Smoking by French general practitioners: behaviour, attitudes and practice

Abstract

Introduction: This paper examines smoking prevalence, sociodemographic factors and the medical practice of French general practitioners. Method: Data from the 1998 cross-sectional national survey of 2,073 GPs. The questionnaire was administered by telephone. A response rate of 67% was attained. Instrumentation included questions about medical practice, sociodemographic characteristics, and health behaviour. Bivariate and multiple logistic regression (MLR) analyses were conducted. Results: Almost one-third (32.1%) of physicians were current smokers. A significantly higher proportion of male (33.9%) were smokers compared to women (25.4%, p<0.001) and men were more likely to be former smokers (49.1% versus 31.7%). Two-thirds of physicians reported recommending nicotine replacement therapy to their patients. MLR shown that former smokers were more likely (OR = 1.51, 95% CI, 1.24–1.83) to indicate that their help in getting patients to quit was not effective compared to smokers. Also, physicians who were ‘dissatisfied’ with the profession were more likely (OR = 0.75, 95% CI, 0.60–0.92) to report their help as not effective than those who were ‘satisfied’. Conclusion: These data support the need for greater professional participation in reducing smoking among general practitioners in France and greater education concerning the vital role of physicians in promoting cessation among the general population.

Abstract

Key points

• This study examines smoking habits among French GP's, intervention practices, and opinions about their ability to help patients quit smoking.

• This study found a smoking prevalence rate of 32.1%, 33.9% of male physicians were smokers and 25.4% of women.

• This study found a smoking prevalence rate of 32.1%, 33.9% of male physicians were smokers and 25.4% of women.

At the time of this survey, approximately 34% of French adults were smokers and the prevalence was higher among men (38%) than women (30%).¹ Since 1990, these rates have been slowly decreasing although they remain quite high among young adults, especially women.² Despite a higher smoking prevalence (though lower daily cigarette consumption), smoking related diseases in France are lower than those in the USA. For example, comparable data shows an incidence rate of smoking attributed deaths per year of 52/1000 for males in France and 3/1000 for females compared to 303/1000 for males and 226/1000 for females in the USA.³ Among women, it is suspected that the differences are primarily the result of the later start of the smoking epidemic in France compared to the USA.⁴ However, it is not clear what accounts for the differences among men.

Most media advertising of tobacco has been restricted in France since 9 July 1976 when the first anti-tobacco law (Loi Veil) was introduced.⁵ In more recent years, the French government has taken an increasingly active role in adopting policies to reduce smoking. A new law (Loi Evin) was enacted in 1991 that extended the advertising prohibition such that by 1993 a total ban in newspapers and magazines was in effect.⁶ Smoking was also prohibited in schools, hospitals, restaurants, public transportation, workplaces, and other public places. The social acceptance and compliance with efforts to control tobacco use is evolving yet smoking is widely accepted in many social environs.⁷

Studies have convincingly shown that physicians can play an effective role in getting their patients to quit smoking.^8,9 In France, physicians have not been as effective in patient smoking cessation interventions, in part because of their own high rate of smoking. Josseran and colleagues found that approximately 34% of French physicians smoked cigarettes in 1994 and that this behaviour was less likely to result in patient compliance with anti-tobacco messages.¹⁰ Slama et al. in a regional study on attitudes and practices of 700 French general practitioners (GPs) found that a majority of physicians were concerned about smoking cessation for patients but were not confident of their ability to help them to quit.¹¹

Countries with historically high prevalence rates have observed that a reduction in physician smoking patterns precedes decreases in the overall population. At the beginning of the 1950s in the USA, 60% of physicians were smokers, a rate that was similar or exceeded that found among the adult male population.¹²

In 1999, the French Ministry of Health announced a new series of public health measures to reduce smoking that included permission to purchase NRT over the counter, greater health education programmes targeting youth populations, and a campaign to engage physicians in smoking cessation efforts with their patients.¹³ Over the counter sales of NRT increased twofold in 2000 compared to 1999 when these products were regulated by prescription.¹⁴ A recent French national survey indicated that 58.9% of French smokers want to quit, but about 1% of former smokers quit because of their physicians' recommendations and even fewer current smokers reported that they were motivated by physicians to quit.¹ These findings indicate a tremendous opportunity for physicians to become more involved in motivating and helping their patients to quit.

In January 2003, a new plan against cancer was introduced by the French Ministry of Health that included 70 measures to address tobacco consumption and public policy including eight measures related specifically to smoking prevention. Of particular importance was the expressed need to develop smoking cessation consultations for outpatients in hospitals and the promotion of ‘minimum advice’ for smokers who consult with health care practitioners.

This present study examines smoking among a more recent and larger nationally representative sample of general practitioners in France regarding their smoking behaviour, intervention practices with patients, and opinions about their ability to help patients quit. These findings could be particularly useful in determining the role of physicians in promoting smoking cessation and supporting tobacco control policies.

Methods

Data from the 1998 Institut National de Prévention et d'Education pour la Santé (formerly Comité Français d'Education pour la Santé, CFES) national survey on GPs included 2,073 respondents. The sample was randomly selected from a nationally representative database of GPs in France maintained by a pharmaceutical industry. Each eligible physician received a prenotification letter describing the purpose of the survey and sponsorship. The response rate was approximately 67%, which is similar to other studies of physicians.^15,16 Information on the survey design and strategies has been published elsewhere.¹⁷

A professional research firm conducted the survey and the questionnaire was administered, after a pilot test, in the fall of 1998 using a computer assisted telephone interviewing (CATI) system. Interviewers received specific instructions on optimal strategies to collect data in an efficient manner from physicians.

Instrumentation included variables related to medical care, sociodemographic characteristics, and opinions about medical practice. Questions on cigarette smoking referred primarily to smoking status, frequency of smoking, smoking cessation methods prescribed, number of tobacco patients seen during the last seven days for smoking cessation, and perceived effectiveness of assisting patients in stopping smoking. Physicians were also asked about their satisfaction with medical practice.

Current smoking and frequency of smoking were dependent variables in this analysis. Smoking was defined as presently smoking cigarettes regularly or occasionally, and the variable consisted of smokers and non-smokers. Non-smokers consisted of former and never smokers. Frequency of smoking referred to the number of cigarettes consumed daily and was divided into 15 and fewer cigarettes per day and greater than 15 cigarettes per day.

The number of tobacco patients seen in last the 7 days was defined as 1 or more and none. Prescribed methods for smoking cessation were categorized as quitting without any assistance (i.e. ‘cold turkey’), use of nicotine replacement therapy (NRT) or other methods. The perceived effectiveness of smoking cessation interventions by physicians was defined as effective versus not effective. Satisfaction with the medical profession was divided into two dichotomous categories satisfied and not satisfied and the type of practice included group and individual modalities.

Statistical analysis consisted of descriptive statistics (i.e. frequency distributions and cross-tabulations) and multivariate logistic regression (MLR) analysis. MLR was conducted to assess the simultaneous effects of multiple predictors on each of the dependent variables. The Statistical Packages for the Social Sciences (SPSS v.10) computer program was used to conduct the analysis.

Results

Approximately 21% of GPs were female and the mean age was 45.1 years. On average male GPs (46.0 years, SD: 7.4 years) were older than their female counterparts (41.5 years, SD: 6.7 years, p<0.001). A majority (56%) of respondents were solo practitioners and 71.9% responded that they were satisfied with practicing medicine (table 1).

The analysis showed that 32.1% of physicians in this survey currently smoked cigarettes and 84.3% of smokers consumed 15 or fewer cigarettes per day (mean = 8.2 cig/day). A significantly higher proportion of male physicians (33.9%) were smokers compared to women (25.4%, p<0.001, table 2) and men were also more likely to be former smokers (49.1% versus 31.7%, p<0.001). There was a statistically significant difference in the prevalence of smoking (28.9%) of younger (<40 years old), and older doctors (33.5%, p<0.05). Satisfaction with medical practice was a significant predictor of the number of cigarettes smoked daily. Physicians who were dissatisfied smoked a higher mean number of cigarettes per day than contented doctors (9.6 cig/day versus 7.5 cig/day, p<0.05).

A majority of physicians (60.1%) in the last 7 days had attended to 1 or more patients who smoked (table 1). Two-thirds of physicians recommended NRT to their patients but non-smokers were significantly (p<0.01) more likely than smokers to recommend use of these products (table 1).

Regression Analysis

As presented in table 3, the multivariate logistical regression (MLR) models included gender, age, practice type, and satisfaction with the medical profession. The analysis revealed that women physicians were less likely then men (OR = 0.68, 95% CI, 0.53–0.87) to be current smokers. Physicians working in group practices were more likely (OR = 1.55, 95% CI, 1.17–2.06) to be smokers in reference to solo practitioners. Doctors who indicated that they were ‘dissatisfied’ with the medical profession were less likely (OR = 0.61, 95% CI, 0.38–0.97) to consume 15 or few cigarettes daily than those who indicated they were ‘satisfied.’ Older physicians were significantly more likely (OR = 1.63, 95% CI, 1.23–2.14) to be former smokers than younger physicians and those who were dissatisfied with the profession were less likely to be former smokers compared to colleagues who were satisfied.

A second set of MLR models focused on physician smoking cessation practices and opinions. Physicians in group practices were less likely than solo practitioners to indicate that they had not seen at least one smoking patient in the last seven days. Group practice physicians were less likely than solo practitioners to believe that their advice to smokers was often effective (OR = 0.45, 95% CI, 0.33–0.62). Medical practitioners who were dissatisfied with the profession were less likely than those who were satisfied (OR = 0 .75, 95% CI, 0.60–0.92) to believe that their advice to smokers to stop smoking was more effective. Current smokers compared to former smokers were more likely (OR = 1.51, 95% CI, 1.24–1.83) to indicate that their help in getting patients to quit was often effective. Group practitioners were more likely (OR = 1.57, 95% CI = 1.15–2.16) than solo practitioners to indicate that stopping ‘cold turkey’ was better than using NRT therapies for smoking cessation as did smokers compared to former smokers (OR = 1.40, 95% CI, 1.11–1.77).

Discussion

The results of this study revealed that 32.1% of French GPs were current smokers. The prevalence of smoking of both male and female physicians was about 3 percentage pointers lower than their counterparts in the French population.¹ A major difference in the smoking behaviour was the number of cigarettes consumed per day: physicians smoked 8.2 cigarettes daily compared to 14.4 among French adults. Compared to the previous CFES cross-sectional national survey by French general practitioners in 1994, the overall prevalence rate decreased from 34% and among men, it declined from 36.1% to 33.9% though increased slightly from 24.9% to 25.4% among women.⁹ Nevetheless, the reductions by gender in smoking between the two time periods were not statistically significant. We also found a reduction in the number of cigarettes smoked daily between 1994 and 1998 among both male (11.2% to 8.4%) and female physicians (8% to 7.4%).

As others have noted, smoking among physicians is relatively high and comparable to the general population in many southern European countries such as Spain, Italy, and Greece.^10,18,19 Countries with historically high prevalence rates have observed that a reduction in physician smoking patterns often precedes a decrease in the population rate.²⁰ However, the downward trend in smoking among physicians over time may be due more to a cohort effect, as younger physicians are less likely to start smoking than their older counterparts, rather than to an increased cessation rate among smokers.²¹ In this respect, the French physician population may be following the same trend.

Less promising is the finding that smoking among women increased slightly from 1994 to 1998. These results suggest that smoking among women physicians in France may reflect the pattern of increased smoking among younger women^1,2,20 and of smoking (35.6%) among female medical students.²² As a result, their rates may not decrease to the same extent as their professional male colleagues although both groups would continue to benefit from greater prevention efforts targeting youth populations.

Interestingly, we found that 40% of general practitioners indicated that they did not see any smoking patients in the last 7 days. This proportion appears high considering that smokers are more likely to require office visits than non-smokers and the fact that approximately one-third of the French population are smokers. We suspect that there may be many missed opportunities to enquire about the smoking habits of patients. Greater efforts should be made to encourage physicians to follow practice guidelines with respect to ascertaining current tobacco use and exposure information from patients.

The MLR findings showed that physicians who smoked were more confident than non-smokers in their perceived ability to help smokers quit and were more likely to recommend quitting ‘cold turkey’ than by using NRT. In this regard, our findings about physicians' perceptions of their ability to help patients is especially germane because perceptions matter as they influence doctors to try to assist patients and may determine the degree of effort they devote to smoking cessation advice or counseling.²³ It may suggest also that physicians who smoke tend to overestimate their ability to help patients quit.

It is not clear why physicians' attitudes regarding the profession are correlated with their views about the effectiveness of their smoking cessation advice. It is possible that less positive opinions about the medical profession or their status within the profession could be reflected in perceptions about their ability to help patients quit smoking. Additional research regarding this matter could help to elucidate the role of professional contentment in motivating patients to quit smoking.

French physicians' recommendations to patients may be influenced by what Deekers et al. referred to as the ‘moralization of judgment’.²¹ That is, they may perceive a philosophical conflict between the right of individuals to exercise their own judgment regarding personal behaviour and their responsibility as physicians.^24,25 Deekers et al. speculated that this perception could impact how physicians perceived their role with respect to patients.

Interestingly, this study found an association between the type of medical practice (solo versus group practice) and smoking cessation practices. The finding that group practitioners were less likely to have seen a smoking patient within the last week or who were also less likely to view their help as often effective may reflect a lesser likelihood to ask about the smoking status of patients and to provide advice and counseling.

Some of the limitations of this study are that although we do not have an exact estimate, there is likely to be some underreporting of smoking among physicians as the findings about smoking status and patient practices are based on self-reports and that smokers may be less likely to participate or answer questions about tobacco consumption. The lack of information about non-respondents precludes assessing possible bias due to non-response. Also, although about half of physicians in France are general practitioners, smoking behaviour and practices may vary among different specialties or settings.²⁶ Some caution should also be exercised in comparing the 1994 CFES survey of general practitioners to the present one because they are both cross-sectional studies and thus the comparisons are only suggestive or indicative rather than confirmatory or causal.

In sum, these data support the need for greater support and education of general practitioners in France concerning their vital role in promoting smoking among the general population. It appears that many physicians may be missing important opportunities to advise patients about smoking cessation strategies and may not be fully aware of the efficacy of NRT.²⁷ In the context of the clinic or office environment, nurses and paraprofessionals could also play an important role in patient education concerning smoking cessation and NRT, and greater efforts should be made to involve them in the process.

Medical associations should take a lead role in addressing this matter not only to reduce tobacco consumption among the general population but also to encourage cessation among physicians. We also recommend that health institutions assume a greater responsibility in promoting smoke-free facilities and practices among their staff and patients. Lastly, we maintain that medical schools should focus more on reducing smoking among medical students and provide the training necessary to conduct patient tobacco education and cessation. Future morbidity and mortality from smoking related diseases in France will be influenced considerably by the behaviour and practices of French physicians.

This paper was supported in part by a grant from the Fogarty International Center Minority International Research and Training Program (PSV) 5T37TW00113-05.

References