Complying with the framework convention for tobacco control: an application of the Abridged SimSmoke model to Israel

To reduce non-communicable deaths, the World Health Organization (WHO) set a voluntary target of a 30 % reduction in smoking rates by 2025 as part of its global Non-Communicable Disease agenda [1]. In addition, the WHO launched a policy package that focuses on selected demand side measures for tobacco products, MPOWER, in 2008 [2]. This package recommends the following policies: Monitor tobacco use and prevention policies, Protect people from tobacco smoke, Offer help to quit tobacco use, Warn about the dangers of tobacco, Enforce bans on tobacco advertising, promotion and sponsorship, and Raise taxes on tobacco. The magnitude of the effect of each MPOWER policy on smoking prevalence varies and depends on the policies implemented, how they are implemented, and the country-level policies that were previously in effect [3].

The Simsmoke Tobacco Control Policy simulation model has been used to estimate the impact of past policies and potential impact of implementing stronger MPOWER policies on smoking prevalence and smoking-attributable deaths among current smokers within a country or state [4‐10]. This model projects the smoking rates over time with the birth and deaths, and the initiation and cessation of smokers. The model predicts well by age and gender for countries that have and have not implemented many strong policies [5‐11]. In a previous application [10], we developed a simplified form of SimSmoke, Abridged SimSmoke, to evaluate the results of implementing past MPOWER-required policies on country-level reductions in smoking-related deaths. A later paper [12] extended Abridged SimSmoke to show the effect of implementing future policies in accordance with MPOWER requirements. The model calculates the impact of the policies on current smokers only, and does not address those who have not yet begun smoking.

The data requirements for Abridged SimSmoke are less than for the original SimSmoke and parallel the smoking prevalence and policy data collected for the biennial WHO MPOWER/WHO Report on the Global Tobacco Epidemic [1]. Like SimSmoke, Abridged SimSmoke projects changes in smoking prevalence and smoking-attributable deaths resulting from the implementation of MPOWER policies. We apply that model to Israel.

Israel ratified the FCTC in August 2005 (apps.who.int/fctc/implementation/database/parties/Israel) [13]. Previously, Israel had implemented multiple tobacco policies including taxation (beginning in 1952 and since expanded), smoke-free air policies (beginning in 1982 and since expanded), restrictions on advertising (beginning in 1983 and since expanded), and health warnings (beginning in 1983 and since expanded). Since signing the FCTC, Israel has legislated and implemented stronger smoke-free air laws and began to subsidize cessation treatment through the National Basket of Services. However, despite the passage of the National Plan for Reduction of Smoking and its Harms in 2011, which was based in part on the FCTC and MPOWER, Israel does not yet meet the complete MPOWER requirements for any of the six recommended MPOWER policies [13, 14]. For example, smoke-free regulations are not well enforced [13, 15‐20] he planned comprehensive ban on advertising, promotion, and marketing did not pass into legislation, and there is no targeted budget for media campaigns or other tobacco control activities [13].

Abridged SimSmoke uses data from a single year to project short-term (5 years) and long-term (40 years) effects on smoking prevalence. It is similar to the original SimSmoke in that it relies on population size smoking prevalence and policy modules to make predictions. Abridged SimSmoke also uses formulas similar to SimSmoke to show a reduction in smoking prevalence in each policy module. Therefore, Abridged SimSmoke predicts the effect of individual and combined policies on smoking prevalence and the number of smokers, which is used to project the number of smoking-attributable deaths among smokers alive in 2014.

The results are presented in Table 2. The table first presents the initial levels for smoking prevalence (by gender) and the total number of smokers. Based on the current level of smoking (24.5 % among males and 13.2 % among females), the number of smokers is 1.1 million in 2014. Based on numbers of smokers, the number of smoking-attributable deaths is projected as 550 thousand (350 thousand male and 200 thousand female) as a lower estimate and 710 thousand (455 thousand male and 255 thousand female) as an upper estimate of the smokers alive in 2014.

The table also shows the effects of each policy individually and in combination. The effects are projected as short and long-term percentage reductions in smoking prevalence and the long-term effects on the number of smokers and smoking-attributable deaths.

With the stronger set of policies consistent with FCTC recommendation as specified in MPOWER, the model projects that smoking prevalence can be reduced by 23 % within 5 years increasing to 34 % within 40 years. As a result, between 187 thousand (120 thousand male and 67 thousand female) and 240 thousand (157 thousand male and 88 thousand female) premature deaths of current smokers alive in 2015 are projected to be averted.

Among smokers alive in 2014 in Israel, between 550 and 710 thousand premature deaths are predicted. This demonstrates the urgent need for strong policies to reduce tobacco use. Abridged Simsmoke provides estimates of reductions in smoking prevalence, and the resultant number of smoking-attributable deaths which would be avoided, if MPOWER policies were fully implemented [12]. With a complete implementation of policies, Israel is predicted to reach the goal of reducing smoking rates by 22 % in the next 5 years and by 34 % in forty years. As a consequence, between 187 thousand and 243 thousand premature deaths will be averted. The model indicates that increased cigarette taxes, stronger and better enforced smoke-free air laws, and media campaigns and marketing restrictions can play an important role. To a lesser extent, bolder and graphic health warnings and more comprehensive provision of cessation treatments will also contribute to decreasing the burden of tobacco-related mortality.

Abridged SimSmoke has been developed based on an extensively validated simulation model and provides support for our estimates. To explicitly consider the predictions of the abridged model, we have compared predictions from Abridged SimSmoke for nine countries that have reached MPOWER goals, including Finland and Italy, to results from the complete SimSmoke model for or those countries and found that the results predicted by Abridged SimSmoke for smoking prevalence and deaths were reasonably close to the reported findings from the complete model [10]. Nevertheless, the findings from Abridged SimSmoke should be interpreted in light of the model’s limitations.

Abridged SimSmoke does not incorporate future changes in demographics or smoking prevalence that may reflect the effect of previously implemented policies. The abridged model utilizes smoking prevalence data from 2014 as the basis for estimates of the number of smokers and expected deaths due to smoking among current smokers. The model only attempts to incorporate the effects of stronger future policies on those who were smokers in that survey year. It does not incorporate changes in smoking trends, including those that may be influenced by changes in policies implemented prior to that year, or by smoking cessation which may occur spontaneously over time, as indicated by patterns of declining smoking prevalence among older adults [21].

Difference in smoking patterns between the Jewish and Arab population (Age adjusted estimates: Jewish males: 23.1 %, Jewish females: 14.0 %, Arab males: 46.6.%, Arab females: 6.1 %) [21] are not addressed, due to a lack of sufficient information to distinguish differences in relative risks and differences in policy effects among these populations. Further research is warranted.

The model does not incorporate the effects of the unborn, youth, and young adults who will initiate smoking in future years (in the absence of strong policies). Therefore, failure to incorporate cessation of current smokers may be offset by the benefits of newly implemented policies on reducing future initiation. While not considered in our analysis, substantial additional gains can be expected through the effects of policies on the young. In particular, SimSmoke attributes more than twice the effect of tax increases on those under the age of 24 as for those ages 25 and above [31, 32]. Further, advertising restrictions are expected to have a 50 % greater effect on those below age 24 [3, 33]. The implementation and enforcement of bans on all smoking in the workplace can also be expected to have an important impact on those newly entering the workforce. These estimates would imply a 25–35 % overall reduction in smoking initiation. We estimate between 550,000 and 710,000 deaths due to smoking. Ginsburg and Geva [34, 35] estimated the number of smoking-attributable deaths at 7,400 per year, implying about 375,000 deaths over 50 years. However, there are important differences between the two methods (e.g., Ginsburg and Geva used deaths by cause vs our use of total mortality, and our upper level estimates are based on higher levels of risks found in the more recent studies [23, 24]). Further study is warranted to examine the composition of deaths by cause and the number of deaths over time by age and gender. In addition, we focused on mortality, but additional costs are associated with morbidity and productivity loss due to premature death. In addition to the 7000 and 7850 smoking-attributable deaths each year, Ginsburg and Geva [34, 35] estimated 319,231 active and 356,601 total smoking attributable hospital days and 1678 million NIS ($482 million) in total health service costs, 0.2 % of GNP. With productivity losses, smoke related costs overall were 3,587 million NIS ($1,030 million), 0.41 % of GNP. We did not include deaths due to second hand-smoke in the model. Globally, secondhand smoke is responsible for 1 % of total mortality and 0.7 % of the worldwide disease burden [36]. We also did not address the economic costs or pain and suffering, caused by tobacco smoke exposure, such as increases in rates of child respiratory disease, low birthweight, sudden infant death syndrome, and pre-term birth [37]. A study of parentally-reported exposure of infants aged 0–2 between 2009–2012 found that 31.5 % (Jewish infants: 24.8 %, Arab infants: 52.0) of infants were at least occasionally exposed to tobacco smoke [21]. A reduction in smoking rates would also decrease exposure of children to smoking.

Another limitation is that the effect of policies is subject to uncertainty. We have provided ranges of the effect sizes in Table 1. In addition, the model only applies to cigarettes and thus does not incorporate other tobacco products such as pipes, cigars, water pipes, and smokeless tobacco. If tax increases and other policies target cigarette smokers exclusively, there may be a substitution toward greater usage of other tobacco products. By directing policies (e.g., media campaigns) at all tobacco products, some of this substitution may be avoided. E-cigarettes potentially provide a lower harm alternative to smoking and therefore may either accelerate or attenuate reductions in net population smoking prevalence depending on how they influence smoking patterns (e.g, initiation and cessation rates) and related policies and practices [38].

Abridged SimSmoke has been developed to use data from the biennial WHO MPOWER Reports. The MPOWER policy data are restricted to a specific set of policies and definitions. The model does not consider policies directed at price minimizing behavior [39], enforcement against smuggling [40], product regulation (e.g., nicotine reduction) [41, 42], plain packaging [43, 44], harm reduction policies (i.e., encouraging substitution of less harmful alternatives to cigarettes) [45, 46] and youth access policies [47]. These other policies have also been shown to be effective and are likely to be needed for Israel to approach the endgame for smoking cigarettes.

We did not consider the costs of the interventions relative to their effectiveness. Studies of specific policies have generally found that tobacco control policies are cost effective [48‐50]. Taxing policy and health warnings, both of which involve minimal costs, are particularly cost effective. Further study is warranted on the costs of implementation and the cost savings that would specifically be incurred in Israel.

Abridged SimSmoke shows that the required MPOWER tobacco control policies will save lives and eventually decrease tobacco use. The model enables the user to consider policies individually and in combination to observe how policies in different combinations lead to differing rates of reductions in smoking prevalence and smoking-attributable deaths. It also translates empirical information into a user-friendly format that can be easily interpreted. This is particularly important given difficulties inherent in tobacco-control policy making, which may include local and global tobacco industry interference [51, 52], lack of public knowledge of the harms of tobacco smoke exposure [12] and underestimation by policy makers regarding public support for smoke-free public places [29, 53].

Abridged SimSmoke for Israel shows an urgent need for strengthening comprehensive and proven effective policies and practices to combat and ultimately eliminate the wholly preventable deaths and chronic disease burdens and disabilities caused by tobacco use, primarily driven by the smoking of cigarettes [54, 55].