Determinants of motivation to quit in smokers screened for the early detection of lung cancer: a qualitative study

Background The promotion of smoking cessation within lung cancer screening could lead to benefits for smoking-related disease and improve cost-effectiveness of screening. Little is known about how smokers respond to lung cancer screening and how this impacts smoking behaviour. We aimed to understand how lung cancer screening influences individual motivations about smoking, including in those who have stopped smoking since screening. Methods Thirty one long-term smokers aged 51–74 took part in semi-structured interviews about smoking. They had been screened with the EarlyCDT-Lung Test (13 positive result; 18 negative) as part of the Early Cancer Detection Test Lung Cancer Scotland Study. They were purposively sampled for interview based on their self-reported post-screening smoking behaviour. Eleven participants had stopped smoking since screening. Verbatim interview transcripts were analysed using thematic analysis. Results Two key overarching themes were interpretations of screening test results and emotional responses to those interpretations. Participants’ understanding of the risk implied by their test result was often inaccurate, for example a negative result interpreted as an ‘all-clear’ from lung cancer and a positive result as meaning lung cancer would definitely develop. Those interpretations led to emotional responses (fear, shock, worry, relief, indifference) influencing motivations about smoking. Other themes included a wake-up call causing changes in perceived risk of smoking-related disease, a feeling that now is the time to stop smoking and family influences. There was no clear pattern in smoking motivations in those who received positive or negative test results. Of those who had stopped smoking, some cited screening experiences as the sole motivation, some cited screening along with other coinciding factors, and others cited non-screening reasons. Cues to change were experienced at different stages of the screening process. Some participants indicated they underwent screening to try and stop smoking, while others expressed little or no desire to stop. Conclusions We observed complex and individualised motivations about smoking following lung cancer screening. To be most effective, smoking cessation support in this context should explore understanding of screening test results and may need to be highly tailored to individual emotional responses to screening. Electronic supplementary material The online version of this article (10.1186/s12889-018-6211-1) contains supplementary material, which is available to authorized users.


Background
Tobacco use is responsible for more than five million deaths a year worldwide [1] and awareness of the link between smoking and lung cancer is high [2]. Lung cancer screening using low-dose computed tomography (CT) is recommended in the USA for those aged 55-80 with a 30 pack year smoking history [3]. Smokers who engage in lung cancer screening may be more motivated to quit [4] and screening could further influence smoking thoughts, motivations and behaviour via a 'teachable moment' or a 'license to smoke'.
Predictors of attempts to stop smoking include having made a past quit attempt, lower cigarette dependence, higher motivation and intention to quit and belief in the harm caused by smoking [5]. Qualitative research has further described how health concerns can lead to quit attempts, often combined with other internal or environmental influences [6]. This work was conducted outside of the context of lung cancer screening.
Of three randomised lung cancer screening studies that have compared smoking in screened and control groups, one observed higher quit rates in the screened group at 2 weeks and at 2 years [7] and two found no long-term effect of lung cancer screening on smoking [8,9]. Groups screened with CT or chest X-ray in the National Lung Screening Trial reported similar smoking cessation rates after 3 years [10]. Screened participants who receive an abnormal screening result appear more likely to quit and less likely to relapse [7,[10][11][12][13][14]. Increasing rates of smoking cessation in screening patients could lead to benefits for smoking-related disease and cost-effectiveness of screening [15]. Professional and medical guidelines recommend the integration of smoking cessation interventions into lung cancer screening programmes [16,17] although pilot studies of this approach provide mixed findings of effectiveness [18][19][20][21][22].
There is a lack of evidence on how support should be provided to most effectively promote smoking abstinence in those screened for lung cancer [23,24]. In particular, little is known about how lung cancer screening influences individual motivations about smoking. Two studies have used qualitative methods to explore smoking in this context. The first study found that nobody within a sample of 35 National Lung Screening Trial participants had stopped smoking more than a year after screening but some had reduced their smoking. Structured interviews were used and factors influencing motivations about smoking were not explored in depth [25]. The second study reported three of 37 participants had quit smoking since screening, one of whom said the offer of screening had changed their thoughts about smoking and another said the finding of nodules motivated them to quit [26]. Others reported a lack of urgency to quit, sometimes citing the monitoring of CT findings as a reason. To add to this evidence that lung cancer screening can both increase and decrease motivation to quit, a qualitative investigation is needed of factors influencing the motivations of those that decide to attempt to stop smoking after screening and those that decide not to. There has been no in-depth study to date of individuals who have stopped smoking after being screened for lung cancer.
Based on previous literature on this topic we expected that lung cancer screening might involve experiences that in some way promote or inhibit attempts to stop smoking. The aim of our study is to explore motivations about smoking in smokers screened for the early detection of lung cancer, including those who stop smoking after screening, to better understand how screening impacts motivations to stop or continue smoking and how cessation support can promote smoking abstinence in this context.

Study design
We conducted a qualitative study as part of the Early Cancer Detection Test -Lung Cancer Scotland (ECLS) study, a randomised controlled trial evaluating the effectiveness of a blood test (EarlyCDT-Lung) to detect lung cancer early [27]. ECLS study participants lived predominantly in the most deprived areas of the three study regions of Greater Glasgow and Clyde, Tayside and Lanarkshire, in Scotland, UK. Participants were at increased risk of lung cancer due to having at least a 20 pack-year smoking history or a family history of lung cancer combined with a smoking history representing an equivalent risk. Blood samples were taken from all participants and those randomised to the screening arm were screened for levels of autoantibodies to lung cancer, which enabled risk stratification for the targeting of chest CT scans. Study materials informed participants that EarlyCDT-Lung detects 40 of every 100 cases of lung cancer and that eight out of every nine people receiving a positive test result do not have lung cancer. Those with a negative test result were notified by a letter stating that between 98 and 99 out of every 100 people with a negative test do not have lung cancer at the time of the test, and inviting them to contact the research centre if they have any questions. Those with a positive test result discussed the implications with a research nurse face-to-face or by telephone. They were informed the CT scan might detect pulmonary nodules and that in the majority of people they are of no health concern. They underwent a chest X-ray and CT scan and, if lung cancer was not diagnosed, they received four further CT scans at 6 month intervals. Smokers did not routinely receive cessation support as part of the ECLS study in order to prevent stigmatisation as a smoker indicated as a potential recruitment barrier by pre-trial focus groups [28]. However, they could be asked by the research nurse at their initial visit if they would like information on smoking cessation or referral on to an appropriate service. Participants were reminded of the importance of visiting their general practitioner if they experience named lung cancer symptoms.

Recruitment and data collection
We sampled participants for the qualitative study from a subset of 1043 ECLS study participants taking part in a nested questionnaire study exploring psychological and behavioural responses to screening. Questionnaires collected self-reported data on current smoking status and recent attempts to stop smoking. A quota sampling approach was adopted with the aim of recruiting ten people who had stopped smoking since screening, ten who had attempted to stop but were still smoking, and ten who had not attempted to stop. Our definitions for each category of this sampling frame and other eligibility criteria are shown in Fig. 1. We aimed to recruit participants from across two ECLS study regions (Lanarkshire had not yet begun recruiting) and participants who received positive and negative EarlyCDT-Lung results. This approach was to ensure a diverse range of screening experiences and behavioural responses were represented in our sample. Within each quota we took a convenience sampling approach: eligible individuals who had most recently returned a follow up questionnaire were sampled in advance of scheduled research visits. They were sent an invitation letter, information leaflet and a contact form to return in a prepaid envelope to express interest in taking part. The leaflet explained that we were investigating what people think about smoking after lung cancer screening and emphasised that the purpose of the study was not to try to stop them smoking. This aimed to avoid discouraging those who did not want to stop smoking from taking part. On return of a contact form a researcher telephoned the participant to explain the study, answer any questions and arrange a convenient time for an interview. Participants completed a consent form before the interview. They were advised that the researcher held no strong feelings about smoking and was simply interested in their thoughts and feelings. Semi-structured, face-to-face interviews began with questions about smoking histories and general ECLS study experiences, then focused on motivations and decisions made about smoking since ECLS study enrolment and explored reasons for those. The interview guide is available in Additional file 1. Interviews also covered topics not reported here: barriers and facilitators to smoking abstinence and attitudes and preferences for smoking cessation support within lung cancer screening. Interviews were audio recorded and transcribed (anonymised) verbatim. All participants were offered a £5 multi-store gift voucher to thank them for participation. They were already receiving a series of identical £5 vouchers for completing the questionnaires so the interview incentive formed part of a larger available remuneration package.

Analysis
Transcripts were analysed in NVivo software using a process of inductive and deductive thematic analysis [29,30]. This involved familiarisation with the data, systematic coding of data, generation of a set of initial codes, sorting of codes into structures containing overarching themes and their subthemes (using separate a priori structures to address distinct aims of the research), reviewing and refining themes and finally, defining and further refining themes to create a coherent and internally consistent account of the data. One coder analysed all the data and a second senior investigator checked samples of transcripts in discussion with the coder at various points throughout the analysis. This enabled the revision of codes and their structures and the development of themes. We continued the coding process until all data had been systematically coded. During this process any new concepts identified prompted the review of all interview transcripts to check for the presence of that concept. We followed the criteria of Seale et al. [31] to ensure quality of the analysis and write up. We report here the subset of data relating to motivations about smoking.

Results
Of 12,210 ECLS study participants, 63 were sent an invitation to the qualitative study, 46 (73%) responded and 31 (49%) were interviewed (Fig. 1). Based on participant preference, 27 interviews took place in participants' homes and four in a private room in a clinical research facility. Average interview length was 40 min (range 16-80). Characteristics of participants and of the source population are shown in Table 1. Interviewees were comparable to the source population on most characteristics. They were, however, less likely to live in the most deprived areas and were more likely to have been intending to stop smoking at ECLS study enrolment. The timing of interviews in relation to ECLS study events are shown in Table 2. Most interviews took place within 5 months of screening and, for those with a positive result, most took place before their 6 month ECLS study schedule CT scan. During their interviews 11 participants reported having stopped smoking since lung cancer screening. The quotes presented were selected as being the most succinct and representative data extract(s) of each theme.
The two key overarching themes extracted in relation to motivations about smoking were interpretations of screening test results and emotional responses to those interpretations.

Interpretations of screening test results
Participants' interpretations of screening test results were a perceptual filter through which screening influenced motivations about smoking. Understanding of the risk implied by test results was often inaccurate; for example, a negative result interpreted as being 'all clear':

Emotional responses to interpretations of screening test results
Emotional responses to the interpretations described above were central to participants' motivations about smoking. These responses included fear, shock, upset, worry, anxiety and guilt: Some participants experienced a desire to change their smoking behaviour following screening and some wanted to continue smoking. Importantly, there were individual differences in the way in which emotional responses impacted motivations about smoking, with no clear pattern according to test results or interpretations of their meaning. Some were motivated to stop smoking by a positive test result and felt they would have continued smoking if the result had been negative. In others the opposite responses were observedthey were motivated to stop smoking by a negative result but would have continued smoking if it had been positive.

Interactions between overarching themes and motivations about smoking
We present three examples that demonstrate the link between participants' interpretations of test results, their emotional responses and their motivations about smoking. The participant in the first example explained how her emotional response to the test result inhibited her ability to understand the risk information provided to her and made her too scared to telephone the study centre to ask questions. She described a vicious circle whereby this emotional response and uncertainty led to her smoke more heavily: P12: Have I got it [cancer], or am I going to get it? If I stop smoking will that change, or will I still get it anyway, because of this gene? So there's a lot of questions, you know. And when you go there for that appointment after it all, you cannae [can not] really take it in, you know you're sort of sitting talking and you think 'I must remember that, I must remember that, I must remember that, ' … and I did get a letter I couldnae [  I: So at which point did you kind of make that decision that you were going to try and stop? P1: Well after I got the result of the CT scan, that was when I decided that I was definitely going to do it. So that was a few weeks after I got the result that I actually stopped, so, I was really really shocked that there was something there. And I've not asked the doctor, I'm going to phone up and ask exactly what this is. I need to phone and ask does he think it is cancer that's there, do you know what I mean? He says it's right here at the front of my lung, but it is only tiny, he said they had to search the scans to actually find it, so it is tiny, six whatever, I don't know if it's centimetres or--(Woman, 54, positive, stopped smoking) These examples show that individuals' responses were different but that their interpretation of their test result and emotional response were key recurring themes in their motivations about smoking.
The two overarching themes presented above describe the key aspects of how screening influenced motivations about smoking. Further themes presented below elaborate on aspects of the 'teachable moment' represented by screening and important social and contextual factors influencing motivations about smoking.
'Teachable moment' Theme: A wake-up call Screening served as a health scare and a wake-up call, prompting thoughts about the threat of lung cancer. This was described in terms of an objective confirmation of the known risks of smoking: P22: Most smokers are very sensible people, you know the risks that you're taking but it's a very concrete thing isn't it when you get a test result like that, well it's concrete in some ways, the reality of what you're doing to yourself. (Woman, 59, positive, not stopped smoking) Another participant expressed surprise that a positive result, interpreted as meaning her life was at risk, hadn't caused her to decide to stop smoking: P2: Even after I found out that I did have a positive result and both lungs have got nodules ... I'm still smoking! I mean I never ever ever thought that I would do that. I thought any time when it comes, I'll have to make a decision and I'll make it and that'll be it, you know, when my life's at risk right away I'll make it and that'll be it and I haven't done that. (Woman, 60, positive, not stopped smoking) A participant expressed a desire for the receipt of a forceful message from a doctor for an even bigger wake-up call and additional motivation to stop smoking: The offer of information about local smoking cessation services by research nurses was not mentioned by participants during interviews or identified as a theme influencing motivations about smoking.

Contextual factors
There were important non-screening factors influencing smoking motivations in this sample. These included age and life stage factors such as becoming a grandparent, respiratory symptoms, and financial factors. These themes are described in Table 3 with example quotes.

Coinciding factors
Some said the screening had motivated them to stop smoking in combination with coinciding non-screening factors: P7: Everything sort of fitted in at the right time for me because ... before I had a wee scare and I kept thinking Lung cancer screening was a novel experience that, in combination with other factors, provided an opportunity that they felt might not be available again. Screening was described as having come at the right time and having 'brought it all in' and 'brought it all to a head' in relation to other motivating factors.
Of those who did not want to stop smoking there was often a lack of detail in their accounts of their motivation to continue smoking as it was not something they had given much thought. Three themes we extracted from these participants' data are described in Table 4 with example quotes.
Cues to change were experienced at different stages of the screening process, not always immediately following a test result. Some people already had a desire to stop or cut down smoking before screening and their screening experiences either reinforced or did not change these desires (Table 3). Participation in lung cancer screening led to other smoking behaviour changes, for example several participants had tried to cut down but had not attempted to stop. One participant had been prompted by their screening experiences to begin using filters in their roll-up cigarettes but had not had any thoughts about stopping smoking. A diagram of themes is shown in Fig. 2.

Discussion
This study is the first to our knowledge to explore in-depth how lung cancer screening influences motivations to quit smoking and the first to purposively recruit individuals that have stopped smoking after lung cancer screening.

Principal findings
Motivations about smoking were closely related to emotional responses to test results, which in turn were linked to how individuals interpreted these test results, often inaccurately. Because people had different levels of understanding about what the test result meant, different emotional responses to those understandings and different pre-existing motivations to change their smoking behaviour, their motivations about smoking were individualised and difficult to predict. We found positive and negative results were both experienced as a reason   to stop smoking or a reason to continue smoking. Screening was a 'wake-up call' to the risks of smoking and created a sense that 'now' was the time to stop smoking. Family members, age-related factors and the existence of multiple coinciding non-screening factors were also influential. The teachable moment was experienced at different stages of the screening process and some had participated in screening in order to try to stop smoking. Those who did not want to stop smoking experienced reassurance from study scans, felt that it was too late to stop, or avoided thinking about stopping.

Strengths and limitations
Qualitative methods allowed an in-depth and nuanced exploration of the patient perspective of the topic. There was a good response rate to study invitations. Rigour was demonstrated by the inclusion of those who had received positive and negative test results and those who had and had not tried to stop smoking, along with the use of an inductive and deductive approach to analysis. Reliability was ensured by digital recording and verbatim transcription of interviews, the use of software to allow systematic coding, and discussion between two researchers during the sorting of codes into structures to generate overarching themes. We demonstrated sensitivity to context by positioning the study within the wider ECLS study and through the use of neutral, non-judgmental language in approaching the topic of smoking in the invitation materials and interviews [32]. Some limitations should also be considered. Participants were likely to be more motivated to stop smoking than the wider smoking population because they had agreed to take part in a screening study, a nested questionnaire study and a further qualitative study. Having a blood sample taken for cancer screening could have been an unusual experience for them and the results may be less generalisable to screening programmes involving just CT without a preceding blood screen. During recruitment to the ECLS study, participants were provided with information about concepts such as randomisation and allocation to the control group, which can cause confusion [28] and could have inhibited understanding of other information such as the meaning of test results. We did not have data on patient self-reported health-literacy or numeracy to explore whether this influenced interpretations of test results. Finally, most participants asked if the interviewer was a smoker and may have adjusted their explanations about smoking after learning that this was not the case. The researcher did, however, take a neutral standpoint on any issues arising and as a visitor to the study regions could be distanced from the local ECLS study clinical processes and engage simply as an interested outside observer of experiences.

Comparison with other data
Individualistic responses to screening test results can help explain why, except for those receiving abnormal results in some studies, consistent patterns in the impact of lung cancer screening on smoking have not been observed to date. Misinterpretations of the degree of lung cancer risk implied by positive or negative lung cancer screening results is a novel finding that highlights complexity in the behavioural impact of lung cancer screening. It suggests that any change in motivation for smoking cessation created by lung cancer screening may sometimes be based on a suboptimal understanding of information provided to screening participants. This provides support for concerns raised about the effects on patients of dichotomising cancer screening test results into 'positive' and 'negative' [33].
The finding that lung cancer screening test results are experienced emotionally, and that this can influence health behaviour, has been reported in previous work. A qualitative study of CT lung-screened smokers reported emotional arousal as one of three key pathways by which screening may influence motivation around smoking cessation [26]. Abnormal lung screening results have been shown to have a short-term adverse impact on emotional outcomes [34][35][36][37] and to promote smoking cessation [38]. Uncertainty management theory can provide a framework around which to understand the relationship between emotional and behavioural responses to screening. Screening test results can change an individuals' level of uncertainty about their health, their appraisal of this uncertainty can elicit a positive or negative emotional response, which can influence smoking behaviour [39]. In smokers undergoing lung cancer screening these emotional responses might create active or passive dissonance with emotional responses to smoking behaviours. In this way, smokers' motivations and decisions about smoking are a response to screening resulting from their appraisal of uncertainty and the emotions that result. In our participants the uncertainty they experienced relates to their understanding of screening test results. Furthermore, their emotional response to uncertainty appraisal can lead to information seeking or avoidance and in the case of information seeking behaviour, this could improve understanding of screening test results.
The themes 'wake-up call' and 'now is the time to stop smoking' support the idea that lung cancer screening can be a teachable moment for smoking cessation. This contrasts with evidence from the National Lung Screening Trial, which reported screening was not a cue to action and high risk perceptions were not related to quitting [25]. This may be due to our study adopting a more in-depth and loosely structured approach to data collection, allowing the wider context of smoking motivations to be explored. For example, we found when lung cancer screening played a role in motivations to stop smoking there were often other important non-screening factors at play. Some had taken part in screening in order to try and stop smoking. This is consistent with a previous qualitative finding that motivation to stop smoking was one of three perceived benefits of lung cancer screening in smokers [40]. We also found some evidence that reassurance from CT scans could reduce motivation to quit, and that some people cut down their smoking or made other changes, both findings that have been reported previously [25,26].

Implications for research, policy and practice
The polarised way in which screening test results were sometimes interpreted in our study (very high/very low risk) was a factor in motivations about smoking. To aid understanding, ECLS study screening test results were communicated as 'positive' or 'negative' along with probabilities using simple frequencies. However, there were still deficits in understanding, highlighting a need for further progress in this area to enable better understanding of lung cancer screening tests and test results by participants. Research to describe the complexities of experiences of uncertainty in lung cancer screening patients who smoke can help to develop communication processes that facilitate desired behavioural responses in the management of that uncertainty. This can include information seeking and smoking quit attempts.
Smokers may experience different emotional responses and motivations about smoking if they understand their risk in a different way. It is therefore important that quantitative studies of the impact of lung cancer screening on smoking account for levels of perceived risk. Any assessment of the overall benefits and harms of a lung screening programme should consider how well test results are understood and how individuals might react emotionally and behaviourally to those results. Furthermore, when considering the emotional harms of screening it should be acknowledged that short-term emotional harms could promote longer term benefits such as smoking cessation.
The findings suggest that smoking cessation advice in lung cancer screening should be tailored according to individual interpretations of, and emotional responses to test results. A randomised trial of male smokers found computer-tailored advice did not result in significantly different abstinence rates than standardised advice following lung cancer screening [18]. Importantly, the advice was tailored to smoking attitudes and behaviour but not to understanding of and emotional response to screening test results. A telephone counselling intervention which aimed to use lung cancer screening test results to increase risk perceptions was effective at promoting cessation in a pilot randomised trial [20]. The stated aim of this strategy was to capitalise on the teachable moment of an abnormal result and to counteract the potential for reduced motivation to quit after a result showing no nodules or abnormalities. A responsive approach such as this, rather than a computer-tailored method, has the flexibility to adapt the advice depending on the attitudes and intentions of the individual, which our study showed can be unpredictable and individualistic. Such interventions should also be flexible in the timing of delivery and should be offered after the test result is delivered so that interpretation of the result can be explored and emotional responses can be further pursued. Family members could also be involved and non-screening factors addressed. The finding that those who did not want to stop smoking sometimes felt it was too late to do so requires further exploration to ensure this perception can be addressed. Further research is needed to explore what type of cessation support lung cancer screening participants who smoke would find most acceptable and useful.

Conclusions
Our study demonstrates individualised and complex motivations about smoking among lung cancer screening participants and the ways in which lung cancer screening can create a 'teachable moment' in motivations about smoking. Emotional and behavioural responses to test results, which can be misinterpreted, varied between individuals. Lung cancer screening presents an opportunity to engage high risk smokers in cessation attempts but cessation support may need to be tailored to an individual's emotional response to their understanding of their test result and take account of the range of factors we have identified to be most effective.
Endnotes 1 unique indentifier 2 gender 3 age (years) 4 EarlyCDT-Lung test result 5 post-screening smoking cessation as described by participants in interviews (stopped smoking/not stopped smoking)

Additional file
Additional file 1: Interview schedule. Question guide for semi-structured interviews. (DOCX 14 kb) Abbreviations CT: Low-dose computed tomography; ECLS: Early Cancer Detection Test -Lung Cancer Scotland supporting recruitment, and NRS Mental Health Network (Laura Hodges, Emma Eliasson, Emma Haigh and Katie Richards) for supporting interview transcription.

Funding
Dundee Cancer Centre Development Fund. Oncimmune Ltd. and Chief Scientist Office, Scottish Government funded the ECLS Study. The funding bodies had no role in the design of the study, collection, analysis, and interpretation of data or in writing the manuscript.

Availability of data and materials
The dataset generated and analysed during the current study is not publicly available because personal narrative could identify individual participants and participants consented for data to be accessed only by the research team.
Authors' contributions BY conceived the original idea for the study, recruited participants, conducted the interviews, transcription, analysis and drafted the paper. RdN, KV, DK, JFRR and RL elaborated the study design and research question. FS and KV sought funding from Dundee Cancer Centre. FMSS and RL sought ethical approval. RdN, KV, DK and JFRR supervised the study. The ECLS study team monitored study progress. RdN supervised and contributed to the analysis and interpretation of data. FMSS and SS were the principal investigators of the ECLS trial; contributed to the design of the main trial and substudies; oversaw the conduct of the study, its analysis and the manuscript preparation. All authors commented critically on drafts of the paper and approved the final version.
Ethics approval and consent to participate Our study complies with the appropriate national research ethics process and R&D and governance processes. The study was externally reviewed by East of Scotland Research Ethics Service REC 1 (reference 13/ES/0024, amendment AM05). Individual written informed consent for participation in the research was obtained from all participants.

Consent for publication
Written consent for publication of anonymised direct quotes and study findings was obtained from all participants.

Competing interests
The authors declare that they have no competing interests.

Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.