as ‘Brandon 20s’). $0.04, p<0.001; Imperial Tobacco: $0.03, 95% CI $0.02 to A large number of records were excluded in December 2012 $0.04, p<0.001). Imperial Tobacco prices appeared to stabilise (n=74) and January 2013 (n=21) because they were of packs in the final phases, while BATA premium brands showed small that did not conform to the minimum pack size dimension spe- increases in each phase, so that on August 2013 the cified in the plain packaging legislation—JPS Nano 20s and Pall inflation-adjusted stick price was $0.04 (95% CI $0.04 to Mall Slims 20s—and were therefore ineligible for legal sale after $0.05, p<0.001) higher than in May–July 2012. Philip Morris 1 December 2012. Prior to December 2012, JPS Nano 20s were premium brands also showed minor increases in the first phases one of the most commonly observed lowest-priced products on followed by a large increase in stick price compared to May– price boards; both brands were replaced with packs of 23s on July 2012 ($0.04, 95% CI $0.03 to $0.05, p<0.001), and then the introduction of plain packaging. continued to increase by approximately $0.01 and $0.02 in the final two phases, respectively. Advertised cigarette prices: most-prominent single cigarette packs Advertised cigarette prices: lowest priced cigarette packs Of the 5473 most-prominent packs observed, two-thirds Of the 5859 lowest-priced packs, 93.8% were value brands and (64.6%) were premium brands, while 20.7% were value brands 83.2% were packs of 20 cigarettes. Table 2 and figure 3 show and 14.7% were mainstream. that inflation-adjusted stick prices of all observed lowest-priced Figure 1 and table 2 show real increases in the stick prices of packs significantly changed over the study period (p<0.001). the most prominently listed value, mainstream and premium Small but significant decreases (both p<0.05) were seen in packs over the study period (all p<0.001). The adjusted average August–October 2012 and November 2012–January 2013 com- stick prices of value brands remained stable until the February – pared to May–July 2012, while the inflation-adjusted price in April 2013 phase, where a $0.02 (95% CI $0.01 to $0.03, February–April 2013 was $0.02 (95% CI $0.02 to $0.03, p<0.001) increase was observed relative to phase one. Prices p<0.001) higher than in the first phase. The stick price then decreased slightly in the latter two phases, but remained signifi- decreased slightly, again in real terms, but remained significantly cantly higher than in May–July 2012 (both p<0.05). higher in the final two phases compared to May–July 2012 Mainstream brands’ stick prices increased in every phase, and (both p<0.001). prices in all phases were significantly higher in real terms than in After packs of 20s, 22s were the next most common pack size May–July 2012 (all at least p<0.01), again with a large relative (7.8%), followed by 25s (5.7%), 23s (2.0%), 26s (1.3%), 21s increase in the February–April 2013 phase ($0.04; 95% CI $0.03 (0.2%) and only two packs of 30s were ever recorded as the

Figure 1 Estimated marginal means of inflation-adjusted stick prices ($Aug2013) of the most prominently advertised value, mainstream and premium single packs by PP/CPI phase (with 95% CIs), adjusted for store type, city and area socioeconomic status.

Scollo M, et al. Tob Control 2015;24:ii82–ii89. doi:10.1136/tobaccocontrol-2014-051950 ii85 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Table 2 Marginal mean inflation-adjusted stick prices ($Aug2013) by PP/CPI phase, store type, city and SES category, estimated using linear mixed models, with post-estimation Wald tests for main effects and within-group comparisons Most-prominently advertised single pack by market segment

Value brands (n=1135) Mainstream brands (n=802) Premium brands (n=3536) All lowest-priced packs (n=5859) Stick price (95% CI) Stick price (95% CI) Stick price (95% CI) Stick price (95% CI)

Plain packaging/Consumer (Wald χ2=61.36, p<0.001) (Wald χ2=117.82, p<0.001) (Wald χ2=822.63, p<0.001) (Wald χ2=291.06, p<0.001) Price Index phase May–July 2012 (Ref) $0.576 ($0.567 to $0.584) $0.694 ($0.683 to $0.704) $0.742 ($0.738 to $0.747) $0.573 ($0.568 to $0.579) August–October 2012 $0.577 ($0.569 to $0.585) $0.704 ($0.694 to $0.713)** $0.748 ($0.744 to $0.752)*** $0.569 ($0.564 to $0.574)* November 2012–January 2013 $0.576 ($0.568 to $0.584) $0.710 ($0.701 to $0.720)*** $0.748 ($0.744 to $0.752)*** $0.568 ($0.564 to $0.573)* February–April 2013 $0.595 ($0.587 to $0.604)*** $0.736 ($0.726 to $0.746)*** $0.777 ($0.773 to $0.782)*** $0.597 ($0.592 to $0.602)*** May–July 2013 $0.588 ($0.578 to $0.597)* $0.746 ($0.735 to $0.757)*** $0.782 ($0.778 to $0.786)*** $0.592 ($0.587 to $0.597)*** August 2013 $0.590 ($0.579 to $0.602)* $0.757 ($0.744 to $0.771)*** $0.793 ($0.787 to $0.798)*** $0.591 ($0.585 to $0.598)*** Store type (Wald χ2=31.62, p<0.001) (Wald χ2=55.58, p<0.001) (Wald χ2=171.63, p<0.001) (Wald χ2=58.52, p<0.001) Supermarket (Ref) $0.552 ($0.540 to $0.564) $0.679 ($0.663 to $0.695) $0.734 ($0.727 to $0.742) $0.556 ($0.548 to $0.563) Convenience store $0.593 ($0.575 to $0.611)*** $0.715 ($0.692 to $0.739)* $0.749 ($0.739 to $0.759)* $0.586 ($0.576 to $0.597)*** Small mixed business† $0.594 ($0.573 to $0.614)** $0.714 ($0.676 to $0.751) $0.767 ($0.754 to $0.779)*** $0.598 ($0.585 to $0.611)*** Petrol station $0.605 ($0.586 to $0.624)*** $0.753 ($0.739 to $0.766)*** $0.794 ($0.787 to $0.800)*** $0.590 ($0.583 to $0.598)*** Newsagent/lottery $0.589 ($0.573 to $0.605)*** $0.680 ($0.656 to $0.704) $0.739 ($0.729 to $0.748) $0.591 ($0.581 to $0.601)*** City (Wald χ2=5.06, p=0.168) (Wald χ2=7.00, p=0.072) (Wald χ2=1.22, p=0.748) (Wald χ2=12.48, p=0.006) Sydney (Ref) $0.576 ($0.563 to $0.589) $0.722 ($0.707 to $0.738) $0.762 ($0.756 to $0.769) $0.591 ($0.584 to $0.598) Melbourne $0.587 ($0.575 to $0.600) $0.718 ($0.704 to $0.732) $0.762 ($0.755 to $0.768) $0.575 ($0.568 to $0.582)** Adelaide $0.566 ($0.545 to $0.587) $0.702 ($0.676 to $0.727) $0.757 ($0.748 to $0.766) $0.578 ($0.567 to $0.588)* Perth $0.592 ($0.577 to $0.608) $0.745 ($0.723 to $0.768) $0.763 ($0.753 to $0.773) $0.573 ($0.563 to $0.584)** Area socioeconomic status (Wald χ2=2.71, p=0.438) (Wald χ2=2.74, p=0.434) (Wald χ2=1.07, p=0.784) (Wald χ2=9.34, p=0.025) Most disadvantage (Ref) $0.583 ($0.568 to $0.598) $0.714 ($0.697 to $0.731) $0.760 ($0.752 to $0.768) $0.571 ($0.563 to $0.580) Moderate disadvantage $0.584 ($0.565 to $0.602) $0.711 ($0.688 to $0.735) $0.762 ($0.753 to $0.771) $0.581 ($0.571 to $0.590) Low disadvantage $0.590 ($0.577 to $0.602) $0.729 ($0.714 to $0.744) $0.759 ($0.752 to $0.766) $0.578 ($0.571 to $0.586) Least disadvantage $0.574 ($0.560 to $0.588) $0.724 ($0.707 to $0.740) $0.764 ($0.757 to $0.771) $0.589 ($0.582 to $0.597)** *p<0.05; **p<0.01; ***p<0.001. †Includes corner stores, milk bars, delis, general stores, etc.

lowest-priced advertised pack (<0.1%). While the proportion of relative to phase one. Following this increase, stick prices of 20s 20s remained relatively stable across the PP/CPI phases stabilised and remained significantly higher than at the start of (minimum 79.8%; maximum 86.8%), the proportion of small the study period in the final two phases (both p<0.001). Packs novel pack sizes (21s, 22s and 23s) more than doubled from of 21s, 22s and 23s showed an almost identical pattern, 7.1% to 15.3% and the proportion of 25s, 26s and 30s although the observed increase in February–April 2013 was observed decreased substantially from 13.1% to 2.0% from the smaller: $0.02 (95% CI $0.01 to $0.03, p<0.001). The first to last phases. This is largely attributable to fewer packs of inflation-adjusted stick price of these small novel packs 25s observed as the lowest-priced advertised packs—26s and decreased slightly in May–July 2013 but remained consistently 30s were consistently rare across the study. higher than in May–July 2012 in the final phases (both Figure 3 and table A2 (see online supplementary appendix) p<0.05). Conversely, the stick price of 25s, 26s, and 30s also show that packs of 20 cigarettes showed a small but signifi- showed a small but significant real increase in August–October cant decrease (p<0.01) in the inflation-adjusted stick price in 2012 ($0.01, 95% CI $0.01 to $0.02, p<0.01), before increas- August–October 2012, followed by a large increase of $0.03 ing in February–April 2013, relative to phase one, by a similar (95% CI $0.02 to $0.03, p<0.001) in February–April 2013 margin to the other pack sizes: $0.03 (95% CI $0.02 to $0.05,

Figure 2 Estimated marginal means of inflation-adjusted stick prices ($Aug2013) of the most prominently advertised premium single packs by manufacturer and PP/CPI phase (with 95% CIs), adjusted for store type, city and area socioeconomic status.

ii86 Scollo M, et al. Tob Control 2015;24:ii82–ii89. doi:10.1136/tobaccocontrol-2014-051950 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Figure 3 Estimated marginal means of inflation-adjusted stick prices ($Aug2013) of the lowest-priced advertised single packs, overall and by pack size category, by PP/CPI phase (with 95% CIs), adjusted for store type, city and area socioeconomic status.

p<0.001), and again in the final phases to $0.05 (95% CI declines in the lead-up to plain packaging implementation, the $0.02 to $0.08, p<0.01) above the May–July 2012 stick price. prices per stick of the lowest-priced pack on the board also Each of the three largest manufacturers accounted for increased in real terms. The increase in prices following the approximately one-third of the observed lowest-priced packs: scheduled increase in excise/customs duty in February 2013 BATA: 32.7%, Imperial Tobacco: 32.8%, Philip Morris: 31.6%. were well above the applicable inflation increases and were seen The remaining 2.9% of packs from other smaller manufacturers across brands from the three largest tobacco manufacturers and were aggregated. Figure 4 and table A2 (see online supplemen- across pack sizes. tary appendix) show that significant main effects of PP/CPI The results of the current study concur with those of a study phase were found for the three largest manufacturers (all of RRPs of all brands from major tobacco manufacturers and p<0.001), while the lowest-priced brands from ‘other’ manufac- importers which also found prices to have increased above infla- turers showed no change in the inflation-adjusted stick price tion.11 Examination of the Tobacco Sub-index of Australia’s CPI over the study period. While the three major manufacturers (based on regular government surveys of prices among multiple each showed a slightly different pattern of change from phase to products in a representative range of retail outlets) also suggests phase, the net result for all three was that a substantial real an increase in prices above the rate of inflation.28 increase in the stick price was seen between November 2012– The current study had several strengths, including careful January 2013 and February–April 2013—$0.03 for BATA and attention to unobtrusive observation, a multistate panel of Imperial Tobacco, and $0.02 for Philip Morris. Further, the diverse store types in a range of SES areas and an extended inflation-adjusted stick price of each brand in August 2013 was fieldwork period before and after plain packaging implementa- significantly higher than in May–July 2012: BATA: $0.02 (95% tion. We excluded tobacconists—from which approximately CI $<0.01 to $0.03, p<0.05); Imperial Tobacco: $0.04 (95% 11% of smokers report purchasing tobacco8—to limit study CI $0.03 to $0.05, p<0.001); Philip Morris: $0.02 (95% CI costs and because their specialised nature meant that the pres- $<0.01 to $0.03, p<0.01). The inflation-adjusted stick prices ervation of unobtrusive observational fieldwork would be diffi- from the three largest manufacturers appeared to converge cult. Prices in tobacconists tend to be lower than in other retail upwards over the study period (figure 4). There was a $0.04 per outlets,17 so the average prices obtained in this study may have stick difference between the highest (Philip Morris) and lowest been lower if this group had been included. On the other (Imperial Tobacco) stick prices in May–July 2012, compared to hand, just over a quarter of the stores initially visited during a $0.01 difference between the highest and lowest (still Philip sampling were excluded because they had no visible price Morris and Imperial Tobacco, respectively) at August 2013. boards. If products in such stores tended to be less discounted (ie, more likely to be sold closer to RRPs), then the average DISCUSSION prices obtained in our sample would be lower than prices The stick price of single packs most prominently promoted on across the Australian market as a whole. It seems unlikely that Australian price boards increased significantly in real terms fol- these exclusions would have materially affected the pattern of lowing the introduction of plain packaging. Increases were par- findings over time. ticularly strong where the most prominently promoted single Other retail studies have opted to record the price of specific products were premium or mainstream brands. After small brands and pack sizes regardless of place on the board or pack

Figure 4 Estimated marginal means of inflation-adjusted stick prices ($Aug2013) of the lowest-priced advertised single packs by manufacturer, and by PP/CPI phase (with 95% CIs), adjusted for store type, city and area socioeconomic status.

Scollo M, et al. Tob Control 2015;24:ii82–ii89. doi:10.1136/tobaccocontrol-2014-051950 ii87 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

– bundling,15 17 whereas our study recorded the most-prominent otherwise have been the case. Other research reported in this and the lowest-priced advertised single cigarette packs from volume12 suggests a significant increase in the use of such brands each store. While this introduced variability in the pack sizes following the introduction of plain packaging legislation corre- recorded, our method enabled the number of eligible stores to sponding with differentially greater increases in reported prices be maximised as the sample was not restricted to those that had paid among those using premium and mainstream brands com- the target product/s listed at the time of store selection. pared to those using value brands The current study shows that Likewise, selected stores were not lost from the sample if they even among value brands, prices of the lowest-price cigarettes removed any particular product from their price boards. The promoted on price boards were significantly higher in real terms strategy enabled the study to responsively reflect changes in 9 months after the introduction of plain packaging in Australia. product offerings among small cheap brands. For example, the most common brand to be listed as the cheapest on the board at JPS Nano the start of the study, 20s, changed to packs of 23s in What this paper adds December 2012 and was thereafter rarely the cheapest listed pack. Had we monitored only JPS Nano throughout the study, our estimates of price changes among packs with the lowest What is already known ▸ advertised purchase cost after plain packaging would have been Tobacco companies vigorously opposed the introduction of ’ confounded by this change in pack size. Australia s plain packaging legislation, claiming that it would The current results do not reflect the average of all the pro- lead to widespread falls in the prices of tobacco products. ▸ ducts on the price board, nor the average price Australian The recommended retail prices did not fall, and there has smokers actually paid for their tobacco. Limited funding necessi- been no fall in prices evident in measures such as the ’ tated the combination of the price collection tasks in this study Tobacco Sub-Index of Australia s Consumer Price Index. with other tasks (timing of pack retrieval24 25 and willingness to 7 What this paper adds sell illicit tobacco ). To maintain the unobtrusive observational ▸ As part of a real-world observational audit of tobacco retail – nature of the study, we constrained ourselves to collecting 2 3 outlets across Australia, this study has shown that the price of key pieces of pricing information. A full and detailed retail audit the most-prominently promoted single packs as well as the of tobacco prices in Australia by market segment and distribu- lowest-priced packs of cigarettes increased in real terms in the tion channel would be costly, since it would examine the retail months immediately after plain packaging implementation. prices of all products offered for sale across all brands, pack sizes and pack bundles (single, multibuy and carton), both via the Internet and in all categories of retail outlets. It would also Contributors MS and MW conceived of this study. MB coordinated the collection collect information about explicit ‘specials’, multi-pack offers of data and undertook data analysis. All authors contributed to the drafting and fi and other discounting of particular brands. Data on prices at nalisation of the manuscript. which products are actually sold could much more simply and Funding This study was supported by Quit Victoria, Cancer Council Australia, cheaply be obtained by the Australian government if tobacco Cancer Council South Australia, Cancer Council Victoria and Action of Smoking and Health (UK) and its partner organisations Cancer Research UK, Smokefree wholesalers were required to collect such information from Southwest, Fresh and Tobacco Free Futures. their customers and report aggregated information directly to Competing interests The authors wish to advise that MS was a technical writer the Government. Regular disclosure would enable ongoing for and MW a member of the Tobacco Working Group of the Australian National monitoring and evaluation using complete and accurate sales Preventive Health Task Force and MW was a member of the Expert Advisory weighted price data to assess the effects of public health policies Committee on Plain Packaging that advised the Australian Department of Health on and programmes across all of Australia.29 Provision of annual research pertaining to the plain packaging legislation. MW holds competitive grant data on sales and prices is required of tobacco manufacturers funding from the Australian National Health and Medical Research Council, US National Institutes of Health, Australian National Preventive Health Agency and and importers under Section 35 of the Smoke-free Environment BUPA Health Foundation. Act 1990 in New Zealand,30 but no such requirements are Ethics approval Institutional Research Review Committee, Cancer Council Victoria. imposed on the tobacco industry in Australia. While not a complete picture of retail prices, the information Provenance and peer review Not commissioned; externally peer reviewed. recorded in this study allowed robust tracking of the price of Open Access This is an Open Access article distributed in accordance with the whatever single pack was most actively being promoted to custo- Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, mers and also the price of whatever product was being advertised and license their derivative works on different terms, provided the original work is to the most price-sensitive smokers. Contrary to predictions by properly cited and the use is non-commercial. See: http://creativecommons.org/ the tobacco industry, our study provides no evidence of down- licenses/by-nc/4.0/ ward pressure on the price of prominently promoted premium single pack products. Prices also increased among the lowest- REFERENCES priced products on price boards. Real increases in price coincided 1 Liberman J, Scollo M, Freeman B, et al. Plain tobacco packaging in Australia: the with the indexation of excise/customs duty on tobacco products, historical and social context. In: Voon T, Mitchell AD, Liberman J, Ayers G, eds. suggesting that, as has been demonstrated elsewhere,31 compan- Public health and plain packaging of cigarettes. Cheltenham, UK: Edward Elgar ies are ‘over-shifting’ tax increases to consumers, taking the Publishing, 2013:30–47. opportunity of each increase in duty to increase their own 2 British American Tobacco Australia. Submission on the tobacco plain packaging bill 2011. Canberra: Department of Health and Ageing Consultation website, 2 June margins and, consequently, sales revenues. 2014, 2011. http://webarchive.nla.gov.au/gov/20130904155836/http://www. It is evident that several new value brands in relatively small yourhealth.gov.au/internet/yourhealth/publishing.nsf/Content/ novel pack sizes were promoted at low introductory prices over bata-response-tpp-oct2011~ppes 2012 prior to the introduction of plain packaging. Such packs 3 Imperial Tobacco Australia Limited. Submission to the Department of Health and Ageing regarding the Tobacco Plain Packaging Bill 2012 (Exposure draft) and were considerably cheaper per stick than packs of the traditional Consultation Paper. Sydney: 2 June 2014, 2011. http://webarchive.nla.gov.au/gov/ 20s and 25s and the increasing prevalence over time of packs in 20130329044623/http://www.yourhealth.gov.au/internet/yourhealth/publishing.nsf/ novel sizes would have resulted in lower prices than would Content/ppit~ncr~loe ii88 Scollo M, et al. Tob Control 2015;24:ii82–ii89. doi:10.1136/tobaccocontrol-2014-051950 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

4 Philip Morris Limited. Commoditising tobacco products through plain packaging will 17 Burton S, Williams K, Fry R, et al. Marketing cigarettes when all else is unavailable: harm public health, violate treaties and does not meet the test of ‘evidence-based evidence of discounting in price-sensitive neighbourhoods. Tob Control 2014;23: policy’. Melbourne: 2 June 2014, 2011. http://webarchive.nla.gov.au/gov/ e24–9. 20130329051831/http://www.yourhealth.gov.au/internet/yourhealth/publishing.nsf/ 18 Dalglish E, McLaughlin D, Dobson A, et al. Cigarette availability and price in low Content/Phillip-Morris-Limited~Phillip-Morris-Limited-2 and high socioeconomic areas. Aust N Z J Public Health 2013;37:371–6. 5 KPMG LLP. Illicit tobacco in Australia: 2013 half year report. Sydney: November 19 Padilla J. The impact of plain packaging of cigarettes in Australia: a simulation 2013, 2013. http://www.bata.com.au/group/sites/BAT_7WYKG8.nsf/ exercise. A report prepared for Philip Morris International. Brussels: LECG Consulting vwPagesWebLive/DO9879X3?opendocument&SKN=1 Belgium, May 2013, 2010. http://webarchive.nla.gov.au/gov/20130904081138/ 6 KPMG LLP. Illicit tobacco in Australia: 2013 full-year report. Sydney: April 2014, http://www.ehealth.gov.au/internet/yourhealth/publishing.nsf/Content/ 2014. http://www.bata.com.au/group/sites/BAT_7WYKG8.nsf/vwPagesWebLive/ Phillip-Morris-Limited-Annex-14 DO9FC38M?opendocument&SKN=1 20 Freeman B. Chapter 11. Tobacco advertising and promotion. Melbourne: Cancer 7 Scollo M, Bayly M, Wakefield M. Availability of illicit tobacco in small retail outlets Council Victoria, 2013. http://www.tobaccoinaustralia.org.au/chapter-11-advertising before and after the implementation of Australian plain packaging legislation. Tob 21 Wakefield M, Zacher M, Scollo M, et al. Brand placement on price boards after Control 2014. doi:10.1136/tobaccocontrol-2013-051353 tobacco display bans: a point-of-sale audit in Melbourne, Australia. Tob Control 8 Scollo M, Zacher M, Durkin S, et al. Early evidence about the predicted unintended 2012;21:589–92. consequences of standardised packaging of tobacco products in Australia: a 22 Scollo M, Occleston J, Bayly M, et al. Tobacco product developments coinciding cross-sectional study of the place of purchase, regular brands and use of illicit with the implementation of plain packaging in Australia. Tob Control 2014. doi:10. tobacco. BMJ Open 2014;4:e005873. 1136/tobaccocontrol-2013-051509. 9 The Department of Health. Tobacco key facts and figures. Canberra: Australian 23 Australian Bureau of Statistics. Technical paper: socio-economic indexes for areas Government, 2014. [updated 17 July 2014]. http://www.health.gov.au/internet/ (SEIFA) 2011. Canberra, Australian Capital Territory: ABS, April 2014, 2013. http:// main/publishing.nsf/Content/tobacco-kff www.abs.gov.au/websitedbs/censushome.nsf/home/seifa2011? 10 Australian Institute of Health and Welfare. Highlights from the 2013 survey: opendocument&navpos=260 Tobacco Smoking. Canberra: AIHW, 2014. [updated 27 July; AIHW cat. no. PHE 24 Wakefield M, Bayly M, Scollo M. Product retrieval time in small tobacco retail 145]. http://www.aihw.gov.au/alcohol-and-other-drugs/ndshs/ outlets before and after the Australian plain packaging policy: real-world study. Tob 11 Scollo M, Bayly M, Wakefield M. Did the recommended retail price of Control 2014;23:70–6. tobacco products fall in Australia following the implementation of plain 25 Bayly M, Scollo M, Wakefield M. No lasting effects of plain packaging on cigarette packaging? Tob Control 2015;24:ii26–32. pack retrieval time in small Australian retail outlets. Tob Control 2015;24: 12 Scollo M, Zacher M, Coomber K, et al. Changes in use of types of tobacco products e108–9. by pack sizes and price segments, prices paid and consumption following the 26 StataCorp. Stata Statistical Software: Release 12 [database on the Internet]. 2011. introduction of plain packaging in Australia. Tob Control 2015;24:ii66–75. 27 Anon. Retail World Annual Report 2013. Market sizes and shares. Retail World 13 Scollo M, Lindorff K, Coomber K, et al. Standardised packaging and new enlarged 2013; December:30. graphic health warnings for tobacco products in Australia—legislative requirements 28 Quit Victoria. Did growth in marketplace prices soften after the introduction of and implementation of the Tobacco Plain Packaging Act 2011 and the Competition standardised packaging of tobacco products in Australia? Melbourne: Cancer and Consumer (Tobacco) Information Standard, 2011. Tob Control 2015;24:ii9–16. Council Victoria, 2014. http://www.cancervic.org.au/plainfacts/browse.asp? 14 Scollo M. Chapter 13. The pricing and taxation of tobacco products in Australia. ContainerID=industryclaimsofreducedpricesandincreasedconsumption Melbourne, Australia: Cancer Council Victoria, 2013. http://www.webcitation.org/ 29 Gartner CE, Chapman SF, Hall WD, et al. Why we need tobacco sales data for good 6QsaCaM3R tobacco control. Med J Aust 2010;192:3–4. 15 Scollo M, Younie S, Wakefield M, et al. Impact of tobacco tax reforms on tobacco 30 Ministry of Health New Zealand. Tobacco returns. 2014. http://www.health.govt.nz/ prices and tobacco use in Australia. Tob Control 2003;12:ii59–66. our-work/preventative-health-wellness/tobacco-control/tobacco-returns 16 McCarthy M, Scully M, Wakefield M. Price discounting of cigarettes in milk bars 31 Gilmore AB. Understanding the vector in order to plan effective tobacco control near secondary schools occurs more frequently in areas with greater socioeconomic policies: an analysis of contemporary tobacco industry materials. Tob Control disadvantage. Aust N Z J Public Health 2011;35:71–4. 2012;21:119–26.

Scollo M, et al. Tob Control 2015;24:ii82–ii89. doi:10.1136/tobaccocontrol-2014-051950 ii89 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

The advertised price of cigarette packs in retail outlets across Australia before and after the implementation of plain packaging: a repeated measures observational study Michelle Scollo, Megan Bayly and Melanie Wakefield

Tob Control 2015 24: ii82-ii89 doi: 10.1136/tobaccocontrol-2014-051950

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii82

These include: Supplementary Supplementary material can be found at: Material http://tobaccocontrol.bmj.com/content/suppl/2015/02/25/tobaccocon trol-2014-051950.DC1.html References This article cites 9 articles, 5 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii82#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Brief report Did the recommended retail price of tobacco products fall in Australia following the implementation of plain packaging? Michelle Scollo, Megan Bayly, Melanie Wakefield

Centre for Behavioural ABSTRACT on seven occasions over that period.5 Excise and Research in Cancer, Cancer Objectives This study aimed to assess the extent of customs duty on tobacco products have been Council Victoria, Melbourne, Victoria, Australia any fall in recommended retail prices (RRPs) of tobacco increased twice yearly in Australia since provision products sold in Australia following the 2012 for automatic indexation of all duties was included Correspondence to implementation of plain packaging. in regulations in 1986 in order to ensure that these Dr Michelle Scollo, Centre for Methods RRPs published in price lists by the New duties kept pace with inflation.5 Indexation Behavioural Research in South Wales Retail Tobacco Traders Association covering occurred on 1 February and 1 August each year in Cancer, Cancer Council Victoria, 615 St Kilda Rd, the months of November 2011, November 2012 and line with increases in the Consumer Price Index Melbourne, VIC 3004, November 2013 were recorded for all listed brands and (CPI) over the previous 6-month periods to 31 Australia; pack sizes. Average prices per stick were computed in December and to 30 June. [email protected] 2013 dollars for cigarette brands in each of 10 possible The 6-monthly changes in tobacco excise/ Received 2 August 2014 pack sizes, for each major tobacco manufacturer, and customs duty each year have prompted tobacco Accepted 19 October 2014 across market segments. Average prices per gram in companies to adjust wholesale prices of all tobacco Australian dollars adjusted to 2013 prices (AUD2013) products each February and August. Companies were calculated for roll-your-own (RYO) tobacco brands have used the opportunity of the increases in duty for each major manufacturer. to also adjust recommended retail prices (RRPs). Results Inflation-adjusted average RRPs per stick in Large retailers commonly buy stock at lower cost November 2013 were on average 6.4% higher than in and sell cigarettes at lower-than-recommended November 2011 and 3.4% higher than in November prices. While RRPs are not compulsory, they 2012. The average RRP per gram of RYO products in provide a guide to small retailers as to prices that November 2013 was 10.2% and 5.4% higher than in can be charged for each product. Increases in the November 2011 and November 2012, respectively. excise/customs duty can be fully passed on to retai- Within cigarette brands, the highest increases in RRP lers and consumers (‘fully-shifted’). Alternatively, from 2011 to 2013 were seen among mainstream and they can be ‘under-shifted’: in this case, the retail premium brands (10.0% and 10.1%, respectively) and price goes up by less than the amount necessitated among packs of 30s (18.3%) and 50s (12.5%). by the increase in excise/customs duty. Or they can Conclusions The RRPs of tobacco products were be ‘over-shifted’—increased in price beyond the higher in real terms 1 year after Australia’s plain amount necessitated by the increase in duty. packaging legislation was implemented. These increases Increases in the Tobacco Sub-Index of Australia’s exceeded increases resulting from Consumer Price Index CPI suggest that biannual adjustments have consist- (CPI) indexation of duty and occurred across all three ently been used by the industry in Australia as an major manufacturers for both factory made and RYO opportunity for over-shifting—so that prices of brands, all three cigarette market segments and all tobacco products have increased in Australia by major pack sizes. more than the overall rate of inflation even over periods where there have been no real increases in excise/customs duty.6 INTRODUCTION Official government statistics collected in order The tobacco industry claimed that the introduction to compute the country’s CPI suggest that overall of plain packaging of tobacco products in Australia tobacco prices did not fall following the introduc- would lead to a drop in the price of tobacco pro- tion of plain packaging.7 Indeed, consistent with ducts as manufacturers would be forced to compete the previous 10 years, they appear to have risen by – on price alone for market share.1 3 It was claimed more than the overall rate of inflation.6 However, it that falls would be particularly marked for must be recognised that the Tobacco Sub-index of Open Access Scan to access more premium brands.3 The tobacco plain packaging the CPI does not provide details on—and may not free content legislation came into effect between October and reflect overall trends taking into account changes in December 2012, with Australian manufacturers —patterns of price increases of different types of required to produce only compliant packs from 1 products and brands. October 2012, and retailers permitted to sell only We aimed to assess whether the RRPs of tobacco compliant packs from 1 December 2012.4 products produced by the three largest tobacco There were no real increases in excise and companies operating in Australia fell following the customs duty on tobacco products in Australia implementation of plain packaging. To do this, we To cite: Scollo M, Bayly M, between a 25% increase in duty on 29 April 2010 compared changes in RRPs of all the products sold Wakefield M. Tob Control and a 12.5% increase in duty on 1 December by these three companies between November 2011 – 2015;24:ii90 ii93. 2013, though 6-monthly duty indexation occurred (immediately after passage of the legislation ii90 Scollo M, et al. Tob Control 2015;24:ii90–ii93. doi:10.1136/tobaccocontrol-2014-051948 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Brief report through Parliament8), to November 2012 (by which time plain RRP by the number of sticks in each cigarette pack (or grams in packs were being purchased by the majority of smokers9) and each RYO product). These prices were then converted into stand- November 2013 (prior to the large increase in duty on 1 ard November 2013 prices to take account of inflation. To do this December 201310 11), a period during which there were no real we multiplied the listed price by the all groups, all capital cities increases (ie, no increases beyond the 6-monthly indexation) in Australian CPI figure for June 2013, divided by the June 2012 excise/customs duty. index figure for November 2012 prices (and by the June 2011 index figure for November 2011 prices).16 In this way, any METHOD observed increases in prices would reflect those made over and RRPs for the majority of tobacco products on the Australian above CPI-linked increases in tobacco excise/customs duty. The market are available in The Australian Retail Tobacconist (ART), a average RRP per stick in 2013 dollars (AUD2013) of cigarettes retailer’s trade magazine that is typically published quarterly. The was then calculated by manufacturer, market segment and pack magazine also lists the pack size and the manufacturer of each size and for the 10 most popular FMC products in Australia.17 brand including all brand extensions; for instance, Pall Mall, Pall The average recommended retail gram price (AUD2013) of RYO Mall Slims and Pall Mall Extra Kings. Factory-made cigarettes tobacco was calculated by manufacturer only, as pouch sizes (FMCs) were categorised into market segments as ‘value’, ‘main- within each brand changed considerably in the examined period. stream’ and ‘premium’ according to the classification adopted by the Retail World grocery magazine.12 Where products were not RESULTS mentioned in Retail World we imputed a market segment category Table 1 shows that the average inflation-adjusted RRPs of cigar- by comparing the stick price of these products to the highest and ettes and RYO tobacco increased from 2011 to 2012, and again lowest prices per stick of products in each category. from 2012 to 2013. The average overall inflation-adjusted We examined the RRPs listed that were applicable in November increase in RRP per stick was 2.9% from 2011 to 2012 and of 2011, 2012 and 2013 for all tobacco products manufactured 3.4% from 2012 to 2013. This varied across manufacturers, by British American Tobacco Australia (BATA), Imperial Tobacco where the increase for Imperial Tobacco brands between 2011 and Philip Morris, and noted that no individual pack price was and 2013 was approximately half that of BATA and Philip – lower in 2013 than it was in 2011 or 2012.13 15 However, many Morris, and by market segment, where the value segment brands were released in new pack sizes in 2013, and we wanted to increased by approximately 3.7 percentage points less than account for these new products as well as for inflation when asses- mainstream and premium segments. sing the extent of changes in price over the 2 years from 2011. To Table 1 also shows that the total number of products (brand– calculate prices per cigarette stick (or per gram of each pack size configurations) increased over time. It can be seen that roll-your-own (RYO) tobacco pouch), we divided each product’s this change predominantly occurred in the value market segment,

Table 1 Average recommended retail price in AUD2013 per stick for factory made cigarettes in 2011–2013 by manufacturer, market segment and pack size Per stick price, AUD2013 (SD; number of products) Per cent change

At November 2011 At November 2012 At November 2013 2011–2012 2012–2013 2011–2013

Brands n=27 n=28 n=27 Brand–pack sizes n=62 n=65 n=71 All cigarette brand–pack sizes $A0.68 ($A0.07; 62) $A0.70 ($A0.08; 65) $A0.72 ($A0.09; 71) 2.9 3.4 6.4 Manufacturer BATA $A0.71 ($A0.08; 29) $A0.74 ($A0.08; 28) $A0.76 ($A0.09; 34) 5.0 2.1 7.2 Imperial $A0.65 ($A0.06; 17) $A0.66 ($A0.08; 19) $A0.67 ($A0.08; 21) 2.4 1.1 3.5 Philip Morris $A0.65 ($A0.04; 16) $A0.66 ($A0.06; 18) $A0.70 ($A0.06; 16) 1.1 6.8 8.0 Market segment Value $A0.60 ($A0.01; 20) $A0.62 ($A0.04; 24) $A0.64 ($A0.04; 31) 2.4 3.8 6.3 Mainstream $A0.65 ($A0.03; 16) $A0.68 ($A0.03; 17) $A0.71 ($A0.04; 16) 4.6 5.1 10.0 Premium $A0.75 ($A0.05; 26) $A0.78 ($A0.05; 24) $A0.82 ($A0.04; 24) 4.9 5.0 10.1 Pack size 20 sticks $A0.70 ($A0.08; 28) $A0.71 ($A0.09; 27) $A0.74 ($A0.10; 25) 2.3 3.4 5.8 21 sticks (0) (0) $A0.65 (<$A0.01; 2) 22 sticks (0) (0) $A0.61 (1) 23 sticks (0) (0) $A0.63 ($A0.06; 2) 25 sticks $A0.68 ($A0.06; 19) $A0.71 ($A0.06; 21) $A0.76 ($A0.08; 21) 4.5 6.6 11.4 26 sticks (0) $A0.57 (1) $A0.63 ($A0.03; 3) 10.2 30 sticks $A0.62 ($A0.02; 5) $A0.65 ($A0.01; 5) $A0.73 ($A0.08; 7) 4.9 12.8 18.3 35 sticks $A0.71 ($A0.08; 3) $A0.74 ($A0.09; 3) $A0.66 (1) 4.5 −11.1* −7.1 40 sticks $A0.60 ($A0.02; 5) $A0.61 ($A0.03; 6) $A0.65 ($A0.04; 7) 2.1 5.2 7.4 50 sticks $A0.59 (<$A0.01; 2) $A0.63 (<$A0.01; 2) $A0.66 ($A0.01; 2) 6.0 6.1 12.5 *Note: The 11.1% reduction in recommended retail prices (RRP) among packs of 35 sticks from 2012 to 2013 did not actually correspond to a decrease in RRP among any individual product. Rather, coinciding with the introduction of plain packaging, two of the three brands available in 35s, Cambridge and Wills, were repackaged from 35 to 30 stick packs. As the only remaining brand in this category, Escort, was a mainstream brand, the per stick RRP of 35s was substantially lower in 2013. The change in RRPs among packs of 30s was correspondingly higher. BATA, British American Tobacco Australia.

Scollo M, et al. Tob Control 2015;24:ii90–ii93. doi:10.1136/tobaccocontrol-2014-051948 ii91 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Brief report

Figure 1 Annual per cent change in recommended retail prices (RRP) for the 10 leading FMC products in Australia and overall for all FMC products within each market segment. FMC, Factory-made cigarettes; BATA, British American Tobacco Australia; IT, Imperial Tobacco; PM: Philip Morris. *RRP for JPS 25s was not listed at November 2013. Source: selection of market-leading brands based on data from Euromonitor International presented in table 10.6.2 at http://www.tobaccoinaustralia.org.au/ chapter-10-tobacco-industry/10-6-market-share-and-brand-share. with the introduction of several small unusual pack sizes (ie, 21, For RYO tobacco products, an overall real increase in per 22, 23 and 26 sticks). Even though the RRP of these packs with gram RRP of 10.2% was observed and the increases from 2011 ‘bonus’ cigarettes was typically the same or lower than the equiva- to 2012 and 2012 to 2013 were of a similar magnitude lent packs of 20 or 25 within that brand, the net effect of the (table 2). The average inflation-adjusted RRP of BATA RYO pro- inclusion of these new cheap packs did not result in lower average ducts appeared to increase substantially in 2012; this was due to per stick RRPs in November 2013. Imperial Tobacco showed the very cheap BATA RYO brand, Samson, being removed from the lowest increase over the examined period, and this can largely be ART listing in 2012. Similarly, the atypically small increase in attributed to new JPS products, a value brand, being introduced at RRP observed for Imperial RYO products from 2011 and 2012 low RRPs in a range of brand extensions and several pack sizes was due to two expensive RYO varieties being including 20s, 23s, 26s and 40s. However, even among value removed from the ART in 2012. Imperial Tobacco brands, RRPs increased in real terms by an average of 5% from 2011 to 2013 (data not shown). DISCUSSION The annual per cent change in RRP per stick of the 10 The tobacco industry claimed that prices of tobacco products leading FMC products (brands and pack sizes) in Australia are would fall in Australia following the introduction of plain shown in figure 1. Figure 1 shows that while the per cent packaging and that this would lead to an increase in con- change observed in both years was variable between market seg- sumption. The major companies operating in Australia intro- ments and individual products, overall, as also seen in table 1, duced several new products in unusual pack sizes in the lead the increase observed in 2012 to 2013 was larger than that of up to plain packaging that were indeed priced cheaper per 2011 to 2012. The variability in per cent change observed in pack and cheaper per stick than existing products on the table 1 between product categories is attributable to variability market.18 However, examination of price lists for cigarettes in increases across brands as well as changes in the product and RYO tobacco show that the RRPs of all other products offerings within categories, that is, the removal and introduction produced by BATA, Imperial Tobacco and Philip Morris of sub-brands and pack sizes. However, the net effect is wide- increased well above inflation from 2011 to 2012, and from spread RRP increases over and above inflation. 2012 to 2013.

Table 2 Average recommend retail price in AUD2013 per gram for 2011–2013 by manufacturer for RYO tobacco Per gram price, AUD2013 (SD; number of products) Per cent change

At November-2011 At November-2012 At November-2013 2011–2012 2012–2013 2011–2013

Brands n=22 n=23 n=22 Brand–pack sizes n=35 n=37 n=36 All RYO pouches $A0.72 ($A0.09; 35) $A0.75 ($A0.04; 37) $A0.79 ($A0.05; 36) 4.5 5.4 10.2 Manufacturer BATA $A0.68 ($A0.09; 8) $A0.76 ($A0.05; 10) $A0.80 ($A0.03; 9) 11.3 4.9 16.8 Imperial $A0.75 ($A0.09; 21) $A0.76 ($A0.04; 21) $A0.80 ($A0.04; 21) 1.2 6.0 7.3 Philip Morris $A0.66 ($A0.03; 6) $A0.71 ($A0.03; 6) $A0.74 ($A0.05; 6) 7.9 4.5 12.7 BATA, British American Tobacco Australia; RYO, roll-your-own. ii92 Scollo M, et al. Tob Control 2015;24:ii90–ii93. doi:10.1136/tobaccocontrol-2014-051948 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Brief report

This analysis is limited to the published RRPs. Trade publica- and license their derivative works on different terms, provided the original work is tions such as the ART can be slow to reflect actual changes in properly cited and the use is non-commercial. See: http://creativecommons.org/ the market. For example, Holiday 22s were available in retail licenses/by-nc/4.0/ 19 outlets throughout 2012 but were not included in the ART REFERENCES until February 2013. RRPs are generally somewhat higher than 1 British American Tobacco Australia. Submission on the Tobacco Plain Packaging Bill prices observed in retail outlets,20 with lower prices observed in 2011. Canberra: Department of Health and Ageing Consultation website, 2 June low-SES neighbourhoods and in those with higher percentages 2014 2011. http://content.webarchive.nla.gov.au/gov/wayback/20130904173432/ 21 http://www.yourhealth.gov.au/internet/yourhealth/publishing.nsf/Content/ of people under the age of 18. However, at least one study 2EE2F6F3EC74F628CA2579540005F68B/$File/British%20American%20Tobacco% suggests that RRPs do tend to follow very similar patterns of 20Australia%20-%20Public%20Submission.pdf increases over time to those observed in retail outlets.22 The 2 Philip Morris Limited. Commoditising tobacco products through plain packaging will increases in RRPs of value brands observed between 2011 and harm public health, violate treaties and does not meet the test of ‘evidence-based ’ 2013 are in line with increases observed in audits of prices of policy . Melbourne, 2 June 2014 2011. http://webarchive.nla.gov.au/gov/ 19 20130329051831/http://www.yourhealth.gov.au/internet/yourhealth/publishing.nsf/ cheap brands promoted in retail outlets. Content/Phillip-Morris-Limited~Phillip-Morris-Limited-2 In conclusion, this analysis shows that average inflation-adjusted 3 Padilla J. The impact of plain packaging of cigarettes in Australia: a simulation increases in RRPs were greater for premium and mainstream exercise. A report prepared for Philip Morris International. Brussels: LECG brands than for value brands, greater for RYO than for FM cigar- Consulting Belgium, May 2013 2010. http://webarchive.nla.gov.au/gov/ 20130904081138/http://www.ehealth.gov.au/internet/yourhealth/publishing.nsf/ ettes, greater among the 25s, 30s and 50s than among other pack Content/Phillip-Morris-Limited-Annex-14 sizes, and greater among Philip Morris brands than among brands 4 Tobacco Plain Packaging Act, Stat. No. 148 (2011). http://www.comlaw.gov.au/ of the other two companies. Increases were particularly large Details/C2013C00190 among the market leading brands used by the majority of 5 Scollo M. Chapter 13. The pricing and taxation of tobacco products in Australia. Australian smokers. Industry pricing strategies will no doubt drive Melbourne, Australia: Cancer Council Victoria, 2013. http://www.webcitation.org/ 6QsaCaM3R changes in market share for various product types in Australia, so 6 Quit Victoria. Did growth in marketplace prices soften after the introduction of 23 consumer surveys such as that reported on in this volume will standardised packaging of tobacco products in Australia? Melbourne: Cancer be needed to compute average changes in prices paid within and Council Victoria, 2014. http://www.cancervic.org.au/plainfacts/browse.asp? between various market segments taking into account real world ContainerID=industryclaimsofreducedpricesandincreasedconsumption 7 Australian Bureau of Statistics. 6401.0 Consumer Price Index, Australia Table 11. retail prices and brand choices. Nonetheless, following the intro- CPI: group, sub-group and expenditure class, index numbers by capital city duction of plain packaging in Australia, it is clear that companies [database on the Internet]. ABS, 2014 [cited March 2014]. http://www.abs.gov.au/ continued to increase the RRPs of tobacco products in all market AUSSTATS/[email protected]/DetailsPage/6401.0Jun%202012?OpenDocument segments above and beyond the rate of inflation. This contrasts 8 Cancer Council Victoria. Plain tobacco packaging in Australia: a time-line of starkly with claims by companies and others opposing tobacco legislative and legal developments. Melbourne: Cancer Council Victoria, 2014. http://www.cancervic.org.au/plainfacts/browse.asp?ContainerID=time plain packaging legislation that prices would fall. lineandinternationaldevelopments 9 Wakefield MA, Hayes L, Durkin S, et al. Introduction effects of the Australian plain packaging policy on adult smokers: a cross-sectional study. BMJ Open 2013;3:pii: e003175. What this paper adds 10 Excise Tariff Amendment (Tobacco) Act 2014, Stat. No 9 (2014). http://www. comlaw.gov.au/Details/C2014A00009 11 Customs Tariff Amendment (Tobacco) Act 2014, Stat. No 8 (2014). http://www. ▸ The tobacco industry claimed that standardised packaging comlaw.gov.au/Details/C2014A00008 would result in a drop in tobacco prices, particularly for 12 Anon. Retail World Annual Report 2013. Market sizes and shares. Retail World premium brands. 2013;December:30. ▸ fl 13 NSW Retail Tobacco Traders’ Association. Price lists−cigarettes. Aust Retail Examination of the in ation-adjusted recommended retail – ’ Tobacconist 2011;83:1 4. prices (RRPs) of all cigarette products from Australia s three 14 NSW Retail Tobacco Traders’ Association. Price lists−cigarettes. Aust Retail largest tobacco manufacturers shows that RRPs in fact Tobacconist 2012;86:1–16. increased by 3.4% on average from 2012 to 2013, with 15 NSW Retail Tobacco Traders’ Association. Price lists−cigarettes. Aust Retail increases greater among premium and mainstream than Tobacconist 2013;89:17. 16 Australian Bureau of Statistics. 6401.0 Consumer Price Index, Australia Table 11. among value brands. Roll-your-own tobacco products showed CPI: group, sub-group and expenditure class, index numbers by capital city an even greater average increase of more than 5.4%. [database on the Internet]. ABS, 2013 [cited March 2014]. http://www.abs.gov.au/ ▸ Tobacco companies continue to increase RRPs of tobacco AUSSTATS/[email protected]/DetailsPage/6401.0Jun%202012?OpenDocument products beyond the rate of inflation in Australia. 17 Freeman B. Chapter 10. The tobacco industry in Australian society. Melbourne: Cancer Council Victoria, 2013. http://www.tobaccoinaustralia.org.au/chapter-10-tobacco-industry 18 Scollo M, Occleston J, Bayly M, et al. Tobacco product developments coinciding Contributors MS designed this study and coordinated collection of data. MB with the implementation of plain packaging in Australia. Tob Control 2014. collected data and undertook data analysis. All authors contributed to the drafting doi:10.1136/tobaccocontrol–2013–051509 and finalisation of the manuscript. 19 Scollo M, Bayly M, Wakefield M. The advertised price of cigarette packs in retail outlets across Australia before and after the implementation of plain packaging: a Funding Cancer Council Victoria. repeated measures observational study. Tob Control 2015;24:ii82–9. Competing interests The authors wish to advise that MS was a technical writer 20 Scollo M, Owen T, Boulter J. Price discounting of cigarettes during the National for and MW a member of the Tobacco Working Group of the Australian National Tobacco Campaign. Canberra: Commonwealth Department of Health and Aged Preventive Health Task Force and MW was a member of the Expert Advisory Care, 2000. http://www.health.gov.au/internet/main/publishing.nsf/Content/ Committee on Plain Packaging that advised the Australian Department of Health on 7E318B2BCB5DAE26CA256F190004524F/$File/tobccamp_2-ch5.pdf research pertaining to the plain packaging legislation. MW holds competitive grant 21 Burton S, Williams K, Fry R, et al. Marketing cigarettes when all else is unavailable: funding from the Australian National Health and Medical Research Council, US evidence of discounting in price-sensitive neighbourhoods. Tob Control 2014;23:e24–9. National Institutes of Health, Australian National Preventive Health Agency and 22 Scollo M, Freeman J, Wakefield M. Ch 7. Early evidence of impact of recent reforms BUPA Health Foundation. to tobacco taxes on tobacco prices and tobacco use in Australia. Canberra: Australian Government Department of Health and Ageing, 2004. http://www.quitnow.gov.au/ Provenance and peer review Not commissioned; externally peer reviewed. internet/quitnow/publishing.nsf/Content/national-tobacco-campaign-lp Open Access This is an Open Access article distributed in accordance with the 23 Scollo M, Zacher M, Coomber K, et al. Changes in use of types of tobacco products Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which by pack sizes and price segments, prices paid and consumption following the permits others to distribute, remix, adapt, build upon this work non-commercially, introduction of plain packaging in Australia. Tob Control 2015;24:ii66–75.

Scollo M, et al. Tob Control 2015;24:ii90–ii93. doi:10.1136/tobaccocontrol-2014-051948 ii93 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Did the recommended retail price of tobacco products fall in Australia following the implementation of plain packaging? Michelle Scollo, Megan Bayly and Melanie Wakefield

Tob Control 2015 24: ii90-ii93 doi: 10.1136/tobaccocontrol-2014-051948

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii90

These include: References This article cites 5 articles, 1 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii90#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Brief report Personal pack display and active smoking at outdoor café strips: assessing the impact of plain packaging 1 year postimplementation Meghan Zacher,1 Megan Bayly,1 Emily Brennan,1 Joanne Dono,2 Caroline Miller,2,3 Sarah Durkin,1 Michelle Scollo,1 Melanie Wakefield1

1Cancer Council Victoria, ABSTRACT placed on top, or in an external case. The current Melbourne, Victoria, Australia Aims We observed tobacco pack display and smoking study extends these findings by assessing whether 2South Australian Health and Medical Research Institute, at outdoor venues over three summers to assess changes these changes were sustained 1 year after PP was Adelaide, South Australia, in their prevalence following Australia’s introduction of introduced. Australia plain tobacco packaging with larger pictorial health 3Discipline of Public Health, warnings. METHODS University of Adelaide, Methods Between January and April 2012 (preplain Details regarding sample selection and data collection Adelaide, South Australia, Australia packaging (PP)), 2013 (early post-PP) and 2014 (1 year for the pre-PP and early post-PP phases have been post-PP), we counted patrons, smokers and tobacco described elsewhere,12and similar methods were Correspondence to packs at cafés, restaurants and bars with outdoor used for the 1 year post-PP phase. Briefly, for the fi Professor Melanie Wake eld, seating. Pack type (branded, plain or unknown) and pre-PP phase we selected street segments (referred to Centre for Behavioural ‘ ’ Research in Cancer, Cancer orientation were noted. Rates of active smoking, pack as café strips ) from a range of socioeconomic areas Council Victoria, 615 St Kilda display and pack orientation were analysed using in Melbourne and Adelaide that were known to have Road, Melbourne, VIC 3004, multilevel Poisson regression. many popular cafés, restaurants and bars. Australia; melanie.wakefield@ Results Prevalence of pack display among patrons Fieldworkers sampled every venue in their assigned cancervic.org.au declined from pre-PP (1 pack per 8.7 patrons) to early café strip/s which had outdoor seating visible from Received 5 June 2014 post-PP (1 pack per 10.4), and remained low 1 year the footpath. New venues were added to the sample Accepted 27 August 2014 post-PP (1 pack per 10.3). This appeared to be driven by if they had opened between phases. a sustained decline in active smoking post-PP (pre-PP: Between mid-October and mid-April of 2011– 8.4% of patrons were smoking; early post-PP: 6.4%; 2012 (pre-PP) and 2012–2013 (early post-PP), field- 1 year post-PP: 6.8%). Notably, active smoking declined workers conducted nine waves of observations at more in venues with children present than in those approximately 2-week intervals, achieving high inter- without. While early post-PP, plain packs were less often rater reliability.2 For the 1 year post-PP phase, five displayed face-up (74.0%) and more often concealed waves of data were collected at approximately 2-week (8.9%) than branded packs pre-PP (face-up: 85.2%; intervals between mid-January and mid-April 2014. concealed: 4.0%), this was not sustained 1 year post-PP At each venue, fieldworkers counted the number (face-up: 85.7%; concealed: 4.4%). Also, external case of seated patrons, patrons smoking, holding or use increased from pre-PP (1.2%) to early post-PP (3.5%), lighting a cigarette (‘active smokers’), and tobacco but returned to pre-PP levels 1 year post-PP (1.9%). packs, noting the pack type in the post-PP phases Conclusions This study demonstrated a sustained (branded, plain or unknown). They also recorded reduction in visibility of tobacco products and smoking in whether children were present, how many packs public, particularly in the presence of children, from pre-PP were oriented face-up with the brand name and to 1 year post-PP. This effect is likely to reduce smoking- variant visible, face-down, standing or on their side related social norms, thereby weakening an important or concealed by an object like a wallet or phone (by influence on smoking uptake and better supporting quit pack type), and how many packs were in an attempts. unknown orientation due to distance or an external case (not recorded by pack type).

INTRODUCTION Statistical analysis In two recent studies,12we monitored prevalence We used data from the fivewavesineachofthe and nature of personal tobacco pack display and three phases that were conducted between January smoking in outdoor areas of cafés, restaurants and April. Preliminary analyses confirmed that Open Access Scan to access more and bars in two Australian cities (Melbourne and restricting the pre-PP and early post-PP periods to free content Adelaide) in order to evaluate whether such beha- the five waves of data did not substantially change viours were more common in a branded environment the results from those previously published for than under plain packaging (PP), which was fully these periods.12Multilevel Poisson models were implemented in Australia on 1 December 2012.3 We employed in Stata 12.14 to test whether outcomes found that prevalence of pack display and active of interest 1 year post-PP were different than smokingdeclinedfrompre-PP(October2011to pre-PP and early post-PP. We report April 2012) to early post-PP (October 2012 to April Bonferroni-adjusted p values to account for mul- To cite: Zacher M, Bayly M, 2013), particularly in the presence of children.2 Packs tiple comparisons. Random intercepts were Brennan E, et al. Tob were also less often oriented face-up early post-PP, included for café strip and venue in all models to – Control 2015;24:ii94 ii97. and were more likely to be concealed with an object adjust SEs for correlations among venues within the ii94 Zacher M, et al. Tob Control 2015;24:ii94–ii97. doi:10.1136/tobaccocontrol-2014-051826 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Brief report same café strip and for multiple observations over time within once. There were 6997 observations from these venues (pre-PP: the same venue. All models adjusted for city, area SES using an n=2189; early post-PP: n=2367; 1 year post-PP: n=2441), Index of Relative Disadvantage,5 presence of children, month though many (n=3050) were not included in regression analyses (January, February and March/April), day and time (weekdays as no patrons were present. Patrons were present at a total of before 16:00, weekdays after 16:00, and weekends), tempera- 3947 venue observations (pre-PP: n=1340; early post-PP: ture (<22°C, 22–27°C, ≥28°C), and wind speed (in km/h). n=1296; 1 year post-PP: n=1311). Fewer venue observations To analyse outcomes as rates, offset terms were used. Number were used in analyses of the rates of packs to active smokers of patrons was the offset term for the rates of packs to patrons and (n=1195), face-up and concealed packs to known-orientation smokers to patrons; at least one patron had to be recorded for an branded (pre-PP) or plain (post-PP) packs (n=1381), and exter- observation to be analysed. Similarly, number of smokers was the nal cases to all packs (n=1470). offset for the rate of packs to smokers, and only observations with one or more smokers present were included. We compared rates Prevalence of pack display and active smoking of face-up orientation and pack concealment among branded The rate of pack display among patrons was lower in the early packs pre-PP to rates among plain packs post-PP.As it was impos- post-PP phase and 1 year post-PP compared with pre-PP, and sible to tell in the post-PP phases whether packs in external cases there was no change between early and 1 year post-PP (table 1). and those packs in unknown orientation (due to distance) were While one in every 8.7 patrons displayed a pack pre-PP, plain or branded, they could not be included in these analyses. this declined to one in every 10.4 patrons early post-PP and one Accordingly, the equivalent packs were excluded from the pre-PP in every 10.3 patrons 1 year post-PP. Prevalence of active data, so that denominators were comparable with the post-PP smoking also declined, from 8.4% of patrons pre-PP to 6.4% phases. Only observations for which at least one known-orienta- early post-PP and remained lower (at 6.8%) 1 year post-PP. The tion branded (pre-PP) or plain (early or 1 year post-PP) pack was rate of pack display relative to active smokers (not shown in recorded were analysed. The rate of external case use was ana- table 1) did not change between pre-PP and early post-PP lysed out of all observed packs; accordingly, at least one pack (Incident Rate Ratio (IRR)=1.03, p=1.000) or between pre-PP had to be observed to be analysed. We also tested whether and 1 year post-PP (IRR=0.95, p=1.000). In each phase, there declines in pack display and active smoking among patrons was one pack observed for every 0.7 active smokers (thus, there between pre-PP and 1 year post-PP were again greater in venues were more packs displayed than there were patrons actively with children present than in those without. Finally, we con- smoking). ducted sensitivity analyses to assess whether excluding venues There was a significant interaction for the rate of active which were not observed in all three phases altered our results. smoking among patrons between pre-PP and 1 year post-PP and the presence of children at a venue (p=0.015), with a greater RESULTS decline in venues with children present (IRR=0.47, p<0.001) In total, 585 unique venues were observed over the course of than in those without (IRR=0.88, p=.058) (figure 1). A similar the study, of which 519 venues had patrons present at least pattern was observed by presence of children for pack display

Table 1 Changes in personal pack display, active smoking and pack orientation between preplain packaging ( January–April 2012), and early ( January–April 2013) and 1 year ( January–April 2014) post-plain packaging Versus pre-PP Versus early post-PP

Outcome (rate of…) Rate Percentage of… Adjusted IRR p Value* Adjusted IRR p Value*

Packs to patrons …patrons who displayed pack Pre-PP 1:8.7 11.5 Ref 1.21 <0.001 Early post-PP 1:10.4 9.7 0.83 <0.001 Ref 1 year post-PP 1:10.3 9.7 0.84 0.001 1.02 1.000 Active smokers to patrons …patrons who were smoking Pre-PP 1:11.9 8.4 Ref 1.28 <0.001 Early post-PP 1:15.7 6.4 0.78 <0.001 Ref 1 year post-PP 1:14.7 6.8 0.85 0.013 1.08 0.607 Face-up packs to all packs†…packs† that were face-up Pre-PP 1:1.2 85.2 Ref 1.15 0.037 Early post-PP 1:1.4 74.0 0.87 0.037 Ref 1 year post-PP 1:1.2 85.7 0.99 1.000 1.14 0.087 Concealed packs to all packs†…packs† that were concealed Pre-PP 1:25.3 4.0 Ref 0.40 <0.001 Early post-PP 1:11.3 8.9 2.48 <0.001 Ref 1 year post-PP 1:22.5 4.4 1.22 1.000 0.49 0.005 External cases to all packs …packs that were in cases Pre-PP 1:83.1 1.2 Ref 0.29 0.001 Early post-PP 1:28.3 3.5 3.44 0.001 Ref 1 year post-PP 1:52.2 1.9 1.36 1.000 0.40 0.012 *Bonferonni-adjusted p-values are reported.. †At pre-plain packaging, ‘all packs’ for the rates of face-up orientation and pack concealment includes branded packs in known orientations (face-up, face-down, standing or concealed); at post-plain packaging, ‘all packs’ includes plain packs in known orientations (face-up, face-down, standing or concealed). PP, plain packaging.

Zacher M, et al. Tob Control 2015;24:ii94–ii97. doi:10.1136/tobaccocontrol-2014-051826 ii95 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Brief report

DISCUSSION The results of this study extend previous findings2 by confirm- ing that declines in personal pack display and active smoking at outdoor public venues were maintained 1 year after the intro- duction of plain tobacco packaging with refreshed and enlarged graphic health warnings (GHWs). While 1 in every 8.7 patrons displayed a tobacco pack pre-PP, this declined to 1 in 10.4 patrons early post-PP and 1 in 10.3 patrons 1 year post-PP. Prevalence of active smoking declined from 8.4% of patrons pre-PP to 6.4% early post-PP and 6.8% 1 year post-PP. These findings are consistent with naturalistic studies which have found that, when told to carry mocked-up plain packs with large GHWs, smokers smoked less around others and were more likely to forgo cigarettes than when carrying their regular branded packs.67It is also notable that, consistent with effects Figure 1 Percentage of patrons who displayed a pack (pack display) observed between pre-PP and early post-PP,2 there were greater and percentage of patrons actively smoking (smoking), by phase and presence of children at venue. declines pre-PP to 1 year post-PP in active smoking at venues where children were present. This suggests an enhancement of social pressure to forego smoking when children are present. Findings reported here are consistent with those of experimen- among patrons, but the interaction was not significant tal and descriptive research which has shown that plain packs are (p=0.211; figure 1). less appealing and convey more negative characteristics of smokers than traditional branded packs.89Smokers may have Pack orientation chosen to light up less often in public after plain packs were The percentage of packs oriented face-up declined from pre-PP introduced to elude being judged by passers-by or to avoid feel- (branded packs: 85.2%) to early post-PP (plain packs: 74.0%), ings of shame or embarrassment. Observed declines in pack but returned to the baseline level 1 year post-PP (plain packs: display and active smoking could also reflect the reduced preva- 85.7%; tables 1 and 2). Similarly, while pack concealment lence of population smoking over time, evident from recently increased between the pre-PP (branded packs: 4.0%) and early released national survey data showing a reduction in 14+ years post-PP (plain packs: 8.9%) phases, concealment resumed its daily smoking from 15.1% in 2010 to 12.8% in 2013.10 baseline level by 1 year post-PP (plain packs: 4.4%). Finally, Changes in pack orientation that were observed early post-PP though prevalence of use of external cigarette cases increased were not sustained 1 year post-PP. These findings may indicate between pre-PP (1.2%) and early post-PP (3.5%), at 1 year that the use of external cases became more of a nuisance over post-PP (1.9%) it was no different to pre-PP. time and/or that the salience of the larger front-of-pack GHWs and impact of the removal of reassuring brand imagery decreased Sensitivity analyses over time (reducing smokers’ motivation to conceal the packs); a Of the 519 venues observed that had patrons present at least wear-out effect that would be expected.11 However, a positive once, 10 were not observed in one or both of the post-PP consequence of these findings is that less pack concealment phases, as they banned smoking in outdoor areas, and an add- 1 year post-PP means that the owner of the pack, other smokers, itional 161 venues were not open for business in all three and non-smokers are more frequently exposed in an incidental phases. Sensitivity analyses excluding all observations from manner to the large graphic warning images. venues which banned smoking outdoors or which were not One limitation of the study is that there was a substantial open for business in all three phases obtained results similar to increase in tobacco excise and customs duties on 1 December those of the main analysis. 2013, which could have independently resulted in smokers

Table 2 Pack orientations, by phase and pack type Pre-plain packaging Early post-plain packaging 1 year post-plain packaging

All packs Branded packs* All packs† Plain packs Branded packs All packs† Plain packs Branded packs

Total N 1164 1112 907 788 33 940 854 43 Orientation (%) Face-up 81.4 85.2 67.5 74.0 84.8 82.0 85.7 86.0 Face-down 8.0 8.4 11.7 12.9 12.1 7.2 7.7 4.7 Standing 2.4 2.5 3.7 4.2 0.0 2.1 2.1 2.3 Concealed 3.8 4.0 9.7 8.9 3.0 5.0 4.4 7.0 External case 1.2 – 3.5 –– 1.9 –– Undetermined 3.3 – 3.9 –– 1.7 –– *At pre-plain packaging, ‘Branded packs’ excludes packs in external cases and undetermined orientations, which would have been recorded as an unknown packaging type in the post-plain packaging phases. This means that the three phases have equivalent denominators. †At post-plain packaging, ‘All packs’ includes plain and branded packs as well as packs of unknown type (early post-PP: n=86; 1 year post-PP: n=43). Of these unknown type packs, the majority were coded as having an undetermined orientation or were in an external case. Some were concealed, and a small number (2 in early post-plain packaging and 3 in 1 year post-plain packaging) had valid data for pack orientation but missing data for the pack type variable. PP, plain packaging. ii96 Zacher M, et al. Tob Control 2015;24:ii94–ii97. doi:10.1136/tobaccocontrol-2014-051826 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Brief report decreasing their consumption or trying to quit in the early Ethics approval This was an unobtrusive observational study counting café months of 2014.12 13 Similarly, the decline in pack display patrons and cigarette packs and was conducted in a public setting. might also reflect lower willingness to risk being asked to share Provenance and peer review Not commissioned; externally peer reviewed. a cigarette as the price of tobacco increases, although the lower Open Access This is an Open Access article distributed in accordance with the display rates when children are present tends to support more Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which of a social pressure interpretation. Another limitation is that we 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 only observed behaviours in two metropolitan areas, and so properly cited and the use is non-commercial. See: http://creativecommons.org/ cannot generalise our results nationally. However, a strength of licenses/by-nc/4.0/ our study is that outcomes are not subject to social desirability bias or misreporting, and reflect behaviours and the visibility of tobacco products and smoking in real-world scenarios. REFERENCES fi The visibility of tobacco use in public places is increasingly 1 Wake eld M, Zacher M, Bayly M, et al. The silent salesman: an observational study of personal tobacco pack display at outdoor café strips in Australia. Tob Control receiving research attention as an indicator of – 14–16 2014;23:339 44. de-normalisation. Lower rates of exposure (particularly 2 Zacher M, Bayly M, Brennan E, et al. Personal tobacco pack display before and among children) to smoking in public is likely to influence after the introduction of plain packaging with larger pictorial health warnings in social norms for tobacco use.17 Also, the reduction in visible Australia: an observational study of outdoor café strips. Addiction – smoking-related cues may aid quit attempts.18 19 Our finding of 2014;109:653 62. 3 Tobacco Plain Packaging Act 2011 (Cth), Stat. No. 148 (2011). http://www.austlii. a sustained reduction 1 year after PP implementation in the fre- edu.au/au/legis/cth/num_act/tppa2011180/ quency of pack display and active smoking in public venues may 4 StataCorp. Stata Statistical Software: Release 12. College Station, TX: StataCorp LP, therefore have positive impacts both for those trying to quit and 2011. ultimately, for reducing smoking uptake. 5 Australia Bureau of Statistics. Technical paper: socio-economic indexes for areas (SEIFA) 2011 Catalogue no. 2033.0.55.001. Canberra, Australia: Australian Bureau of Statistics, 2013. 6 Moodie CS, Mackintosh AM. Young adult women smokers’ response to using plain What this paper adds cigarette packaging: a naturalistic approach. BMJ Open 2013;3:e002402. 7 Moodie C, Mackintosh AM, Hastings G, et al. Young adult smokers’ perceptions of plain packaging: a pilot naturalistic study. Tob Control 2011;20:367–73. ▸ Rates of personal tobacco pack display and active smoking 8 Moodie C, Stead M, Bauld L, et al. Plain Tobacco Packaging: A Systematic Review. Stirling, Scotland: University of Stirling, 2012.http://phrc.lshtm.ac.uk/project_ among patrons in outdoor areas of cafes, restaurants and 2011-2016_006.html bars remained lower 1 year after the implementation of 9 Stead M, Moodie C, Angus K, et al. Is consumer response to plain/standardised plain packaging in Australia compared to the summer before tobacco packaging consistent with framework convention on tobacco control implementation. guidelines? A systematic review of quantitative studies. PLoS ONE 2013; ▸ Though packs were less likely to be face-up and more likely 8:e75919. 10 Australian Institute of Health and Welfare. Tobacco smoking (NDSHS 2013 key to be concealed or in external cases immediately following findings). Bruce, ACT: Australian Institute of Health and Welfare, 2013. [23 Jul implementation, these orientations returned to baseline 2014]. http://www.aihw.gov.au/alcohol-and-other-drugs/ndshs/2013/tobacco/ levels 1 year post-implementation. 11 Hammond D. Health warning messages on tobacco products: A review. Tobacco Control 2011;20:327–37. 12 Dunlop SM, Cotter TF, Perez DA. Impact of the 2010 tobacco tax increase in Australia on short-term smoking cessation: a continuous tracking survey. Med J Aust 2011;195:469–72. Acknowledgements The authors thank staff of Cancer Council Victoria, Cancer 13 Dunlop SM, Perez D, Cotter T. Australian smokers’ and recent quitters’ responses to Council South Australia, and South Australian Health and Medical Research Institute the increasing price of cigarettes in the context of a tobacco tax increase. Addiction who undertook data collection. 2011;106:1687–95. Contributors MW and MS conceived the study; MZ, MB, EB and SD improved its 14 Patel V, Nowostawski M, Thomson G, et al. Developing a smartphone ‘app’ for design; MZ, MB, JD and CM coordinated aspects of data collection; MZ undertook public health research: the example of measuring observed smoking in vehicles. data analysis; MW and MZ drafted the manuscript; all authors revised the J Epidemiol Community Health 2013;67:446–52. manuscript and have read and approved the final version. 15 Thomson G, Russell M, Jenkin G, et al. Informing outdoor smokefree policy: methods for measuring the proportion of people smoking in outdoor public areas. Funding This study was funded by Cancer Council Victoria, Cancer Council South Health Place 2013;20:19–24. Australia, South Australian Health and Medical Research Institute, and Quit Victoria. 16 Pearson AL, Nutsford D, Thomson G. Measuring visual exposure to smoking Competing interests The authors wish to advise that MW was a member and behaviours: a viewshed analysis of smoking at outdoor bars and cafes across a MS a technical writer for the Tobacco Working Group of the Australian National capital city’s downtown area. BMC Public Health 2014;14:300. Preventive Health Task Force and MW was a member of the Expert Advisory 17 Alesci NL, Forster JL, Blaine T. Smoking visibility, perceived acceptability, and Committee on Plain Packaging that advised the Australian Department of Health on frequency in various locations among youth and adults. Prev Med 2003;36: research pertaining to the plain packaging legislation. MW, SD and EB hold 272–81. competitive grant funding from the Australian National Health and Medical Research 18 Erblich J, Montgomery GH. Cue-induced cigarette cravings and smoking cessation: Council, MW and CM hold such funding from Cancer Council South Australia and the role of expectancies. Nicotine Tob Res 2012;14:809–15. MW holds such funding from the US National Institutes of Health, Australian 19 Ferguson SG, Shiffman S. The relevance and treatment of cue-induced cravings in National Preventive Health Agency and BUPA Health Foundation. tobacco dependence. J Subst Abuse Treat 2009;36:235–43.

Zacher M, et al. Tob Control 2015;24:ii94–ii97. doi:10.1136/tobaccocontrol-2014-051826 ii97 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Personal pack display and active smoking at outdoor café strips: assessing the impact of plain packaging 1 year postimplementation Meghan Zacher, Megan Bayly, Emily Brennan, Joanne Dono, Caroline Miller, Sarah Durkin, Michelle Scollo and Melanie Wakefield

Tob Control 2015 24: ii94-ii97 doi: 10.1136/tobaccocontrol-2014-051826

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii94

These include: References This article cites 14 articles, 6 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii94#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research letter Did smokers shift from small implementation of the legislation.12 ‘Where did you buy or get the cigarettes However, studies of retail outlets conducted or tobacco you are currently smoking?’ mixed businesses to discount across Australia over the period of imple- For analyses, we coded store types into outlets following the mentation of legislation found no lasting seven categories: supermarkets, tobacco- 13 introduction of plain packaging effects on retail serving time and a study nists, small mixed businesses (including conducted in the Australian state of convenience stores, small grocery shops, in Australia? A national cross- Victoria 1 year before (2011) and 1 year milk bars, delis and newsagents/news- sectional survey after (2013) introduction detected no stands), petrol stations, the internet, duty changes among current smokers in usual free or from overseas and other (including Opponents of plain packaging (PP) in place of purchase of tobacco products.14 informal sellers, bars and pubs, bottle – – Australia1 4 and elsewhere5 8 have claimed We aimed to repeat and extend this analysis shops and liquor stores, vending machines that the legislation would create confusion using a large national data set. and ‘given as a gift’). for retailers attempting to locate packs of We used data from continuous national Data were weighted to the national uniform appearance. This, it was claimed, cross-sectional telephone surveys conducted smoker/recent quitter population by would result in increases in pack retrieval between 9 April 2012 and 30 March 2014 mobile phone status, gender, age by educa- and overall serving times,910impatience with about 100 respondents per week aged tion and state of residence as described in with queuing among customers, and a con- 18–69 years, described fully in Wakefield Wa k e field et al15 (this volume). We under- sequent shift in patronage, sales and profits et al15 (this volume). took logistic regression analyses to assess from small mixed businesses to large dis- All respondents were asked ‘Do you changes in purchase channel between count outlets such as supermarkets.10 11 currently smoke factory-made (FM) cigar- pre-PP (April–September 2012, n=2193), Surveys of retailers funded by tobacco ettes only, roll-your-own (RYO) cigarettes the transition to PP phase (October and industry groups reported perceived negative only, both, or neither of these?’ Cigarette November 2012, n=765) and the PP effects among retailers shortly after smokers (FM and RYO) were asked, phase (December 2012–March 2014,

Table 1 Place of purchase of cigarette smokers’ current packs–percentages and results of logistic regression models (n=7950) Differences between PP phases—unadjusted Differences between PP phases—adjusted Linear trend models† models†‡ post-PP—adjusted models Time period Per cent OR 95% CI Per cent OR 95% CI OR

Supermarket Pre-PP 57.9 1.00 58.1 1.00 Transition 55.9 0.92 0.76 to 1.11 55.9 0.91 0.75 to 1.11 PP 57.4 0.98 0.88 to 1.10 57.4 0.97 0.87 to 1.09 1.00 Tobacconist Pre-PP 13.6 1.00 13.8 1.00 Transition 13.8 1.01 0.78 to 1.32 13.7 1.00 0.76 to 1.30 PP 14.2 1.05 0.89 to 1.23 14.1 1.03 0.87 to 1.22 1.00 Small mixed business Pre-PP 13.7 1.00 13.6 1.00 Transition 15.7 1.17 0.90 to 1.54 15.6 1.17 0.90 to 1.54 PP 15.4 1.15 0.98 to 1.35 15.5 1.16 0.99 to 1.37 1.01 Petrol station Pre-PP 10.9 1.00 10.7 1.00 Transition 12.0 1.11 0.82 to 1.50 12.1 1.16 0.85 to 1.58 PP 10.2 0.93 0.77 to 1.12 10.3 0.96 0.79 to 1.16 0.99 Internet Pre-PP <0.1 1.00 –– Transition 0 –– ––– PP 0.1 8.29 0.97 to 71.07 –– – 0.92 Duty free or overseas Pre-PP 1.9 1.00 1.8 1.00 Transition 0.7 0.36* 0.14 to 0.90 0.7 0.38* 0.15 to 0.97 PP 0.6 0.30*** 0.18 to 0.51 0.6 0.32*** 0.19 to 0.53 0.94 Other source Pre-PP 2.0 1.00 2.0 1.00 Transition 2.0 1.04 0.55 to 1.98 2.1 1.05 0.55 to 2.00 PP 2.0 1.02 0.69 to 1.52 2.0 1.03 0.69 to 1.52 0.99 ***p≤0.001, **p<0.01, *p<0.05. †Of n=8679 cigarette smokers, those who did not provide their current brand name (n=283) or their current pack size (n=195) were not asked about their place of purchase. We also excluded those who did not know where their current cigarettes/tobacco was purchased (n=61), refused to provide the source or responded not applicable (n=6), mistakenly skipped the question due to coding of previous brand or pack size responses (n=11) or did not have SES information (n=173). Analysed n=7950 (pre-PP n=2079; transition n=707; PP n=5164). ‡Models controlled for sex, age group, SES and education, with the exception of the model predicting ‘Internet purchase’ which did not control for respondent characteristics due to low cell sizes resulting in collinearity issues. PP, plain packaging; SES, socioeconomic status. ii98 Tob Control March 2015 Vol 24 No S2 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research letter n=5721). We also tested whether there Acknowledgements The authors thank Ms Megan ‘evidence-based policy’. Melbourne, 2011. http:// were linear changes over the months Bayly for data checking and review. webarchive.nla.gov.au/gov/20130329051831/http:// www.yourhealth.gov.au/internet/yourhealth/ Contributors MS and MW designed this study. MZ during the PP phase, as unintended conse- publishing.nsf/Content/Phillip-Morris-Limited~Phillip- cleaned data files, undertook analysis and reported quences of PP could occur gradually. We Morris-Limited-2 (accessed Oct 2014). results. MS drafted the manuscript with contributions controlled for sex, age (18–29, 30–49 and 3 Imperial Tobacco Australia Limited. Submission to the from all authors. KC undertook further analysis. All – Department of Health and Ageing regarding the 50 69 years), area-based socioeconomic authors approved the final manuscript. status16 and education. tobacco plain packaging bill 2012 (exposure draft) and Funding The National Plain Packaging survey was consultation paper. Sydney, 2011. http://webarchive. As shown in table 1, odds of reporting funded under a contract with the Australian nla.gov.au/gov/20130329044623/http://www. purchase from major channels such as Government Department of Health and Ageing. yourhealth.gov.au/internet/yourhealth/publishing.nsf/ supermarkets, tobacconists, small mixed Competing interests The authors wish to advise Content/ppit~ncr~loe (accessed 10 Jan 2012). businesses and petrol stations did not that MS was a technical writer for and MW a member 4 Alliance of Australian Retailers. Submission on the of the Tobacco Working Group of the Australian exposure draft of tobacco plain packaging bill and change between the pre-PP and PP phases. House Standing Commitee on Health and Ageing The adjusted proportion whose latest pack National Preventive Health Task Force and MW was a member of the Expert Advisory Committee on Plain Inquiry. 2011. https://www.australianretailers.com.au/ was purchased duty free or overseas Packaging that advised the Australian Department of 5 British American Tobacco (BAT). UK Standardised declined from 1.8% pre-PP to 0.7% in the Health on research pertaining to the plain packaging Packaging Consultation: Response of British transition phase and 0.6% during PP. legislation. MW holds competitive grant funding from American Tobacco UK Limited. 2012. http://www. the Australian National Health and Medical Research bat.com/group/sites/uk__3mnfen.nsf/ Purchase of packs over the internet or from vwPagesWebLive/DO8WZC5E/$FILE/medMD8WZC6J. ‘other sources’ was negligible throughout Council, US National Institutes of Health, Australian National Preventive Health Agency and BUPA Health pdf?openelement (accessed May 2014). the survey. No gradual changes in place of Foundation. 6 Japan Tobacco International ( JTI). Response to the purchase were detected during PP (table 1). Department of Health’s consultation on the Ethics approval The survey was approved by the standardised packaging of tobacco products. 2012. Excise/customs duty on tobacco pro- Cancer Council Victoria Human Ethics Committee http://www.jti.com/how-we-do-business/ ducts in Australia increased by 12.5% on (HREC 0018). key-regulatory-submissions/ (accessed May 2014). 1 December 2013. The resulting increases Provenance and peer review Not commissioned; 7 Oireachtas Joint Committee on Health and Children. in prices may have prompted some consu- externally peer reviewed. Public health (standardised packaging of tobacco) mers to shift from convenience outlets bill presentations and submissions. Dublin: Houses of the Oireachtas, 2014. http://www.oireachtas.ie/ (where tobacco products tend to be sold parliament/oireachtasbusiness/committees_list/ at or close to recommended retail prices) health-and-children/submissionsandpresentations/ to outlets such as supermarkets and tobac- (accessed May 2014). conists (where tobacco products tend to 8 British American Tobacco (New Zealand) Ltd. 17 Proposal to introduce plain packaging of tobacco be more heavily discounted). When we products in New Zealand: submission by British excluded data from cases collected after American Tobacco (New Zealand) Limited. Open Access this tax increase, the percentage of Scan to access more Wellington: Parliament of New Zealand, 2014. http:// respondents who reported purchasing free content www.parliament.nz/en-nz/pb/sc/documents/evidence? from small mixed businesses increased sig- Custom=00dbhoh_bill12969_1&Criteria. fi Open Access This is an Open Access article PageNumber=2 (accessed May 2014). ni cantly between pre-PP (13.6%) and PP distributed in accordance with the Creative Commons 9 Deloitte. Potential impact on retailers from the year 1 (15.8%; adjusted OR=1.19, 95% Attribution Non Commercial (CC BY-NC 4.0) license, introduction of plain tobacco packaging. Sydney: CI 1.01 to 1.41, p=0.039) and increased which permits others to distribute, remix, adapt, build Alliance of Australian Retailers, 2011. https://www. linearly over the months during PP year 1 upon this work non-commercially, and license their australianretailers.com.au/research.php derivative works on different terms, provided the 10 Tobacco Retailers Alliance. Submission to Chantler (adjusted OR=1.03, 95% CI 1.01 to independent review or plain packaging. London: — original work is properly cited and the use is non- 1.06, p=0.038) results not in table. commercial. See: http://creativecommons.org/licenses/ Kings College, 2014. http://www.kcl.ac.uk/health/ This survey provides no evidence of a by-nc/4.0/ packaging-docs.aspx shift to overseas or duty-free purchase. This 11 Deloitte. Plain packaging and channel shift. Sydney: was to be expected given the reduction in Alliance of Australian Retailers, 2011. https://www. australianretailers.com.au/downloads/pdf/deloitte/ the limit for import of duty-free cigarettes Potential_impact_of_channel_shift.pdf —from 200 to 50 cigarettes per person— 12 Australian Association of Convenience Stores. Impact 18 — that came into force in September 2012. To cite Scollo M, Coomber K, Zacher M, et al. Tob of plain packaging on small retailers wave 2. The results of this large national study Control 2015;24:ii98–ii100. 2013. http://www.aacs.org.au/wp-content/uploads/ fi 2013/10/Impact-of-Plain-Packaging-on-Small- con rm those of our earlier study con- Received 15 December 2014 Retailers-W2-Final-Report.pdf ducted in Victoria which also found no Accepted 12 January 2015 13 Bayly M, Scollo M, Wakefield M. No lasting effects decline in percentages of smokers purchas- Tob Control 2015;24:ii98–ii100. of plain packaging on cigarette pack retrieval time in ing from convenience outlets following the doi:10.1136/tobaccocontrol-2014-052199 small Australian retail outlets. Tob Control 2015;24: – introduction of PP.14 Findings of our study e108 9. 14 Scollo M, Zacher M, Durkin S, et al. Early evidence corroborate sales data from international about the predicted unintended consequences of market research company, Euromonitor, REFERENCES standardised packaging of tobacco products in which suggest no major shifts in channel 1 British American Tobacco Australia. Submission on the Australia: a cross-sectional study of the place of of purchase between 2012 and 2013.19 tobacco plain packaging bill 2011.Canberra: purchase, regular brands and use of illicit tobacco. Department of Health and Ageing Consultation website, BMJ Open 2014;4:e005873. Michelle Scollo, Kerri Coomber, Meghan Zacher, 2011. http://content.webarchive.nla.gov.au/gov/wayback/ 15 Wakefield M, Coomber K, Zacher M, et al. Australian Melanie Wakefield 20130904173432/http://www.yourhealth.gov.au/ adult smokers’ responses to plain packaging with internet/yourhealth/publishing.nsf/Content/ larger graphic health warnings 1 year after Centre for Behavioural Research in Cancer, Cancer 2EE2F6F3EC74F628CA2579540005F68B/$File/British% implementation: results from a national cross-sectional Council Victoria, Melbourne, Victoria, Australia 20American%20Tobacco%20Australia%20-%20Public tracking survey. Tob Control 2015;24:ii17–25. Correspondence to Dr Michelle Scollo, Centre for %20Submission.pdf (accessed Oct 2014). 16 Australian Bureau of Statistics. Technical paper: Behavioural Research in Cancer, Cancer Council 2 Philip Morris Limited. Commoditising tobacco socio-economic indexes for areas (SEIFA) 2011. Victoria, 615 St Kilda Rd, Melbourne, VIC 3004, products through plain packaging will harm public Canberra, Australian Capital Territory: ABS, 2013. http:// Australia; [email protected] health, violate treaties and does not meet the test of www.abs.gov.au/websitedbs/censushome.nsf/home/

Tob Control March 2015 Vol 24 No S2 ii99 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research letter

seifa2011?opendocument&navpos=260 (accessed Apr Victoria, 2014. http://www.tobaccoinaustralia.org.au/ 19 Freeman B. Chapter 10. The tobacco industry in 2014). chapter-13-taxation/13-3-the-price-of-tobacco- Australian society, Section 10.3 Retail value and 17 Scollo M, Bayly M. Section 13.3 The price of tobacco products-in-australia. volume, Updated September 2014. Melbourne: products in Australia in Chapter 13. The pricing and 18 Customs By-Law No. 1228133, (2012). http://www. Cancer Council Victoria, 2013. http://www. taxation of tobacco products in Australia (December customs.gov.au/webdata/resources/files/ tobaccoinaustralia.org.au/chapter-10-tobacco-industry 2014 update). Melbourne, Australia: Cancer Council BL1228133web.pdf

ii100 Tob Control March 2015 Vol 24 No S2 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Did smokers shift from small mixed businesses to discount outlets following the introduction of plain packaging in Australia? A national cross-sectional survey Michelle Scollo, Kerri Coomber, Meghan Zacher and Melanie Wakefield

Tob Control 2015 24: ii98-ii100 doi: 10.1136/tobaccocontrol-2014-052199

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii98

These include: References This article cites 1 articles, 0 of which you can access for free at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii98#BIBL Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

ii32 Durkin S, et al. Tob Control 2015;24:ii26–ii32. doi:10.1136/tobaccocontrol-2014-052058 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Lighter Side

Tob Control 2015 24: ii32

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii32.citation

These include: Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/ Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com Research paper

Wakefield M, et al. Tob Control 2015;24:ii17–ii25. doi:10.1136/tobaccocontrol-2014-052050 ii25 Downloaded from http://tobaccocontrol.bmj.com/ on March 18, 2015 - Published by group.bmj.com

Lighter Side

Tob Control 2015 24: ii25

Updated information and services can be found at: http://tobaccocontrol.bmj.com/content/24/Suppl_2/ii25.citation

These include: Open Access 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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Open access (193)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/