Full implementation of the Framework Convention on Tobacco Control (FCTC) will result in a ‘new world order’ in tobacco control. Many countries are only just beginning implementation, and no country has implemented it in its entirety, but a few countries (Australia, Canada, Sweden and a few states in the United States) are beginning to look beyond this treaty. What might be the next steps beyond the FCTC strategies to reduce smoking prevalence?
Some strategies represent modest extensions of FCTC provisions. They pose few risks and would not be costly to implement. Arguably, there is sufficient evidence from similar policies in other contexts to warrant their implementation on a cost–benefit analysis. The policies we would put into this category are: retail display bans; plain packaging of cigarettes (or 95% health warnings on all surfaces); bans on smoking in cars; and mandatory screening of anti-smoking counter-advertising prior to movies containing smoking.
Retail display bans and mandatory plain packaging of tobacco products are simple extensions of advertising bans. It is accepted widely that cigarette packs provide brand imagery, convey promotional messages, are displayed to maximize imagery and positioned at eye level in the most visible position in shops [1,2]. Restricting packaging and placement are accepted by the public for other products, such as ‘prescription only’ medications .
Many jurisdictions ban distracting activities while driving, such as using a mobile phone. Smoking while driving has been linked to traffic collisions , and a good case can be made that lighting a cigarette, smoking it and disposing of it are unnecessary distractions that increase risk  and expose other car occupants to high levels of second-hand smoke .
Adolescents' exposure to movies containing positive smoking scenes has been found to be associated with greater likelihood of smoking initiation . Research also suggests that this effect can be reduced by screening anti-smoking counter advertising prior to these movies .
The net effect of these measures, in addition to those in the FCTC such as tax increases, is likely to be a slow but steady downward trend in smoking prevalence. However, more radical approaches need to be considered to achieve a faster impact on smoking prevalence and related harm. In this category we place: harm reduction with low nitrosamine smokeless tobacco (LNSLT) and/or high dose recreational clean nicotine products ; improving the regulatory structure ; restrictions on where tobacco can be sold and the number of outlets ; regulated market models, i.e. moving retail sales from the open market to a government controlled monopoly ; and smoker-licensing schemes that require smokers to obtain a licence to purchase a restricted quantity of tobacco always accompanied by advice as to its harmfulness .
Some of these strategies could potentially increase population nicotine use (e.g. harm reduction), and others may be costly and/or face major obstacles to their implementation (e.g. the regulated market model or smoker-licensing schemes), but they also have the potential to produce a much greater reduction in harm, particularly if they were combined with an explicit goal to remove combustible tobacco products from the market; for example, over a 10-year period .
Because of their greater risk and probable resistance to their use, stronger evidence would be required to support their introduction. We believe that the academic community needs to move quickly to define, agree upon and evaluate each of these strategies and design a comprehensive surveillance system to monitor their impacts if implemented.
As with the ‘low-risk’ options, evidence can be borrowed from other contexts to support many of these strategies. For example, some jurisdictions have a regulated alcohol market or restrictions on the types and density of outlets that are allowed to retail alcoholic beverages. Gun licences restrict who can purchase firearms and medical practitioner prescriptions restrict access to some medications . However, better evidence will be needed. These strategies do not lend themselves readily to double-blind randomized controlled trials, but a mix of quasi-experimental studies and small-scale trials should suffice.
For harm reduction, we will first need to test consumer understanding of differential risk statements on labelling or health warnings. We will then need to test market the most promising ones on relevant products, and assess the incentive effects of tax differentials for tobacco products based on their harmfulness to users. In countries where sale of LNSLT products are banned, trials could involve sale as ‘tobacconist only’ products in specialist tobacconists in limited areas. This would allow an assessment of the potential interest in these products among smokers and assessment of the prevalence of ‘dual use’ of LNSLT and cigarettes.
Reducing the nicotine content of cigarettes has also been proposed as a harm reduction measure which would gradually reduce the addictiveness of cigarettes . This is not a strategy we currently support, because it seems likely that only at the point where there is minimal nicotine in cigarettes would one see a benefit. Prior to that it is likely that exposure to tar would increase as smokers try to extract their usual nicotine boost, and taking the last step to minimal nicotine content would suffer the same problems as an outright ban on sale of smoked tobacco. Nevertheless, research should be carried out to replicate the only small-scale trial of this carried out to date , to confirm if this strategy would reduce nicotine addiction and demand at population level.
Developing a more sensible regulatory structure is desirable, but difficult, given that in most countries cigarettes are regulated by one body and NRT by another, usually medicines regulators. The introduction of cigarette-like nicotine delivery systems such as the e-cigarette provides an opportunity to develop a set of criteria (such as safety, nicotine delivery, attractiveness) against which its availability can be co-regulated alongside cigarettes.
Reducing the outlet density of tobacco retailers could be trialled in a jurisdiction with an existing tobacco retail licence scheme or registry. Options to test this could involve providing incentives for retailers to surrender their licences or to revoke licences for all outlets within a specified radius of a school . A wider trial could involve classification of all currently licensed retailers into different licence categories, e.g. major general retailers (supermarkets), minor general retailers (convenience stores and service stations) and specialist tobacconists, with each category of licence gradually phased out over a 3–5-year period.
Such experiments are essential if we are serious about an end to the smoking epidemic within the life-times of current adults. The most optimistic projections of future smoking prevalence indicate that we will not achieve this goal by simply doing more of the same [17,18].