Patterns of smoking behaviour in patients following cardiac surgery. A prospective study

Abstract

Objectives. To describe patterns of smoking behaviour in patients undergoing cardiac surgery. Design. A prospective population-based study of patients undergoing cardiac surgery between September 2004 and September 2005. Smoking behaviour and socio-demographic variables were obtained by questionnaires at baseline, six and 12 months after surgery. Results. At baseline 534 patients (median age 69 years, 23% females) were included, 89% responded after six and 12 months. At baseline 14% (n=74) were current smokers, 59% (n=316) were former smokers and 27% (n=143) had never smoked. At six months 8% were current smokers (n=36) while 9% (n=40) were current smokers at 12 months. A total of nine smokers had relapsed at 12 months. Conclusion. About half of the current smokers gave up smoking after cardiac surgery. Changes in smoking behaviour were most likely to occur during the first six months. This supports that smoking cessation interventions should continue after discharge. Cardiac surgery can serve as a teachable moment; an opportunity to encourage patients to give up smoking and prevent relapses among those who stopped smoking before surgery.

Key words::

• cardiac surgery
• smoking
• smoking cessation
• secondary prevention

Smokers requiring surgery are forced to abstinence from tobacco during their hospital stay and nicotine withdrawal symptoms may be minimal in the postoperative period suggesting that this may be an excellent opportunity for smokers to attempt sustained abstinence (1,2).

Current smoking at the time of coronary artery bypass surgery (CABG) is one of the strongest predictors for mortality after CABG (3–5). The beneficial effects of giving up smoking on cardiac patients’ mortality and morbidity are well established (6). The effect of smoking cessation has also been related to other aspects of health status, such as the patients’ health-related quality of life (HRQOL) (7). Current smoking has been found to predict poor quality of life after CABG (8).

Most of the knowledge relating to smoking among patients with various manifestations of heart disease has been gathered from patients suffering from either myocardial infarction (MI) or angina pectoris (AP). The literature on smoking cessation following cardiac surgery is scarce, and knowledge from other groups of cardiac patients may not be applied to patients who have undergone CABG (3) or heart valve surgery (9). Behaviour change associated with different cardiac procedures might differ and influence outcome (10). Specific features related to surgery itself may represent an additional challenge concerning smoking and cardiac surgery. Smokers constitute an important risk group of patients prone to develop a range of postoperative complications including impaired wound and bone healing and life-threatening pulmonary and cardiovascular complications (11).

The purpose of the present study was to describe patterns of smoking behaviour in patients undergoing cardiac surgery. We explored patterns of smoking before and after cardiac surgery, patients’ intention to stop smoking and the occurrence of smoking relapses after cardiac surgery.

Methods

Design and patients

The data were obtained from a prospective study designed to explore chronic pain and HRQOL before and six and 12 months after cardiac surgery (12). This was an observational study; hence no systematic interventions concerning smoking were undertaken. The Regional Committee for Medical Research Ethics, Health Region IV, Norway approved the study. Written informed consent was obtained from each patient at baseline.

All patients undergoing cardiac surgery at the Department of Cardiothoracic Surgery, St. Olav's Hospital, Norway, between September 2004 and September 2005 were assessed consecutively for inclusion in the study. Of the 631 patients undergoing cardiac surgery, 97 patients were excluded (12). Thus, 534 patients were included in the study at baseline (84.6%). Baseline questionnaires were given to the patients for self-administration after being admitted to the clinic. Six and 12 months after surgery the patients were sent questionnaires by post. Non-responders were sent one postal reminder.

Measures

Demographic and medical information were collected from an institutional database where pre-, intra- and early postoperative variables are recorded prospectively. Study-specific questionnaires concerning socio-demographic and medical variables were constructed. These questionnaires included questions on smoking behaviour.

Smoking status

Information about smoking status was based on self-report. Smoking status was assessed at baseline, and six and 12 months after surgery. All patients were asked whether they currently or previously had smoked tobacco. Current and former smokers were asked to report how many cigarettes they smoked daily and for how many years they had smoked. At baseline current smokers were asked about their intention to quit.

Patients were categorised as (1) never smokers, (2) former smokers or (3) current smokers. Current smokers were defined as those who reported daily or occasionally smoking at hospital admission. “Pack-years” was defined as number of years of daily smoking multiplied by number of cigarettes smoked daily divided by 20. Patients who had stopped smoking before surgery, but reported smoking at follow-up were defined as relapsers.

Statistical analysis

Data are given as means (SD), medians (ranges) and numbers (%) respectively. Never smokers, former smokers and current smokers were compared using one-way analysis of variance (ANOVA) or Kruskal-Wallis test when appropriate for continuous variables, and Fisher's exact test for categorical variables. The Mann-Whitney U-test or the Student's t-test was used to compare continuous variables between two groups. McNemar's exact test was used to compare the proportion of smokers before and after surgery. Statistical calculations were performed using SPSS for Windows^® version 15.0 (SPSS Inc., Chicago, Illinois, USA). Two-sided p-values <0.05 were considered statistical significant.

Results

Patients

Patient demographics and details about the surgical procedure are given in Table I. Smoking status was available on 533 of the 534 patients included at baseline. After six months 462 of 521 eligible patients responded (88.6%), while 465 patients responded after 12 months (89.4%). Analysis of responders and non-responders at six and 12 months did not reveal systematic differences concerning the baseline data (data not shown).

Table I. Demographic and clinical characteristics of patients at baseline by smoking status before surgery (n=533).

	Smoking exposure
Variable	Current (n=74)	Former (n=316)	Never (n=143)	P-value^a
Age at operation (years), median	63.4 (23.8–81.2)	69.9 (24.9–86.8)	70.4 (30.0–85.6)	<0.001^b
Gender, n (%)				0.001^c
Male	59 (80)	259 (82)	94 (66)
Female	15 (20)	57 (18)	49 (34)
Education, n (%)				0.858^b
Primary education	34 (47)	140 (45)	69 (49)
Secondary education	24 (33)	122 (39)	45 (32)
Higher education	15 (20)	52 (16)	26 (19)
Marital status, n (%)				0.004^c
Cohabited/married	50 (68)	254 (80)	96 (68)
Living alone	24 (32)	62 (20)	46 (32)
Previous cardiac surgery, n (%)	2 (3)	29 (9)	9 (6)	0.142^c
BMI (kg/m^2), mean	26.6 (4.5)	27.0 (3.7)	26.8 (3.9)	0.730^d
Diabetes mellitus, n (%)	7 (10)	39 (12)	24 (17)	0.272^c
Lung disease, n (%)	10 (13.5)	41 (13.0)	14 (9.7)	0.593^c
EuroSCORE, median	3.8 (0–9)	4.0 (0–13)	5.0 (0–12)	0.091^b
NYHA-class I/II/III/IV, n	5/18/45/6	11/74/194/37	9/32/89/12	0.429^b
Emergency surgery, n (%)	0 (0)	2 (0.6)	1 (1)	0.998^c
XC time (min), median	38.0 (12–273)	42.0 (5–188)	43.0 (11–180)	0.632^b
CPB time (min), median	65.0 (23–377)	68.5 (16–441)	67.0 (23–283)	0.892^b
Surgery, n (%)				0.190^c
Isolated CABG	49 (66)	226 (72)	89 (62)
Isolated HVS	2 (3)	19 (6)	15 (11)
HVS + CABG	13 (18)	39 (12)	23 (16)
Miscellaneous	10 (13)	32 (10)	16 (11)
Pack-years, median	20.8 (0.75–122.5)	18.8 (0.05–165.0)	–	0.372^e
12 Months Mortality, n (%)	4 (5)	6 (2)	4 (3)	0.207^c

BMI, body mass index; NYHA, New York Heart Association; XC, cross clamp; CPB, cardiopulmonary bypass; CABG, coronary artery bypass grafting; HVS, heart valve surgery. Descriptive data given as frequencies (%), means (sd) and medians (range) for each group of smoking exposure.

^aP-values calculated for comparisons of smokers, former smokers and never smokers.

^bKruskal-Wallis test.

^cFisher's exact test.

^dOne-way ANOVA, ^eMann-Whitney U-test.

Patterns of smoking behaviour

Baseline. Before surgery 74 patients (14%) were current smokers, 316 patients (59%) were categorised as former smokers and 143 of the patients (27%) had never smoked (Figure 1). Among former smokers 70 patients (22%, [70/316]) had stopped <12 months before surgery, while 247 patients (78%, [247/316]) had stopped 12 months or more before surgery. Among current smokers, 69% (48/70) reported intention to quit, 7% (5/70) did not intend to quit, while 24% (17/70) were ambiguous (four missing). At baseline current smokers were significantly younger than former smokers and never smokers (median age 63.4 versus 69.9 and 70.4 years, p<0.001). More females than males were never smokers, while more males than females were former smokers (p=0.001). There were also differences in smoking status related to marital status (p=0.004, Table I).

Figure 1. Patterns of smoking before, six and 12 months after cardiac surgery. ^aSmoking status available on 533 of 534 patients included at baseline. ^bSmoking status available on 457 of 462 patients responding at six months. ^cSmoking status available on 459 of 465 patients responding at 12 months. ^§One smoker at 12 month reported to be never smoker at baseline. Hence the total number of smokers at 12 months was 40. Never smokers are not included in the figure (n=143). Some patients missing at six months responded at 12 months and opposite.

Smoking status six months after surgery. At six months 8% were defined as current smokers (36/457, Figure 1). Seven former smokers reported that they had relapsed at six months (7/278, 39 missing). Among current smokers at baseline, 49% reported that they had stopped smoking during the last six months (28/57, 17 missing).

Smoking status 12 months after surgery. At 12 months 9% were defined as current smokers (40/459, Figure 1). Only four patients (4/459) reported that they had stopped smoking during the last six months. Of former smokers who stopped smoking during the last 12 months before surgery, two had relapsed between six and 12 months after surgery. The total number of patients who reported relapsing after surgery was nine (3% [9/280], 37 missing). Most patients relapsing had stopped smoking less than 12 months before surgery (14% [8/58], 12 missing).

The prevalence of smoking was reduced from 14% before surgery to 9% after 12 months (McNemar's exact test, p=0.007). Among patients who intended to quit smoking before surgery, 50% (22/44, four missing) reported to be smoke-free 12 months after surgery, while 40% (4/10, seven missing) among the ambiguous patients were smoke-free. In addition, one patient (20%, [1/5]) who did not intend to quit before surgery had stopped smoking 12 months after surgery.

Discussion

This prospective population-based study showed that most patients who undergo cardiac surgery do not smoke at the time of surgery. However, a majority (59%) of the patients had a history of smoking. Moreover, about half of the current smokers gave up smoking after cardiac surgery, while some former smokers relapsed. Changes in smoking behaviour were most likely to occur during the first six months after surgery.

Our study supports that cardiac surgery can serve as a “teachable moment” to modify smoking behaviour; it is an opportunity for patients to give up smoking and prevent relapses for those who stopped smoking before surgery (13). In the present study the prevalence of smoking was 14% before cardiac surgery while reduced to 9% after one year. In contrast, Lindsay et al. found no changes in the prevalence of smoking before and after CABG-surgery with 25% current smokers both before, 16 months and seven years after surgery (5). This study population was younger than ours, with a mean age of 58 years at the time of surgery. As most other studies we found that current smokers at the time of surgery were younger than former and never smokers (9,14).

A Swedish study found 13% smoking at the time of surgery as compared to 7.5% two years after CABG (15). A Greek study (4) reported a very high proportion of active smokers at the time of CABG-surgery (62.3%). This was explained by a younger population and the lack of official antismoking policy in Greece. Mixed results concerning the prevalence of smoking in patients undergoing cardiac surgery illustrate difficulties comparing smoking patterns in different studies with different populations. Many countries, Norway included, have implemented restrictive laws regulating tobacco consumption. Beliefs about smoking and attitudes towards smoking have been changing, especially in Western high-income countries (11). Smoking in the Norwegian adult population has gradually decreased from 33% in 1998 to 21% in 2008 (16). Hence, up-to-date research is important to describe smoking patterns among different patient groups as well. In addition to differences between countries, different hospital routines concerning smoking cessation may influence the outcomes.

Smokers who undergo cardiac surgery need to be counselled specifically about smoking cessation. Ideally patients should stop smoking before surgery to prevent postoperative complications (11). Being stressed about the forthcoming surgery may be a hinder for smoking-cessation, as many view smoking as a tool to manage stress (2). The need to perform studies on patients after cardiac surgery, as compared to cardiac patients in general is supported by Hajek et al. (1). They found that patients who had had CABG were twice as likely to return to smoking as patients who had had a MI 12 month after a smoking cessation intervention. It may be that advices must be targeted specifically after cardiac surgery, compared to cardiac patients in general. Vachenauer et al. (17) concluded that many continue smoking as long as they are free from symptoms of AP. Undergoing cardiac surgery can produce the false perception that illness is under control and the patients are cured (18). However mixed results have been obtained concerning this, and other studies have found that patients after CABG are more likely to abstain from smoking (10,14).

A recent European survey on secondary and primary prevention in coronary patients, found that only one in seven patients was advised to attend smoking cessation service, and only one-third of those actually did so (14). Poor secondary prevention after CABG has been found by others as well (19). Gabriel et al. (20) described an “information gap” in patients undergoing cardiac surgery; for most patients the relationship between tobacco use and cardiovascular disease was not clear. Moreover, a majority of the patients were not informed about this relationship.

Concerning the results of the current study, it seems very important to offer intense counselling during the first six months after surgery, when most patients relapsed. This is also supported in a systematic review on smoking cessation interventions in hospitalised patients concluding that only interventions with follow-up after discharge are effective (21).

Interestingly many patients who were ambiguous about their intention to quit reported that they had stopped smoking 12 months after surgery. A recent study (22) found that people seem to overestimate their capacity to control their own impulses, like smoking. We found that the intention to quit smoking did not correspond to the smoking behaviour reported by patients after cardiac surgery. A better understanding of the relationship between intention to quit and actual behaviour may be helpful when tailoring smoking cessation advices and programs to patients after cardiac surgery.

The validity of self-reported smoking is often questioned because of the widespread belief that smokers are inclined to underestimate the amount smoked or to deny smoking at all (23). However, Patrick et al. (23) concluded that self-reports of smoking are accurate in studies of different populations. The prospective design protected against recall- bias concerning smoking related to the cardiac surgery. However, former smokers were asked about their smoking behaviour in the past, hence recall bias can not be completely ruled out. Due to the small group of current smokers, the sample size was too small to perform further subgroup analysis. Another concern was the number of patients missing at follow-up. The response-rate at six and 12 months was high compared to other postal surveys (5,8,17). Contacting patients by telephone may have reduced the patients lost at follow-up, but may also have introduced an additional telephone-bias (24).

In conclusion, about half of the current smokers at the time of surgery gave up smoking after surgery. Intention to quit smoking before surgery did not correspond to the smoking behaviour after surgery. Understanding smoking patterns is necessary to develop potential interventions to support smoking cessation in patients undergoing cardiac surgery. The complexity of smoking patterns support that the results from other groups of cardiac patients should not be directly applied on patients undergoing cardiac surgery.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.