Introduction

Top Abstract Introduction Methods Results Discussion References

Smoking remains the single most important remediable cause of premature death in the Western world. For the first time in 25 years, its incidence is rising in British men aged 20-24 and women aged 25-34. ^{1 2} It is estimated that 2% of smokers will quit if they are advised to do so by a doctor.³ Doctors are often exhorted to advise all smokers to quit each time they attend for health care on the assumption that repeated interventions will result in additional quitters among the remaining smokers.^4-6 However, some doctors believe that this routine repetition is frustrating and ineffective.⁷ A previous qualitative study of health promotion showed that patients resent doctors dictating to them about lifestyle change.⁸ The stages of change model of behaviour change shows that action oriented advice for those who are not ready to change is at best unhelpful, and could even entrench unhealthy behaviour. ^{9 10}

To make the most of opportunities for smoking intervention that arise in normal health care, it may be important to understand patients' perceptions of the acceptability of interventions they have received. Few studies have examined patients' experiences of opportunistic antismoking interventions. Since judging acceptability involves understanding patients' feelings, ideas, perceptions, and unique experiences, we believed that qualitative research methods would be best suited to this purpose.¹¹ We therefore planned to explore smokers' in-depth accounts of their interactions with the health services about smoking for evidence of possible unintended effects of antismoking counselling and for ideas about interventions that patients might find acceptable. We believed that a typology of smokers could be constructed from these accounts, and that this might help doctors in providing effective opportunistic antismoking interventions.

	Methods

Top Abstract Introduction Methods Results Discussion References

Subjects and setting
Interviews with current smokers and smokers who had recently quit were conducted as part of the evaluation of the Welsh smoking intervention study, which took place in 21 general practices in south Wales.¹² Forty two of the 536 smokers who were opportunistically recruited into the primary care, controlled trial aspect of this research were interviewed. Sampling was purposeful, in that we set out to obtain interviews from subjects with a broad range of sociodemographic characteristics that were potentially relevant to the study question. Of the 42 subjects interviewed, 24 were women; six were aged 20-29, 13 were 30-39, 12 were 40-49, six were 59-59, and five were over 60; 19 had no educational qualifications, eight had O levels or GCSEs, two had A levels, six had a degree or diploma, and seven had a vocational qualification. Twenty subjects were in social class I-IIIN and 22 in class IIIM-IV; 10 subjects had recently stopped smoking, and the remaining 32 were ongoing smokers.¹³ The study was approved by relevant local research ethics committees.

Interviews
We used a semistructured interview guide that had been piloted previously. Topics included initial smoking, attempts to quit, thoughts about future smoking, past experiences with the health services, and the most appropriate way for health services to help the subject and other smokers. The schedule was open ended, and interviewers followed up other issues that were raised by subjects. Subjects were encouraged to say what they really felt and not to worry about whether or not this would be acceptable to the interviewer. Interviews lasting 20-75 minutes were conducted in the subjects' homes; they were audio taped and then transcribed. Twenty four interviews were conducted by a social scientist and 18 by a general practitioner, who was known by the subjects to be a doctor. We stopped the interview phase of the study when no new themes were emerging.

Analysis
All three authors and the research assistant were involved in the initial coding of 73 categories. Analysis progressed through stages of data reduction, data display, and drawing conclusions.¹⁷ Continuing discussions between the three authors, rereading of interviews, and construction of data matrices for each interview resulted in the identification 30 themes.

	Results

Top Abstract Introduction Methods Results Discussion References

Because of the remarkable similarity in the accounts of those who had quit and those who continued to smoke, data from interviews with both these groups were pooled. Interviews conducted by the general practitioner and the social scientist contained a similar proportion of accounts that were critical of the health services. Thus, the suggestion that subjects would be less frank when interviewed by the general practitioner was not supported. The main themes relevant to subjects' interactions with the health services are given in the box.

Smokers' evaluations
Subjects did not need to be told what to do about smoking since they had already made their own evaluations about their habit. A typical response was that of a 40 year old woman: "Well I'm telling myself the ... same thing. I mean it's a waste of money, you are ruining your health, it's obviously so many years off your life, things like before you could walk for miles and miles ... and now you are out of breath. ... I'm telling myself all these things, the problems like the smell of it, the expense and things like that. ... I know it all."

Doctors' powers of persuasion
Most subjects were sceptical about the power of doctors to influence smoking behaviour, especially since smokers already knew the risks they were taking with their health. Half stated that quitting is "down to the individual." A 40 year old man stated that: "Everyone knows the dangers of smoking now. It's not like it's a top secret. ... If that smoker don't want to stop smoking, the doctor could be three hours talking to him and he'll walk out of the surgery and have a fag and thank God for that. I think everyone has heard of the consequences of what smoking does to you ... so I can't see there is any good in going into great detail about it, because a smoker already knows it causes heart, cancer, whatever."

Centring on the patient
If doctors are to raise the topic of smoking opportunistically, most subjects stated that good practice involves using a respectful tone, sensitivity to the patient's receptivity, understanding the patient as an individual, being supportive, and, most frequently, not "preaching." Approaching the subject in any of these ways was taken as support for the view that doctors should adopt a patient centred approach to talking about smoking in the consultation.

Anticipating antismoking advice
Over half the subjects anticipated that they would receive advice about smoking when attending for health care. Some shrugged this off, while others experienced irritation and guilt and saw these interventions as an inappropriate invasion of their privacy. Some modified their help seeking behaviour as a consequence of anticipated medical responses to their smoking, generally by changing their usual doctor. Two subjects, however, gave accounts of repercussions that were potentially dangerous.

Types of smokers
Three broad types of smoker were identified, primarily according to how they reacted to advice from doctors to quit smoking. A "contrary" group tended to be less convinced of the merits of giving up, smoked more in response to being told to quit, and anticipated "ritualistic" advice from health professionals. They were sceptical about the power of doctors' words to influence them and reported that they were already saturated with antismoking information. They were more likely to recount negative experiences of interacting with doctors about smoking, to change help seeking behaviour because of these negative experiences, and were more likely to assert that quitting smoking was down to the individual.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Many subjects were sceptical about the power of the doctor's words to influence smoking habits, and they made the point that the negative effects of smoking were already well known to established smokers. These findings are common in published reports.^20-22 Most subjects felt that giving up was ultimately down to the individual, a finding that also emerged from the study of Stott and Pill on perceptions of health promotion in working class women.⁸

Many patients who were clearly not ready to quit anticipated that they would be advised to do this by doctors. When this happened, they responded by simply shrugging it off, feeling guilty, getting annoyed, or changing their help seeking behaviour. Two subjects gave accounts of putting their health in danger by not attending for needed medical help because they feared the doctor would talk to them about stopping smoking. While it is important to make the most of opportunities for effective health promotion during a consultation, doctors should not assume that repeating antismoking advice over and over again for all smokers will continue to be of benefit. The oft repeated exhortation that doctors should advise their patients to stop smoking whenever they see them deserves careful reconsideration.

Interventions that patients found acceptable took account of their receptiveness; were conveyed in a respectful tone; avoided preaching; showed support and caring; and attempted to understand them as a unique individual. These findings agree with those of a similar study of participants in an American randomised trial of antismoking interventions: they most appreciated doctors who provided a caring, individualised approach.²³ The importance of a caring, sustained relationship between doctor and patient to the acceptability of lifestyle advice from doctors was also highlighted in the study of Stott and Pill.⁸ However, a few participants in the present study felt that "scaring" patientsespecially those who had not been smoking for longmight have some advantage.

Typologies of smokers have been constructed before, but these have been based on factor analysis of questionnaire data. ^{24 25} A review of qualitative reports on smoking shows that this is the first attempt to construct a typology of smokers based on their reported interactions with health services. The risks of damaging the doctor-patient relationship through antismoking advice seems greatest with those smokers who fit into the contrary and self blaming categories. Considering how the patient views himself or herself as a smoker and how he or she is likely to react to differing styles of intervention may be useful to doctors when talking to patients about smoking.