The aim of this study was to evaluate Berlin's medical students' smoking habits, knowledge about smoking and attitudes toward smoking cessation counseling. Our investigation found several important results which are worth to discuss.
One quarter of all medical students surveyed in our study were current smokers, a rate similar to that of the general population . It seems that medical students' undergraduate education about the hazards of smoking have relatively little impact on smoking behavior . Various cross-sectional investigations have suggested that there is an alarming worldwide trend for smoking rates to increase during students' time at medical schools [29–34]. To discourage smoking among medical students, it is essential to introduce teaching on tobacco dependence and cessation early in the course of medical school. Tobacco curriculum should continue throughout the entire medical degree as it is difficult to determine whether this effect directly reflects students' seniority, age or both . Contrary to findings from other studies of medical students, the current smoking status did not vary significantly between genders . Moreover, smoking prevalence among women was higher in the present study than reported by most studies in other countries [17, 37]. However, these findings are consistent with other studies conducted in German medical schools [38, 39].
Although the majority of medical students correctly identified rates of smoking-attributable lung cancer and COPD, they lacked sufficient knowledge about tobacco and its effects. At least one-fifth of the participants underestimated the rate of smoking-related COPD. Furthermore, students in our study, as in the Göttingen sample , greatly underestimated smoking-related mortality and disease - smokers gave significantly less appropriate answers than non-smokers. This finding is consistent with other studies in this field [16, 17]. Underestimation of smoking-attributable morbidity and mortality could have a negative impact on medical students' efforts to counsel smoking patients in the future. The belief that smoking is not life-threatening or, at least, not too hazardous might undermine future physicians' promotion of smoking cessation.
Of all the harmful substances contained within tobacco smoke, a large proportion of students in our study believed that nicotine alone is responsible for coronary artery disease (Figure 2). These results are consistent with the study by Raupach and colleagues  and of particular interest given that many of German medical textbooks use misleading terms for the health effects of smoking . The words "smoking" and "nicotine" are used synonymously within the context of cardiovascular risk factors. This misuse  erroneously suggests a casual relationship between nicotine and coronary heart disease and may explain German general practitioners' hesitancy to recommend NRT.
Students in our study knew little about cessation techniques. As in the Göttingen sample , subjects rated "willpower alone" as the most effective of all tobacco cessation methods, rating it above NRT alone or cognitive behavioral support programs plus NRT, although the former has been shown to be effective and safe  and the latter has demonstrated optimal cessation outcomes [42–44]. Moreover, students rated advice from a general practitioner similarly to self-help material and acupuncture, despite the fact that evidence does not support the efficacy of acupuncture as a smoking cessation treatment [43, 45] and research has shown that GP consultations with patients yield one-year cessation rates of 3-10% [42, 46]. This underestimation of physicians' ability to promote smoking cessation may adversely affect their professional practice later in life. Future general practitioners who attach little importance to physicians' advice are unlikely to make an effort to provide smoking-prevention counseling once they have become general practitioners themselves . The finding that smokers and non-smokers assess the effectiveness of cessation methods differently  could be replicated to a certain extent in our study. Reasons for this may lie in subjects' different levels of education in the two studies (academic versus non-academic sample).
Apart from knowledge about effectiveness of different cessation methods and smoking-related morbidity and mortality, a person's experience and smoking status may have an influence on the counseling of smoking patients. In our sample, students' perceptions of the effect of smoking on longevity differed with respect to their personal smoking habits. These results are consistent with the study by Raupach and colleagues . The results of the British Doctors' Study  indicate that a nonsmoker's chance of living to the age of 90 years (24%) is six times greater than that observed in smokers (4%). The two questionnaire items related to this study assessed students' personal experiences rather than their knowledge about smokers' and nonsmokers' life expectancies. More smokers than nonsmokers in our sample stated that they personally knew lifelong smokers (Figure 4). Research suggests that the smoking habits of parents may have an influence on whether or not a medical students smokes [34, 39, 49]. Coming from families or communities with higher smoking prevalence could increase one's chances of personally knowing a 90-year-old lifelong smoker. In addition, cognitive dissonance may also play a role for smokers . However, it may also reflect an excessively optimistic view of smoking held by smoking medical students, which might eventually undermine their own commitment to promoting smoking cessation among their patients.
A large proportion of students thought that they did not have adequate skills to counsel patients about smoking. In fact, only half of them reported actually having recommended smoking cessation to a patient, possibly due to a perceived lack of competence pertaining to clinical behavior. Similar trends have been found among practicing doctors. Although 70% of smokers visit a general practitioner annually, most are not advised or assisted in smoking cessation matters . Differences found between perceived counseling skills of smokers and non-smokers were surprising because research suggests that more nonsmokers than smokers are active in smoking cessation counseling [52–55]. This difference may be due to the addictive nature of smoking. Smokers may feel more apt to put themselves in a smoking patients' position than non-smokers. A large number of smokers in our study indicated that they wished to stop smoking and about half of them had made one or more quit attempts. Further research is needed to explore this consideration. The finding that female nonsmokers rated their competence in tobacco cessation counseling significantly lower than their male colleagues does not reflect actual differences but rather a possible negatively distorted self-perception towards reality [56–59]. Adequate training may help overcome this misperception and increase female medical students' self-confidence in their ability to provide smoking cessation advice or counseling.
The current study is subject to certain methodological limitations. First, our sample only consists of fifth-year medical students. Therefore, a comparison between students in preclinical and clinical years regarding smoking habits, smoking-related knowledge and students' perceived competence was not possible. Second, smoking status of subjects was assessed only by means of self-report, potentially rendering our results less reliable. However, the use of confirmatory carbon monoxide or cotinine tests was impracticable for such a large sample. Because the survey was anonymous and completely voluntary, one can assume that smoking status was reliably captured. Third, the design of our study was cross-sectional and this form of research can only provide a snapshot of the situation in the sample.
Nevertheless, the results of our study support the findings of Raupach and colleagues  and indicate an urgent need to better equip medical students to treat smoking patients. One way to counteract their insufficient knowledge is to provide adequate education in the medical curriculum, especially because medical school is an ideal time for training in smoking cessation techniques . Roche and colleagues demonstrated significantly improved skills of medical students in smoking intervention after such training. This effect was not dependent on the mode of delivery . Smoking-related knowledge of medical students in Hong Kong increased after a three hour seminar on tobacco .
Research suggests that role-playing, computer-assisted instructions, group discussions , and simulated patients  are useful methods in developing smoking cessation intervention skills. For this reason, a tobacco module should be integrated into the curriculum of every medical school, thus providing medical professionals with universal training in nicotine dependence intervention and smokers with healthcare professionals skilled to adequately assist them in their quit attempt.