Evidence and Tobacco

Action on Smoking and Health told us evidence for banning "powerwall" retail tobacco displays was overwhelming. They said legal challenges against such retail display bans in other jurisdictions failed "as the health benefits of eliminating advertising are overwhelming." Indeed, "[a] total ban of tobacco product displays, together with other measures, is required to achieve the goals and objectives set in the five-year plan of Ministry of Health for tobacco control." The evidence is overwhelming and a ban is necessary. Ok.

So it's fun to read Professor Richard Edwards's plenary address to the Oceania Tobacco Control Conference in Brisbane of last October.

[W]hen Janine Paynter and I carried out a systematic review in 2009 of the evidence on point of sale (PoS) displays, there were some obvious weaknesses and gaps. For example, all but one of the observational studies investigating links between exposure to PoS and children’s smoking were cross-sectional (and hence have limitations in determining direction of causality) and many were carried out in California, which had some important differences in the policy context limiting the application of the findings to other settings. Experimental studies are hampered in many ways e.g. by the difficulty of replicating ubiquitous and recurring exposures like PoS displays in an experimental design. Use of self-reports of the impact of PoS on smoking, purchase and quitting in a variety of study designs have some obvious possible limitations. There was no peer-reviewed evidence from the evaluation of PoS bans in jurisdictions like Canadian States, Iceland and Thailand. Even now there is only published evidence from one jurisdiction, Ireland, and then from a somewhat poorly resourced evaluation carried out after the event with funding cobbled together from sources mainly from outside of the country. 

So a total ban on retail displays was required, but there wasn't really any evidence for it. As for plain packaging, which is likely on the horizon for New Zealand:

Plain packs have not been implemented, so evidence of the probable population impact must come from experimental studies, focus groups, surveys and so on; rather than rigorous controlled studies of the impact of the actual intervention in the real setting, as would be the ideal.

We can learn from those kinds of studies; we just need to be a bit cautious in extrapolating to population-level effects from them. It would be surprising if evidence coming from those studies would be sufficient to say anything like "The evidence for plain packaging is overwhelming; it's necessary, along with other measures, to achieve objectives...". So keep an eye out for ASH's eventual submissions on eventual plain packaging proposals.

Edwards' whole address is worth reading. Some highlights:

• Comparing the 1996 and 2006 censuses, there's been a drop in smoking prevalence among 14-15 year olds but no change in uptake among young adults. Excise tax did rise substantially over the period. Edwards says the "evidence base" for successful interventions in the older cohort is weak, "and the ethical and moral framework to justify some interventions is less clear cut." We'll have to remember that when interventions are proposed.


• He cites increased "social smoking" among young adults - folks that might have a cigarette while out drinking, but otherwise don't smoke. I'd find it a bit surprising if that level of smoking resulted in substantial negative health effects. Says Edwards "The frequency of XS alcohol consumption, and its role in promoting uptake and maintenance of smoking and undermining quitting, suggests co-interventions may be needed and that we cannot tackle smoking in isolation." So anti-alcohol policy may be part of anti-tobacco policy...stay tuned.


• Edwards worries about FOREST's framing of tobacco control as an issue of freedom versus authoritarianism; he later suggests emphasizing "how tobacco control measures are pro-freedom by freeing smokers from an unwanted addiction, and by protecting our children from the risk of addiction and premature death." I can buy arguments about the conflicting freedoms of smokers and non-smokers in public spaces, but framing pure paternalism as being freedom-promoting is Orwellian. 


• Despite spending a fair bit of time talking about the importance of de-normalizing tobacco, he then says:

  We need to be very careful that interventions do not stigmatise smokers. Once again this involves keeping in close touch with how smokers are feeling through in-depth research. These quotes show how the experience of stigma among smokers and practice of stigmatising behaviours can be very real. This reduces support among smokes (and also among non-smokers) for tobacco control and the tobacco free vision, and may drive smokers together in a sort of Dunkirk spirit against the perceived assault from a marginalising society or harden the determination of smokers to smoke, as encapsulated in this last quote. We should be anti-smoking, but never anti-smoker.

• He wonders whether achieving tobacco-free goals given broader societal failings:

  Can we achieve tobacco free goals among Pacific, among Māori, among aboriginal communities without wealth redistribution, without equalisation of power? The answer is uncertain, but it is certainly a strong argument that we in tobacco control need also to be at the forefront of campaigns for wider social and political justice. The additional advantage is of course that addressing the broader social and economic inequalities will also benefit a whole range of other public health issues such as excessive alcohol use, accidents and obesity and the related morbidity and mortality.

  I'm reminded of Edwards' colleagues' argument for massive income redistribution in order to reduce disparities in life expectancy. But I love that he wants explicitly to link large-scale intervention in your private life in anti-tobacco policy with large-scale intervention in the economy and social justice. It helps us remember that the pro-freedom or pro-intervention framing is the right one.