Teachers Perception on Tobacco Free School in Bangladesh

Introduction: Comprehensive tobacco-free school policy benefits everyone by reducing exposure to second-hand smoking and creating a system that reinforces tobacco-free norm and attitudes, which in turn affect tobacco use having an impact on the health of the community. Objective: This study aims to assess the perception of teachers on tobacco free school environments and factors supporting it. Methods: A two-stage cluster sampling technique was used for the selection of schools with selection probability proportional to enrolment size followed by stratified random sampling of government and private schools. Data were collected from 559 secondary school teachers using a structured questionnaire. Results: Estimation indicated that 48.3% (95% CI: 44.5%, 52.6%) of the teachers were less supportive and 51.7% (95% CI: 47.4%, 55.5%) were more supportive for smoking free school. Logistic regression analysis revealed that non tobacco user teachers were more likely to be supportive (OR=1. 891, 95% CI: 1.197, 2.986) for tobacco free school. However, no statistically significant association was found between perceived supportive tobacco free school and age, sex, level of education, type of school and family size (p>0.05). Exposure to second hand smoking and curriculum content of tobacco issues were not significantly related with tobacco free school. Conclusion: Comprehensive school based programme with participation of school personnel and community can effectively implement tobacco free school programmes.

of smoking among adolescence 5 .Study in Bangladesh found that smoking by teachers appeared to be a strong predictor of students smoking behaviour 6 .Students smoking behaviour is closely linked to the school environment especially the teacher's use of tobacco.Although Article 16 of the Framework Convention on Tobacco Control (FCTC) exists to prohibit sale of tobacco products to minors and Article 8 allows countries to enact laws to protect public from exposure to tobacco smoke which covers schools 7 .However, enforcement of law is weak, especially in schools, where most countries have designated as a tobacco free premises, having a written school policy on teacher smoking within and outside school premises is almost non-existent.Teachers may be prohibited from smoking within the class or within the school, but it does not discourage them outside the school premises.A Malaysian survey in 1994 reported that 20% of high school teachers were current smokers 8 .Tobacco free school policy exists in many countries and it is vitally important that managers of schools and teachers i m p l e m e n t this policy s t r i c t l y .Unfortunately, teachers themselves, who are role models for their students, smoke and many admit smoking on school premises.A study in Denmark found that almost 90% of students have seen a teacher smoking within school premises although a school no-smoking policy exists 13 , butis poorly enforced.Tobacco use among teachers in school influences smoking behaviour of school students and having a strict tobacco-free policy in school reduces smoking among school students 9,14 .Teachers play an important role in implementing this policy and yet there is a paucity ofresearch on the teacher's perceptionof this policy.The objective of this study was to examine the determinants of teacher's perception on tobacco free school policy in Bangladesh.

Materials and Methods:
Study design and sampling procedure: This was a school-based cross-sectional study conducted in all administrative divisions of Bangladesh.Amultistage cluster sampling technique was adoptedto select 30 districts using probability proportionalto enrolment(PPE), i.e. schools with a high numberof students were more likely to be selected thanschools with a small number of students.In thesecond stage, all the schools in the district weredivided into government and private, andone school from each selected district was then selected randomly.All the available teachers on the day ofthe survey were included in the sample.Instruments development and data collectionprocedure: A structured questionnaire was developed consistingof socio-demographic characteristics, smoking habits and its pattern, perception on smoking free school, exposure to second hand smoking, tobacco related curriculum school.Each item had a multiple-choice question with asingle answer.There was no skipping or branchingpattern of any question.Data werecollected by using an anonymous self-administeredquestionnaire.Informed verbal consent from theschool authority was obtained after explaining thepurpose of the study.The questionnaire wasdistributed to the teachers after explaining thepurpose of the study and the instructions to fill inthe questionnaire.1).

Perceptions on strategies for reducing smoking
Table 2 refers to the teacher's perception on different strategic options for smoking free school.There were ten questions with four answer options.Each question was scaled at 1 for 'strongly agree', 2 for 'agree', 3 for 'disagree' and 4 for 'strongly disagree'.
Positive or supportive attitude toward smoking free school was indicated by higher score.However, the last question of imposing penalty for sale of tobacco product in the campus was reverse score that is 4 marks were given for 'strongly agree' and so on.Eighty to 90% of the respondents were strongly disagreeing on smoking in the hostel (89.2%), smoking in the dining hall (88.4%), smoking by less than 18 years (84.9%),smoking in all the buildings in schools(84.8%),smoking in the canteen (82.9%) and sale tobacco product in the schools(80.4%).Seventy to 80% of the respondents strongly disagreed for giving cigarette, bidi or tobacco product to persons aged less than 18 years (78.2%),tobacco advertisement in the educational campus (77.3%) and participation and donation to fund for school cultural show, prize etc, by the tobacco company (68.9%).Onlyhalf (51.5%) of the teachersstrongly agreesfor imposing penalty for sale of tobacco product in school campuses'.After excluding missing data, log transformation was done for skewed distributions of score on perception.Then the data were classified into "less supportive" for less than 50 Th percentile and "supportive" for score more than the 50th percentile.Estimation indicated that 48.3% (95% CI:44.5%,52.6%) of the teachers were less supportive and 51.7% (95% CI: 47.4%, 55.5%) were strongly supportive for tobaccofree school.

Factors influencing smoking free school environment: Logistic regression analysis
To identify the factors influencing perceived smoking free school environment, a logistic regression model was fitted with supporting smoking free school (dichotomous) being the dependent variable, and the selected independent variables on socio-demographic characteristics, environmental variables showing statistical significance in the chi-square analyses.Although several factors were significantly associated with the supportive smoking free environment in bi-variate analysis, in logistics analysis, only current smoking were appeared to be an important influencing factor for smoking free school environment (p<0.05).The result showed that nonsmoker teachers were more likely to be supporting (OR=1.891, 95% CI: 1.197, 2.986) for smoking free school.However, no statistically significant association was found between perceived supportive smoking free school and age, sex, level of education, type of school and family size (p>0.05).

Discussion:
Enforcement of tobacco free premises in any environment has been shown to reduce the prevalence and consumption of tobacco use 15,16 .This directly reduces environmental tobacco smoke (ETS) and, its health effects.Having a tobacco free school policy not only reduces prevalence of tobacco use and increases cessation rates 17 , but also has an additional health effect of reducing environmental tobacco smoke exposure.This reduction in environmental tobacco smoke has been shown to improve health related quality of life among non-smokers 18 .In our study, 32.8% respondents have encountered situations where others have smoked in their presence in schools in the past 7 days, but this was not significantly associated with support for a tobacco free school policy.
In bi-variate analysis, it was found that female teachers showedstrongly supportive attitude towards smoking free school compared to their male counterparts.This is reflective of tobacco use among females in this study group, with only 2.4% of female teachers being tobacco users.This is unlikely related to gender specific issues but likely due to support for tobacco free school policy among non-tobacco users.
A study among college students in the United States reported that non-tobacco users were more supportive of a tobacco-free school policy than tobacco users 19 .
In another study, current smokers were only 0.30 times as likely to favour smoke-free school grounds (OR = 0.30, 95% CI 0.20-0.46)compared to nonsmokers 20 .A similar result was observed in our study, indicating that non tobacco users were 1.891 times higher supportive for tobacco free school policy.Teachers are trained to educate the students.They are the effective socio-behavioural tool to transmit their knowledge in order to change the lifestyle of students.
Higher level of education enables one to have greater access or awareness of knowledge of the health effects of tobacco use and hence support for a tobacco free school.Although a direct causal effect linking higher education and support for tobacco free school policy remains to be clarified, some studies have iterated an indirect causality, where higher education increases knowledge of the health effects of tobacco use 21,22 .Level of education is also related to socioeconomic factors.Among teachers, a higher education level translates to have a better teaching post with better salary and income, leading to a higher socioeconomic status.The 2002 International Tobacco Control (ITC) Four Country survey found that awareness on health effects of smoking was better in those from a higher socioeconomic status 21 .However, our study did not find any statistically significant supportive attitude towards tobacco free school (p>0.05)among the highest educated teachers.This study was not without its limitations.Although the respondents were anonymous, many still fear repercussions if they did not support tobacco free school policy as it is advocated by the Government of Bangladesh, hence the data were skewed towards support for this policy.This is also reflected by the low figure of current tobacco user among teachers, which may be an under reported figure.Future studies are needed to confirm these findings.

Conclusion:
De-normalizing tobacco use through positive role modelling is one important step mediated by school personnel.Though, the study finding revealed that only non-smoker teachers are more likely to be supportive tobacco free school, but other demographic, socio-cultural and environmental factors might be associated with smoking in school, that need to be explored by further study.Our study recommended that prohibition of tobacco use should be comprehensive involving all the school personnel.Effectiveness in preventing and controlling tobacco use can be enhanced by increasing the consistency and visibility of its enforcement of law and community participation.
o r c e d school policy significantly reduces the p r e v a l e n c e of daily smoking in s c h o o l s 9 -11 .T o b a c c o free school policy also i n f l u e n c e s t e a c h e r ' s behaviour in school.It was reported that in schools which apply non-smoking r e g u l a t i o n or advocate a n t itobacco, the p e r c e n t a g e of teachers s m o k i n g in front of pupils is significantly l o w e r 1 2 .

Table 2 Percentage distributions of teachers by perceptions on strategies for tobacco free school
Incomplete and inconsistent data were discardedand were not included for final analysis.Finally,data from 559 respondents were used for analysis.Data analyses were done using SPSS version 19 software (Statistical Package for Social Sciences).Simple frequencies, bi-variate, multivariateanalyses were done as appropriate.Statisticalsignificance was