October 9, 2017
Stanton A. Glantz, PhD
Assuming ecigs will cut smoking does not prove that ecigs will save lives; new paper an exercise in tautology
The paper, “Potential deaths averted in USA by replacing cigarettes with e-cigarettes” by David Levy et al. published in Tobacco Control on October 2, 2017, attracted a moderate amount of attention with its conclusion that “Compared with the Status Quo, replacement of cigarette by e-cigarette use over a 10-year period yields 6.6 million fewer premature deaths with 86.7 million fewer life years lost in the Optimistic Scenario. … Our projections show that a strategy of replacing cigarette smoking with vaping would yield substantial life year gains, …”
This is a pretty impressive result until you consider that the Optimistic Scenario is based on a series of assumptions that are of which are inconsistent with empirical evidence to date:
- Cigarette smoking prevalence drops from 17% to 5% in 10 years (from 19.3% to 4.6% in men and from $14.1% to 4.6% in women between 2016 and 2026).
- The existence of e-cigarettes does not, on average, depress quitting cigarettes.
- There is no relapse from e-cigarette use to cigarette smoking.
- No youth who initiate with e-cigarettes progress to cigarette use.
- No dual use of cigarettes and e-cigarettes.
- The evidence free claim that e-cigarettes are 5% as dangerous as cigarettes.
(These assumptions were not clearly stated in the main paper; we figured them out based on the appendix and by examining the Excel spreadsheet of the model that the authors shared with us in response to an email. Their Pessimessitic Scenario is based on similar assumptions, albeit with a higher e-cig risk.)
It is important to recognize that the authors don’t even pretend that their estimates are based on reality. In their Discussion they state: “Unlike previous models of e-cigarette use, our model was not developed to predict future e-cigarette and cigarette use based on past trends. Rather the aim was to examine a hypothetical endgame strategy of reducing cigarette use through switching to e-cigarettes. ”
I used their model to investigate the lives saved if cigarette use dropped along the same trajectory that the authors assumed because of increased smokefree laws, anti-smoking media, higher taxes and achieving smokefree movies without anyone even using an e-cigarette. Doing so (without e-cigarettes) would have saved even more lives – 7.0 million fewer lives lost and 91.4 million fewer life years lost.
The centrality of the first assumption is demonstrated by the fact that, using the Levy et al model (which they provided me as a spreadsheet) requires that e-cigarettes be more dangerous than cigarettes (about 120-130% more dangerous) to end up having a net negative population health impact. This is because the assume huge drops in smoking cigarettes.
In other words, all that this paper shows is that if a lot fewer people smoke, there will be a lot fewer premature deaths from smoking. It contributes absolutely nothing to the discussion of the efficacy of e-cigarettes for harm reduction.
It would be like someone asking what would happen to melting ice sheets if we were to be able to bring greenhouse gasses back to pre-industrial levels in 10 years. The calculation is possible but the results would not contribute anything to public policy because, like the Levy paper, the assumptions are discordant with reality.
Finally, in the Conclusion of their paper Levy et al state that “tobacco industry documents reveal an industry strategy of ‘divide and conquer’ focused primarily on fostering divisions within the tobacco control community regarding modified risk products. Indeed, the tobacco control community has had divided approaches to e-cigarettes, and in the process may have lost focus on cigarettes, the most deadly form of nicotine delivery ... Most important, an e-cigarette substitution strategy provides the justification to redouble efforts to target cigarette use, as called for by the WHO Framework Convention for Tobacco Control.” The authors are, of course, welcome to their opinion. But nothing in this paper provides an empirical support for this argument.
Comments
More problems with Levy model
Posted on behalf of Matthew Peters from <a class="js-about-module-title module__title__link" href="https://en.wikipedia.org/wiki/Macquarie_University" title="More at Wikipedia ";Macquarie University in Sydney:
The data, interpretations and implications of the data modelling exercise conducted by Levy et al#_edn1" name="_ednref1" title="";[1] should not go unchallenged. Regardless of the number and confidence of the opinions voiced, and the observation of lower levels of selected toxicants in e-cigarette users that are alluded to, there is great uncertainty about the extent to which harm might be reduced by the exclusive use of electronic cigarette rather than combustible tobacco. On this background, to describe one of two models, a 95% harm reduction as optimistic and the second, still a substantial, hopeful estimate of 60% reduction as pessimistic betrays a bias at the outset. The use of this “pessimistic” descriptor would to a casual reader imply that the truth lay, inevitably, somewhere between the two estimates.
Then there is the detail of the model. Firstly, the use of Holford projections#_edn2" name="_ednref2" title="";[2] overestimated 2015 smoking rates in the US by at least 10% compared to CDC data#_edn3" name="_ednref3" title="";[3] - underestimating the recent rate of decline in smoking prevalence in men and women between 2005 and 2015 by one-third. A higher base rate and slower rate of decline exaggerate tobacco-related harms in the status quo – naturally favouring each of the modelled scenarios. Starting with lower, more accurate estimates of current smoking and rates of decline would also increase the counterbalancing harms from initiation in non-smokers.
There are other obvious problems. In the status quo, 20% of boys and 14.8% of girls would commence smoking between ages 15 and 25. To achieve the effects modelled, 83% of the 15 year olds boys who would have commenced smoking and 70% of the pre-destined female smokers would have to be diverted to exclusive e-cigarette use. In the optimistic target, requiring the most complete diversion, there is no allowance for harms of “off target” initiation of e-cigarette use i.e. initiation in a young person otherwise destined not to smoke. This is irrational given that e-cigarette use in young people trebles smoking rates after allowance for predictors of smoking uptake#_edn4" name="_ednref4" title="";[4]. The harder e-cigarette use is pushed the more off-target use is likely unless magic is in operation. This detail is critical. Under the published “pessimistic” model for 15 year old girls, there is already a negative effect in terms of years of life lost; not a benefit.
It goes on. Ten years after the commencement of the model in 15 year olds, no individual appears to have ceased the use of e-cigarettes. In the real world, discontinuation is the norm, rather than the exception. Even in England, with ready availability and a favourable public discourse, a minority of smokers choose e-cigarettes to aid smoking cessation#_edn5" name="_ednref5" title="";[5]. Assumed smoking cessation rates with e-cigarettes are greatly in excess of what is achieved in clinical trials and standard clinical practice. How an intervention that is rejected longer-term in most smokers and ineffective in short-term use will achieve the 5%/10% smoking rates as modelled is unclear but it does not affect the calculations in the model as under both scenarios the outcome in smoking rates is predetermined. The viability of the steps needed to achieve this outcome are not questioned.
It is not difficult to see where this is heading. The purported benefits could only ever be achieved if electronic cigarettes or other novel products were freely available, including to adolescents at or before the typical age of initiation of combustible tobacco use, and had low pricing/taxation amidst a generally liberal approach to their use in public places. This would need to be supported by strong and effective marketing, attractive packaging and advertising standards that permitted uncertain/unproven claims about long-term safety and health benefits. Products would need to be researched and designed that were more rewarding/addictive than combustible tobacco to prevent switch and interchange with smoking. One wonders what responsible corporate entity would develop such a product, potentially harming many of its consumers whilst removing their easy and free choice to cease its use.
References:
#_ednref1" name="_edn1" title="";[1] Levy DT, Borland R, Lindblom EN, et al. Potential deaths averted in USA by replacing cigarettes with e-cigarettes. Tobacco Control (online first; doi: 10.1136/tobaccocontrol-2017-053759)
#_ednref2" name="_edn2" title="";[2] Holford TR, Levy DT, McKay LA, et al. Patterns of birth cohort-specific smoking histories, 1965-2009. Am J Prev Med 2014;46:e31–7.
#_ednref3" name="_edn3" title="";[3] Jamal A, King BA, Neff LJ, Whitmill J, Babb SD, Graffunder CM. Current Cigarette Smoking Among Adults — United States, 2005–2015. MMWR Morb Mortal Wkly Rep 2016;65:1205–1211.
#_ednref4" name="_edn4" title="";[4] Soneji S, Barrington-Trimis JL, Wills TA et al. Association Between Initial Use of e-Cigarettes and Subsequent Cigarette Smoking Among Adolescents and Young Adults. A Systematic Review and Meta-analysis. JAMA Pediatr 2017; 171:788-797.
#_ednref5" name="_edn5" title="";[5] West R, Brown J. Latest trends on smoking in England from the Smoking Toolkit Study. http://www.smokinginengland.info/latest-statistics/";http://www.smokingi... (Accessed Oct 13 2017)
Add new comment