Smoking Knowledge, Attitudes, Behavior, and Associated Factors Among Chinese Male Surgeons

A survey of male surgeons’ knowledge, attitudes, and behavior toward smoking was conducted from July 2004 to October 2004 in six cities in China: Guangzhou, Chengdu, Wuhan, Tianjin, Harbin, and Lanzhou. These cities represent six different regions in China: south, southwest, central, north, northeast, and northwest. In each city, five hospitals were randomly selected within the provincial, city, and district levels. A total of 949 surgeons participated in the survey: 13.3% of the surgeons were female but only 1.6% (n = 2) of them were current smokers, which is consistent with female smoking rates in China [5, 16]. Consequently, this study focused only on the smoking knowledge, attitudes, and behavior of the 823 male surgeons. When the term “respondents” is used, it refers only to the male surgeons in the study.

This study was approved by the Institutional Review Boards of the Chinese Center for Disease Control and Prevention and the University of California, Berkeley. Participants were contacted by telephone and asked to participate in a face-to-face interview and complete a self-administered questionnaire. The face-to-face interviews were conducted by staff members of the Chinese Center for Disease Control and Prevention. The questionnaire consisted of four sections with 112 total questions. The first section asked questions about the male surgeons’ smoking cessation counseling behavior. The second section asked questions about their smoking and their quitting smoking behavior. The third section contained 52 scaled questions that assessed the surgeons’ knowledge about their role in smoking prevention and control, smoking-related and passive smoking-related diseases, and methods of smoking cessation. The fourth section contained six demographic questions (age, years of practice, education, position, number of patients per day, and monthly income). To ensure the quality of data collection, 10% of the questionnaires were examined by a quality control team in each city every day.

All analyses were conducted with STATA, version 9.0 (Stata Corporation, College Station TX, 2003). A descriptive analysis was performed for demographic features (Table 1). Differences in proportion were assessed by the Pearson chi-square test (Table 2). Previous studies have shown that lack of knowledge and inappropriate attitudes toward smoking can influence physicians’ smoking behavior [17, 18], but these studies were not conducted specifically with Chinese male surgeons. The present study used three logistic regression models to analyze factors related to the participating Chinese male surgeons’ smoking status, smoking behavior in front of patients, and smoking cessation activity (Tables 3–5). Odds ratios (OR) and the corresponding 95% confidence intervals (CI) were computed to assess the strength of association. For the logistic regression analyses, the respondents’ smoking status and smoking behavior among current smokers were dichotomized into current smokers (daily smokers and occasional smokers) and nonsmokers (past smokers and never smokers), and into smoking and not smoking in front of patients, respectively. Smoking cessation activity was also dichotomized into (1) always (always or often) and not always (sometimes or rarely or never) asks if a patient smokes, and (2) always (always or often) and not always (sometimes or rarely or never) advises smokers to quit smoking. Analyses accounted for clustering at the city and type of hospital.

This study focused on four groups of explanatory variables in the logistic regression models: (1) professional background, which includes the surgeons’ years of practice, education, position, and city; (2) knowledge about the relationship between active smoking and health, and between passive smoking and lung cancer; (3) attitudes regarding whether “health professionals should serve as role models for patients” and “all indoor smoking in hospitals should be prohibited”; and (4) smoking behavior, measured both as smoking in front of patients and the number of cigarettes smoked per day. The key variables of interest are in the last three groups, as they are more modifiable than demographic characteristics. The knowledge variables previously had been shown to differ significantly between physician smokers and nonsmokers but had not been analyzed for their effect on smoking cessation practices [5]. Responses related to knowledge and attitudes were grouped into “agree (related)” and “probably or not agree (probably related, not related, unknown)” and “oppose (strongly oppose, oppose, neutral)” and “support (supportive, strongly supportive),” respectively. A P value of <0.05 was considered statistically significant, with all P values being two-tailed.

Table 1 shows the demographic data for the 823 male surgeons participating in this study. Their mean age was 38 years, and 59.2% (n = 487) had worked more than 10 years. Most (n = 669; 81.3%) had at least a bachelor’s medical degree, and 33.8% (n = 278) held an associate or senior position. Among all of the respondents, 19.3% were from Guangzhou, 17.6% from Chengdu, 19.4% from Wuhan, 12.9% from Tianjin, 13.7% from Harbin, and 17% from Lanzhou. The chi-square analysis indicates that significant differences in smoking status were found between subgroups of age, years of practice, education, position, and city.

Table 2 describes the knowledge and attitudes of the male surgeons toward smoking. Current smokers were significantly less likely than nonsmokers to believe that smoking is harmful to health (49.7% versus 70.7%; P < 0.05) and that passive smoking is related to lung cancer (82.3% versus 89.8%; P < 0.05). Current smokers also were significantly less likely (P < 0.05) than nonsmokers to support the views that (1) health professionals should serve as role models for their patients (73.7% versus 87.1%; P < 0.05) and (2) all indoor smoking in hospitals should be prohibited (68.6% versus 86.9%; P < 0.05).

Current smokers who had ever smoked in front of patients, compared to those who never had done so, were significantly less likely to believe that smoking is harmful to health (42.7% versus 55.1%; P < 0.05) and that all indoor smoking in hospitals should be prohibited (61.3% versus 74.2%; P < 0.05). Current smokers who had ever smoked in front of patients, compared to those who never had done so, were less likely (although not significantly) to know that passive smoking is related to lung cancer or to believe that health professionals should serve as role models for patients (81.5% versus 82.7 %; P > 0.05; 71.0% versus 76.1%; P > 0.05).

Smoking status (Table 3) was significantly associated (P < 0.05) with the respondents’ professional background (education, position, city), and also with their knowledge and attitudes toward smoking. Male surgeons who did not agree that active smoking is harmful to one’s health (OR = 2.48) or that passive smoking is related to lung cancer (OR = 2.01) were more likely to smoke than those who agreed with these statements. Male surgeons who did not agree that health professionals should serve as role models for their patients (OR = 1.74) or that indoor smoking in hospitals should be prohibited (OR = 3.09) were more likely to smoke than those who agreed with the respective statements.

This study found that professional background (years of practice), knowledge, and attitudes toward smoking, as well as smoking intensity, were significant predictors (P < 0.05) of having ever smoked in front of patients (Table 4). Male surgeon smokers who had practiced more than ten years were more likely to have smoked in front of patients. Male surgeons who did not completely agree that smoking is harmful to health were more likely (OR = 1.64) to have smoked in front of patients than those who completely agreed. Male surgeons who did not believe hospitals should be completely smoke-free were more likely (OR = 1.81) to have smoked in front of patients than those who believed that hospitals should be completely smoke-free. Male surgeons who smoked more than a half pack of cigarettes daily were more likely (OR = 2.67) to have smoked in front of patients than those who smoked a half pack or less of cigarettes daily. Knowledge about smoking harms and attitudes about role modeling were not significant predictors of having ever smoked in front of patients.

In this study, only 25% of the participating Chinese male surgeons always asked patients about their smoking status. Among those surgeons who never or seldom asked their patients about smoking status, the main reason given for not asking was “patient’s condition is not related to smoking” (45.3%). In addition, only 27% of the male surgeons always advised their patients to quit smoking. Among those respondents who never or seldom advised their patients to quit, the main reason given was the belief that “they could not influence a patient” (39.4%).

This study found (Table 5) that the male surgeons’ “ask” behavior was significantly associated (P < 0.05) with their professional background (education, position, city) and their knowledge about smoking. Those who completely agreed with the statement “smoking is harmful to health” were 1.45 times more likely (OR = 0.69) to always ask if a patient smokes than those who did not completely agree with this statement. Knowledge that passive smoking is related to lung cancer and attitudes about role modeling and smoking bans in hospitals were not significantly associated with the male surgeons’ “ask” behavior.