Prevalence of smoking in physicians and medical students, and the generation effect in the Netherlands

Sm. Sci.Med. Vol. 36,No. 6,pp.817-822, 1993 Printed in Great Britain. All rights reserved

0277-9536/93 $6.00 + 0.00 Copyright 0 1993 Pergamon Press Ltd

PREVALENCE OF SMOKING IN PHYSICIANS MEDICAL STUDENTS, AND THE GENERATION IN THE NETHERLANDS HELEEN M. DEKKER,‘* and

AND EFFECT

CASPAR W. N. LOOMAN,’ HANS P. ADRIAANSE~ PAUL

J.

VAN DER MAAS't

‘Department of Public Health and Social Medicine (M.G.Z.), Erasmus University, P.O. Box 1738, 3000 DR Rotterdam, The Netherlands and ‘Department of Health Education, University of Limburg, Maastricht, The Netherlands Abstract-This study investigates smoking habits and attitudes towards smoking in general practitioners, consultants at a university hospital, medical students and students of health policy and management (H.P.M.). An anonymous, self-administered postal survey was used. Thirty-eight percent of the general practitioners, 27% of the consultants, 18% of the medical students and 31% of the H.P.M. students are current smokers. The prevalence of smoking was found to be higher in the male general practitioners and the male H.P.M. students than in the general male population. The prevalence of smoking was lower in female physicians and students than in their male counterparts and also lower than in the general female population. Medical students are not inclined to start smoking: a strong generation effect can be observed. This will reinforce the current downward trend in the prevalence of smoking in Dutch physicians. The doctors were found to have a suboptimal level of knowledge about methods of smoking cessation and about the association between smoking and health disorders. Most Dutch physicians, especially those who smoke, fail to perceive their role as an example to the

general population concerning smoking behaviour. Medical students were found to have even less recognition of their future exemplary role. Kqv words-smoking,

exemplary

role, physicians,

medical

INTRODUCTION smoking is a major threat to public health, the reduction of tobacco consumption is generally acknowledged as an extremely important goal for preventive actions. Physicians and other workers in the public health sector are in a position to play an important role in reducing the consumption of tobacco [l]. Many physicians regularly witness the injurious effects of smoking, and therefore one might expect to find low prevalence of smoking in this profession. In recent years there have been several studies of smoking prevalence among medical students and physicians in Great Britain and the United States. These studies were conducted at different times, in different age groups and using different definitions of smoking, some were local, some statewide, some national. While these studies are not entirely comparable, there is no doubt that there is a downward trend in smoking prevalence in these groups. In Great Britain smoking prevalence was found to be 10% in general practitioners [2] and 3% in consultants [3]. As

*To whom reprint requests should be directed. tTo whom correspondence should be addressed. IData extracted from a survey conducted in 1989 by the Dutch Institute for Public Opinion polls (N.I.P.O.) for The Netherlands Foundation on Smoking and Health. 817

students,

The Netherlands

In physicians in California, United States [4], it was found to be 8%. In Great Britain and the United States a decrease in the prevalence of smoking occurred earlier in physicians than in the general population [5]; this indicates that physicians have a leading role vis-&vis the general population with regard to smoking. Smoking rates in medical students have been found to vary, from 12% in Great Britain [6] to 5% in the United States [7, 81. In The Netherlands there have been several surveys of the smoking habits and attitudes towards smoking of general practitioners [2,9], but there is no comparable information concerning consultants and health care students. In 1989 the overall Dutch population showed a smoking prevalence of 36% in men and 29% in women [lo]. Smoking was less prevalent in the highest socio-economic bracket: 31 and 25%,$ respectively. This paper reports the results from a postal survey of smoking habits and attitudes towards smoking in 4 selected groups of present and prospective health professionals in the Rotterdam area. The survey was designed to obtain information about the smoking habits of health professionals in order to compare these with smoking behaviour in the overall population, and to investigate the physicians’ perception of their exemplary role.

818

HELEEN

M.

METHODS

Subjects Subjects were drawn from 4 groups: general practitioners, consultants, medical students and students of health policy and management (H.P.M.). General practitioners in the Rotterdam area were contacted through the address file of the local General Practitioners Association (n = 320). All 270 consultants working in the University Hospital of Rotterdam were contacted with the help of the administrative department. Two hundred subjects were randomly selected from the address files of medical students at the Erasmus University of Rotterdam (population n = 1515), and likewise a further 200 from the H.P.M. students (population n = 332). Questionnaire The data were collected in 1989, by means of self-administered questionnaires. These contained 34 questions concerning past and present smoking habits, future intentions concerning smoking, knowledge and attitudes about tobacco use, and the responsibilities of health sector workers in this matter. Most questions were translated from a standard questionnaire, developed by a working group of the World Health Organization [I l] (questionnaires may be obtained on request from the authors). Procedure The questionnaire was mailed to all selected subjects; they could return it anonymously to the Department of Public Health and Social Medicine at the Erasmus University of Rotterdam. A reply card was enclosed with the questionnaire. which could be returned separately, in order to identify nonrespondents without violating the anonymity guarantee. Non-respondents received a reminder and a new questionnaire after 5 weeks (second round). Response The response rate was 65% in the first round and 43% in the second round, resulting in an overall

Table

I. Smoking

behaviour

of general

Ge”eLd practitioners Smokine

status

Current smokers Current smokers (index I)t Current smokers (index 2)t Former smokers Never smoked *Health Policy and tIndex figures show the total Dutch age differences. as 1.00.

practittoners,

DEKKER et ul

rate of 80% (n total = 789). Completion rates in the first and second rounds combined were 82”/0 for the general practitioners, 73% for the consultants, 80% for the medical students and 85% for the H.P.M. students. A comparison of the smoking behaviour of early and late respondents showed that the percentage of smokers was lower in the former group than in the latter (27 and 42% respectively). The largest difference was found in the consultants and medical students. No difference was found in smoking prevalence between early and late respondents in the H.P.M. students.

RESULTS

Smoking

habits

Smoking prevalence. Smoking status was determined by one question: “Do you smoke now?” This includes both regular smokers and occasional smokers. It was found that 30% of the overall sample were smokers, ranging from 18% of the medical students to 38% of the general practitioners. The smoking behaviour of male and female physicians and students is shown in Table 1. Male and female respondents showed a difference in smoking behaviour: 35% of the male respondents and 21% of the female respondents were current smokers. Higher rates were found for men in each of the 4 groups. The difference between the sexes in general practitioners and consultants was statistically significant (loglinear regression, P < 0.05). when corrected for age differences. The largest difference between male and female respondents was found in the consultants: 33% of the male consultants were smokers, against only 12% of the females, This 12% was the lowest percentage of female smokers found in the 4 groups. The lowest percentage of male smokers was found in medical students. at 19%. The highest percentage of both male and female smokers was found in the H.P.M. students, at 43 and 27%. respectively. Former smokers were more prevalent in men, except for the group of consultants. The percentage of former smokers was higher in the

consultants. medical students (questionnatre data)

and

H.P.M.’

The Netherlands,

H P.M.’ students

Medical students

Consullants

students,

19X9.

Total

M n =221

F n =42

M II = 144

F n =52

M n = 72

F n =8X

M n = 37

F n = 133

M II = 474

F 11= 315

4 I% 1.01

24% 0.69

33% 0.80

12% 0.35

19% 0.56

16% 0.49

43% I .22

27% 0.81

35% 0.90

2I% 0.63

I.13

0.82

0.89

0.41

0.80

0.69

1.70

1.12

I 05

0.83

34% 25%

28% 48%

3I% 36%

38% 50%

13% 68%

3% 81%

16% 41%

15% 58%

29% 36%

17% 62%

Management. the ratio between the percentage of current smokers tn the group concerned and the percentage of current smokers in population (index I) and the highest socio-economic bracket in the general Dutch population (index 2). corrected for The percentage of current smokers in the total Dutch population and the highest socio-economic bracket were taken

Smoking among physicians and students

819

Table 2. Percentages of current and former smokers and those who have never smoked among general practitioners, consultants, medical students and H.P.M. students divided into age groups, The Netherlands, 1989 (questionnaire data) Smokine Current Former

status

d 19 years n =39

20-29 years n = 272

3&39 years n = 241

4&49 years n=164

SO+ years n = 67

Mean n = 789

smokers

10

24

38

34

2s

8

30

II

23

39

52

24

82

65

39

21

23

46

smokers

Never smoked

physicians than in the students. Of the medical students the majority had never smoked. Table 1 also shows a comparison of the percentages of smokers in the respondents with that in the general Dutch population and that in the population in the highest socio-economic bracket. The figures are corrected for age differences by means of indirect standardization [12]. On the whole, the percentage of smokers in the male respondents is similar to that in the total Dutch population, while in the female respondents the figure is lower. Here again we see striking differences between groups. Compared with the general population, the differences in smoking prevalence in female consultants, and male and female medical students were statistically significant (loglinear regression, P < 0.05), corrected for age differences. Age group. The distribution of smoking behaviour by age groups is presented in Table 2. The percentage of smokers increased with age: 10% in the group aged up to 19 years to 38% in the age group 3&39 years. From this point the percentage decreases again. The percentage of former smokers increases with age, ranging from 8 to 52%. The percentage of those who never smoked decreases with age, with a sharp decline after the age of 29. The mean age at which respondents started smoking was 18 for general practitioners and 19 for consultants. Students who currently smoked had started at the age of 17. Tobacco consumption. Most of the current smokers used manufactured and/or hand-rolled cigarettes (65%). Twenty-one percent of the general practitioners, 15% of the consultants, 14% of the medical students and 27% of the H.P.M. students were current cigarette smokers. Those who preferred manufactured cigarettes usually chose filter-tipped cigarettes (84%). Thirty-eight percent of the current smokers smoked a pipe, cigars and/or cigarillos. These were almost exclusively men. There appears to be a tendency among current smokers to shift from cigarettes to cigars and cigarillos. This is illustrated by the fact that 50% of the male physicians who

*Converted into number of cigarettes according to the formula (1 I]: number of manufactured cigarettes per week + number of handrolled cigarettes per week + 3 x number of pipes per week + 3 x number of cigars per week + 3 x number of cigarillos per week = mean number in cigarettes per week.

smoked used a pipe, cigars and/or cigarillos, compared with 43% of the male students. Only a small percentage of pipe, cigar and cigarillo smokers was found in former smokers. Among the current smokers, medical students were the largest consumers: on average 77* cigarettes a week. General practitioners, consultants and H.P.M. students smoked on average 62,60 and 57 cigarettes per week, respectively. The largest proportion (11%) of heavy smokers (i.e. over 20* cigarettes per day) was found in general practitioners. Average tobacco consumption per smoker was higher in women than in men (72 and 56,* respectively), except for the medical students. Male medical students consumed almost the same amount of tobacco as their female counterparts. Reasons for current smoking. The main reasons for starting to smoke were that friends smoked and that respondents enjoyed smoking. Students were especially influenced by smoking behaviour of friends. ‘A sense of belonging to the group’ appeared to be a reason to start smoking for 34% of the respondents. The majority of current smokers continued the habit because they enjoyed it. Joining in for the sake of company also played an important role. Attempts to stop smoking. Even though the majority of the current smokers consumed the same amount of tobacco as in the previous year, they reported that they had considered stopping smoking. In each group of respondents over 50% of the smokers had seriously tried to stop smoking at least once. Nearly a third of the smokers indicated that they had cut down in comparison with the previous year. Future smoking behaviour. The physicians’ and students’ expectations for the future with regard to their smoking habits indicate that a drastic decrease in smoking among the physicians may be expected. Over 50% of the respondents expected to have stopped smoking within 5 years. This percentage was 57% for men and 42% for women. About 25% of the smokers also reported that in 5 years they would probably not be smoking every day. Overall, the majority of smokers stated that they hoped to stop in the near future. But even if most of them would unfortunately not succeed in breaking their habit it is important to know that. Ninety-five percent of the non-smokers stated that they would not start smoking within 5 years. This means that the smoking

820

HELEEN M. DEKKER et al.

behaviour of the present generation of medical students may present a picture of more favourable smoking habits in a future generation of physicians. Attitude

of physicians

towards patients’

smoking

Nearly all general practitioners did not allow smoking in their waiting rooms. If a patient was found to be smoking, in spite of this, three quarters of the general practitioners would raise objections. Consultants were less concerned about such infringements. Sixty-two percent of general practitioners asked about smoking behaviour when taking a patient’s history. A quarter of them always informed patients who smoked about the deleterious effects of tobacco and another 46% often gave information. In general, the consultants paid less attention to the smoking behaviour of their patients than did the general practitioners. Nearly all the medical students engaged in clinical work asked patients about their smoking habits. The majority of physicians advised patients with symptoms or diagnoses of smoking-related diseases to stop smoking. Advice to stop smoking was given less often if patients themselves raised the question of smoking. Only 7% of physicians would advise patients against smoking in cases where there was no particular reason to do so. Only a quarter of the general practitioners indicated that they currently had sufficient knowledge for counselling patients who were willing to stop. The proportion of consultants who considered their knowledge to be adequate was even lower. Knowledge

of the dangers

of tobacco

The noxious effects of smoking caused nearly three quarters of the respondents to be moderately to very concerned for their own health. Current smokers were clearly less concerned (55%). Physicians and students knew that there is an important association between smoking and the following common diseases: cancer of the lung; coronary vascular disease; chronic bronchitis; peripheral vascular disease; emphysema; and carcinoma of the larynx. The association between smoking and leukoplakia was not well known; 12% of the physicians were completely unaware of such an association. The respondents knew that neonatal mortality and peptic ulcer are somehow associated with smoking. There seemed to

Table

3. Ge”eKil practitioners

and consultants

indicated

be considerable ignorance of the association between cancer of the bladder and smoking: 43% of the physicians mentioned no assocation or indicated that they did not know. The H.P.M. students knew less about the association between smoking and these disorders. Perception

Attitudes controlling

agreeing

with the statement,

towards

legal

the tobacco

and

other

regulations

.for

problem

The government can introduce legal measures to reduce tobacco consumption, and to limit sales and advertising. The opinion of the respondents about legal measures concerning smoking are shown in Table 4. Most respondents were in favour of the legal restriction of smoking in public places. Although smokers were significantly less supportive to this

their agreement behaviour

or

disagreement

By my own health behaviour, I consciously set a good example to my patients. My smoking behaviour is unimportant in giving advice. Telling patients what is good for them is important. My health behaviour is hardly noticed by my patients, so it will have no influence on them. My life-style outside my work is irrelevant to my patients. of respondents

role

The respondents were asked whether they agreed with several statements about their possible exemplary role with regard to smoking and other aspects of life style. Most respondents felt that convincing people to stop smoking is not a physician’s responsibility. Female respondents especially held this opinion. A considerable number of the physicians did not acknowledge the desirability of an exemplary role concerning smoking with respect to patients. The students were even less convinced about this exemplary role than the physicians. A summary of the results is presented in Table 3. Physicians who smoke, particularly, think that they do not need to set a good example to their patients by their own health behaviour. This can also be shown by giving advice: telling patients what is good for them is the most important thing. There are however important differences according to smoking habits (64% in current smokers, 29% in those who have never smoked). Fifty-eight percent thought that their health behaviour will have no influence on their patients, as it is hardly noticed by them. On the other hand, more than half of the physicians did maintain that their life-style outside their work was to a certain extent relevant to their patients. Overall, one has the impression that physicians rate the importance of their own health behaviour rather low. This is especialy the case for smokers.

statement

Percentages

of an exemplary

The Netherlands.

with

of the statements

on

health

Current smokers n = 149

FWIIler smokers n = 140

Never smoked n = I45

Total n = 434

36 64

58 36

55 29

49 44

72

51

50

58

53

39

37

43

1989, (questionnaire

data).

Smoking among physicians and students Table 4. General

practitioners,

consultants,

medical students and H.P.M. students indicated with each of the statements on legal measures

Statement Smoking in A total ban The sale of The selling Percentages

public places should be restricted. on the advertising for tobacco is desirable. tobacco products should be banned for persons under sixteen. price of tobacco products should be increased considerably. of respondents

strongly

agreeing with the statement,

The Netherlands,

measure than their non-smoking colleagues, still 74% of the smokers agreed. About half of the respondents is strongly in favour of a total ban on the advertising of tobacco products. A ban on the sale of tobacco products to persons under 16 was almost equally supported by half of the respondents. Only about one third of the respondents was strongly in favour of a considerable increase in the selling price of tobacco products. Current smokers were significantly less enthusiastic about such a measure than non-smokers. On the whole physicians were more in favour of legal measures than were students. DISCUSSION

There is probably substantial under-estimation in reported consumption of tobacco [13]. Therefore, only approximate conclusions can be drawn about the subjects’ tobacco consumption. In 1983 56% of Dutch general practitioners were self-reported smokers [14]. This was more than the percentage of smokers in the general population at that time. In spite of the downward trend, the proportion of male general practitioners in the Rotterdam area who smoke [2, 141 is still higher than the proportion of men in the general population who smoke and particularly higher than that in the population of the highest socio-economic bracket. Male general practitioners do not set a positive example, unlike their female colleagues. Lower prevalence of smoking in students in comparison with the general population suggests that in the near future a new generation of physicians will better be able to fulfil an exemplary role as far as smoking behaviour is concerned. The smoking behaviour of Dutch physicians compares unfavourably with that of their colleagues in many other countries [24,9]. The downward trend in smoking among physicians is due to a higher percentage of physicians who have never smoked, rather than to an increased percentage of those who have given up smoking. Nowadays, younger physicians are less inclined to start smoking than their predecessors. Their older colleagues who smoke are having problems giving up the habit, just like people who do not belong to the medical profession. The good intentions of physicians to stop smoking within the next 5 years should perhaps be taken seriously, but one should not be too optimistic about the results. For this reason, the disappearance

821 the extent to which they agreed or disagreed

Current smokers n =231

Former smokers n = 187

Never smoked n = 357

Total n = 775

74 52 49 28

90 57 45 37

90 48 47 39

85 51 47 35

1989, (questionnaire

data).

of the smoking physician is mainly a generation effect, coinciding with a dramatic change in the social climate with respect to tobacco consumption. The shift from cigarettes to other types of tobacco is at best a partial improvement, while cancer risks for cigar and pipe smokers are still higher than those for non-smokers [ 151.Relative risks of the different patterns of smoking have been investigated several times. Although the results varied greatly among studies, it is obvious that the relative risk depends on the type of tobacco product, the amount of tobacco consumption and the degree of inhalation [ 151. The current smoking rate in Dutch medical students is still much higher than in American medical students [7, 81. Dutch medical schools should consider action aimed at reducing smoking rates in students. The mean age at which respondents started smoking is between 17 and 19. These figures are considerably higher than in many other countries [16]. Differences among countries suggest social and cultural differences. A possible explanation would be that in the types of secondary school that prepare for scientific education, smoking is quite unusual. Studying at university implies for most students moving to another town. This will bring them in a position with perhaps more stress and less external control than at their parents’ home, inducing some of them to start smoking. Smoking is not a ‘neutral’ subject, especially for the young: for some it is socially reprehensible, for others, especially in the Dutch Students Clubs, smoking is an accepted and usual form of behaviour. Not more than half of the physicians recognize their exemplary role with respect to their patients concerning smoking. Here, physicians who smoke obviously find themselves in a conflict situation. As physicians should know, smoking is at the moment the most important preventable risk factor with regard to health [17]. On the one hand, a physician who smokes must advise patients against smoking; on the other hand, the physician is setting a bad example him/herself. The fact that not more physicians see their own behaviour as exemplary for their patients may have to do with a widespread fear in The Netherlands to ‘moralize’, to judge and to influence other people’s behaviour. This factor may also be responsible for the fact that not more than half of the physicians are in favour of a ban on advertising for tobacco products and sales restrictions. Also,

HELEEN M. DEKKER ef nl.

822

respondents may consider this type of measure futile. This would not explain however why a considerable increase in selling price, which is proven to be effective, finds so little support. A serious approach to the habit requires sufficient knowledge of the determinants of smoking and possibilities for its treatment. Health professionals should receive specific training on how to help patients who want to stop smoking; smoking prevention should be included in the medical curricula. We have already seen a downward trend in the prevalence of smoking in male physicians in The Netherlands [2, 14, 181. The findings reported here, concerning the current smoking habits of medical students, justify the assumption that the proportion of Dutch physicians who smoke will continue to decrease considerably: in the future the medical profession will better fulfil its exemplary role with regard to smoking. This probably will contribute to a further reduction of smoking in the general public, which, in its turn, will lead to a further improvement of public health. Acknowledgements-We

would

like to thank

Dr Roberto

Masironi of the World Health Organization, Geneva, for making the questionnaire available and for his kind support. We are grateful to the local General Practitioners Association (Rijnmond region), the Board of the University Hospital of Rotterdam, the Faculty Board of the Erasmus University of Rotterdam, Faculty of Medicine and Health Sciences, for the realization of this research and to the respondents for their kind cooperation.

J.

C.

van.

Roken:

Medisch Contact 9, 259, 2. Adriaanse H. P.. van

Volksvijand

nummer

&&n.

1981.

Reek J. and van Nederlandse huisartsen

1989.

prevalence.

Am.

J. pub/.

Hlth

79,

Tessier J. F., Freour P., Crofton J. and Kombou L. Smoking habits and attitudes of medical students towards smoking and antismoking campaigns in fourteen european countries. Eur. J. Epidemiol. 5, 311-321, 1989.

Maynard L., Goldberg R., Ockene J., Levy B., Howe J. and Dalen J. Behaviors and attitudes among medical

students concerning cigarette smoking and alcohol consumption. J. med. Educ. 61, 921-922, 1986. Conard S., Hughes P., Baldwin D. C., Achenbach K. E. and Sheehan D. V. Substance use by fourth-year students at 13 U.S. Medical Schools. J. med. Educ. 63, 747-758,

1988.

Adriaanse H. P. and van worldwide. A review of tobacco consumption in 1951-1983. In Tobacco e

Reek J. Physicians’ smoking 100 surveys on physicians’ 27 countries in the period

Personale Sanitaria, Proceeding Inlemational Symposium Tobacco Smoking Among the He&h Professionals (Edited bv Marcolongo A. and

Andreatta contra i 10 Stichting Stichting 1990. II World

L.); pp. 135-150. Lega-Italiana pe; la Lotta Tumori, Venice, 1986. Volksgezondheid en Roken. Jaaruerslag 1989. Volksgezondheid en Roken, ‘s-Gravenhage,

Health Organization. Guidelines for Ihe Conduct (f Tobacco Smoking Suruyys Among He&h Prqfissionals. WHOjSMO 84.1.. Geneva, 1984.

and 12 Breslow N. and Day N. E. Indirect standardization multiplicative models. J. Chronic Dis. UI, 289.-303. 1975. 13 van Reek J., Hagen H., Adriaanse H. and van Iwaarden T. Changes in Dutch tobacco production since World War II: implications for an anti-smoking policy. Hlth Policy 11, 233-238, 1989. 14. Adriaanse H. P., van Reek J. and Metsemakers J. Smoking behavior of Dutch general practitioners in the period 1977-1983. Stand. J.prim. Hlth Core4, 151 156. 15. Gsell 0. R. and Abelin relation to lung cancer Cancer

Rabier J. R. in de periode

Rookgedrag 1983-1988. Ned. Tijd.whr. Geneeskd. 134, 1502~1505, 1990. 3. Davies P. D. 0. and Ra_jan K. Attitudes to smoking and smoking habit among the staff of a hospital. Thor& 44, 378-381, 4. Fortman

smoking

152-157,

1986.

REFERENCES I. Es

6

cigarette

1989.

S. P.. Sallis J. F., Magnus P. M. and Farquhar J. W. Attitudes and practices of physicians regarding hypertension and smoking: the Stanford five city project. Prev. Med. 14, 70-80. 1985. 5. Pierce J. P. International comparisons of trends in

Inst. 48,

T. Cigar and pipe smoking in and excess mortality. J. nom.

1795-1803,

1972.

16. Vlajinac H., Adanja B., Jarebinski M. Cigarette smoking among medical students in Belgrade related to parental smoking habits. Sot. Sci. Med. 29, 891-894. 1989. 17. Coebergh J. W. W. Tabaksgebruik-Genot. hinder, ziekte en beleid. In Epidemiologie en Gezondheidsbeleid (Edited by van der Maas P. J., Hofman A. and Dekker E.), pp. 74489. Samson Stafleu. Alphen aan den Rijn, 1989. H. P. and van Reek J. Ar/s en Roken: 18. Adriaanse Consumprie, Opinies en Attituden. (GVO Cahier 26) Rijksuniversiteit Limburg. vakgroep GVO, Maastricht, 1987.