Tobacco Use and Smoking Cessation Practices among Physicians in Developing Countries: A Literature Review (1987–2010)

Physicians have a key role to play in combating tobacco use and reducing the tobacco induced harm to health. However, there is a paucity of information about tobacco-use and cessation among physicians in developing countries. To assess the need for and nature of smoking cessation services among physicians in developing countries, a detailed literature review of studies published in English, between 1987 and 2010 was carried out. The electronic databases Medline and Pub Med were searched for published studies. The findings show that there are regional variations in the current smoking prevalence, quitting intentions, and cessation services among physicians. Smoking prevalence (median) was highest in Central/Eastern Europe (37%), followed by Africa (29%), Central and South America (25%) and Asia (17.5%). There were significant gender differences in smoking prevalence across studies, with higher prevalence among males than females. Smoking at work or in front of patients was commonly practiced by physicians in some countries. Asking about smoking status or advising patients to quit smoking was not common practice among the physicians, especially among smoker physicians. Organized smoking cessation programs for physicians did not exist in all of these regions. This review suggests that while smoking of physicians varies across different developing regions; prevalence rates tend to be higher than among physicians in developed countries. Quitting rates were low among the physicians, and the delivery of advice on quitting smoking was not common across the studies. To promote tobacco control and increase cessation in populations, there is a need to build physicians’ capacity so that they can engage in tobacco use prevention and cessation activities.


Introduction
Tobacco smoking is the leading cause of preventable death world-wide. Currently, tobacco is responsible for 5 million deaths annually. This annual toll is expected to increase to 10 million within the next 20-30 years, with 70% of deaths occurring in developing countries (World Health Organization (WHO)) [1]. Discouraging initiation and reducing the use of tobacco is currently among the most important public health strategies that countries can implement. Because of their close contact with the public, and the fact that physicians are role models, opinion leaders and often decision makers for healthcare policy, physicians can play a key role in efforts to reduce smoking. Research has shown that smoking cessation interventions by physicians are both efficacious and cost-effective with regard to patient smoking outcomes [2]. Smoking cessation in large number of smokers, as could be delivered by physicians, could reduce the epidemic of tobacco deaths in the next few decades [3].
In developed countries physicians play a key role in promoting smoking cessation [4], and recommendations to engage physicians more actively in the smoking cessation effort go back decades [5]. In developing countries, however, physicians are less involved in tobacco control and smoking cessation efforts [6], perhaps because a substantial proportion of physicians in these countries are smokers themselves. In a recent review, Abdullah et al. reported that smoking prevalence among Chinese male physicians ranged from 26% to 61% [7]. A high smoking prevalence was also reported among physicians in several other developing countries. For example, the prevalence was as high as 40% [8] among Bosnian and 50% [9] among Pakistani male physicians. During the last two decades, a number of studies have addressed tobacco use among physicians in developing countries, but many have been published in local journals and were not accessible internationally. At the same time, there have been significant changes in the tobacco control policies across the nations during the last decade. An understanding of the differences in the findings from earlier studies to that of newer studies would provide insights about the changing trajectories of tobacco use by and cessation counseling delivered by physicians in developing countries.
To provide a summary of research on tobacco use and cessation counseling practices involving physicians in developing countries, we reviewed literatures published during the last three decades.

Methods
A literature review of all articles published in peer-reviewed journals that included data on prevalence of physician smoking in developing countries was conducted. Developing countries (low and middle income economies), as defined by the World Bank [10], were grouped under four regions: Africa, Asia, South/Central America and Central/Eastern Europe. A significant amount of literature was available from developed nations but was not included in this review because of the focus on developing countries. The literature review began with searches on Medline and PubMed using four combinations of medical subject headings (MeSH): physician (including subheadings: physician's role; physician's practice patterns; and -physicians‖ as a keyword) and tobacco (including: tobacco, smokeless; tobacco use cessation; and -tobacco‖ as a keyword); physician and smoking (including: smoking cessation and -smoking‖ as a keyword); -doctor‖ (as keyword only) and smoking; and -doctor‖ and tobacco. We restricted our review to materials published in the last 33 years (1987-2010), to provide a review since the tobacco use related information was available in the literature.
All English language articles for which a full-text version could be found were included in the review if they reported prevalence of smoking among physicians or health care workers. Abstracts for publications for which full-text versions were unavailable were also combed for prevalence data. For foreign language articles, an attempt was made to acquire English versions by contacting the corresponding authors. If no response was forthcoming, we similarly reported the prevalence data from the available English abstract. Where no English abstract was available, we excluded the article after an attempt to contact the corresponding author. We also scanned the references of articles reviewed for additional materials, but this did not yield many articles that met the criteria of our study. Manuscripts were arranged by region and in descending order of year of publication. We used publication date rather than study date for consistency, since publication dates were always available and study dates were not. Although a thorough review of grey literature was not conducted, we did use the WHO country reports providing prevalence data on physician smoking. Studies that were conducted among medical students were not included. Also studies that were conducted among healthcare workers but did not include physicians were excluded.
We assigned each manuscript a reference number based on the criteria mentioned above. We included the response rate for the studies where available and rounded this to the nearest whole number. We also provided sample size, indicating the total number of respondents (not the number originally surveyed). We provided smoking prevalence, with gender breakdown where available, and also information about the number of cigarettes smoked per day. Where physician smoking prevalence was not reported separately from other health workers, we reported the combined figure. We also noted other key findings as appropriate.

General Descriptions
We identified 59 full-text published studies that fit our criteria. Only one study [11] addressed more than one country in the same paper. We reviewed 10 additional abstracts where full-text English articles were not available. We also included the data from 10 countries assessed in the WHO's Tobacco or Health Report, and 13 WHO country fact sheets. In total, we had 92 separate sources of data on smoking among physicians. The regional breakdown for all sources was as follows: nine from Africa, 28 from Asia, 31 from Central/Eastern Europe, and 24 from Central/South America. Sample sizes ranged from as high as 20,848 [12] to as low as 63 [13] across studies. As shown in Table 1, most of these studies were cross-sectional in nature and used postal or self-administered questionnaire surveys to collect data. Only one study, in Russia, conducted in-person interviews [14]. For twelve sources, methods of data collection were unreported or unclear.

Prevalence of Tobacco Use
Overall smoking prevalence rates across studies ranged from as high as 63% in the Philippines [1] and 62% in Mexico [15], to as low as 3% in Nigeria [16]. In eight studies, the overall smoking prevalence rate was under 10% and in four studies it was 50% or above.

Gender Difference
In fifty seven studies that reported gender-specific tobacco use prevalence, smoking prevalence was higher among male than female physicians. In addition, at least in Europe, male smokers smoked more cigarettes per day than their female counterparts [8,17]. Among both African and Asian physicians, the gender difference was especially pronounced, with most studies reporting the smoking prevalence rate among females as below 5%, while males smoked as much or more than males in other regions. In Eastern and Central Europe, females smoked notably less than males in most of the studies. However, in a 1996 study, the daily smoking rate was higher among Bosnian female physicians (55%) than their male counterparts (50%) [18]. In Central/South America, the female rates, although still lower, track more closely with male rates. In a Mexican study, females actually smoked more than males: 39% vs. 22% [19].

Number of Cigarettes Smoked per Day
Twenty articles reported the number of cigarettes smoked per day. However, methods varied: some reported average daily cigarette consumption while others reported a range (e.g., <10 cigarettes per day, 10-20 cigarettes per day, etc.). Others studies categorized smokers as light, moderate or heavy, with varying definitions, making comparisons among studies somewhat difficult. Some trends do emerge, however, from the data, especially in regional comparisons.        Based on two Asian studies, one third [42] and one half [45] of subjects smoked more than 20 cigarettes per day, while two other studies reported a daily rate of 10-20 cigarettes [1,43]. Another study in China reported an average intake of 10 cigarettes per day [41]. Based on two studies in Africa, more than half of the physician smoked more than 15 cigarettes per day [28,30]. In most studies in Central and Eastern Europe, physicians smoked between 10 and 20 cigarettes per day. A 2006 study in Ukraine [27] found that 66% of physicians smoked less than a half-pack per day. Central and South American physicians had notably lower rates of smoking: in all but one study, the average numbers of cigarettes smoked daily were below 10. In one Brazilian study [22], the rate was 14 cigarettes per day.

Regional Variations in Smoking
The smoking prevalence rate among physicians in Central and South America ranged from 6.4% [25] to 62% [65] (Median: 25%). The prevalence in Central/Eastern Europe ranged from 14% among Ukrainian physicians [27] to 59% in a 1993 Latvian study [1] (Median: 37%). In Asia the prevalence rate ranged from as low as 9.2% among Lao physicians [31] to as high as 50% among Mongolian military doctors [1] (Median: 17.5%). Prevalence rates among African physicians who smoked ranged from 3% at two Nigerian teaching hospitals [16] to 40% in both Egypt and Algeria during a multi-country survey [11] (Median: 29%). Again, the lack of continuity in methods and the broad timeframe makes regional comparisons tenuous. Additionally, prevalence of smoking by physicians may differ heavily between regions of the same country, so single data points may sharply under or over-estimate country prevalence [25].

Smoking in Front of Patients
The proportion of physicians who smoked in front of patients or at the clinic/hospitals was as high as 70% among physicians in Senegal [28], 66% in Costa Rica [63], and 50% among Egyptian physicians [15]. In a study by Jiang et al. [40], one third of Chinese physicians smoked in front of patients and almost all smoked during their work shift. A few studies showed physicians with more restraint, with a majority of smoker physicians in some regions avoiding smoking in front of patients or banning it in the workplace [29,55,64].

Quitting Smoking among Physicians
Several studies reported quit rates, which were mainly natural quit rates (i.e., cold turkey) among physicians (i.e., former smoking, defined as smoking in the past but not at the time of survey). Twenty-five studies reported the prevalence of former smokers across studies, as shown in Table 2. Former smoking rates ranged from as low as 2% among Chinese physicians [41] to as high as 40% among Costa Rican physicians [63].

Intention to Quit and Past Quitting Attempt
Twenty-two articles reported on desire to quit or previous quit attempts among physicians who smoked. As many as 60% of Indian physicians who smoked had tried to quit in the past year [20]; one-third of Chinese physicians who smoked wanted to quit [42] and 80% of South African physicians thought of quitting at some point in the future [29]. In Central and Eastern Europe, between 50% [48] and 62% [56] of smoking physicians had tried to quit, and 68-75% expressed an intention to quit smoking [52,55]. In Latin America, 73% of physicians who smoked had tried to quit in the past [19,22] and 76% expressed an intention to quit [61].

Smoking Cessation Counseling Practices among Physicians
In most studies where the question was raised, physicians and other health care workers felt that helping patients to quit was part of their job [13,29,47]. In Bosnia-Herzegovina [8] and Senegal [16], physicians felt that patients were likely to listen to their smoking cessation advice. However, in Turkey [56] and China [42], the physicians saw their role as minor in the patient's decision to quit.

Asking about Tobacco Use
In seven studies, at least three-quarters of physicians, more frequently among non-smoker physicians, reported that they questioned their patients about tobacco use or knew their smoking status [16,27,29,47,61,63]. In a Turkish study [49], physicians asked about smoking almost half the time, while only 30% of Cuban physicians always asked about smoking [60].

Advising Patients to Quit Smoking
The prevalence of advice to quit differed across studies. In Estonia, nearly all physicians (96%) reported advising patients to quit [56]. In China, 70% of physicians had counseled patients to quit within the last year [42]. About half the physicians in Malaysia [45], Guatemala [61], and Russia [13] advised patients to quit, while less than one third of physicians in a Chinese study [6] did so. In South Africa [29] and Malaysia [45], advice to quit depended upon the patient having a condition associated with smoking. The most commonly used methods to encourage patients to quit were brief counseling and education about the dangers of smoking in several studies [6,27,45,63].

Provision of Counseling/Pharmacological Therapy and Follow-up Arrangements
Giving cessation counseling and making follow-up arrangements were not common across studies. For example, less than one-quarter of Ukrainian physicians provided advice on quitting and just 10% provided nicotine replacement therapy (NRT) or referrals to quitting clinics [27]. In a Turkish study, less than 5% of physicians always arranged follow-up visits or assisted patients with a cessation plan [49]. Only 1% of Nigerian physicians reportedly had prescribed pharmacotherapy for smoking cessation [16]. Providing behavioral counseling or prescribing NRT was also not common among Costa Rican [63], Russian [13], and Guatemalan physicians [61]. Two studies showed more promising results: in Estonia, about two-thirds of surveyed physicians spent time counseling smokers [56]; and in Brazil, slightly more than half of the surveyed physicians provided either a prescription or a referral along with advice to quit [22]. Factors that influenced physicians' engagement in smoking cessation included physicians' smoking status [40,44,61], their attitudes towards smoking [20,40,44,48,49], and confidence level in their smoking cessation skills [6,20].

Interventions to Promote Smoking Cessation among Physicians
Literature describing programs promoting smoking cessation among physicians is scarce in developing countries. Only three studies provided brief information on the nature of such cessation interventions. Mirra and Rosemberg [25] reported that in Brazil there is few tobacco use reduction or smoking cessation programs that targeted physicians, but details on those programs were not available. Barnoya reported that physician smokers who had quit in the past had primarily used the -cold turkey‖ method, and none had tried NRT [61]. Viegas [21] reported, also in Brazil, that about three-quarters of smoking physicians were advised by their own peer doctors to quit. A few studies suggested different measures to reduce tobacco use among physicians [25,27,29,41,55,63]. These include introducing tobacco control education in the medical school; a policy initiative to ban smoking in health facilities; training physicians on smoking cessation as part of national action plan; an awareness campaign customized to physicians; and quitting support services customized for physicians.

Discussion and Conclusions
The study summarizes studies from across a number of developing countries on tobacco smoking and smoking cessation practices among the physicians. We found a substantial and informative body of literature on smoking among physicians in developing countries. A higher number of studies were identified from Central/Eastern Europe and Central/South America than from Asia and Africa. A significant number of studies were conducted starting in 2000, which might be due to the aggressive tobacco control activities that were initiated by the WHO and other advocacy groups. We found that the focused region or country to study tobacco use among physicians varied between the studies conducted during 1980s and those conducted in the last decade. For example, no study among physicians was reported in China during 1980s while from 1999 onwards nine studies were reported. This likely reflects the fact that smoking was so socially acceptable in China a decade ago that even the public health workforce did not feel the need to conduct surveys of physicians smoking behavior and cessation practices. However, regardless of the origin or period of the study, tobacco use among physicians in many developing countries remains a major public health concern.
We found variation in smoking rates across the studies that are related to several factors including methodological differences in data collection, differing sample sizes across studies, differing time periods, and variations in the smoking culture across the countries. Overall, our findings strongly suggest that the prevalence of smoking among physicians in developing countries is high. This high prevalence of smoking among physicians has implications for the general population because continued role modeling of smoking by physicians undermines the messages that smoking is harmful and that quitting is important [4]. Studies find that non-smoking physicians are more successful in getting their patients to attempt to quit than smoking physicians [67]. Moreover, physicians who smoke may increase public skepticism about quitting, with people inclined to ask why they should stop smoking when their doctor continues to smoke [68].
Many smoker physicians in this review had tried to quit; many were successful in their past quitting efforts, while some were unsuccessful. Identification of factors in well-designed studies that were associated with success and failure could guide policy makers in designing appropriate interventions. Many physicians initiated smoking while they were at medical schools [20,40,46]; therefore, there is a need to design interventions that would discourage tobacco use initiation at the course of their studies at medical schools. In most of the studies, male physicians smoked more than their female counterparts. The difference was greater in Asian studies, suggesting cultural sensitivity in approving smoking by professional women in Asian countries. Although the regional differences may reflect cultural differences regarding smoking by women, continuous efforts should be maintained to keep the prevalence of female smokers low in these countries in response to aggressive promotion of tobacco use among Asian women by the tobacco industry [69]. At the same time, there is a need for targeted programs to promote cessation among male physicians who smoke [6]. Although ex-smokers are few in number, they can serve as role models in encouraging quitting and can provide social support to physicians who want to quit [7,33]. Other innovative approaches such as mobilizing key leaders to promote smoking cessation could also play a major role in encouraging physicians to quit smoking [7].
Smoking in front of patients or at worksites was prevalent in all the regions. This practice might be related to the lack of awareness about their professional role and reinforcement of smoke-free hospital policies. It might also relate to the doctors' physical dependency on nicotine, myths about smoking that tobacco use represent high social status and inappropriate attitudes towards smoking cessation [7,70]. Strengthening the implementation of smoke-free hospital policies in combination with providing smoking cessation training and support for health care workers who smoke could reduce smoking within hospitals.
The review found that many physicians, regardless of their smoking status, are missing opportunities to advise or counsel patients on quitting smoking. In a study by Zou et al. [6], more positive beliefs and higher confidence level were associated with asking and advising patients to quit. This indicates the need for professional training to address local doctors' beliefs, attitudes and confidence levels in relation to smoking cessation. The review also revealed that only a minority of physicians prescribed medications such as NRT, which provide effective pharmacological support for smoking cessation. The low use of pharmacotherapy might be explained by doctors' doubts about the effectiveness of such products [6,7] or the products' unavailability in the relevant country, or financial barriers in buying NRT.
Although reducing smoking among physicians and involving them in tobacco control activities is critical in combating tobacco-induced morbidity and mortality in these developing regions of the world, this review suggests that the evidence-based smoking cessation programs targeting physicians are scarce. This reflects the lack of interest among policy makers and researchers to focus on reduction of tobacco use among physicians. Evidence from both developing [27,48] and developed [71] nations showed that the rate of asking about smoking and advising to quit smoking was uncommon among smoking physicians. Therefore, organized programs to address tobacco use among physicians should be a priority for developing countries, which would have an impact on the smoking rate at the population level. As other allied health professionals (i.e., nurses, physiotherapist, and community health workers) also play a critical role in the healthcare delivery in developing countries and have frequent contacts with the public, engaging the allied health professionals in the tobacco use reduction and cessation effort would be useful. For example, community health workers played a key role in the delivery of other preventive health programs (i.e., immunization, breastfeeding) and treatment for malaria in low income countries, engaging them in the tobacco control effort would have great population health impact.
As evidence-based smoking cessation interventions are already available in developed countries, testing and adopting them in developing country settings should be a priority. Involvement of professional societies, medical associations, tobacco-control advocates and other influential local leaders and hospital executives, non-governmental organizations, and international agencies is necessary to reach physicians with cessation interventions and encourage them to act as tobacco-control advocates. Intervention programs should address influencing factors such as physician characteristics (e.g., smoking status, perception of self as role model), structural factors (e.g., time, and reimbursement), and cessation-specific knowledge and skills. As social context [72] and myths [73] may play roles in sustaining smoking behavior and quitting effort, future intervention programs should also address these issues. The WHO FCTC has dramatically altered the global tobacco control environment and public health professionals should utilize this opportunity to engage physicians in the tobacco control effort and address tobacco use by the physicians. Global tobacco control efforts should highlight tobacco use by the physicians as one of the major priorities and work with professional societies and government bodies to address the issue in an organized manner. Regular data collection from physicians using standard questionnaires, such as Global Health Professionals Survey [74], would be useful to monitor the tobacco use trends over time and assess the impact of any tobacco control policy or targeted interventions among physicians.
The review has both strengths and limitations. To the best of our knowledge, it is the first comprehensive review on the topic to focus on developing countries. Methodological differences across studies (i.e., sample sizes, how the data were collected, gender ratio of the collected subjects, the definition used to define current smokers, and the date in which the studies were conducted) were unavoidable limitations, as was our access to published data only. Finally, the findings we summarized are based on the reports in the published papers. No attempts were made to validate the findings or conclusions of the reported studies. Therefore, these limitations should be considered when extrapolating the findings of the current paper. Also, this review focused mainly on cigarette smoking as most of the available reports was based on cigarette smoking among the physicians. However, this should be noted that tobacco use encompasses more than cigarette smoking. Future studies should consider physician's use of and counseling behaviors regarding other forms of tobacco (smoked and smokeless) use.
Overall, this review suggests that while smoking habits of physicians vary across different developing regions; prevalence rates tend to be higher than among physicians in developed countries. Also, quitting rates were low among the physicians, and the delivery of advice on quitting smoking was not common across the studies. Although the number of studies from each of the global regions under consideration was unbalanced in the review, this may reflect patterns of tobacco control activities targeting physicians in each region. Future review might focus on a single geographic region and explore studies by cultural norms to document culturally sensitive factors associated with physician smoking or delivery of cessation intervention. Future research would benefit from the exploration of the factors discussed in this review to determine the facilitating factors to reduce tobacco smoke among physicians and to promote physician-led smoking cessation and tobacco use reduction intervention.