Smoking has long been identified as a risk factor for coronary heart disease (CHD) in the West (U.S. Department of Health and Human Services [USDHHS], 2004). The harmful effects of smoking on coronary heart disease in Asians have also been supported by research (Asia Pacific Cohort Studies Collaboration, 2005). The prevalence of cigarette smoking in Taiwanese men was 39.56% in 2006, significantly higher than that in women (4.12%) principally due to culture norms (Taiwan Bureau of Health Promotion, 2007). The high rate of smoking among males may partially explain why heart disease is the third leading cause of death for males in Taiwan (Department of Health, Executive Yuan, 2008).
Smoking cessation has promising effects in terms of decreased subsequent clinical events and mortality for CHD patients (USDHHS, 2004). However, in Western countries, about 60% of CHD patients who received no intervention continued smoking or relapsed (van Berkel, Boersma, Roos-Hesselink, Erdman, & Simoons, 1999). In Taiwan, the few studies that have examined smoking cessation rates in CHD patients are outdated and focused only on cardiac surgery patients (Chung, 1994). As non-surgical treatment options have increased in prevalence in recent decades, smoking cessation rates may vary based on treatment option taken. A US study showed that the abstinence rate for CHD patients differed by types of procedures: 55% for coronary artery bypass surgery (CABG), 25% for angioplasty patients, and 14% for those who received angiography (Crouse & Hagaman, 1991). As most CHD patients in Taiwan receive angioplasty and medical treatment rather than CABG, it is anticipated that a substantial proportion of male CHD patients in Taiwan may still smoke.
Very little is known about factors associated with successful smoking cessation and maintenance of abstinence in Taiwanese CHD patients. Hospitalization for cardiac events may be a stimulus to smoking cessation and the period of hospitalization can, therefore, be seen as a teachable moment. In order to use the opportunity of hospitalization to deliver effective smoking cessation programs for those patients who still smoke in Taiwan, healthcare providers, especially nurses, must first understand what factors motivate and impede patients in the smoking cessation process.
Factors that contribute to smoking cessation have been examined in several studies conducted in Western countries. Subjects who are older or in a higher socioeconomic status are more likely to quit smoking. Heavy smokers are more likely to have less initial cessation as well as less long-term abstinence (Atterbring et al., 2004; Velicer, Redding, Sun, & Prochaska, 2007). Experiencing several quit attempts and maintaining quit attempts for relatively long periods of time are both related significantly to successful abstinence (Velicer et al., 2007). Nicotine dependence is negatively associated with short-term cessation rates (6 months) (Johnson, Budz, Mackay, & Miller, 1999).
Disease-related factors have also been identified as having a significant relationship to the smoking cessation process. More severe diagnoses, more complicated treatments, and longer hospital stays have all been found to increase smoking abstinence rates (Crouse & Hagaman, 1991; Rigotti, Singer, Mulley, & Thibault, 1991). Research findings also indicate that outcome expectancy and selfefficacy are related to the smoking cessation process. Outcome expectancies are strongly related to a motivation to quit smoking and may have a link to smoking abstinence (Norman, Velicer, Fava, & Prochaska, 1998). Self-efficacy in not smoking is associated with short-term abstinence (Gulliver, Hughes, Solomon, & Dey, 1995).
Social influences also play important roles in the smoking cessation process. Social support may be related to smoking abstinence and the length of maintenance of abstinence (Gulliver et al., 1995). In terms of a social contagion (defined as the number of smokers in a network), the more smokers there are in a person's social network, the more difficult it is for that person to remain abstinent (Johnson et al., 1999).
Additionally, physical environment influences smoking abstinence. A smoking ban in worksites has been found to have some impact on reducing daily cigarette consumption rates (Farrelly, Evans, & Sfekas, 1999). Furthermore, smoking restrictions at home may be associated with daily cigarette consumption, motivation to quit, and number of quit attempts (Farkas, Gilpin, Distefan, & Pierce, 1999).
A conceptual framework was developed for this study (Figure 1) based on a literature review and framed within the ecological model proposed by Sallis and Owen (1997), which stresses that behaviors are not only influenced by intrapersonal factors and social and cultural variables, but also significantly affected by an individual's physical environment. Smoking status after hospital discharge was predicted by three sets of variables: personal characteristics, social influences, and environmental restrictions. Personal characteristics included demographics, smoking and quitting histories, degree of nicotine dependence, disease severity and cognitive processes. Demographics included age and education level. Smoking and quitting histories consisted of the number of cigarettes smoked per day and the number of previous quit attempts. Disease severity was assessed based on diagnosis, treatment prescribed, and length of hospital stay. Cognitive processes included self efficacy in not smoking and outcome expectancy for smoking. Social influences on smoking cessation consisted of social contagion and social support. Environmental restrictions on smoking included smoking bans in the worksite and at home.
Because diverse smoking status patterns may be present following hospital discharge, subjects included in this study were classified into one of four smoking categories, namely continuous smokers, attempters, continuous abstainers, or non-continuous abstainers (Figure 2). Continuous smokers were those who abstained for less than 24 hours after discharge. Attempters were those who abstained from smoking for at least 24 hours within 3 months after discharge. Continuous abstainers were those who had not smoked a single puff of a cigarette for the entire 3 months after discharge. Non-continuous abstainers were those who had abstained for at least 24 hours, but who did not remain abstinent for the entire 3 months.
The aim of this study was to identify factors proposed in the conceptual model which differentiated attempters from continuous smokers and which differentiated continuous abstainers from non-continuous abstainers.
The research questions for this study were: (1) Which factors proposed in the conceptual framework differentiated attempters from continuous smokers? (2) Which factors proposed in the conceptual framework differentiated continuous abstainers from non-continuous abstainers?
A longitudinal correlational design using two structured questionnaires and two telephone interviews was employed.
Subjects were recruited from the cardiac units of five hospitals in the Taipei area, from March 2004 to October 2004. A total smoking ban policy was in effect in every hospital building. Oral advice on smoking cessation for CHD patients was provided by healthcare providers in each hospital. The criteria for subject inclusion were (1) a diagnosis of CHD (including angina, unstable angina, or acute myocardial infarction) and having been admitted to a cardiac unit in one of the five target hospitals, (2) male, (3) able to read Chinese, and (4) smoking one or more cigarettes per day prior to hospitalization. Because smoking is more prevalent in males than females in Taiwan, only male subjects were included due to availability considerations in recruiting a sufficient number of subjects for this study. Sample size was estimated based on the requirement of at least 10 cases per predictor for logistic regression (Harrell, Lee, Califf, Pryor, & Rosati, 1984). Given that education and treatment were treated as dummy coded variables and each had three levels, a minimum of 170 subjects were required for the total of 17 predictors represented in the conceptual framework. To ensure that an adequate sample size was achieved, we recruited 250 subjects for an anticipated 30% potential loss of subjects at 3-month follow-up.
Potential subjects were identified through a review of medical charts in the cardiac units during hospitalization. Those who received angiography and angioplasty were met on the day of hospital discharge or one day after completion of their procedure. Those diagnosed with acute myocardial infarction or unstable angina and those receiving CABG were approached after their physical conditions had stabilized.
Subjects completed the first questionnaire during their hospital stay. Questions included information on their background, the smoking and quitting histories before hospital admission, nicotine dependence, self-efficacy in smoking abstinence, and outcome expectancy for smoking. At 6 weeks and 3 months, respectively, after hospital discharge, smoking status was obtained through telephone interviews. After the phone contact at 3 months after hospital discharge, a second questionnaire was mailed to subjects. This questionnaire contained questions about social contagions, social support, and smoking restrictions in the environment. A reminder phone call was made if subjects had not returned the mailed questionnaire after 10 days.
Variables and their measurements
All measures and items that were originally developed in English were translated into Chinese and backtranslated into English independently by bilingual translators using an iterative process. This process was used to ensure conceptual and semantic equivalence between the English and Chinese versions. In addition, the face validity of all newly developed Chinese versions of measures and items was assessed by 10 respondents with characteristics similar to those in the proposed sample.
Smoking status after hospital discharge was assessed through two phone interviews at 6 weeks and at 3 months after discharge. Questions in the first phone interview asked subjects to recall the past 6 weeks after hospital discharge and to answer whether they had smoked at least one puff of a cigarette and, if so, when they had smoked their first cigarette during this period. The same questions were asked again at 3 months after discharge.
Demographic measures included age and education level. Smoking history included the average number of cigarettes smoked per day one month prior to hospitalization. Quit attempt was defined as a single period of at least 7 days of abstinence occurring within the past two years. The degree of nicotine dependence before hospitalization was assessed using the Fagerstrom test for nicotine dependence scale (FTND; Heatherton, Kozlowski, Frecker, & Fagerstrom, 1991). Information used to assess CHD severity included the medical diagnosis, cardiac procedures, and duration of the hospital stay; all obtained through a review of subject medical charts.
Outcome expectancy for smoking was measured by the decisional balance scale (Velicer, DiClemente, Prochaska, & Branderburg, 1985), designed to assess decision making for smoking behavior. The decisional balance scale includes a 10-item pro subscale and the 10-item con subscale. Five-point importance response formats were employed. Higher scores on the pro subscale indicate greater perceived positive consequences of smoking and higher scores on the con subscale indicate greater perceived negative consequences of smoking. The internal alpha coefficient was reported as .87 for the pro scale and .90 for the con scale (Velicer et al., 1985). Internal consistency reliability for the present study was .79 for the pro scale and .85 for the con scale.
Self-efficacy in not smoking was assessed using the short form of the Situational Self-efficacy Measure which was originally developed and tested by Fava, Rossi, Velicer, and Prochaska in 1991 (as cited in Fava, Velicer, & Prochaska, 1995). This scale was designed to reflect the confidence of the individual not to smoke across an array of risk situations. The scale consists of three subscales, positive affective/social situations, negative affective situations, and habitual/craving situations. Each subscale is a 3-item, 5-point scale. Higher scores indicate higher confidence not to smoke. The internal reliability alpha coefficients for this study were similar to that reported by Fava et al. in 1991 (as cited in Fava et al., 1995) (.95 for the total score, .88 for the positive affective/social situation, .92 for the negative affective situation, and .84 for the habitual/craving situation). Because these three subscales were highly correlated (r = .80-.83), the total score was used to represent selfefficacy for not smoking in the data analysis.
Smoking-related social support was measured by the short form of the Partner Interaction Questionnaire (PIQ-20; Cohen & Lichtenstein, 1990), designed to measure the frequency of smoking-related interactions between partners when one is trying to quit smoking. The PIQ-20 has two subscales, one for positive and one for negative behavior. Each subscale consists of 10 items. The response format was a five-point scale ranging from never (0) to very often (4). Good internal reliability was shown by a Cronbach's alpha of .89 for the positive subscale and .85 for negative subscale (Cohen & Lichtenstein, 1990). In the present study, to capture influences from different sources of support behaviors in the home, subjects were asked to estimate the frequency of support behavior of all other household members (including spouse and others living at the subject's place of residence). In addition, the option “doesn't apply as I have not tried to quit” was added to the original five-point scale in consideration that those who had not tried to quit for at least 24 hour might not be able to answer the five items in the positive behaviors of family support. For the present study, Cronbach's alpha values were .87 for the positive subscale and .92 for the negative subscale.
Social contagion, the proportion of people in the subject's social network who also smoked within 3 months after discharge, was measured by four items, three of which were adopted from Pan's work (2001) and one developed by the investigator. Four items were used to assess environmental restrictions on smoking. Three items measured level of prohibition of smoking in the workplace. One item, adopted from the work of Farkas and colleagues (1999) (“Which statement best describes the rules on smoking in your home?”) was used to measure the degree of restriction on smoking at home.
Permission of the relevant university's Institutional Review Board and ethics committee in the five target hospitals was obtained. All subjects received oral and written information and provided written consent. Collected data were stored in a cabinet secured by a combination lock and information identifying subjects was destroyed immediately after data collection was finished.
Descriptive statistics were used to describe characteristics of the entire sample. Research questions were analyzed by hierarchical logistic regression, which indicates how the predictors together accounted for variance in the dependent variable by considering both theoretical differences and the temporal order of the three sets of predictors. In addition, a backward stepwise logistic regression was used to reduce the number of predictors in the relatively small sample and to obtain a parsimonious model. The p level for removal was set at > .15 instead of > .05 based on Bendel and Afifi's suggestion (1977) to avoid excluding important variables from the model. Three sets of variables were sequentially entered in the model as follows: (1) personal characteristics (age, education, number of cigarettes smoked per day, number of previous quit attempts, degree of nicotine dependence, diagnosis, treatment, length of hospital stay, outcome expectancy and self-efficacy), (2) social influence factors (social support and social contagion), and (3) environmental restrictions (workplace and home). Each set of variables was added to the variables that remained in the previous step if the p value for the model's Chi-square value between the current step and the previous step was less than .15. After entering variables in each step, backward stepwise elimination was used to delete newly added predictors that did not meet the p < .15 criterion. Variables for which the p value was less than .15 were then retained in all future steps. After entering three sets of variables, backward elimination with the p < .15 criterion was used to obtain the most parsimonious model.
Of 338 patients invited to participate, 250 subjects agreed to participate as subjects in the study. Compared to those who did not participate in the study, those who agreed to participate were significantly younger, with higher education levels and more severe heart disease diagnoses.
At 3 months after discharge, 191 of the 250 recruited subjects finished the study. After estimating data for missing items on scales, a total of 148 subjects were determined as having provided complete data. Possible differences between those with complete data and those without were examined using the variables of age, educational level, the number of cigarettes smoked before hospitalization, diagnosis, treatment, and the length of hospital stay. No significant difference was found between these two groups (all p values > .20). The following descriptive statistics and multivariate analyses are based on the final sample size of 148.
Demographic Characteristics and Clinical Measures
The mean age of subjects was 55.3 years. Just over three-quarters (78.4%) of subjects had finished their education to the junior or senior high school level. Numbers of cigarettes smoked per day ranged from 2.5 to 80 during the one month prior to hospitalization, with a mean of 23.66 cigarettes smoked per day. In the previous 2 years, most patients (69.6%) had not stopped smoking for even one week. Subjects scored an average of 5.0 on the FTND scale, indicating moderate dependence on nicotine (Fagerstrom, Heatherton, & Kozlowski, 1991). Based on admission diagnoses, almost half of the subjects (46.6%) were diagnosed with angina; about one-fifth had unstable angina and one-third had AMI (acute myocardial infarction). Of the four treatments, most patients received angioplasty with a stent (70.3%), while the smallest group received medicine only (7.4%). The average hospital stay length was 5.77 days.
About 44% of subjects (n = 65) remained abstinent 3 months after hospital discharge and are considered “continuous abstainers” (i.e., not a single cigarette puff taken during this period). Thirty-three subjects (22.3%) were continuous smokers, that is, they never stopped smoking for at least 24 hours. Fifty subjects (33.8%) were attempters, that is, they had tried to quit but did not remain abstinent for the entire 3 months. Among those who did smoke within 3 months (attempters and continuous smokers), almost one-half of the subjects smoked their first cigarette during the first 24 hours, and another 30% of subjects smoked their first cigarette within 2 weeks after discharge.
Attempters vs. continuous smokers
Among those variables in the most parsimonious model (Table 1) that differentiated attempters from continuous smokers, self-efficacy in not smoking and positive family support behaviors were significant predictors. The odds of being an attempter increased by 10% for each unit of self-efficacy. With regard to family support for not smoking, the odds of being an attempter increased by 15% for each unit of positive family support.
Continuous abstainers vs. non-continuous abstainers
The most parsimonious model (Table 2) discriminating continuous abstainers from non-continuous abstainers included number of previous quit attempts, length of hospital stay, self-efficacy, positive and negative family support, and smoking bans. Among those variables, length of hospital stay, self-efficacy for not smoking, and positive and negative family support were significant predictors. The odds of being a continuous abstainer increased by 22% for each additional day of hospital stay, and increased by 9% for every additional unit of self-efficacy. With regard to family support for not smoking, the odds of being a continuous abstainer increased by 28% for each one unit increase of positive family support and decreased by 25% for each one unit increase of negative family support.
The cessation rate (43.9%) in this CHD sample is in line with several studies conducted in Western countries that found cessation rates of 31%-59.4% among cardiac patients with no intervention at 3-6 months after hospital discharge (Johnson et al., 1999; Marshall, 1990; Rigotti et al., 1991). However, it was lower than findings from other studies for cardiac patients who had been hospitalized in Taiwan (61%-92.6%) (Chung, 1994; Huang, 1990). The differences might be because those two previous studies involved only AMI and CABG patients and did not include patients with less-severe disease, who were also considered in the current study.
All patients were forced to stop smoking for at least 20 hours during hospitalization, but not every patient took advantage of this forced abstinence to sustain cessation after discharge. Among those who restarted smoking within 3 months following hospital discharge, about 80% started smoking again within 2 weeks. This result suggests that the period of initial cessation was very short and implies that the problem with Taiwanese men stopping smoking may be how to initiate and maintain a period of early abstinence rather than prevent late relapse.
Self-efficacy emerged as an important predictor in both research questions. As previous studies strongly suggested (Stuart, Borland, & McMurray, 1994), self-efficacy was positively related to making quit attempts, which was defined as stopping smoking for at least 24 hours. This study's findings also replicated those of other studies showing that self-efficacy is associated with short-term smoking abstinence (Gulliver et al., 1995). These results reflect Bandura's suggestion (1997) that self-efficacy can affect both initiation and persistence of the behavior required to lead to outcomes.
The length of hospital stay was a significant predictor in discriminating continuous abstainers from non-continuous abstainers. Results on the association between length of hospital stay and smoking from previous studies either in CHD patients or in general patients are not conclusive. In a univariate analysis, but not a multivariate analysis, Rigotti and associates (1991) showed that length of hospital stay was associated with making a quit attempt but not with abstinence at the one-year follow-up. Other investigators (Rigotti et al., 2000) found that patients hospitalized for a longer time were more likely to continue to refrain from smoking at one and six months after discharge.
Length of hospital stay may also reflect disease severity because more-serious diseases require longer periods of treatment. Another plausible reason may be related to hospital circumstances. There is a no-smoking policy in all hospitals, where no cigarettes are sold. Although patients can still smoke outside the hospital building, access to the outside is difficult. Thus, those who stay in the hospital longer are forced to be abstinent longer than those who have shorter stays. Therefore, those staying longer might gain more confidence in remaining abstinent. In addition, they might experience the most severe withdrawal symptoms during their hospital stay. All of the patients in the present study were prohibited from smoking for at least 20 hours or longer after admission regardless of treatment/procedure received. Withdrawal symptoms typically begin at 2-12 hours and reach a peak 48-72 hours after smoking the last cigarette (Hughes, Higgins, & Bickel, 1994). Rigotti et al. (2000) found that from 24 to 48 hours after admission, 89% of patients had at least one symptom and 55% of them developed cigarette cravings. Thus, it is possible that those with longer hospital stays may have had fewer withdrawal symptoms after discharge. In turn, patients with fewer withdrawal syndromes may face fewer problems maintaining abstinence longer.
Those who received more-positive support behaviors from their families were more likely to attempt to quit smoking. This result is supported by the finding of Roski, Schmid, and Lando (1996). Also, in the present study, among attempters, those who had higher positive support and lower negative support were more likely to become continuous abstainers. These results support the findings of previous work (Cohen & Lichtenstein, 1990). These results also echo the results of some studies suggesting that positive and negative support behaviors might exert different influences on smoking cessation processes at different times (Roski et al., 1996).
Two cautions should be noted. Subjects were asked to recall the frequency of support behaviors in the past 3 months, but it is possible that they actually reported only the more recent support behaviors. That is, social support might not reflect the process of how household members helped or hindered smoking cessation during the entire 3 months but rather may reflect their recent smoking status near the end of the 3 months. Another point of caution is related to the addition in this study of one extra option “doesn't apply as I have not tried to quit” to the five items in the subscale of positive family support. This extra option was treated as “never” and was scored as “0.” This change is logical because no support would be offered by family members to subjects who checked this option. However, this change might have biased the results toward less positive support for continuous smokers.
Several limitations bear consideration. Generalization of results is limited by selection bias. First, those who were younger, with higher educational levels, and with more severe heart disease diagnoses were more likely to participate in a study like this than those who were not. Second, compared with those who finished the study, those who had less severe diagnoses were more likely to have dropped out of the study. Apossible limitation is that self-reporting may involve some deception, inferring the smoking cessation rate may be an overestimate. Also, in light of the fact that many new smoking cessation clinics have been established with the support of the Department of Health in Taiwan, it is possible that subjects changed their smoking behavior based on factors other than variables considered in this study.
Conclusions and Implications
Considering that 23% of patients had made no attempt to quit smoking, healthcare providers need to deliver a clear and strong message to patients to raise their motivation to quit smoking by emphasizing that smoking is a major cause of their heart disease, and that stopping smoking can reduce morbidity and mortality. The first 2 weeks after hospital discharge, especially the first 24 hours, is the most critical period because a substantial number of subjects may taste their first cigarette during this time. Intensive follow-up contacts and support from healthcare providers within 2 weeks of discharge may help sustain patients' motivation to remain abstinent and to resume abstinence again if they have already started smoking again. Follow-up contacts may include contacts by telephone or during a clinic or hospital visit. Discussing smoking cessation with patients should involve family members. It is possible that during hospitalization, family members, especially wives and others who do not smoke, may have criticized patients for having a smoking habit. Healthcare providers need to express their understanding of family members who try to persuade patients to stop smoking and encourage them to use more positive support behaviors to help patients. Changes in smoking behaviors in this study were primarily related to perceived self-efficacy for not smoking and family support behaviors. Any form of intervention for smoking cessation for male patients with coronary heart disease should address these two factors.
Based on the investigator's observations in the hospital, about 10% of smoking patients had low literacy levels. Such patients may be under-studied and may be the most deficient in resources to aid in their quitting smoking. Future research should include this population using in-person interviews. A qualitative study may also be needed to explore why and how early relapse occurs within the first 2 weeks after hospital discharge. While coronary heart disease tends to present itself in women at older ages than in men, the upward trend in smoking among females makes the health consequences of smoking and issues related to smoking cessation among female patients topics that should also be considered in future studies.
Asia Pacific Cohort Studies Collaboration. (2005). Smoking, quitting, and the risk of cardiovascular disease among women and men in the Asia-Pacific region. International Journal of Epidemiology, 34
Atterbring, M. F., Hartford, M., Hjalmarson, A., Caidahl, K., Karlsson, T., & Herlitz, J. (2004). Smoking habits and predictors of continued smoking in patients with acute coronary syndromes. Journal of Advanced Nursing, 46
Bandura, A. (1997). Self-efficacy: The exercise of control.
New York: Freeman.
Bendel, R. B., & Afifi, A. A. (1977). Comparison of stopping rules in forward regression. Journal of the American Statistical Association, 72,
Chung, Y. C. (1994). The study of factors correlated with life adjustment of coronary artery bypass graft outpatient.
Unpublished master's thesis, National Defense Medical Center, Taipei, Taiwan.
Cohen, S., & Lichtenstein, E. (1990). Partner behaviors that support quitting smoking. Journal of Consulting and Clinical Psychology, 58
Crouse, J. R., & Hagaman, A. P. (1991). Smoking cessation in relation to cardiac procedures. American Journal of Epidemiology, 134
Department of Health, Executive Yuan, R. O. C. (Taiwan). (2008, January 7). Statistics of causes of death Volumn I.
Retrieved January 20, 2008, from http://www.doh.gov.tw/EN2006/DM/DM2_p01.aspx?class_no=390&now_fod_list_no=9256&level_no=2&doc_no=51991
Fagerstrom, K. O., Heatherton, T. F., & Kozlowski, L. T. (1991). Nicotine addiction and its assessment. Ear, Nose and Throat Journal, 69
Farkas, A. J., Gilpin, E. A., Distefan, J. M., & Pierce, J. P. (1999). The effects of household and workplace smoking restriction on quitting behaviours. Tobacco Control, 8
Farrelly, M. C., Evans, W. N., & Sfekas, A. E. S. (1999). The impact of workplace smoking bans: Results from a national survey. Tobacco Control, 8
Fava, J. L., Velicer, W. F., & Prochaska, J. O. (1995). Applying the transtheoretical model to a representative sample of smokers. Addictive Behaviors, 20
Gulliver, S. B., Hughes, J. R., Solomon, L. J., & Dey, A. N. (1995). An investigation of self-efficacy, partner support and daily stress as predictors of relapse to smoking in self-quitters. Addiction, 90
Harrell, F. E., Lee, K. L., Califf, R. M., Pryor, D. B., & Rosati, R. A. (1984). Regression modeling strategies for improved prognostic prediction. Statistics in Medicine, 3
Heatherton, T. F., Kozlowski, L. T., Frecker, R. C., & Fagerstrom, K. O. (1991). The Fagerstrom test for nicotine dependence: A revision of the Fagerstrom tolerance questionnaire. British Journal of Addiction, 86
Huang, S. F. (1990). Exploring psychosocial adaptation and related factors among AMI patients after hospital discharge.
Unpublished mater's thesis, National Taiwan University, Taipei, ROC.
Hughes, J. R., Higgins, S. T., & Bickel, W. K. (1994). Nicotine withdrawal versus other drug withdrawal syndromes: Similarities and dissimilarities. Addiction, 89
Johnson, J. L., Budz, B., Mackay, M., & Miller, C. (1999). Evaluation of a nurse-delivered smoking cessation intervention for hospitalized patients with cardiac disease. Heart & Lung, 28
Marshall, P. (1990). “Just one more!” A study into the smoking attitude and behaviour of patients following first myocardial infarction. International Journal of Nursing Studies, 27
Norman, G. J., Velicer, W. F., Fava, J. L., & Prochaska, J. O. (1998). Dynamic typology clustering within the stages of change for smoking cessation. Addictive Behaviors, 23
Pan, L.Y. (2001). Stage of smoking cessation and its related factors of male ever smokers.
Unpublished doctoral dissertation, National Taiwan University, Taipei, ROC.
Rigotti, N. A., Arnsten, J. H., McKool, K. M., Wood-Reid, K., Pasternak, R. C., & Singer, D. E. (2000). Smoking by patients in a smoke-free hospital: Prevalence, predictors, and implications. Preventive Medicine, 31
Rigotti, N. A., Singer, D. E., Mulley, A. G., & Thibault, G. E. (1991). Smoking cessation following admission to a coronary care unit. Journal of General Internal Medicine, 6
Roski, J., Schmid, L. A., & Lando, H. A. (1996). Long-term association of helpful and harmful spousal behaviors with smoking cessation. Addictive Behaviors, 21
Sallis, J. F., & Owen, N. (1997). Ecological models. In K. Glanz, K. F. M. Lewis, & K. B. Rimer (Eds.), Health behavior and health education. Theory, research, and practice
(pp. 403-424). San Francisco: Jossey-Bass.
Stuart, K., Borland, R., & McMurray, N. (1994). Self-efficacy, health locus of control, and smoking cessation. Addictive Behaviors, 19
Taiwan Bureau of Health Promotion. (2007). Taiwan tobacco control 2007 annual report: 100% smoke free environment.
U.S. Department of Health and Human Services. (2004). The health consequences of smoking: A report of the Surgeon General.
Atlanta, GA: Author.
van Berkel, T. F. M., Boersma, H., Roos-Hesselink, J. W., Erdman, R. A. M., & Simoons, M. L. (1999). Impact of smoking cessation and smoking interventions in patients with coronary heart disease. European Heart Journal, 20
Velicer, W. F., DiClemente, C. C., Prochaska, J. O., & Branderburg, N. (1985). Decisional balance measure for assessing and predicting smoking status. Journal of Personality and Social Psychology, 48
Velicer, W. F., Redding, C. A., Sun, X., & Prochaska, J. O. (2007). Demographic variables, smoking variables, and outcome across five studies. Health Psychology, 26