Objective To determine the coverage of smoking restriction policies in indoor workplaces in China and to assess the relationships between these restrictive policies and secondhand smoke (SHS) exposure and smoking behaviours.
Methods A cross-sectional household survey was conducted in six counties in Sichuan, Jiangxi and Henan provinces in 2004. Using a standardised questionnaire, information on demographic characteristics, knowledge, attitudes and behaviours related to smoking and SHS exposure was collected through face-to-face interviews by trained local investigators among 12 036 respondents. Of respondents, 2698 individuals worked mainly indoors and were included in data analysis.
Results Only 28.5% of respondents reported that indoor workplaces had a smoke-free policy. Even when respondents reported smoke-free policies, 41.1% smokers reported that they were non-compliant with policies and smoked at work. In addition, 32.0% of non-smokers reported being exposed to SHS at work despite smoke-free policies. Non-smokers who reported no smoking restriction policies were 3.7 times more likely to be exposed to SHS than those working in smoke-free workplaces (adjusted OR 3.7, 95% CI 1.3 to 10.1). On average, respondents complying with smoke-free policies smoked 3.8 fewer cigarettes than those reporting no policies in their workplaces at a marginally non-significant level (p=0.06) (adjusted mean difference −3.8, 95% CI −8.0 to 0.5).
Conclusions In China, few workplaces have implemented policies to restrict smoking, and, even in workplaces that have policies, workers report exposure to SHS while at their places of employment. Many workers report a lack of compliance with smoke-free policies. China needs better implementation of SHS policies to promote compliance. Working to improve implementation of smoke-free policies would promote cessation since Chinese smokers who were compliant with these efforts reported smoking fewer cigarettes per day.
- smoking restriction policy
- secondhand smoke
- environmental tobacco smoke
- public policy
Statistics from Altmetric.com
It has been well documented that exposure to secondhand smoke (SHS) causes coronary heart disease, lung cancer and chronic respiratory infections.1 To protect non-smokers from SHS exposure, the Framework Convention on Tobacco Control (FCTC) mandates that all signatories adopt and implement legislative measures in indoor workplaces. The implementation of smoke-free policies has shown significant effects on reducing SHS exposure,2–4 respiratory symptoms,5 6 and the numbers of hospital admissions for acute coronary syndrome.7–9 In addition, smoke-free policies have been shown to reduce tobacco consumption and increase smoking cessation.10–12
It is well known that China is the world largest producer and consumer of tobacco products. According to the national survey in 2002, about 300 million smokers are living in China and about 5 million non-smokers are exposed to SHS.13 Among these passive smokers, 67% are exposed to SHS in public places and 35% in workplaces.14 Two recent studies reported quite high air nicotine concentrations in hospitals, schools, governmental buildings and other workplaces.15 16 Monitoring of airborne nicotine conducted in urban and rural areas of China detected SHS in 91% of locations sampled, indicating frequent exposure in public places. The median concentrations of air nicotine in hospitals, schools and governmental buildings were 0.17 μg/m3, 0.27 μg/m3 and 0.48 21 μg/m3, respectively. This monitoring took place in selected county areas of Sichuan, Jiangxi and Henan provinces and their capital cities, as well as in Beijing.15 A second study, also indicating high levels of SHS exposure, reported a median air nicotine concentration of 1.21 μg/m3 in 14 health department offices in 10 provinces throughout the country.16
Since the 1990s, local Chinese governments have taken action to protect non-smokers from SHS exposure in public places. By the end of 2006, 46% of cities had passed regulations to restrict or ban smoking in hospitals, schools, libraries, public transportation vehicles, waiting rooms and other places.17 However, most of these laws did not specifically cover offices and other indoor workplaces. A majority of the current smoke-free policies in workplaces are adopted voluntarily and are not enforced by regulation or legislation. To date, knowledge of the coverage of these voluntary workplace smoking restriction policies and their performance protecting people from SHS exposure and reducing smoking has been limited, especially in the counties and more rural areas of China. In China, a county is a lower level government administrative unit that is next to city in a province. Counties usually have populations of about 500 000 residents who reside in towns and villages. Roughly 70% of the Chinese population resides in counties. An average Chinese city can have a population of a few million, while an average Chinese province can have a population of 50 000 000.
The purpose of this study is twofold: to determine the coverage of workplace smoking restriction policies among indoor workers in six counties in China, and to assess the relationship between restrictive workplace smoking policies and SHS exposure among non-smokers as well as cigarette consumption and intention to quit among smokers.
This study is based on the baseline household survey of the Fogarty International Center (FIC) Initiative, ‘International Tobacco and Health Research Capacity Building Program’, which aimed to reduce SHS exposure in county areas.18 Six counties in three provinces were selected to participate in this project. The six counties were: Mianzhu and Xichong, from Sichuan province; Anyi and Hukou, from Jiangxi province; and Xin'an and Yangshi, from Henan province. The locations were selected based on the willingness of their government officials to participate. They were also selected in an attempt to have a geographic distribution that reflected different regions of China. Sichuan province is located in the southwest region, Jiangxi province is located in the southeast region and Henan province is located in the central region.
Sample selection and data collection
In each county, the population was stratified into two groups: urban and rural residents. A three-stage cluster random sampling method was adopted to enrol participants. At the first stage, villages or neighbourhoods were randomly selected from rural and urban areas. At the second stage, households were randomly selected from selected villages or neighbourhoods. At the third stage, the adult in selected households whose birth date was closest to the survey date was asked to take part in this survey.
From July 2004 to September 2004, trained local investigators administered face-to-face interviews to 14 400 selected samples. A total of 12 036 participants, ages 18 to 69, completed the interview for an overall response rate of 83.6%.19 To evaluate the indoor workplace smoking restriction policy, a subsample of 2704 respondents who reported working indoors was eligible for data analysis. Women smokers were excluded because only six women reported smoking. Therefore, 2698 indoor workers were interviewed. They included 796 men who smoked and 1902 men and women who were non-smokers.
In this study, information on participants' demographic characteristics, knowledge, attitudes and behaviours related to smoking and SHS exposure was collected through a standardised questionnaire. Smokers were defined as individuals who responded to the questions ‘Have you smoked over 100 cigarettes in your entire life?’ and ‘Did you smoke in the last month?’ affirmatively. Non-smokers included individuals who had never smoked and individuals who reported being former smokers.
The presence of a restrictive workplace smoking policy was measured by the questions: ‘Is there any official policy restricting smoking in any way?’ and ‘Which statement best describes your workplace smoking restriction situation?’. Respondents who answered that smoking was not allowed in any indoor areas were considered to be working under a smoke-free policy. Respondents who answered that smoking was allowed in some indoor areas were considered to be working under a restricted policy. Respondents, who reported no restrictions on smoking, or answered that smoking was allowed in all indoor areas, were considered to be working under an unrestricted policy.
Among smokers who reported working in smoke-free workplaces, compliance with policies was evaluated by the question: ‘Do you usually smoke in the workplace?’, with respondents who answered ‘never’ considered to comply with the smoke-free policy and those who answered ‘often’ or ‘sometimes’ considered not to comply.
Among non-smokers, SHS exposure in workplaces was evaluated by asking: ‘Are you usually exposed to SHS in the workplace?’, with respondents who answered ‘often’ or ‘sometimes’ considered to be exposed to SHS and those who answered ‘never’ considered to be not exposed. Among smokers, respondents who answered ‘yes’ to the question ‘Do you plan to quit smoking?’ were considered to have an intention to quit. Cigarette consumption was indicated by cigarettes smoked per day, which was measured by the question: ‘How many cigarettes do you usually smoke daily?’.
In addition to these primary independent or dependent variables, smokers were asked whether there was any smoking restriction policy in their homes, and if other family members or doctors had advised them to quit smoking. In addition, addiction to nicotine was measured by the question: ‘How soon after first waking up did you smoke your first cigarette?’, with respondents who answered ‘within 30 min’ defined as addicted.
The coverage of smoking restriction policies was described by several demographic and smoking status variables. The demographic variables include gender, residence of living (urban and rural area), province (Sichuan, Jiangxi and Henan), age (20–39 and 40–69 years), educational attainment (junior high school or below, senior high school and college or above) and occupation (government official or scientific personnel, health worker or teacher and labourer or employee in service departments). χ2 tests were conducted to assess if there was difference in reporting smoking restriction policies across these groups.
To assess the relationship between smoking restriction policies and SHS exposure in workplaces among non-smokers and the relationships between policies and cigarette consumption among smokers and quit intention, two dummy variables were created to indicate the three types of smoking policies. Logistic regression and linear regression models were fitted for binary (SHS exposure and intention to quit) and continuous (cigarette consumption) outcomes, respectively. In assessing the relationship between the smoking restriction policy and SHS exposure, the considered covariates included region, province, age, gender, education and occupation. In assessing the relationship between policies and cigarette consumption and quit intention among smokers, the considered covariates included home smoking restriction policy, advice to quit, addiction to nicotine and demographic characteristics. The variable selection strategy for final multivariate regression models is the 10% change in estimate method with forcing of the demographic variables into the model.20
In addition, to examine the relationships between smoke-free policies, cigarette consumption and quit intention by compliance status, a subgroup of participants, smokers who worked under smoke-free policies or in unrestricted workplaces, was drawn from the total study participants. These participants were classified into three groups: those complying with smoke-free policies, those not complying with smoke-free policies and those working in unrestricted workplaces. In the subgroup analysis, two dummy variables were created to indicate the three groups. Logistic and linear regression models were fitted as mentioned above.
Due to the stratified cluster sampling design and unequal selection probability for each sampling unit, weighted estimates were acquired by assigning a weight (defined as the inverse of selection probability) to each observation. All data were analysed using SAS (SAS, Cary, North Carolina, USA) callable SUDAAN V.9.0.1 (RTI, Research Triangle Park, North Carolina, USA). Weighted frequency, mean, OR and mean difference were calculated by the CROSSTAB, DESCRIPT, RLOGIST and REGRESS procedures, respectively.
Of the 2698 respondents who remained in the analysis, 61.8% were women and 38.2% were men; 31.9% were smokers and 68.1% were non-smokers. Only 28.5% (31.5% for non-smokers and 22.3% for smokers) of respondents reported that their indoor workplaces had smoke-free policies and 60.3% (59.5% for non-smokers and 62.2% for smokers) reported there were no smoking restrictions in indoor workplaces at all. Compared with smokers, non-smokers were more likely to report a smoke-free policy (p=0.04).
Reported workplace smoking restriction policies by demographics
Tables 1 and 2 summarise the proportion of non-smokers and smokers who reported working in smoke-free or smoking-restricted environments by demographic characteristics. For non-smokers (table 1), those living in urban areas and Henan province, those with high education levels and those with certain occupations (doctors, teachers, government officials, scientific personnel) were significantly more likely to report smoke-free or smoking-restricted policies in work environments. For smokers (table 2), those living in Henan province and those with certain occupations (doctors, teachers, government officials scientific personnel) were significantly more likely to report smoke-free or smoking-restrictive policies.
Workplace smoking restriction policy and SHS exposure among non-smokers
Table 3 shows 67.3% of non-smokers reported they were exposed to SHS in unrestricted indoor workplaces. The percentages of non-smokers who reported SHS exposure in restricted and smoke-free indoor workplaces were 62.3% and 32.0%, respectively.
After adjusting for region, province, age, gender, education and occupation, non-smokers without any smoking restriction in their workplaces were 3.7 times more likely to be exposed to SHS than those working in a smoke-free workplace (OR 3.7, 95% CI 1.3 to 10.1). Non-smokers reporting a restricted policy in their workplaces were 2.3 times more likely to be exposed to SHS than those working in a smoke-free workplace (OR 2.3, 95% CI 1.1 to 4.7).
Workplace smoking restriction policy and smoking behaviour among smokers
As figure 1 indicates, about 90.9% of smokers smoked in unrestricted indoor workplaces. The percentages of smokers who reported smoking in restricted and smoke-free indoor workplaces were about 83.5% and 41.1%, respectively.
Table 4 shows that among those smokers reporting an unrestricted policy, the mean number of cigarettes smoked per day was 15.1. For those reporting a restrictive or smoke-free policy, the means were 14.2 and 10.3 cigarettes daily, respectively. It indicates that those working under a smoke-free policy smoked 4.4 fewer cigarettes per day than those working in unrestricted working environments (95% CI −9.0 to −0.7). However, the difference diminished and was not statistically significant, after adjusting for demographic characteristics and other confounding factors (mean difference 1.4; 95% CI −5.6 to 2.8).
About 41.4% of smokers working in an unrestricted workplace expressed an intention to quit smoking (table 4). The percentages for those reporting a restrictive and a smoke-free policy were 44.3% and 49.2%, respectively. Multivariate analysis did not show significant effect of smoking restrictions on motivating intention to quit. After adjusting for demographic characteristics and other potential confounders, the ORs of having intention to quit were 1.2 (95% CI 0.4 to 3.7) and 0.8 (95% CI 0.4 to 1.8) for those reporting a restricted policy and those reporting a smoke-free policy, respectively.
Table 5 shows the relationship between smoke-free policy and daily cigarette consumption and quit intention among smokers by compliance status. Those smokers complying with smoke-free policies smoked 3.8 (mean difference −3.8, 95% CI −8.0 to 0.5) fewer cigarettes per day than those who worked in unrestricted indoor workplaces, at a marginally non-significance level (p=0.06). Although the compliant respondents were more likely to show a desire to quit, the results are not statistically significant (OR 1.3, 95% CI 0.4 to 4.0).
This study shows that smoking in indoor workplaces throughout China is very prevalent and the proportion of indoor workers covered by smoke-free policies is very low, especially in the county areas where most of the population resides. Only 28.5% of respondents reported working indoors under a smoke-free policy and 60.3% of respondents reported that there was no policy restricting smoking in their workplaces. By contrast, about 77% of US respondents to a similar survey reported working under a smoke-free workplace policy in 2003.21 In addition, since the Ireland became the first country to ban smoking in all indoor workplaces in 2004, over 10 countries have passed similar national laws or regulations. Banning smoking in workplaces has been extensively recognised as a vital measure in tobacco control by countries around the globe. Since a high prevalence of SHS exposure has been documented in workplaces across China, more efforts, including public education campaigns to build awareness of the dangers of secondhand smoke, are needed as policy actions are being undertaken to reduce smoking in workplaces.
Consistent with previous studies,22–24 this study shows disparities in reported workplace smoking restriction policies across different educational and occupational groups throughout China. While there was little difference in exposure to SHS by gender, it is well known that the vast majority of smokers in China are men: it remains a cultural norm for men to smoke.25 People with higher education or living in urban areas were more likely to report smoking restriction policies in their workplaces. Health workers and teachers were more likely to report smoking restriction policies than other occupations, although only 35% of them reported to have a smoke-free policy in their workplaces. These differences reflect the fact that medical facilities, governmental buildings and schools are more likely to voluntarily or mandatorily adopt smoke-free policies. A major concern is that these disparities may ultimately be translated into increased health disparities across socioeconomic groups. In developing an overarching national smoking restriction policy, banning smoking at all indoor workplaces should be required and considered the best practice to reduce disparities and protect all Chinese citizens.
The ultimate objective of smoking restriction policy development is not to simply pass a law or regulation to cut individual cigarette consumption, but it is to properly implement and enforce the policy effectively in order to protect non-smokers from SHS exposure as well. Monitoring compliance with existing policies is a major component of policy implementation. This study shows that compliance with smoking restriction policies is poor in China. More than 40% of smokers working under a smoke-free policy reported sometimes smoking in their workplaces. About one-third of non-smokers reported being exposed to SHS in smoke-free workplaces. There may be several underlying reasons for this low compliance rate. First, advocacy and publicity are not always sufficient before passing and implementing individual smoke-free policies at workplaces in China. As a result, there is little effort to implement the policies and there is little information or education about why policies are necessary. Evidence exists that misconceptions about SHS policies are extensive.25 In addition, employees have reported that they feel regulations have been enacted without their input and they do not understand their responsibilities or how to contribute to implementation of the policy. In addition, a lack of tobacco control support activities, such as health education or smoking cessation programmes, is a reality in China. Smokers may know they will be fined if they violate a regulation, but they may not adapt their behaviours because they do not understand the extent to which their behaviour may harm the health of themselves or others. In short, they don't understand the need to quit smoking. In this study, it often was expressed that the ‘government must take the lead in order for effective tobacco control to take hold in China’. The current ambiguous stance of the government towards tobacco control has projected the message that tobacco control is voluntary. The government's ambivalent stance severely weakens the ability of other institutions, such as workplaces to properly implement and strictly enforce smoking bans.25 Strong government leadership is needed to promote compliance and send a powerful message to other institutions to take tobacco control seriously.
Our study confirms workplace smoking restriction policies were related to significantly reduced self-reported SHS exposure in workplaces and an indoor smoke-free policy has a much stronger association with SHS exposure than does a restricted one. Similarly, Borland and colleagues found non-smokers working in areas where there was only a work area ban were 2.8 more likely to be exposed to SHS than those working in a smoke-free workplace. Non-smokers who reported no smoking policies were over eight times more likely to be exposed to SHS than those respondents working in smoke-free workplaces.26 The positive impact of smoke-free policies on SHS exposure has been supported by self-reporting epidemiological studies as well as scientific studies measuring air pollutants. For example, a study in China found that the median air nicotine concentrations in workplaces without smoking restriction was about five times higher than in workplaces with restrictions.16 In addition, recent studies have found that air nicotine or particle matter concentrations dropped dramatically after the implementation of smoke-free policies in restaurants and other places.27 28
It has been reported that workplace smoking restrictions can reduce SHS exposure and lead to a decrease in the prevalence of smoking and number of cigarettes consumed daily by smokers. The more restrictive policies have the greatest impact on smoking behaviour.23 29 Due to the small sample size (N=689), our study results show a marginally non-significant (p=0.06) effect of workplace smoking ban in reducing smokers' daily consumption of cigarettes. On average, respondents complying with smoke-free policies smoked 3.8 fewer cigarettes than those reporting no policy in their workplace. However, because the data are cross-sectional, we cannot deduce whether smoking policies (1) caused changes in smoking behaviour, (2) caused heavier and more addicted smokers to leave work, or (3) were more likely implemented at workplaces with lighter smokers. Recent evidence is consistent with the first hypothesis. In a recent longitudinal study, people who worked in environments that had smoke-free policies in place at the baseline and the end of an 8-year follow-up were 2.3 times more likely than people not working in such environments to have quit by the end of follow-up period. Furthermore, respondents who continued smoking reported a decline in their average daily consumption of 3.85 cigarettes.29
No significant association between smoke restriction policy and smokers' intention to quit was found in this study. The low compliance rate may explain this phenomenon. In general, a smoke-free policy should motivate smokers to quit by making it more difficult to consume cigarettes or other tobacco products. Smokers must go outside their workplace building to smoke. In addition, a smoking ban conveys a message that smoking is not a socially acceptable behaviour. Ultimately, this results in a change in social norms. However, this effect is undermined when the smoke-free policy is violated by large numbers of smokers who continue to smoke in smoke-free buildings. In addition, non-compliance may jeopardise a smoker's success in quitting because they are surrounded by others who continue to smoke.30
This study is one of the few that has investigated smoking restriction policies in China. It was conducted among participants randomly selected from the general population in six counties so it provides a clear picture of coverage of workplace smoking restrictions. In addition, by assessing smoking restriction policies in indoor workplaces, this study will contribute to providing supportive evidence for policy development in county areas where smoking and SHS are more prevalent and knowledge about smoking restrictions is limited. In most previous cross-sectional studies, potential confounders other than demographic characteristics were not considered when assessing the association of policy with smoking behaviour. Our study adjusted for factors such as home smoking restriction policy, addiction and advice from doctors as well as respondents' demographic characteristics.
Despite these strengths, this study has several limitations that should be noted. First, workplace smoking restriction policy data are self-reported without verification from employers or workplaces. Respondents' definitions of an official workplace smoking policy may vary and the validity of self-reports is unknown. The second limitation is the lack of temporality, an inherent character of all cross-sectional studies. We cannot infer a causal relationship between workplace smoking restriction policies and employees' smoking behaviour. In addition, this study did not measure the duration of smoking restriction policies. We hypothesise that the longer smoke-free policies remain in place, the stronger their impact on smokers' behaviour. Thus, the lack of measurement of this variable may lead to an underestimation of the impact of the policies. Lastly, we excluded women who smoke from the data analysis because of the low smoking prevalence in women and a very small number of women smokers defined in this study. More research is needed to understand the impact of workplace smoking restriction policies on the smoking behaviour of women. In addition, the confirmation of a low smoking prevalence among women raises a call for more research to understand the differences between smoking prevalence rates for men and women in China.
In 2005, China ratified the FCTC, indicating that the Chinese government has made a serious commitment to taking legislative action to reducing or eliminating SHS in workplaces. Given the fact that more restrictive policies are more effective and most workers support the implementation of smoke-free policies in their workplaces (data not shown in tables), a 100% smoke-free policy should be given priority consideration in future legislation. At present, because workplace smoking restriction policies are often instituted without any public understanding or participation, the existing policies have not been well enforced. To improve compliance with smoke-free policies, supportive activities, such as health education and quit assistance programmes should be developed along with policies banning smoking at the workplace.
What this paper adds
This paper provides information concerning the extent of voluntary workplace smoking restrictions among indoor workers in smaller urban and rural areas of China: places rarely studied.
The paper also assesses the relationship between these restrictive policies and the perceived amount of exposure to secondhand smoke (SHS) reported among non-smokers as well as self-reported behaviour changes by smokers.
This study also shows disparities in reported workplace smoking restriction policies across different educational and occupational groups throughout China. A major concern is that these disparities may ultimately be translated into increased health disparities across socioeconomic groups.
Moving towards an overarching national smoking restriction policy, banning smoking at all indoor workplaces should be required and considered the best practice to reduce disparities and protect all Chinese citizens.
Funding This work was supported by grants from the Fogarty International Center of the National Institutes of Health, USA (grant no. R01-HL-73699).
Competing interests None.
Ethics approval Ethics approval was obtained through the Johns Hopkins Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.
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.