Results

The total sample of 122 220 persons with complete data included 45 995 ever smokers of whom 4336 were former smokers. Table 1 provides the numbers of current smokers by age group and the percentage distributions as those who had never contemplated quitting at the time of interview, those intending to quit, and relapsed quitters. Overall, about 71.8% of current smokers had no intention of stopping smoking; 16.6% of current smokers said they intended to do so but had not taken any action, and 11.6% had once quit but relapsed by the time of the survey. The patterns were similar in men and women with regard to cessation intent: in men, 71.8% not intending to quit, 16.6% intending to quit, and 11.6% relapsed. For women, the corresponding figures were 71.5%, 16.3%, and 11.6%, respectively. Stated intent to quit was associated with age group among males. In general the older male smokers had a higher proportion of quitters, and a lower proportion of persons not intending to quit. The profile was not associated with age group among females.

There were differences in the intent-to-quit profile across groups with different educational levels and occupations (figs 1 and 2). With increasing educational level, the proportion of respondents not intending to quit declined, from 64.1% to 46.3%, and the percentages of respondents who had quit or intended to quit increased. The profile was most favourable for those with college education, but even for this group almost half had never considered giving up their smoking. Of the occupational groups considered, 69.1% of farmers, 64.5% of workers, 62.8% of floating people, and 62.0% of officers did not intend to give up smoking (fig 2). The medical workers and teachers had slightly higher rates of cessation, but about half did not intend to give up smoking.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Percentage distribution by education of relapsed former smokers and former smokers at survey by duration of quitting.      

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Percentage distribution by occupation of relapsed former smokers and former smokers at survey by duration of quitting.

About 20% of ever smokers had stopped smoking at least once, but about half had relapsed. The prevalence rate of reported cessation increased with age (fig 3) above the ages of 25–29 years. The survey questions allowed us to classify the quitters into strata of duration of successful cessation: relapse, < 6 months, 6–24 months, > 24 months, and unknown duration of quitting. Overall, the corresponding percentages of the former smokers were 20.8%, 16.6%, and 37.2%, respectively, and 25.4 % did not give a duration of quitting.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

National projections for numbers of smokers by smoking status, intent to quit, and duration of quitting, based on 1996 national prevalence survey. M, males, F, females.

Patterns of cessation varied with educational level (fig 1). The percentage of persons who had ever taken action to quit was highest—about 33%—for the most educated participants compared with all other groups, for which this percentage was around 20%. Long term quitting was also more common among the most educated (7.1%). There was little variation across the four less educated groups. The proportions of relapsed quitters and former smokers were comparable in all of the groups.

Similar analyses were carried out among selected occupational groups (fig 2). Among these occupations, farmers, workers, and floating people had the lowest rates of quitting. Medical workers, teachers, and officers had the highest rates of quitting, including both former smokers and relapsed current smokers, and the highest proportions intending to quit. However, the percentages of former smokers were still low in both groups.

Of participants from urban and rural areas, 11.7% and 11.4%, respectively, had tried to quit. Relapse rates were quite similar, at 9.0% (urban) and 9.7% (rural), respectively.

Respondents who had quit were asked why and allowed to select reasons from a list. Illness was the most common reason overall, selected by 47.3% of participants who had quit. The second most common reason was illness prevention (34.1%). Family pressure and economic considerations were given by 15% and 11% of responders, respectively. Only a few chose health education and environmental limitation. The reasons for quitting varied with age (table 2). Not surprisingly, the presence of illness was selected more frequently with increasing age. Younger persons more often gave family pressure as the reason for quitting. Costs and the consequences of health education were offered less frequently as the explanation, with restrictions on smoking cited far less frequently. There were also different distributions of reasons for quitting among educational levels, occupations, and income. In general, more smokers with lower educational levels gave up smoking because of illness and cost. For example, among smokers with no schooling, 65% gave illness as a reason, and 13% said cost considerations. The same figures among smokers with at least college education level were 37.9% and 5%, respectively. The trend was the same by income level. Among smokers earning less than 50 RBM per month, 57% gave illness as a reason, while 19% said cost. Those same figures among smokers earning more than 1000 RBM per month were 33% and 0%, respectively. Among farmers, 66% of quitters listed illness, while 16% said cost. The same figures among medical workers were 22% and 0%, respectively, and 40% chose prevention of disease.

Discussion

In approaching these survey data, we have viewed smoking cessation not as a dichotomous process but rather as a continuum involving steps that begin with contemplation and culminate with successful cessation. While our questions do not precisely follow those used to establish stages in the stages of change model, the parallel is close and we link to the model as a framework for interpreting the data or reported intent to quit. We found that the majority of China's current smokers (72%) are not intending to quit and are presumably in the precontemplation stage. About 20% have taken action to quit, but half had relapsed. The findings show that China faces an enormous challenge in persuading smokers to stop smoking. Of the approximately 320 million ever smokers in China, only 11 million—less than 4%—have stopped smoking for at least two years, and the majority—205 million—have never contemplated quitting. The number who have relapsed exceeds the number of former smokers who have successfully quit. These findings should figure centrally in formulating smoking cessation programmes, whether at the population or individual levels.

These findings contrast sharply with the high proportions of former smokers in the USA, the UK, and other developed countries.1 In fact, in the USA the number of former smokers is now approximately equal to the number of current smokers.13 Also in the USA, most smokers report wanting to quit and the majority have made at least one attempt to quit.12 13 By contrast, only a small proportion of smokers in China have successfully quit and the relapse rate is high (table 1 and fig 3). The stages of change model places smokers into five categories with regard to cessation: precontemplation, contemplation, preparation, and action, maintenance or relapse.9 While the 1996 survey questions did not have questions to separate contemplation and preparation, the results show that the majority of smokers are not intending to quit (table 1). Recent US data by contrast show that the majority of smokers report being either in the contemplation or preparation stages.14The findings in China indicate that cessation programmes need to focus, at least initially, in moving more smokers to consider quitting—that is, beyond the stage of precontemplation.

Lack of knowledge of the harmful effects of smoking may partially explain the high proportion of participants who do not intend to quit. The 1996 survey showed that knowledge of the many causal associations of smoking with disease was limited. The majority of smokers in the national prevalence survey did not acknowledge the hazard of smoking2; 68.6% of smokers said that the risk of smoking was negligible or small and 8.1% did not know about any risk. Knowledge of specific diseases caused by smoking was also low, with only a few percent aware that smoking causes heart disease. In the USA and other countries where cessation rates are higher than in China, the majority of smokers acknowledge the health risks.12 15Approaches for increasing knowledge among smokers on the health risks of smoking should be developed and tested in China.

The survey also showed that cost did not deter most smokers from smoking. Only about 10% of smokers identified cost as a reason for quitting, and taxes on cigarettes in China are low in comparison with other countries.16 Participants with lower income, lower educational level, youths, workers, and farmers more often reported cost as a reason for quitting. Price increases appear warranted as a strategy to encourage cessation, even though smokers are already spending around a quarter of their income on cigarettes. One recent analysis suggests that the effect of a price increase in China on smoking might be greater than in developed countries and that there should be a net fiscal gain by the government.16 17Experience from other countries shows that smoking by youth is particularly sensitive to price and increased prices would be expected to deter young people in China from smoking. It would be timely for the Chinese government to evaluate its current tobacco taxes and consider raising them. In other countries, increasing prices do reduce smoking somewhat, particularly among youth.18 However, the relevance of these data to China is uncertain and research specific to China on this topic is needed.

Studies in other countries have shown the key role that physicians can play in smoking cessation19 and strategies have been devised to encourage anti-smoking counselling by physicians.19 Smoking by physicians in China contrasts sharply with many western countries where very few physicians now smoke. Medical workers who smoke are a major barrier to cessation efforts because people will not believe health education messages when they see physicians and other health care workers smoking. In China, the smoking patterns of physicians are similar to those of other occupational groups, although in this analysis we found that medical workers have a somewhat more favourable quitting pattern than others (fig 2). It is doubtful that Chinese physicians who smoke offer strong advice about cessation to patients. Researchers need to know more about the messages given to Chinese patients by their physicians on smoking and professional medical societies in China should discourage smoking by their members.

The high relapse rate indicates a need for effective methods for Chinese smokers for maintaining cessation. Some current methods use traditional Chinese medicine including acupuncture and herbs, with many reports on cessation methods in the Chinese literature.20Nicotine replacement therapy and other pharmacologic approaches have not yet been widely used and these methods need specific evaluation in China.

A potential limitation of this survey is its use of self reporting of cigarette smoking without biochemical validation. In other countries, comparison of self report with biochemical markers has shown that some persons say they are former smokers when nicotine levels are consistent with active smoking.12 One recent study of southeast Asian immigrants in central Ohio showed misclassification of self reported smoking, with salivary cotinine as the standard.21 The tendency to underreport was greater among women. The relevance of these data to our findings is uncertain but replication of a study of this design in China would be useful. Additionally, we have interpreted the survey data within the context of the stages of change model. The appropriateness of this model to smoking in China has not been addressed specifically.

The survey findings suggest that a two pronged national strategy to increase smoking cessation is needed. First, smokers must be mobilised to contemplate quitting and then to take action. Second, effective and evaluated smoking cessation strategies need to be developed. Within the framework of a comprehensive model for prevention, many opportunities exist for programmatic action: media campaigns to increase knowledge, increased tobacco taxes, education of health care professionals to strengthen their role in encouraging and sustaining cessation, enhancing support from family and friends, and developing and evaluating smoking cessation methods. Policies now in place leave these opportunities largely unaddressed. The current package warning, “Smoking is harmful for your health,” has little impact, according to this survey.