Systematic review
Interventions for Tobacco Control in Low- and Middle-income countries: Evidence from Randomised and Quasi-randomised Studies
Systematic review
|Updated
We systematically reviewed the literature to identify randomized and quasi-randomised studies of interventions for tobacco control implemented in low and middle-income countries (LMIC).
Download
Key message
In several high-income countries, there has been an increase in public awareness of the harm caused by smoking tobacco, and a general decrease in smoking rates. Low and middle-income countries (LMIC) on the other hand, remain a large and vulnerable market for tobacco products. The growth in smoking rates is followed ten to twenty years later by an increase in the incidence of non communicable diseases. It is therefore important that efforts to control the consumption of tobacco in LMIC are strengthened. We systematically reviewed the literature to identify randomized and quasi-randomised studies of interventions for tobacco control implemented in LMIC.
We included 45 studies conducted in various low- and middle-income countries. The interventions were broadly on offering help to quit smoking and included pharmacotherapy, health education targeting smoking pregnant women or their husbands, or at the community or primary health care. Studies on school- based interventions and one study on warnings on the dangers of smoking tobacco were also included.
In low- and middle income countries:
- Nicotine replacement therapy and buproprion may help smokers to stop smoking and probably reduces smoking rates.
- Health education that targets smoking pregnant women probably helps them to stop smoking, and may result in one or more quit attempts or a reduction in the amount of smoking.
- We are uncertain of the effect of health education at the primary care or community level on smoking cessation; however health education may decrease overall smoking rates.
- School- based interventions probably prevent progression to regular smoking among experimenters or non smokers. These interventions may reduce overall smoking rates and improve life skills and probably improve knowledge, attitudes and beliefs about the effects of tobacco smoking. We are uncertain if school-based interventions prevent experimentation with cigarettes.
Summary
Background
In many high-income countries, there has been an increase in public awareness of the harm caused by smoking tobacco, and a general decrease in smoking rates. Low and middle-income countries (LMIC) on the other hand, remain a large and vulnerable market for tobacco products. The growth in smoking rates is followed ten to twenty years later by an increase in the incidence of non communicable diseases. It is therefore important that efforts to control the consumption of tobacco in LMIC are strengthened.
The Norwegian Cancer Society commissioned the Norwegian Knowledge Centre for the Health Services to evaluate interventions to prevent and reduce the use of tobacco in low- and middle-income countries. Based on the literature emerging from randomised and non-randomised studies carried out in these countries, this review answered the questions: Which interventions are effective in preventing the use of tobacco? For those already using tobacco products, which interventions are effective in stopping the use of tobacco?
Method
We systematically searched the CENTRAL Cochrane database for references from the Cochrane Tobacco Addiction Group Specialised Register. The specialised register at the time of the search (June 2009, updated March 2011) was populated by studies identified from MEDLINE, EMBASE, PsycLIT/PsycINFO, Science Citations Index (SCI) and Social Science Citations Index (SSCI) via Web of Science, hand searching, and Conference abstracts. In addition, we searched MEDLINE Ovid, EMBASE and PsycLIT/PsycINFO. We also searched the reference lists of all eligible articles for any additional relevant articles. Two reviewers screened references according to the pre-specified inclusion criteria listed below:
Study design: Randomised controlled trials, Quasi-randomised controlled trials (e.g. controlled before-and after studies, interrupted time-series)
Population: All people, including those that smoke or use tobacco products; or are exposed to tobacco smoke .
Setting: Low- and middle-income countries
Intervention: We used the MPOWER framework (WHO 2003) to include interventions that: M onitor tobacco use and prevention policies, P rotect people from tobacco smoke, O ffer help to quit tobacco use, W arn about the dangers of tobacco, E nforce bans on tobacco advertising, promotion or sponsorship , R aise taxes on tobacco. We included other interventions to reduce the supply of tobacco and cigarettes and interventions to prevent tobacco uptake in schools.
Comparison: No intervention, delayed intervention, general information on smoking prevention distributed to all participants, or one intervention compared to another intervention
Outcome: Primary Outcome – smoking quit rates; rates of smoking initiation among non smokers. Secondary outcomes such as changes in smoking behavior, prevalence of quit attempts, change in knowledge about smoking, change in cigarette sales, self efficacy, and adverse effects
Language: No restrictions
We extracted data from included studies and assessed the risk of bias. A meta-analysis was conducted where feasible and a narrative summary where the diversity of the included studies did not permit a meta-analysis. We used the GRADE instrument to assess our confidence in the effect estimates.
Results
Out of the 45 included studies, 26 were randomized controlled trials, 18 quasi randomized trials and 1 controlled before and after study. We found no studies that used an interrupted times series design. The studies were conducted in Asia (n=26), Europe (n=6), Latin America (n=5) and Africa (n=8).The interventions were broadly on offering help to quit smoking and included pharmacotherapy (n=7), interventions targeting smoking pregnant women or their husbands (n=3), and advise and support for smoking cessation delivered in the community or through primary care services (n=16). Other studies involved interventions among school children (n=18) and warnings on the dangers of smoking tobacco (n=1).
There was low quality evidence that nicotine replacement therapy (NRT) and buproprion are more effective than placebo to help smokers to stop smoking (3 RCTs, N=440, RR 2.03 95%CI 1.30–3.19) and probably reduces smoking rates. NRT may be more effective than naltrexone in helping smokers to quit (1 RCT, N=171, RR 7.21 95%CI 2.18–23.83). We are uncertain if NRT helps more smokers to stop smoking than clonidine (1 RCT, N=171, RR 1.85 95% 0.89–3.83). We are uncertain if NRT when combined with psychological techniques helps smokers stop smoking. We are uncertain if NRT combined with psychological techniques helps smokers to stop smoking more than psychological techniques alone (1 RCT, N=23, RR 1.83, 95%CI 0.60–5.61).
Health education that targets smoking pregnant women probably helps pregnant women to stop smoking in the short term (1 RCT, N=492, RR 1.80 95%CI 1.21–2.67) and may result in one or more quit attempts or a reduction in the amount of smoking. We are uncertain if health education targeting smoking husbands of pregnant women helps the husbands to stop smoking (1 RCT, N=758, RR 1.43 95%CI 0.77–2.66). However, health education may result in one or more quit attempts or a reduction in the amount of smoking. We are uncertain of the effect of health education at the primary care or community level on smoking cessation (4 RCTs, N=836, RR 2.14 95%CI 0.77–5.95; 5 non RCTs, N=40854, RR 1.06 95%CI 0.86–1.31). However, health education may decrease overall smoking rates, and may help increase self efficacy, and improve knowledge and attitudes. We are uncertain if high intensity health education is more effective than low intensity education for smoking cessation. High intensity education may however lead to a larger decrease in smoking rates than low intensity education.
School- based interventions probably prevent progression to regular smoking among experimenters or non smokers. These interventions may reduce overall smoking rates and improve life skills. School based interventions probably improve knowledge, attitudes and beliefs about the effects of tobacco smoking. We are uncertain if school-based interventions prevent experimentation with cigarettes.
Discussion
Our findings indicate that NRT or buproprion as well as health education for pregnant women may help smokers in LMIC to quit. However these findings were from few, small studies and it is not clear how long these effects last. We are uncertain of the effect of health education delivered at the primary or community level for smoking cessation. This finding requires cautious interpretation as the studies were heterogeneous and overall reported few events. These findings may be seen to support the notion that health education needs to be carefully orchestrated and directed at different levels in a relevant conceptual model. The findings from the school based interventions could partly be explained by the fact that the included studies were mostly among high school students, a period when experimentation with tobacco is likely to occur. We did not assess if interventions that targeted younger students had a better outcome than those targeting older students who may have already experimented with cigarettes or become regular smokers.
Conclusion
In low- and middle income countries, nicotine replacement therapy and buproprion may help smokers to stop smoking and probably reduces smoking rates. Health education that targets smoking pregnant women probably helps them to stop smoking, and may result in one or more quit attempts or a reduction in the amount of smoking. We are uncertain of the effect of health education at the primary care or community level on smoking cessation, but this may decrease overall smoking rates.
School- based interventions probably prevent progression to regular smoking among experimenters or non smokers. These interventions may reduce overall smoking rates and improve life skills. School based interventions probably improve knowledge, attitudes and beliefs about the effects of tobacco smoking. We are uncertain if school-based interventions prevent experimentation with cigarettes.
However the evidence base is not very strong as most of the included studies were small, implemented over short periods and at times addressed different questions. There is a need for more rigorous studies conducted in LMICs, perhaps with a particular focus on delivery strategies of therapies that have been successful in high income settings. Some interventions such as those targeting the supply of tobacco, enforcing bans on tobacco advertising or raising taxes require further evaluation, especially in LMIC where the legislation and enforcement of tobacco control varies widely.