Objective To evaluate the correlation between the implementation of tobacco control policies, particularly smoke-free bans at work and in public places, and smoking prevalence in private venues in the 27 countries of the European Union.
Design Ecological study with the country as the unit of analysis.
Data sources Data analysis of tobacco control activities in European countries in 2007 as compiled in the Tobacco Control Scale (TCS) and information on the level of smoking permissiveness in houses and cars from the Special Eurobarometer on Tobacco conducted in 2009.
Analysis Spearman rank-correlation coefficients (rsp) and their 95% confidence intervals (CIs) were calculated.
Results The correlation between the TCS score and the prevalence of smoking in private venues (houses and cars) where smoking inside was always allowed was close to zero. A similar lack of association was observed between the TCS score of specific bans at work and in public places and smoking rules inside houses and cars. There was a non-significant direct correlation between the TCS score and the prevalence of smoke-free houses (rsp=0.21, 95% CI −0.19 to 0.55) and a non-significant inverse correlation with smoking allowed in certain rooms inside the house (rsp=−0.34; 95% CI −0.64 to 0.05).
Conclusions Smoke-free legislation in workplaces and public places is not correlated with increased smoking prevalence in private venues (houses and cars) at an ecological level.
- Secondhand smoke
- Public policy
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Once the Framework Convention on Tobacco Control was ratified,1 several European countries implemented smoke-free bans in workplaces and in other public places to protect the non-smoking population from the harmful effects of second-hand smoke (SHS) exposure. Comprehensive smoke-free policies decreases SHS exposure and associated health hazards in both non-smokers and smokers, and increase the likelihood among smokers of quitting or reducing cigarette consumption.2–8 However, the tobacco industry and the hospitality sector, during the debate about implementation of smoke-free policies in different countries, argued that the restriction of smoking in public places would displace tobacco consumption to private venues, particularly at home. Theoretically, according to this argument, exposure to SHS among children would have increased after the implementation of smoke-free legislation. To our knowledge, only one study,9 conducted in the USA, supports this hypothesis, while other studies show no displacement of smoking prevalence toward private house after the implementation of smoking bans.2 ,10 ,11 Moreover, the increase of the overall tobacco control measures may improve the support to smoking bans in public venues and to the denormalisation of tobacco consumption4 ,5 ,7 ,8; this could encourage the adoption of voluntary smoke-free homes and cars.
The objective of this study was to evaluate the correlation between the implementation of tobacco control policies, particularly smoke-free bans in work and public places, and smoking prevalence in private venues (inside houses and cars) in the 27 countries of the European Union.
This ecological study was based on data obtained from different sources, with the country as the unit of analysis. We used data on tobacco control activities in European countries in 2007 as compiled in the Tobacco Control Scale (TCS).12 The TCS provides a score for each country reflecting the level of implementation of smoke-free rules according to six selected cost-effective policies.13 An expert working group from the European Network for Smoking Prevention (ENSP) developed the TCS by means of a questionnaire sent to ENSP correspondents within each country. The score for each policy was weighted by its reported effectiveness based on existing research and the discussion of a panel of experts on tobacco control.
We also obtained information on smoking rules inside homes and cars from the Special Eurobarometer (No. 332) on Tobacco (Eurobarometer 72.3).14 The Eurobarometer is a cross-sectional study conducted for the European Commission by TNS Opinion & Social in the 27 countries of the European Union plus the three additional candidate countries in 2009 (Croatia, Turkey and the Republic of Macedonia). The fieldwork was performed in October 2009. In each country, interviews were conducted face-to-face at people's houses, and in the appropriate national language. The sample was weighted for socio-demographic variables. The final sample (n=30 292) was representative of the population ages 15 years and above in each country (about 1000 participants in each country, except in Cyprus, Luxemburg, and Malta, which had approximately 500 respondents each).14
Tobacco use inside the houses
Data on smoking inside the houses were obtained using the Eurobarometer question: ‘Which statement best describes smoking situation inside your house?’ The possible answers were: ‘smoking is not allowed at all inside the house,’ ‘smoking is allowed only in certain rooms inside the house’ and ‘smoking is allowed everywhere inside the house’.
Tobacco use inside the cars
Information about smoking rules inside private cars was obtained using the Eurobarometer question: ‘Do you allow smoking in your car?’. The possible answers were ‘smoking is never allowed in my car’; ‘smoking is allowed sometimes in my car’ and ‘smoking is allowed all the time in my car’.
Tobacco control policies
To quantify the grade and effort of implementation of tobacco control policies in European countries, we used data from the 2007 TCS.12 The six policies considered in the TCS and their corresponding scores are as follows: price increases through higher taxes on tobacco products (maximum 30 points); bans/restrictions on smoking in public and workplaces (maximum 22 points); better consumer information, including public information campaigns, media coverage and publicising of research findings (maximum 15 points); comprehensive bans on the advertising and promotion of all tobacco products, logos and brand names (maximum 13 points); large, direct health-warning labels on cigarette boxes and other products (maximum 10 points); and treatment to help dependent smokers to quit, including increased access to medications (maximum 10 points). The maximum score of the TCS is 100 points, indicating full implementation of all strategies considered.
We analysed the correlation between the TCS score and smoking rules inside the houses and cars by means of Spearman rank-correlation coefficients (rsp) and calculated the 95% confidence intervals (CIs) of the Spearman coefficients. We also analysed the correlation between the score of bans/restrictions on smoking in workplaces and public places and smoking rules inside houses and cars.
The overall prevalence of Europeans who did not allow smoking at all inside the homes was 61%. The lowest prevalence was observed in Greece (39%) and the highest one in Finland (95%). The prevalence of Europeans who never allowed smoking in their cars was 52%, ranging from 17% in Bulgaria to 78% in Finland.
The TCS score substantially varied among countries, with Austria having the lowest (35/100) and UK the highest country score (93/100). The countries with a higher prevalence of smoke-free houses and cars (Finland and Sweden) showed reasonably high TCS scores (58 and 61 points, respectively). In the countries with lower TCS scores (Greece and Austria; scores ≤36), the prevalence of smoke-free houses and cars was relatively low (39% and 54% in houses, respectively, and 24% and 54% in cars, respectively).
There was a direct non-significant correlation between the overall TCS score and the prevalence of smoke-free houses (rsp=0.21, 95%CI −0.19 to 0.55) and an inverse non-significant correlation with the norm of allowing smoking in certain rooms inside the houses (rsp=−0.34; 95% CI −0.64 to 0.05). The correlations were close to zero between the overall TCS score and the prevalence of houses where smoking was allowed everywhere inside the houses and the prevalence of cars where smoking was allowed all the time in the cars. A similar lack of correlation was observed between the TCS score of specific bans at work and in public places and smoking rules inside houses and cars, with no pattern according to the degree of rules about smoking inside the house or car (table 1).
At an ecological level, there is no correlation between the implementation of smoke-free legislation at work and in public places and an increase in prevalence of smoking in private venues (houses and cars) in the European countries. If the hypothesis argued by the tobacco industry and hospitality sector were true, we would have expected to find a positive correlation. Moreover, more developed smoke-free policies (as measured by the TCS), particularly at work and in public places, were positively correlated, although non-significantly, with a high prevalence of smoke-free houses.
The results from this ecological analysis are in line with those of other studies based on data at the individual level before and after national smoke-free bans at workplaces and in public places. A study conducted in the USA based on individual data15 showed that living in a county with smoke-free legislation in public places was associated with smoke-free houses for smokers and non-smokers (ORs=7.76 and 4.16, respectively), with a dose–response relationship with the level of the smoking ban of the county. Another study conducted in England among children (4–15 years old) with reported data and salivary cotinine as a biomarker of SHS exposure16 showed that the English smoke-free law did not produce a shift in SHS exposure to homes. The same study also found an increase in the proportion of children living in smoke-free houses between 1998 and 2008 (from 64% to 80%).16 Similarly, in Scotland, among primary schoolchildren (around 11 years old), there was a reduction in salivary cotinine concentration10 and an increase in the proportion of children reporting a completely smoke-free house17 after passage of the Scottish smoke-free legislation. This finding likely reflects the impact of smoke-free legislation on both increased smoking cessation of parents and reduced acceptability of smoking at home.
The main limitation of our study derives from its ecological design.18 We have no information about the intensity of SHS exposure19 at the individual level in the houses where smoking continues to be allowed. However, even though we do not infer any relationship at the individual level, our results are in agreement with other studies with data at the individual level.15 ,16 ,17 Also, some discrepancies exist in the prevalence estimates between data from Eurobarometer and data from representative national surveys.20 Nevertheless, the design of the Eurobarometer was the same for all countries, which increases comparability across countries at the ecological level. Moreover, the sample size was satisfactorily large and representative by country, and the interviews were face-to-face. Another potential limitation is derived from the lack of information about the stage of the tobacco epidemic among different countries.21 Countries at late stages have low prevalence of smokers, and hence, the likelihood of smoke-free homes is higher, even in the absence of strong tobacco control policies. Data from Finland and Sweden illustrate this possible paradox. The time interval between the TCS (from 2007) and the Eurobarometer survey (from the end of 2009) is a strength of our study because it provides an adequate time frame (2 years) for observing the potential effects of tobacco control policies on smoking behaviour (potential displacement of smoking to private venues). Moreover, the score of the policy on smoking bans in public and workplaces of the TCS in the period of our study, between 2007 and 2010, did not show appreciable changes: the scores increased in 14 countries, did not change in nine countries, and decreased in five countries.22
Current evidence at the ecological and individual levels indicates no increase in the prevalence of smoking in private venues (houses and cars) following the enactment of comprehensive smoke-free legislation, in particular comprehensive smoking bans in work and public places. However, promotion of smoke-free houses and cars should be a priority for public health interventions. Such private venues, particularly homes,23–25 are the main settings where children are exposed to SHS and children are the most vulnerable to the harmful effects of SHS exposure.
What is already known on this topic
The tobacco industry and the hospitality sector have hypothesised that legislation for smoke-free public places will displace the tobacco consumption to private venues, particularly the homes.
What this paper adds
Our findings confirm that smoke-free legislation in work and public places is not correlated with increased smoking in private venues (houses and cars) at the ecological level.
Promotion of smoke-free houses and cars remains a priority for public health interventions because the main sources of SHS exposure among children are private venues.
Contributors JMMS conceived the study, prepared the database, analysed the data and drafted the manuscript. JMMS, EF, CB, MF, SG, and CLV contributed substantially to the conception, design, and interpretation of data. All authors contributed to the manuscript and approved its final version.
Funding This project was funded by Instituto de Salud Carlos III, Government of Spain (RTICC RD12/0036/0053and PI081436) and Ministry of Universities and Research, Government of Catalonia (grant 2009SGR192). SG and CLV are supported by the Italian Ministry of Health, the Italian Association for Cancer Research (AIRC, No. 10068), and the Italian League Against Cancer (LILT), Milan. The funding organisations had no role in the study design, data collection and analysis, interpretation, writing the report or the decision to submit it for publication.
Competing interests All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no financial relationships with any organisations that might have an interest in the submitted work in the previous 3 years; no other relationships or activities that could appear to have influenced the submitted work.
Provenance and peer review Not commissioned; internally peer reviewed.