Exposure to Second-Hand Smoke in Public Places and Barriers to the Implementation of Smoke-Free Regulations in The Gambia: A Population-Based Survey

Introduction: Second-hand smoke is associated with more than 1.2 million deaths per year among non-smokers. Smoking in public places is prohibited in The Gambia but there is no information on the level of exposure to second-hand smoke among adolescents and adults 15–64 years. The aim of this study was to assess the level and predictors of exposure to second-hand smoke in public places and compliance with smoke-free regulations in The Gambia. Methods: A population-based survey was conducted in an established Health and Demographic Surveillance System (HDSS). A total of 4547 participants (15–64 years) from households within the Farafenni HDSS were interviewed at their homes but only 3343 were included in our analysis. Factors associated with exposure to second-hand smoke in public places were assessed by three different multivariable regression models. Results: Exposure to tobacco smoke in public places was high (66.1%), and higher in men (79.9%) than women (58.7%). Besides being male, less education, lower household income, urban residence and not aware of smoke-free regulations were strongly associated with exposure to second-hand smoke. Conclusion: Despite existing smoke-free regulations, reported exposure to second-hand smoke remains high in public places in The Gambia. The Ministry of Health should continue to strengthen their advocacy and sensitization programs to ensure smoke-free regulations are fully implemented. Some population subgroups are at a higher risk of exposure and could be targeted by interventions; and settings where these subgroups are exposed should be targeted by enforcement efforts.


Introduction 40
Exposure to second-hand tobacco smoke (hereafter referred to as 41 second-hand smoke) is associated with more than 1.2 million deaths per 42 year worldwide among non-smokers; about 11 million disability-ad-43 justed life years (DALYs) are lost due to second-hand smoke exposure, 44 with 61% of them occurring among children [1]. In 2004, 40% of children, 45 33% of men and 35% of women in 192 countries were exposed to second-46 hand smoke indoors [1]. There is no safe level of exposure to tobacco 47 smoke [2]. Article 8 of the World Health Organization Framework Con-48 vention on Tobacco Control (WHO FCTC) recommends eliminating ex-49 posure to tobacco smoke in workplaces and public places, both indoor 50 and outdoor, and this includes public transport [2]. Article 20 (Research,51 surveillance, and exchange of information) stresses the importance of sur-52 veillance for assessing the tobacco epidemic, monitoring tobacco use and 53 the related prevention policies [2]. 54 The Gambia's 2017 Global Youth Tobacco Survey (GYTS) found that 55 61.8% of students aged 13-15 years were exposed to second-hand smoke 56 in enclosed public places [3]. A recent survey in secondary schools also 57 reported a high prevalence of second-hand smoke exposure among 58 young people aged 12-20 years in The sample size was estimated assuming the prevalence of smoking was 115 20%, based on a national prevalence of 15.6% [11], a margin of error of 116 5%, and a design effect of 1.5 [12]. In addition, since the response rate 117 from the last national STEP survey in The Gambia was 79% and to adjust 118 for non-response and increase the power, the sample size was increased 119 by 20%. This resulted in a sample size of 4,613 which was increased to 120 6,000 to allow for emigration and non-response.

121
A multistage sampling technique was applied using the three level 122 of residence within the HDSS (urban, semi urban, and rural). At the first 123 stage, a total of 42 urban, semi and rural villages were randomly selected. 124 Stage two involved sampling of households within the selected villages. 125 There are 3727 compounds and 6037 households within the Farafenni 126 HDSS. The average household size is 10 and on average there are five 15-127 64 years old individuals per household; a total of 2532 households were 128 randomly selected from the HDSS database. In stage three, one to five 129 people between the ages of 15-64 years were initially randomly selected 130 per household proportional to the size of the household, for a total of 131 6000. Because of seasonal migration, some selected individuals were un-132 available. Therefore, one month after starting the data collection process, 133 2856 additional individuals were randomly selected, resulting a total 134 sample of 8856 individuals. We were only able to interview 4547 but our 135 analysis was restricted to 3343 participants with complete information. 136

Exposure and outcome measures 137
The outcome variable was self-reported exposure to second-hand 138 smoke in public places. For indoor exposure, a participant was defined as 139 exposed if he/she responded yes to the question: Did anyone smoke in your 140 presence, inside any enclosed public place, other than your home (such as school, 141 shops, restaurants, police stations, hospitals, public institutions/offices, shopping 142 malls, video clubs/movie theatres etc) during the past 30 days? A similar ques-143 tion relating to outdoor public places was asked: Did anyone smoke in your 144 presence, at any outdoor public place other than your home (such as car park, 145 football field, markets, bantabas, wrestling arenas) during the past 30 days? 146 The exposure variables were demographic variables including age, 147 gender, ethnicity, residence, education, marital status, and household in-148 come. We also included awareness of regulation on smoking in public 149 places which was assessed by asking the following question: Are you 150 aware of the law prohibiting smoking in public places in The Gambia? 151

Data management and analysis 152
Immediately after the data collection, the data were cleaned, coded, 153 and analysed using Stata 16. Descriptive statistics was used to describe 154 the socio-demographic characteristics of the respondents, the level of 155 compliance with smoke free regulations and the prevalence of exposure 156 in both indoor and outdoor public places. We estimated the prevalence 157 of exposure to second-hand smoke in both indoor and outdoor public 158 places by our variables of interests as highlighted in section 2.4. Multivar-159 iable logistic regression analysis was performed to determine the socio-160 demographic factors associated with exposure to second-hand smoke in 161 indoor and outdoor public places. We did not conduct multiple imputa-162 tion for the missing data and hence conducted complete case analysis in 163 all our regression models. Because of inherent difference in policy impli-164 cations, three separate models for indoor, outdoor, and indoor and/or 165 outdoor public places were run. We used confidence intervals and p-val-166 ues in both the descriptive and multivariable regression analysis to deter-167 mine level of significance. 168

Results 169
Out of the 8856 sampled participants, 4,547 responded to the survey 170 while 16 declined to participate. The rest were either absent at the time of 171 our visits or were not reached due to the early suspension of fieldwork 172 activities because of COVID 19. Our analysis was restricted to 3,343 indi-173 viduals with complete information on the key variables. The mean age 174 was 34.6±14.5 years; more than two thirds (64.9%) of the respondents 175 were females, 42.0% lived in urban areas, 48.9.% in rural areas and 9.1% 176 in semi-urban areas ( The prevalence of exposure to second-hand cigarette smoking was 192 higher ) in outdoor (61.3%) than indoor (52.8%) public places. Overall,193 66.1% of the participants were exposed, with little variation by age (Fig-194 ure 1). Levels of exposure were significantly higher in males than in fe-195 males in all settings (indoor exposure 63.7% in males vs 46.9% in females, 196 p<0.001; outdoor exposure: 75.6% in males vs 53.5% in females, p< 0.001).

197
In addition, exposure was higher among urban than rural residents, both 198 indoors (

Factors associated with exposure to second-hand smoke in indoor and 210
outdoor public places 211 Gender, ethnicity, residence, education, income level and level of 212 awareness of the regulation on smoking in public were strongly associ-213 ated with exposure to second-hand smoke in public places (Table 2). In 214 model I (exposure to second-hand smoking in indoor public places), 215 males were significantly more likely to be exposed than females 216 (AOR=1.78 95% CI: 1.51-2.10). Moreover, urban residents were almost 217 three times more likely to be exposed than rural residents (AOR=2.68, CI: 218 2.24-3.21). Those with higher education and those from households with 219 higher income were less likely to be exposed in indoor public places (Ta-220 ble 2). The odds of exposure to second-hand smoke were significantly 221 higher among those unaware of smoking regulations (AOR =1.46. CI:1.25-222 1.71). Models II (outdoor public places) and III (both indoor and outdoor) 223 produced similar results. 224

226
Reference= reference category in regression n analysis 227

Discussion 228
Exposure to second-hand smoke in public places, despite the exist-229 ing tobacco control regulations that prohibit smoking in public, remains 230 high in The Gambia. Awareness of such regulations was moderate while 231 it was low for the regulations protecting children. This can serve as a bar-232 rier to the implementation of tobacco control regulations. Interestingly, 233 most respondents, including smokers, supported the law against smok-234 ing in public places and other smoke free regulations. The factors strongly 235 associated with exposure to second-hand smoke in both indoor and out-236 door public places included male gender, lower education, lower house-237 hold income, urban residence, ethnicity, and lack of awareness of smoke-238 free regulations. 239 Exposure to tobacco smoke was high and, although previous sur-240 veys focused on different age groups [3,4], exposure was similarly high 241 in this study. A cross sectional survey in secondary schools in The Gam-242 bia found that 59.2% of 12-20 years old were exposed to second hand-243 smoke in indoor public places and 61.4% in outdoor public places [4].

244
Preliminary analysis of the 2017 GYTS revealed that 61.8% of students 245 (13-15 years old) were exposed to second hand tobacco smoke in enclosed 246 public places [3]. In our study, women were much less exposed in both 247 indoor and outdoor public places, a finding that is similar to results from 248 South Africa where females were less likely to be exposed in all locations 249 except their homes [13]. In India, males were also more likely to be ex-250 posed in public places, but females were more likely to be exposed at 251 home [14].

252
In a previous study in The Gambia, higher maternal education and 253 living with parents was significantly associated with lower risks of expo-254 sure while higher level of education among fathers was associated with 255 increased odds of exposure in public places [4], suggesting highly edu-256 cated mothers protected their children from exposure to cigarette smoke. 257 In Farafenni, higher education was also a strong protector of being ex-258 posed to second-hand smoke, suggesting increasing public awareness is 259 a good strategy. Our findings also concur with a study in Mongolia and 260 China where among non-smoking women, those with a higher education 261 were less likely to be exposed to second-hand tobacco smoke compared 262 to those with a lower level of education [13].

263
Individuals with a lower income were more likely to be exposed to 264 tobacco smoke in both indoor and outdoor public places. Lower socio-265 economic status was associated with exposure to second-hand smoke in 266 households in the Republic of Korea [15]. A study on exposure to domes-267 tic second-hand smoke in 15 low-and-middle-income countries reported 268 that exposure to second-hand smoke, in most countries, was higher 269 among the socioeconomically disadvantaged [16]. These findings on so-270 cio-economic status, fairly consistent in several studies, including the cur-271 rent research, reflect the greater risk of second-hand smoke exposure 272 among the less affluent because they are more likely to work or live in 273 settings with smokers and, for public places, poorer enforcement of exist-274 ing regulations. 275 Women were less likely to be exposed to second-hand smoke than 276 men. As suggested in the literature, women have a very important role in 277 protecting themselves, children, and other family members from expo-278 sure to cigarette smoke [17]. They may also be less likely to visit public 279 places (such as workplaces) where smokers are present. Women can be 280 powerful advocates for health protection measures and should be in-281 volved in tobacco control at local and national level. The Ministry of 282 Health of The Gambia has a long-standing tradition of using organised 283 women groups and traditional communicators (most of them women) in 284 transmitting key health messages to the population. This could be further 285 strengthened and integrated as part of the national campaign to sensitise 286 communities on the dangers of second-hand tobacco smoke and the to-287 bacco control regulations. 288 Education was a strong protector of being exposed to second-hand 289 smoke which may suggest that measures raising awareness may have an 290 important role to play, although it is also likely that more highly educated 291 individuals are less likely to be living or working with other smokers. The 292 results of the multivariable regression analysis show that non-smokers 293 unaware of the regulations are more likely to be exposed to cigarette 294 smoke in public places. Our findings also suggest that those with a lower 295 level of education are more likely to be exposed to cigarette in public 296 places compared with participants with a high school or college/univer-297 sity education. As people's knowledge of and belief in the adverse health 298 effects of tobacco use increase, the likelihood of using tobacco decreases 299 and their support for protective policies increases [18]. Members of a 300 well-informed society are also more likely to take up and support policies 301 for a tobacco-free environment. Therefore, the Ministry of Health and 302 partners should strengthen their community sensitisation programs to 303 raise awareness of smoke free regulations. 304

Strengths and limitations 305
This is the first study on exposure to second-hand smoke in public 306 places among adults in The Gambia. The study has identified both the 307 population subgroups at a higher risk of exposure and settings where 308 these subgroups are being exposed and thus can be used to guide policy 309 and enforcement efforts. These research findings can be used to inform 310 the development of effective policies and strategies for the control of to-311 bacco and non-communicable diseases in The Gambia, which can be 312 adapted by other low-and middle-income countries. They can also be 313 used as a baseline to monitor over time compliance with the smoke free 314 regulations. 315 There are some limitations, including the low male response rate, 316 mainly due to high seasonal migration, which may have underestimated 317 overall exposure as men are more likely to smoke and to be exposed to 318 environmental tobacco smoke. The early suspension of field activities due 319 to the COVID 19 pandemic is another limitation as the final sample size 320 was smaller than expected. In addition, the study was based on survey 321 interview data and all the information on exposure to second-hand to-322 bacco smoke was self-reported. An objective measurement of exposure to 323 second-hand smoke, e.g., urine and/or blood analysis to assess cotinine 324 (nicotine metabolite) was not performed. Participants can therefore un-325 derreport or over report their exposure status. Future studies in The Gam-326 bia and in other African countries should explore objective measurement 327 of exposure. 328

Conclusion 329
Our findings illustrate that exposure of both smokers and non-smok-330 ers to second-hand smoke is high in The Gambia despite smoke free reg-331 ulations, indicating there are still gaps in tobacco control, especially on 332 the implementation of smoke free laws. People aware of the smoke free 333 regulations and women with a higher level of education were less likely 334 to be exposed to second-hand smoke in both indoor and outdoor public 335 places. The Ministry of Health and its partners, including the members of 336 the Tobacco Control Committee, should continue to strengthen their ad-337 vocacy and sensitisation programs to ensure smoke free regulations are 338 fully implemented and enforced.