Sweden SimSmoke: the effect of tobacco control policies on smoking and snus prevalence and attributable deaths

Abstract

Background: This study examines the effect of past tobacco control policies and projects the effect of future policies on smoking and snus use prevalence and associated premature mortality in Sweden. Methods: The established SimSmoke model was adapted with population, smoking rates and tobacco control policy data from Sweden. SimSmoke evaluates the effect of taxes, smoke-free air, mass media, marketing bans, warning labels, cessation treatment and youth access policies on smoking and snus prevalence and the number of deaths attributable to smoking and snus use by gender from 2010 to 2040. Results: Sweden SimSmoke estimates that significant inroads to reducing smoking and snus prevalence and premature mortality can be achieved through tax increases, especially when combined with other policies. Smoking prevalence can be decreased by as much as 26% in the first few years, reaching a 37% reduction within 30 years. Without effective tobacco control policies, almost 54 500 lives will be lost in Sweden due to tobacco use by the year 2040. Conclusion: Besides presenting the benefits of a comprehensive tobacco control strategy, the model identifies gaps in surveillance and evaluation that can help better focus tobacco control policy in Sweden.

Introduction

With more than 5 million deaths each year attributable to smoking,¹ the World Health Organization (WHO) has set out the Framework Convention for Tobacco Control (FCTC) and the MPOWER report² has defined a set of policies that are consistent with the FCTC. MPOWER suggests that each nation impose taxes on cigarettes that constitute at least 70% of the retail price; require large, bold and graphic health warnings; provide broad access to cessation treatments; conduct a well-funded tobacco control campaign and implement and enforce comprehensive smoke-free indoor air laws and advertising/marketing restrictions. Sweden was one of the first countries to sign (2003) and ratify (2005) the FCTC.

Sweden has had a long history of tobacco control dating back to 1975, when advertising restrictions and other policies were first implemented. Since 1994, Sweden has increased taxes on cigarettes, implemented smoke-free air policies, increased access to cessation treatments, restricted marketing and strengthened health warnings. Sweden is also the only European nation besides Norway to allow commercialization of snus, a smokeless tobacco product that has been legally sold since 1976. Most tobacco control laws and programmes apply to snus as well as cigarettes. While some have claimed that snus use has been responsible for a large reduction in cigarette use and lung cancer deaths,^3–5 others claim that these reductions could have occurred with stricter tobacco control policies.^6,7 With smokeless tobacco use increasing in many countries,^8–10 it is important to consider the potential role of policies in affecting that use.

To examine the potential role of policies in furthering the aims of tobacco control, this study uses a modified version of the SimSmoke tobacco control policy simulation model (Sweden SimSmoke). SimSmoke simultaneously considers a broad array of public policies and has been validated for many countries.^11–17 While several models have examined the hypothetical effect of allowing smokeless tobacco use,^18,19 no study has considered actual smokeless tobacco use along with cigarettes. Sweden SimSmoke applies data from Sweden, and is used here to examine the effect of implementing FCTC-consistent policies on the prevalence of and deaths attributable to smoking and snus use in Sweden.

Methods

SimSmoke includes population, smoking, tobacco-attributable deaths and policy modules.^11,13–17 The model has been extended to distinguish users of cigarettes only, snus only and combined (‘dual’) cigarette and snus users. SimSmoke begins in a baseline year with the population divided into current, former and never smokers and snus users by age and gender. Sweden SimSmoke begins in 2004 based on the availability of data and stability of policies for that period. A discrete time, first-order Markov process is assumed to project population growth and tobacco use rates from the base year to future years. Population growth evolves through births and deaths, and smoking and snus rates evolve through smoking initiation, cessation and relapse. Smoking rates may shift due to changes in tobacco control policies. Smoking-attributable deaths are calculated as the excess mortality risk of current and former tobacco users relative to never users. Data sources are summarized in table 1.

Table 1.

Data used in Sweden SimSmoke

Input	Source	Specifications
I. Population
A. Population	Statistics Sweden website (http://pxweb2.stat.fi/Dialog/varval.asp?)	Breakdowns by age and gender groups
B. Fertility rates	Statistics Sweden website (http://pxweb2.stat.fi/Dialog/varval.asp?)	Breakdowns by age and gender groups
C. Mortality rates	Statistics Sweden website (http://pxweb2.stat.fi/Dialog/varval.asp?)	Breakdowns by age and gender groups
II. Smoking and attributable deaths
A. Baseline smoking rates	2004 Health on Equal Terms Survey	Breakdown of current, former and never smokers by age and gender groups
B. Initiation rates	Change in smoking rates between contiguous age groups	Breakdowns by age and gender groups
C. First-year cessation rates	2004 Health on Equal Terms Survey	Breakdowns by age and gender groups
D. Relapse rates	USDHHS (1989) and other studies	Breakdowns by age
E. Relative risks of current and ex-smokers	Cancer Prevention Study II (NCI 1997)	Breakdowns by age and gender
III. Policies
A. Taxes	Statistics Sweden (http://www.scb.se)	Prices and taxes
B. Smoke-free air laws	WHO website and tobacco control staff in Sweden	Types of laws (worksite, restaurant and other places) and their stringency
C. Media and other educational campaigns	WHO website and tobacco control staff in Sweden	Classification based on expenditures per capita and audience
D. Marketing bans	WHO website and tobacco control staff in Sweden	Extent of bans
E. Warning labels	WHO website and tobacco control staff	Strictness of labels
F. Cessation treatment policies	WHO website and tobacco control staff	Financial reimbursement, quitlines and brief interventions
G. Youth access	WHO website and tobacco control staff	Enforcement checks, penalties, publicity, self-service and vending machine bans

Population and smoking data

SimSmoke was adapted with gender and age-specific population data from Statistics Sweden.²⁰

Smoking and snus sole and dual use prevalence data for 2004 through 2010 were available from the Health on Equal Terms of the National Public Health Survey. Smoking and snus prevalence was based on participant self-report as never, former or current tobacco use. Based on the percent of smokers who quit in the past year, we apply a cessation rate of 5% for cigarette sole and dual use, and 3% for snus only. Because that rate does not incorporate relapse, we apply a 50% relapse rate, as consistent with previous studies.^21,22 Data on relapse was not available for Sweden, so we use US relapse rates for cigarette smokers^23,24 and assume those same rates for single and dual cigarette and snus use.

Initiation rates at each age are measured as the difference between the smoking rate at that age year and the rate at the previous age year. We allowed initiation through age 30 for both genders, as snus initiation and switching occur until these ages.

Smoking and attributable deaths

Because smoking history and the standard of living in Sweden are similar to the United States, we use relative risk estimates from the US Cancer Prevention Study II.^25,26 For ex-smokers, we allow relative risks to decline at the rate observed in US studies.²⁵ Based on a literature review and the advice of an expert panel,²⁷ mortality relative risks for snus users are set at 1.1 for ages 35–49 and at 1.05 for ages ≥50. We assume the same risks for dual users as for cigarette only users. The relative risks of snus sole and dual use are assumed to decline with years quit at the same rate as for smokers.

Tobacco control policies

The policy parameters in SimSmoke are based on thorough reviews of the literature coupled with the advice of an expert panel. Policy effect sizes are applied as percent reductions to the smoking prevalence in the year in which the policy is implemented and are applied to initiation and cessation rates in future years if the policy is sustained. Table 2 summarizes policies and potential effect sizes in Sweden. The effect of a policy depends on its current level that is based on information in the MPOWER report² with corroboration from Swedish tobacco control correspondents.

Table 2.

Policies, description and effect sizes of the SimSmoke model and policies in Sweden

Policy	Description	Potential percentage effecta,b	Policies in Sweden
Tax policy
Tax policy	Cigarette price index, taxes measured in absolute terms	Through price elasticity: −0.3 ages 15–24 −0.2 ages 25–34 −0.1 ages 35 and above	Immediate change in taxes and price by raising excise taxes to 70% of retail price. Price effects are assumed to be the same as in the US model (28,29). Elasticities are half as great for snus as for cigarettes and assume no cross price effects. The effects of changing both the snus and cigarette price are applied for dual users. A consumer price index (CPI), cigarette price index (1985–2009) and snus price index (1985–2009) were obtained for Sweden. Seventy-three percent of cigarette price is taxes, of which 52% is specific taxes (39.2% ad valorem and 12.6% specific). The excise tax rate on snus in 2009 was 38%, having increased from 25% in 2006
Smoke-free air laws
Worksite total ban	Ban in all areas	6% with variations by age and gender	Sweden is considered to have had mid-level worksite laws, but strong laws in other public places since 1993. In 2006, smoke-free air laws were extended to bars and restaurants (set at 50% for both), with an exclusion for ventilated areas. Enforcement level is set to 8 out of 10 for 200830
Worksite ban, except ventilated areas	Smoking restricted to ventilated areas in all indoor workplaces	4% with variations by age and gender
Restaurant total ban	Ban in all indoor restaurants in all areas	1% effect
Restaurant ban, except separate areas	Ban in all restaurants, except in designated areas	0.5% effect
Other places total ban	Ban in 3 of 4 (malls, retail stores, public transportation and elevators)	1% effect
Enforcement and publicity	Government agency is designated to enforce and publicize the laws	Effects reduced by as much as 50% if 0 enforcement
Mass media campaigns
Highly publicized campaign	Campaign publicized heavily on TV (at least 2 months of the year) and at least some other media	3.25% effect (doubled when accompanied by local programmes)	Sweden has been spending about US $0.50 per person since before 2004; therefore, a medium-intensity campaign is assigned. The same campaign level is applied to snus, with the exception that a high-intensity campaign was implemented in 2009
Moderately publicized campaign	Campaign publicized sporadically on TV and in at least some other media, and a local programme	1.8% effect (doubled when accompanied by local programmes)
Marketing bans
Comprehensive marketing ban	Ban is applied to television, radio, print, billboard, in-store displays, sponsorships and free samples	5% reduction in prevalence, 6% reduction in initiation and 3% increase in cessation rates	A ban on most types of direct and indirect advertising has been in place since 2003 in Sweden, with the sale of cigarettes in packages of less than 20 prohibited in 2006. Therefore, we categorize Sweden as having a complete advertising ban until 2003, and partial ban on marketing (50% total advertising and 50% marketing) until 2006, increasing to 75% marketing and 25% advertising in 2007. Enforcement is set at 8 for all years30
Total advertising ban	Ban is applied to all media television, radio, print and billboard	3% reduction in prevalence, 4% reduction in initiation and 2% increase in cessation rates
Enforcement and publicity	Government agency is designated to enforce the laws	Effects reduced by as much as 50% if 0 enforcement
Health warnings
Strong	Labels are large, bold and graphic. Covers at least 50% of the display area and includes all seven MPOWER warning criteria	1% reduction in prevalence and initiation and 5% increase in cessation rate	Sweden has had moderate health warnings since 200431
Moderate	Labels cover one-third of package, not bold or graphic	0.75% reduction in prevalence, 0.5% reduction in initiation rates and 2% increase in cessation rates
Cessation treatment policy
Cessation treatment policy	Complete availability and reimbursement of pharmacotherapy, quitlines and brief interventions	4.75% reduction in prevalence, 39% increase in first-year cessation rate	Nicotine replacement treatment has been available in general stores in Sweden since 2004; yet before 2006, it was only sold by prescription in pharmacies.32 Bupropion has been provided with a prescription since 1996. Financial coverage of treatments has been provided in most places since 2008 for primary care facilities, hospitals and offices of health professionals and in some community centres. Sweden has had a national, active quitline since 1998
Youth access restrictions
Strongly enforced and publicized	Compliance checks are conducted regularly, penalties are heavy, and with publicity is strong, vending machine and self-service bans	30% reduction for age <16 in prevalence and initiation only, 20% reduction for ages 16–17 in prevalence and initiation only	Youth access policy incorporates enforcement, publicity and self-service and vending machine bans. In Sweden, youth access is considered to be enforced at a medium level with no bans on vending machine or self-service displays
Well enforced	Compliance checks are conducted sporadically, penalties are potent and little publicity	15% reduction for age <16 in prevalence and initiation only, 10% reduction for ages 16–17 in prevalence and initiation only

a: Unless otherwise specified, the same percentage effect is applied as a percentage reduction in the prevalence and initiation rate and a percentage increase in the cessation rate, and is applied to all ages and both genders. The effect sizes are shown relative to the absence of any policy.

b: Unless synergies are specified, the effect of a second policy simultaneously implemented is reduced by (1 − the effect of the first policy).

Model outcomes

SimSmoke estimates the effects over time for two primary outcomes: smoking/snus prevalence and smoking (and/or snus)-attributable deaths. Prevalence is projected for the population ages 16–85 separately by gender. The model estimates these outcomes for the tracking period (2004–10) and projects future outcomes through 2040. We validate model projections against survey estimates through 2008, due to the instability of smoking behaviours after the 2008 economic recession.³³

We project the effect of stricter tobacco control policies in isolation and combined. In comparing the effect of policies with the status quo (where tobacco control policies are maintained at their 2010 level), we focus on the relative change in smoking (or snus) prevalence. For smoking and snus-attributable deaths, we calculate lives saved as the difference between the number of deaths under the new policy and the number of deaths under the status quo.

Results

Predictions of smoking and snus prevalence from 2004 to 2010

Between 2004 and 2010, the average price of cigarettes increased by 30% and snus prices increased by 85%, along with stricter restaurant bans and some increase in marketing restrictions. During that period, SimSmoke predicts smoking falls from 14.3 to 12.6% for males and from 23.3 to 20.9% for females, dual use falls from 10.8 to 9.7% for males and from 3.0 to 2.7% for females and snus only falls from 15.6 to 14.6% for males and from 3.5 to 3.3% for females.

Comparing model projections with data from the 2008 Health on Equal Terms of the National Public Health Survey, overall male (female) smoking prevalence fell 13.8% (11.1%) compared with 9.1% (8.2%) predicted by SimSmoke. With regard to exclusive snus use, male prevalence fell 10% compared with 7% predicted by SimSmoke; female prevalence rose slightly, whereas SimSmoke predicted a decline. Thus, declines for snus were over-predicted for women, whereas declines for snus and cigarettes for men and cigarettes for women were under-predicted. However, according to the data, much of the decline through 2008 was reversed with the recession in late 2008 to 2010.

Role of future policies in reducing future smoking and snus use prevalence and deaths

The estimates of smoking and snus use prevalence by gender under the status quo and under varying policy scenarios are shown in table 3 for each of the tobacco use groups. Table 4 displays tobacco-attributable deaths. If tobacco control policies remain at their 2010 levels, as in the status quo scenario, male tobacco use prevalence is projected to decline between 2010 and 2040 from 12.6 to 7.9% for smokers, from 14.6 to 13.5% for snus users and from 9.7 to 8.7% for dual users. Female smoking prevalence falls from 20.9% in 2010 to 16.1% in 2040 and from 2.7 to 2.6% for cigarette and snus use, whereas the prevalence of snus only use increases from 3.3% in 2010 to 3.6% in 2040 (table 3).

Table 3.

Male and female smoking, snus and dual use prevalence, ages 16–85, Sweden, 2010–40

Gender	Males				Females
Year	2010	2011	2020	2040	2010	2011	2020	2040
Cigarettes only
Status quo	12.6%	12.4%	10.9%	7.9%	20.9%	20.7%	19.1%	16.1%
Newly implemented policiesa
Raise excise taxes to 70% of retail price	12.6%	11.0%	9.4%	6.3%	20.9%	18.1%	16.3%	12.8%
Complete smoke-free	12.6%	11.9%	10.5%	7.6%	20.9%	19.9%	18.4%	15.4%
Comprehensive marketing ban	12.6%	12.3%	10.8%	7.8%	20.9%	20.5%	19.0%	15.9%
High-Intensity tobacco control campaign	12.6%	11.9%	10.4%	7.5%	20.9%	20.0%	18.3%	15.3%
Strong health warnings	12.6%	12.4%	10.8%	7.8%	20.9%	20.6%	19.0%	16.0%
Strong youth access enforcement	12.6%	12.4%	10.8%	7.7%	20.9%	20.6%	18.7%	15.0%
Cessation treatment policies	12.6%	12.3%	10.8%	7.7%	20.9%	20.6%	18.9%	15.8%
All of the above policies implemented	12.6%	10.0%	8.3%	5.3%	20.9%	16.5%	14.1%	10.2%
Snus only
Status quo	14.6%	14.5%	14.4%	13.5%	3.3%	3.3%	3.5%	3.6%
Newly implemented policiesa
Raise excise taxes to 70% of retail price	14.6%	13.4%	13.1%	11.9%	3.3%	3.0%	3.1%	3.1%
Complete smoke-free	14.6%	14.5%	14.4%	13.5%	3.3%	3.3%	3.5%	3.6%
Comprehensive marketing ban	14.6%	14.4%	14.2%	13.3%	3.3%	3.3%	3.4%	3.5%
High-intensity tobacco control campaign	14.6%	14.1%	13.8%	12.9%	3.3%	3.2%	3.3%	3.4%
Strong health warnings	14.6%	14.5%	14.3%	13.4%	3.3%	3.3%	3.4%	3.5%
Strong youth access enforcement	14.6%	14.5%	14.1%	12.8%	3.3%	3.3%	3.4%	3.4%
Cessation treatment policies	14.6%	14.5%	14.2%	13.3%	3.3%	3.3%	3.4%	3.5%
All of the above policies implemented	14.6%	12.7%	12.0%	10.4%	3.3%	2.9%	2.9%	2.8%
Dual users – cigarettes and snus
Status quo	9.7%	9.7%	9.3%	8.7%	2.7%	2.7%	2.7%	2.6%
Newly implemented policiesa
Raise excise taxes to 70% of retail price	9.7%	8.3%	7.6%	6.5%	2.7%	2.3%	2.2%	2.0%
Complete smoke-free	9.7%	9.3%	8.9%	8.3%	2.7%	2.6%	2.6%	2.5%
Comprehensive marketing ban	9.7%	9.6%	9.2%	8.6%	2.7%	2.7%	2.6%	2.6%
High-intensity tobacco control campaign	9.7%	9.3%	8.9%	8.2%	2.7%	2.6%	2.5%	2.5%
Strong health warnings	9.7%	9.7%	9.2%	8.6%	2.7%	2.7%	2.6%	2.6%
Strong youth access enforcement	9.7%	9.7%	9.1%	8.3%	2.7%	2.7%	2.6%	2.3%
Cessation treatment policies	9.7%	9.6%	9.2%	8.5%	2.7%	2.7%	2.6%	2.6%
All of the above policies implemented	9.7%	7.6%	6.6%	5.3%	2.7%	2.1%	1.8%	1.5%
% Change in smoking prevalence from status quob
Cigarettes only
Raise excise taxes to 70% of retail price		−11.5%	−13.7%	−20.0%		−12.4%	−15.0%	−20.9%
Complete smoke-free		−3.6%	−3.8%	−4.3%		−3.6%	−3.9%	−4.4%
Comprehensive marketing ban		−0.9%	−1.0%	−1.1%		−0.9%	−1.0%	−1.2%
High-intensity tobacco control campaign		−3.6%	−4.2%	−4.8%		−3.6%	−4.3%	−5.1%
Strong health warnings		−0.2%	−0.6%	−1.0%		−0.2%	−0.6%	−1.0%
Strong youth access enforcement		−0.1%	−0.9%	−2.9%		−0.3%	−2.2%	−7.2%
Cessation treatment policies		−0.5%	−1.3%	−1.8%		−0.5%	−1.2%	−1.8%
All of the above policies		−19.4%	−23.8%	−32.8%		−20.4%	−26.1%	−36.8%
Snus only
Raise excise taxes to 70% of retail price		−8.2%	−9.2%	−11.4%		−8.5%	−9.7%	−11.8%
Complete smoke-free		0.0%	0.1%	0.1%		0.0%	0.1%	0.3%
Comprehensive marketing ban		−0.9%	−1.0%	−1.1%		−0.9%	−1.0%	−1.1%
High-intensity tobacco control campaign		−3.2%	−3.8%	−4.4%		−3.2%	−3.8%	−4.3%
Strong health warnings		−0.2%	−0.5%	−0.9%		−0.2%	−0.5%	−0.9%
Strong youth access enforcement		−0.3%	−1.8%	−4.9%		−0.3%	−1.6%	−3.8%
Cessation treatment policies		−0.5%	−1.0%	−1.6%		−0.5%	−1.0%	−1.6%
All of the above policies		−12.8%	−16.5%	−22.7%		−13.2%	−16.7%	−21.9%
Dual users – cigarettes and snus
Raise excise taxes to 70% of retail price		−14.1%	−18.2%	−25.0%		−15.2%	−19.2%	−25.5%
Complete smoke-free		−3.6%	−4.1%	−4.7%		−3.6%	−4.1%	−4.6%
Comprehensive marketing ban		−0.9%	−1.0%	−1.2%		−0.9%	−1.0%	−1.2%
High-intensity tobacco control campaign		−3.6%	−4.5%	−5.4%		−3.6%	−4.5%	−5.4%
Strong health warnings		−0.2%	−0.7%	−1.2%		−0.2%	−0.7%	−1.1%
Strong youth access enforcement		−0.3%	−1.8%	−4.6%		−0.6%	−3.9%	−10.6%
Cessation treatment policies		−0.5%	−1.4%	−2.2%		−0.5%	−1.4%	−2.1%
All of the above policies		−21.9%	−29.4%	−39.6%		−23.1%	−31.7%	−43.8%

a: Policies are implemented at FCTC-consistent levels in 2010 and maintained at that level through 2040.

b: Percent changes measured as the relative change from the status quo level [e.g. (Prevalence w/policy in 2020 − Prevalence w/status quo in 2020)/Prevalence w/status quo in 2020].

Table 4.

Male and female smoking-attributable deaths, snus-attributable deaths and dual use-attributable deaths, ages 16–85, Sweden, 2010–40

	Males				Females
	2010	2020	2040	Cumulativea	2010	2020	2040	Cumulativea
Cigarettes only
Status quo	7075	7046	5122	196 703	5723	6150	5862	185148
Newly implemented policiesb
Raise excise taxes to 70% of retail price	7075	6817	4694	187 101	5723	5934	5353	174 878
Complete smoke-free	7075	6954	4972	193 021	5723	6063	5675	181 128
Comprehensive marketing ban	7075	7023	5086	195 797	5723	6128	5818	184 161
High-intensity tobacco control campaign	7075	6952	4943	192 645	5723	6062	5638	180 704
Strong health warnings	7075	7035	5089	196 111	5723	6140	5821	184 494
Strong youth access enforcement	7075	7046	5117	196 675	5723	6150	5854	185 102
Cessation treatment policies	7075	7021	5045	195 356	5723	6127	5765	183 652
All of the above policies implemented	7075	6595	4267	177 496	5723	5723	4813	164 334
Snus only
Status quo	212	293	422	10 039	27	41	81	1568
Newly implemented policiesb
Raise excise taxes to 70% of retail price	212	286	396	9613	27	40	76	1494
Complete smoke-free	212	293	422	10 040	27	41	81	1568
Comprehensive marketing ban	212	292	418	9977	27	41	80	1557
High-intensity tobacco control campaign	212	289	406	9790	27	40	78	1525
Strong health warnings	212	292	419	10 003	27	41	80	1561
Strong youth access enforcement	212	293	421	10 035	27	41	81	1567
Cessation treatment policies	212	292	416	9958	27	41	79	1554
All of the above policies implemented	212	280	369	9206	27	39	70	1424
Dual users – cigarettes and snus
Status quo	1954	2565	3025	80 025	246	366	512	12 143
Newly implemented policiesb
Raise excise taxes to 70% of retail price	1954	2460	2651	74 042	246	352	451	11 218
Complete smoke-free	1954	2524	2909	77 931	246	361	492	11 805
Comprehensive marketing ban	1954	2555	2998	79 522	246	365	508	12 062
High-intensity tobacco control campaign	1954	2523	2884	77 674	246	360	487	11 763
Strong health warnings	1954	2559	2995	79 613	246	366	507	12 076
Strong youth access enforcement	1954	2565	3016	79 972	246	366	510	12 131
Cessation treatment policies	1954	2551	2955	79 068	246	364	500	11 986
All of the above policies implemented	1954	2359	2314	68 404	246	337	391	10 299
Lives saved relative to the status quo
Cigarettes only
Raise excise taxes to 70% of retail price		229	428	9602		216	510	10 271
Complete smoke-free		92	150	3681		87	188	4020
Comprehensive marketing ban		23	35	906		22	44	987
High-intensity tobacco control campaign		94	179	4058		89	225	4444
Strong health warnings		11	33	592		11	42	654
Strong youth access enforcement		0	5	28		0	8	46
Cessation treatment policies		25	77	1347		23	97	1496
All of the above policies		451	855	19 207		427	1049	20 815
Snus only
Raise excise taxes to 70% of retail price		7	26	427		1	7	93
Complete smoke-free		0	0	0		0	0	0
Comprehensive marketing ban		1	4	63		0	1	11
High-intensity tobacco control campaign		4	16	250		1	3	42
Strong health warnings		0	3	36		0	1	6
Strong youth access enforcement		0	1	4		0	0	1
Cessation treatment policies		1	6	82		0	1	14
All of the above policies		12	53	834		2	11	143
Dual users – cigarettes and snus
Raise excise taxes to 70% of retail price		105	374	5983		15	61	925
Complete smoke-free		41	116	2093		6	20	338
Comprehensive marketing ban		10	27	502		1	5	81
High-intensity tobacco control campaign		41	141	2351		6	25	380
Strong health warnings		6	30	412		1	5	67
Strong youth access enforcement		0	9	52		0	2	11
Cessation treatment policies		13	70	957		2	12	157
All of the above policies		206	711	11 621		29	121	1843

a: Cumulative = sum of deaths over all years 2011–40 (including years not listed).

b: Policies are implemented at FCTC-consistent levels in 2010 and maintained at that level through 2040.

Relative to the status quo, increasing specific taxes to 70% of price is projected to decrease male smoking rates by 20.0% by 2040. Increasing the snus tax to 70% is projected to reduce snus use by 11.4% by 2040. Increasing both the cigarette tax and snus tax to 70% is projected to reduce dual use by about 25.0% by 2040 (table 3). SimSmoke predicts that of the seven policies alone, increasing the cigarette and snus tax leads to the highest number of cumulative deaths averted between 2011 and 2040, with 9602 male and 10 271 female deaths of smokers only and 5983 male and 925 female deaths of dual users averted by 2040. For those who use snus alone, the cumulative number of deaths averted is 520 (427 male and 93 female) (table 4).

As seen in table 4, a high-intensity tobacco control campaign averts the second largest number of deaths by 2040 in all tobacco use groups. For the ranking of smoking and dual-use-attributable deaths averted (from greatest to least), this policy is followed by smoke-free air laws, cessation treatment policies, a comprehensive marketing ban, strong health warnings and, finally, strong youth access enforcement. Among snus users, the ranking of policies is the same, albeit smoke-free air laws avert the least number of deaths (table 4). The policy ranking is consistent among males and females. A final scenario considers the combined effects of the seven policies. Relative to the status quo scenario, combined policies are projected to avert 40 022 deaths from 2011 to 2040 among smokers of both genders, 977 deaths among snus-only users, and 13 464 deaths among dual users, or a total of 54 463 overall deaths averted.

Discussion

We apply population, smoking prevalence and policy data for Sweden to the established SimSmoke model. In addition to being the first study to consider the potential effects of tobacco control policy in Sweden, to our knowledge, this is the first modelling paper that rigorously examines both smokeless tobacco and cigarette use.

While Sweden has implemented some tobacco control policies in recent years, there is still scope to strengthen tobacco control policies consistent with the FCTC. We estimated that smoking prevalence can be decreased by as much as 26% in the first few years, increasing to a 35% reduction by 30 years with implementation of MPOWER policies. The snus rate can also be decreased by up to 17% within 30 years, but this estimate is made with less confidence due to the limited data available on the effects of policy on snus use. A large increase in cigarette and snus taxes alone would substantially reduce the number of lives lost to tobacco. While our results indicate that implementing MPOWER policies will have a substantial impact on smoking rates and tobacco-attributable deaths, and a modest impact on snus use, there are several caveats that must be considered when interpreting the results.

While this is the first simulation study examining the effect of policies on snus use, our findings on the effect of MPOWER policies on smoking prevalence in Sweden are consistent with those from other European nations. Studies for countries with strong tobacco control policies find that policies have played an important role.^12–17 The models for these countries predicted well, although the model slightly under-predicted the reduction in smoking rates in the UK and Ireland, both with the strongest policies in the European Union.

The mortality risks for smokers are based on US studies, but rates may differ in Sweden. Although relative risks for dual and sole cigarette users we assumed the same, dual users may have lower relative risks due to the lower quantity typically smoked. SimSmoke would then overestimate the number of smoking-attributable deaths. However, some recent studies suggest that snus use may increase the risks of diabetes,³⁴ heart disease,³⁵ maternal and child health outcomes³⁶ and various cancers.³⁷ In developing models for other countries, it will be important to consider differences across countries in the risks profiles of the various types of smokeless tobacco used^38,39 and the extent of switching between cigarettes and smokeless tobacco.^40,41

The projections also exclude the deaths averted due to reductions in secondhand smoke exposure. Snus use may reduce some of this exposure. We also caution that the risks may be overstated for Sweden because of the high rate of some day smoking, which we do not distinguish from everyday use. Similarly, some day snus users merit attention, particularly with regard to trends in dual use.

The policy estimates also depend on underlying assumptions and estimated effect sizes. Knowledge of the effects of each policy varies.⁴² In previous work,^11,43 we have estimated that the effects of cigarette taxes can be expected to vary by about 25% around best estimates, but by 50% around the estimates for other policies (with an upper limit of 100% variation around cessation treatment and youth access policies). Moreover, due to a lack of studies, we were not able to predict how the tobacco industry would respond or adapt to changes in tobacco control policies. In particular, while only 34% of respondents in Sweden have heard of e-cigarettes, and of those, merely 17% know what these products are, e-cigarettes may aid smokers in quitting, thus reducing the smoking prevalence and attributable deaths to a greater extent than predicted by the model.⁴⁴ In addition, we were not able to explicitly incorporate network effects through the workplace, peers and parents.

Because our model began in 2004, we were not able to distinguish the effect of policies on cigarettes vs. snus use. Studies examining the effect of tobacco control policies on snus use are few, and the estimated effects of policy on snus incorporated in SimSmoke reflect this uncertainty. Even tobacco taxation, while subject to considerable research, merits further attention regarding the effect of cigarette price on snus use and the effect of snus price on cigarette use. We have assumed that smoke-free air laws in Sweden do not apply to snus, as its use does not generate smoke. Smoke-free laws could encourage substitution of snus for cigarettes, thereby reducing the incentive for smokers to quit in response to smoke-free laws. Alternatively, some or all of this effect might be offset by the anti-tobacco norms created by the laws.

In general, it will be important to distinguish the effect of snus use on cigarette use and dual use, and distinguish the effects of policies directed specifically at snus use from those targeting cigarette use. Nonetheless, this study suggests that policies may be needed that are specifically directed at snus use, as smoking rates have stalled in recent years, with many joint users of snus and cigarettes. In particular, snus use has been increasing among females, and Sweden has rates of smoking above other nations that have recently implemented strong policies.^12,15 Furthermore, some recent evidence indicates that the Swedish experience may not apply to other nations.⁴⁵ Trends in smokeless tobacco and cigarette use are likely to depend on a nation’s stage in the tobacco epidemic.

In summary, the Sweden SimSmoke results highlight the relative contribution of policies to reducing the tobacco health burden. The model predicts that many premature deaths can be averted by implementing large increases in cigarette and snus taxes, especially when combined with other policies. While the results should be viewed as preliminary; Sweden SimSmoke helps identify the information needed to develop sound policies toward smokeless tobacco use.

Funding

This is a deliverable of the PPACTE project ‘Pricing Policy and Control of Tobacco in Europe’ partly funded by the European Commission through FP7-HEALTH-F2-2009-223323. L.M.C. is a PhD Scholar in Health Services Research partly funded by the Health Research Board in Ireland under grant number PhD/2007/16. This publication was also made possible by Grant Number U01-CA97450-02 from the National Cancer Institute of the National Institutes of Health as part of the Cancer Intervention and Surveillance Modeling Network (CISNET). D.T.L. and A.M.N. were also supported in part by the National Institute On Drug Abuse of the National Institutes of Health under Award Number R01DA036497 to D.T.L.

Conflicts of interest: None declared.

Key points.

• SimSmoke applies data from Sweden to examine the effect of implementing stricter policies that would be fully consistent with the FCTC on the prevalence of and deaths attributable to smoking and snus use in Sweden.

• Sweden SimSmoke highlights the relative contribution of policies to reducing the tobacco health burden. The model predicts that many premature deaths can be averted by implementing large increases in cigarette and snus taxes, especially when combined with other policies.

• The structure of the model helps identify the information still needed to develop sound policies toward smokeless tobacco and cigarette use.