INTRODUCTION
The act of smoking is a widespread phenomenon that has significant implications for public health since it is closely associated with the occurrence of major non-communicable illnesses and premature mortality on a worldwide scale1. Approximately one-third of the worldwide population engages in tobacco use, with 80% of these individuals residing in low- and middle-income nations2. Research findings indicate that there is a significant correlation between smoking and an elevated likelihood of developing heart disease, with the risk being three times higher. Similarly, the risk of stroke is 1.5 times higher among smokers. Additionally, smoking is associated with a 1.4 times increased risk of chronic respiratory illness and a substantial 12 times higher chance of developing lung cancer, as shown by several studies3-7. Each year, smoking kills 7 million globally, including about 1.2 million passive smokers2. The World Health Organization (WHO) has implemented measures to regulate and manage this highly preventable issue in public health through the Framework Convention on Tobacco Control (FCTC). The FCTC offers technical guidance to nations globally on the implementation of protective measures against tobacco consumption8.
In 2003, the Government of Bangladesh demonstrated its commitment to the global anti-tobacco movement by joining the World Health Organization Framework Convention on Tobacco Control (WHO FCTC). Subsequently, in 2005, the government further solidified its stance by enacting the Bangladesh Tobacco Control Act9. Subsequently, Bangladesh has effectively implemented measures to regulate tobacco use, as shown by the findings of the Global Adult Tobacco Survey (GATS) in 201710. Based on the survey findings, there was an observed decrease in the incidence of active smokers by roughly 8%, and a corresponding fall in the occurrence of passive smokers by an estimated 10%–30% during an 8-year period10. The previous study indicated that public transport saw the least change in terms of exposure to secondhand smoke, perhaps due to the existing legislation in Bangladesh that prohibits smoking in public transport9,10. A research investigation carried out in Dhaka city, Bangladesh, revealed a notably high prevalence of smoking among bus drivers, estimated at around 93%. Furthermore, the study provided data suggesting that stress may contribute to an increase in the adoption of this habit11. According to the study of Goon and Bipasha11, around 38% of bus drivers were seen engaging in smoking behavior during periods of traffic congestion, so contravening the legislation pertaining to tobacco use11. The presented research has prompted the need to examine the underlying issues related to the opposition to smoking restrictions, which provides a barrier to the goal of attaining a tobacco-free Bangladesh by 204010. Given the absence of prior research addressing this urgent issue, we undertook a cross-sectional study in Dhaka, Bangladesh, a city where the predominant mode of transportation for its residents is buses and human haulers. The study aimed to explore the underlying causes of non-compliant attitudes towards the smoke-free policy among the professional male bus and human-hauler drivers of Dhaka City.
METHODS
Study design, setting, and population
This was a cross-sectional study, and the data collection was conducted for 6 months (January to June 2020). The study participants included drivers who had been driving buses and human haulers for common public transportation. The study was carried out in Dhaka City, the capital of Bangladesh, covering both Dhaka South City Corporation and Dhaka North City Corporation areas. Eight major bus and human hauler stations from these city corporations were selected conveniently to reach out to the study participants for data collection (Figure 1). The following inclusion criteria have been followed to recruit the study participants for this study: 1) adult male (aged ≥18 years); 2) having a valid professional driving license authorized by the Bangladesh Road Transport Authority12; 3) smoked at least 100 cigarettes in their lifetime; and 4) having no current plan to refrain from smoking. For recruiting the study participants, the study used a systematic random sampling technique (every 3rd driver waiting in the queue of departure from the respective bus station; if anyone did not want to participate then we approached or waited for the subsequent 3rd driver) in the selected bus and human hauler stations. In total, 460 drivers were enrolled in the study and interviewed for the data collection. The proportional allocation method was used for recruiting bus and human hauler drivers based on the number of BRTA-authorized buses and human haulers roaming inside Dhaka City, and, ultimately, 406 bus drivers and 54 human hauler drivers were interviewed.
Data collection
A semi-structured questionnaire was used to collect information from the study participants. A team of trained data collectors interviewed the study participants face-to-face inside the vehicles or at sites near their vehicles, where no third person was present. A field test was carried out before the final data collection to evaluate the comprehension, feasibility, length, and appropriateness of the questionnaire. For the independent variables, comprehensive information on demographics (age, marital status, number of family members), socioeconomic status (number of bread earners in the family, monthly family income), education level, employment characteristics (licensing method, owning vehicle, salary disbursement, driving, frequency of driving, distance of daily travel, daily working hours, etc.), and lifestyle characteristics (duration of smoking in the lifetime, use of smokeless tobacco, use of alcohol, use of cannabis, practice of smoking while in the vehicle, etc.) were obtained from each participant. In terms of the outcome variable, non-compliance was defined as a participant’s confession during the interview that he had smoked inside the bus within the previous 30 days while driving or waiting for the passengers at the bus stations, whereas compliance was defined as the exact opposite. The principal investigator and co-investigators of this study supervised the whole data collection process. The data quality was assured by regular field visits and validity check interviews during data collection.
Data analysis
The R project for statistical computing software (version 4.1.3) was used to conduct the data analysis for this study13. Descriptive statistics were carried out for the participants’ demographics, employment, and lifestyle characteristics. The univariate logistic regression was carried out to create the unadjusted models, and the results were reported as crude odds ratios (ORs), 95% confidence intervals (CIs), and corresponding p-values. For the adjusted model, backward stepwise logistic regression was carried out, including all the variables used in the unadjusted models, and the results were presented by adjusted odds ratios (AORs), 95% CI, and corresponding p-values. Values of p<0.05 were considered statistically significant for both models.
RESULTS
Figure 1 shows that out of 460 drivers, only 20.9% were compliant with the smoke-free policy while on duty, indicating that they did not smoke while on public transport or in the bus station, and the rest were non-compliant (79.1%). The mean age of the participants was 30.6 years (SD=7.9), and the mean education was 5.9 years (SD=3.3).
Effect of sociodemographic characteristics on compliance with the SFP
Table 1 shows the association of sociodemographic characteristics on the level of compliance of the participants, along with the unadjusted odds ratios and their significance level with the smoke-free policy (SFP). The study showed no significant association between non-compliance with SFP and the participants’ sociodemographic characteristics. However, it is worth mentioning that non-compliant drivers constituted the largest share in each age group, with the age group of 26–40 years having the highest number (n=219, or 60.2% of the total number of non-compliant participants). Compared to the participants with <6 years of education, those with 6–10 years of education had 24% less odds (OR=0.76; 95% CI: 0.47–1.21, p=0.25) of being non-compliant, whereas participants with >10 years of education had 29% less odds (OR=0.71; 95% CI: 0.29–1.78, p=0.47). Having more than 4 members (OR=1.57; 95% CI: 0.87–2.83, p=0.13) and multiple earners in the family (OR=1.68: 95% CI: 0.80–3.55, p=0.17) showed evidence of increasing the odds of being non-compliant with the SFP.
Table 1
Effect of employment characteristics on compliance with the SFP
The study investigated the participants’ employment characteristics, such as proper licensing (obtaining the driving license in a proper channel without any means of corruption), salary type, the daily distance of travel, daily working and driving hours, vehicle ownership, and trade union membership. As shown in Table 2, several factors were significantly associated with being non-compliant with SFP. Evidence showed that having a proper license increased the odds (OR=2.3; 95% CI: 1.37–3.85, p=0.001) of being non-compliant with the SFP. Drivers with a monthly or commission-based salary agreement had a 67% less chance (OR=0.33; 95% CI: 0.15–0.70, p=0.003) of being non-compliant compared to those getting daily wages. The drivers who traveled >100 km were 1.89 times more likely (OR=1.89; 95% CI: 1.2–2.98, p=0.006) to be non-compliant compared to others. Moreover, participants working >12 hours a day were 2.71 times more likely (OR=2.71; 95% CI: 1.65–4.42, p<0.001) to be non-compliant with the SFP.
Table 2
Characteristics | Compliance level | OR (95% CI) | p | |
---|---|---|---|---|
Compliant (N=96) n (%) | Non-compliant (N=364) n (%) | |||
Obtained license properlya | ||||
No (Ref.) | 30 (33.3) | 60 (66.7) | 1 | |
Yes | 66 (17.8) | 304 (82.2) | 2.3 (1.37–3.85) | 0.001* |
Own the vehicle | ||||
No (Ref.) | 94 (20.9) | 356 (79.1) | 1 | |
Yes | 2 (20.0) | 8 (80.0) | 1.06 (0.22–5.05) | 0.95 |
Salary type | ||||
Daily wage (Ref.) | 36 (17.3) | 172 (82.7) | 1 | |
Trip basis | 45 (21.6) | 163 (78.4) | 0.77 (0.47–1.24) | 0.28 |
Monthly/commission-based | 13 (38.2) | 21 (61.8) | 0.33 (0.15–0.70) | 0.003* |
Trade union membership | ||||
No (Ref.) | 80 (21.4) | 293 (78.6) | 1 | |
Yes | 16 (18.4) | 71 (81.6) | 1.21 (0.66–2.19) | 0.53 |
Driving experiences (years) | ||||
≤5 (Ref.) | 41 (20.4) | 160 (79.6) | 1 | |
6–10 | 34 (21.5) | 124 (78.5) | 0.93 (0.56–1.56) | 0.80 |
>10 | 21 (20.8) | 80 (79.2) | 0.97 (0.54–1.76) | 0.93 |
Drive daily | ||||
No (Ref.) | 89 (20.8) | 339 (79.2) | 1 | |
Yes | 7 (21.9) | 25 (78.1) | 0.94 (0.39–2.23) | 0.89 |
Monthly driving days | ||||
≤15 (Ref.) | 60 (20.1) | 238 (79.9) | 1 | |
>15 | 36 (22.2) | 126 (77.8) | 0.88 (0.55–1.41) | 0.59 |
Daily distance travelled (km) | ||||
≤100 (Ref.) | 46 (27.9) | 119 (72.1) | 1 | |
>100 | 50 (16.9) | 245 (83.1) | 1.89 (1.20–2.98) | 0.006* |
Daily working hours | ||||
≤12 (Ref.) | 36 (35.3) | 66 (64.7) | 1 | |
>12 | 60 (16.8) | 298 (83.2) | 2.71 (1.65–4.42) | <0.001* |
Driving shift | ||||
Morning to night (Ref.) | 86 (19.9) | 346 (80.1) | 1 | |
Half day | 10 (35.7) | 18 (64.3) | 0.44 (0.19–1.0) | 0.05 |
Effect of lifestyle and behavior-related characteristics on compliance with the SFP
Table 3 shows that the lifestyle and behavior of the drivers played a major role in determining their smoking habits and compliance level. The age when the participants first started to smoke was significantly associated with their compliance level. Study participants who started smoking after the age of 18 years were less likely (OR=0.39; 95% CI: 0.19–0.81, p=0.01) to be non-compliant compared to those who started smoking at a younger age. Moreover, the habit of using smokeless tobacco (OR=3.11; 95% CI: 1.63–5.93, p<0.001) and cannabis (OR=3.66; 95% CI: 1.77–7.55, p<0.001) increased the odds of being non-compliant with SFP significantly. A considerable proportion of the drivers smoked when there was no passenger in the vehicle and during stops (OR=7.14; 95% CI: 2.08–24.51, p=0.001). In addition, those who faced difficulty refraining from smoking in forbidden places showed 8.0 times higher odds (OR=8.0; 95% CI: 1.92–33.47, p=0.004) of being non-compliant with the policy against smoking, and those who could never give up smoking first cigarette in the morning had 4.84 times higher odds (OR=4.84; 95% CI: 2.72–8.59, p<0.001). Chain smoking is an important factor in this regard. The drivers who smoked >10 cigarettes per day were 5.78 times more likely (OR=5.78; 95% CI: 3.32–10.04, p<0.001) to show non-compliance. Those who heard about the tobacco control law (OR= 0.62; 95% CI: 0.39–0.98, p=0.04) and attended the program on it (OR=0.32; 95% CI: 0.19–0.53, p<0.001) were less likely to be non-compliant with the anti-smoking policy.
Table 3
Stepwise regression
The study used the backward stepwise logistic regression model (Table 4) to identify factors with a significant effect on the compliance level of the professional drivers towards the SFP. Compared to participants getting daily wages, those who received trip-based payment had 72% less chances (AOR=0.28; 95% CI: 0.08–0.95, p=0.04) of being non-compliant. Adjusted analysis showed that those who smoked when the bus was stopped and no passenger was present had 10.23 times the odds (AOR=10.23; 95% CI: 2.44–42.89, p=0.001) of showing non-compliance. People showing particular addiction towards smoking the first cigarette in the morning had 3.91 times higher odds (AOR=3.91; 95% CI: 1.05–14.57, p=0.04) than others. Knowledge of tobacco control law also showed significant results after controlling for the factors. Participants knowing the anti-smoking law had 76% reduced odds (AOR=0.24; 95% CI: 0.07–0.81, p=0.02) of being non-compliant with SFP.
Table 4
Factors | Reference category | AOR | 95% CI | p |
---|---|---|---|---|
Income type (trip-based payment) | Daily wage | 0.28 | 0.08–0.95 | 0.04* |
Income type (monthly/commission-based) | Daily wage | 1.43 | 0.11–18.41 | 0.78 |
Driving hours in a day (>12 hours) | ≤12 hours | 2.78 | 0.85–9.06 | 0.09 |
Smokes inside the vehicle, when stopped and passengers present (Yes) | No | 10.23 | 2.44–42.89 | 0.001* |
The cigarette cannot give up (the first in the morning) | Any | 3.91 | 1.05–14.57 | 0.04* |
Knew tobacco control law of Bangladesh (Yes) | No | 0.24 | 0.07–0.81 | 0.02* |
DISCUSSION
This study aimed to measure the association between compliance with the smoke-free policy (SFP) and sociodemographic, employment, lifestyle, and behavior-related characteristics of the bus and human-hauler drivers in Dhaka city, to explain the reasons and extent of the habit of smoking among the mentioned population. Similar studies have been conducted in different populations, predominantly students; however, we could not find any other compliance study conducted on the same population as ours to compare the results. Even though no association was evident between any of the sociodemographic characteristics and non-compliance with SFP, we can safely assume that at least some of them played non-apparent roles in making the participants non-compliant with SFP. Smoking is considered to relieve stress in those who are habituated to or addicted to it. Sociodemographic conditions that possibly create stress and anxiety among the drivers could be a potential reason for the growing non-compliance. Maintaining a large family, mainly when the monthly income is low, is a cause of stress for the earning members of the household. We found higher odds of non-compliance among married drivers and those with four or more family members. Participants with more educational experience could understand the harms of smoking better than the rest, resulting in fewer chances of being non-compliant with the anti-smoking policy. The study also revealed that knowledge about the tobacco control law and attending any program significantly decrease the chances of non-compliance, possibly due to the fact that the contents of the awareness program are not appropriate for this group of the population; participants attend the program only for the refreshments and poor monitoring of the awareness program implementing organization.
Working conditions may significantly affect one’s adherence to smoking. Driving long distances daily means long working hours for the drivers. Bus and human-hauler drivers of Dhaka city, along with their helping hands, have to work from dawn till midnight most days, staying within the vehicle throughout the day, during which smoking is a distraction from the monotonous work. To them, smoking works as a stimulant and often as a relief from stress for those who are habituated to it. It may help the drivers to improve their focus and performance, as explained by Lasebikan et al.14 in their study. This could cause higher odds of non-compliance among drivers covering long distances or working for long hours at a time. Likewise, this could also be the reason why drivers with the habit of smoking inside the vehicles, either during break or working hours, showed significant association with non-compliance to the anti-smoking policy.
A few of the behaviors of smokers could be explained by a possible addiction to nicotine. Being heavily habituated or addicted to smoking compels drivers towards rule-breaking no matter how hard they try otherwise or how conscious they are about it. Smoking since early childhood and more than ten cigarettes daily could point towards a strong smoking habit. When asked, the smokers also disclosed that they found it difficult to refrain from smoking in prohibited places and hated the idea of quitting smoking. Another ongoing study by us on the same population unveiled a significant association between these behaviors and nicotine dependence, which could give rise to non-compliant behavior among smokers. Moreover, a study by Thankappan et al.15 found that those who had started smoking from teenage years had a significant association with nicotine dependence.
Other studies have provided evidence that addiction to cannabis and alcohol is frequently correlated to addiction to smoking as well16,17. In addition, our study also revealed a significant association between smokeless tobacco consumption and non-compliance towards SFP. All of these traits strongly suggest that the participants are more inclined towards smoking and less worried about breaking the rules.
Strength and limitations
The strength of this study is the large sample size of the drivers and coverage of both the City Corporation Area of Dhaka City, which clearly shows the drivers’ compliance with the smoke-free policy. The study is one of its kind in South Asia and opened a new dimension for the program implementors to decide where to intervene to mitigate the burden of smoking. One of the study’s limitations is that no female bus or human-hauler drivers were found during data collection in Dhaka City. As this is a cross-sectional study, there might be some recall bias from the participants during the interview.
Implications
The smoke-free policy compliance can be increased by taking productive merger initiatives from the National Tobacco Control Cell of Bangladesh, Bangladesh Road Transport Authority, Drivers’ Trade Unions, and the Vehicle Owners’ Association. Implementing a secured salary structure for the drivers and setting the working hours to not more than 8 hours per day should be implemented strictly, not only for increasing compliance with the SFP but also for the better health of the drivers as well as the passengers of the transport. The government of Bangladesh and NGOs should start a nicotine dependence reduction program. The initiators of the tobacco awareness campaign should be more careful while preparing their content; they should design their content considering the sociodemographic characteristics of their program participants. Therefore, intervention programs from the government and donor organizations to increase compliance with SFP are highly encouraged.
CONCLUSIONS
Based on the current study’s results, it is apparent that the majority of the public transport drivers were non-compliant. The most commonly used salary disbursement method in Bangladesh, i.e. the daily wage basis and trip-based basis payment, had a direct influence on being non-compliant with SFP among this population. Those who had a high dependence on cigarettes had features of non-compliance. Unexpectedly, those who had the knowledge of the tobacco control law, were more likely to break the rules.