Lung cancer is the leading cause of all cancer deaths. Almost 25% of all cancer deaths are caused by lung cancer [1]. Following the results of the National Lung Cancer Screening Trial, the U.S. Preventive Services Task Force now recommends annual lung cancer screening with low-dose computed tomography (LDCT) in high-risk individuals, and the recent European Union position statement also recommended planning for the implementation of LDCT screening throughout Europe to reduce lung cancer mortality [2,3]. Recently, a Dutch-Belgian lung-cancer screening trial also has shown that lung cancer screening could reduce lung cancer deaths [4].

The objective of lung cancer screening is to detect lung cancer early thereby improving survival rates, but it is well acknowledged that screening alone may not save as many lives as expected without smoking cessation intervention on screening participants [5]. Smoking still remains the leading cause of lung cancer; thus, the most primitive way to reduce lung cancer mortality is to prevent people from smoking [6].

In this context, lung cancer screening provides an excellent opportunity for smoking cessation intervention as participants in screening are generally followed up for a number of years and are likely to be more concerned with their health than the eligible nonparticipants in screening. It is suggested that lung cancer screening could provide a teachable opportunity for smoking cessation as participation in screening may provoke a greater perception and awareness of potential risk from smoking [7].

Korean Lung Cancer Screening Project (K-LUCAS) which was conducted from 2017 to 2018 to study the feasibility of implementation of population-based lung cancer screening in Korea. In K-LUCAS, all current smokers participating in screening were provided with in-person mandatory smoking cessation counseling and changes in smoking status were followed up 6 months after initial screening. K-LUCAS reported that motivation of lung cancer screening participants to quit smoking increased by 9.2% and the smoking cessation rate was 24.7% at 6-month follow-up based on telephone survey [8].

This study aimed to analyze the factors associated with smoking cessation after participating in the lung cancer screening program. This includes the specialties of physicians providing counseling, whether or not a screening unit is operated by public smoking cessation program, and whether or not counseling is provided simultaneously with screening results.

Lung cancer screening provides a golden opportunity for smoking cessation intervention as participants in screening are likely to be more concerned with their health than nonparticipants. Lung cancer screening provides a teachable opportunity for smoking cessation as participation in lung cancer screening may provoke a greater perception and awareness of potential risk from smoking [7]. Previous study indicates that an increase of smoking cessation is a very important factor to improve the cost-effectiveness of lung cancer screening, which would be critical in any national screening programs [12]. Previous randomized controlled lung cancer screening trials have indeed shown that smoking cessation intervention incorporated in a lung cancer screening program can be effective in motivating screening participants to quit smoking. The Danish Lung Cancer Screening Trial (DLCST) reported smoking cessation rates of 10.9% at 1-year after baseline screening [13]. The UK Lung Cancer Screening Trial (UKLS) has shown abstinence rates of 23.6% at the 2-years point [14]. In a Dutch-Belgian randomized lung cancer screening trial (NELSON), 14.5% of participants in the LDCT screening arm quit smoking 2 years after baseline screening although this was lower than that of the controlled arm (no screening, 19.1%) [15]. The smoking abstinence rates were 20.8% at 4 years after baseline screening in the Italian Lung Study (ITALUNG) and 12.8% at 2 years after baseline screening in the German Lung Cancer Screening Trial (LUSI) [16,17].

This study has provided further evidence that smoking cessation intervention incorporated in lung cancer screening program can be effective in motivating currently smoking participants to quit smoking, consistent with previous findings. In K-LUCAS, the 1-week point abstinence rates and 6-month continuous abstinence rates were 24.3% and 10.6% respectively, which were much higher than the average smoking cessation rate of 3.8% after increasing tobacco price by 80% in January 2015 in Korea [18]. Based on the survey, more than 80% of quitters in the K-LUCAS stated that participating in lung cancer screening motivated them to quit smoking.

There were some differences between the factors that indicate smoking cessation success measured by 1-week point abstinence and by continuous abstinence after 6 months of lung cancer screening participation. Family history of lung cancer and public smoking cessation program participation are significantly related to higher success in smoking cessation measured by 1-week point abstinence, but not significant in continuous abstinence. Most of the studies regarding smoking cessation reported lower continuous abstinence rates than point abstinence rates [19]. The achievement of continuous abstinence is difficult because every smoking episode after intervention was counted as failure. Thus, the robust indicator of continuous abstinence might not show a significant difference between the factors associated with success of smoking cessation. However, professions providing smoking cessation counselling showed a significant difference in smoking cessation success between family medicine doctors and internal medicine doctors. This phenomenon seems to have occurred because participants who have the will to quit smoking might continue to visit family medicine clinics that mainly provide smoking cessation counseling. Previous studies reported one of the important factors to lead smoking cessation is how many times smokers periodically visited their physicians for quitting smoking [20,21]. It seems that one of the key factor related to success of smoking cessation is not who provides smoking cessation counseling, but how long the care continuity is maintained after initial smoking cessation counseling.

Another interesting finding of this study involves a significant difference in the abstinence rates reported between screening units. The smoking cessation intervention was more effective in some screening units than others. The difference in abstinence rates was as large as 18.8% between the screening units.

Most interestingly, we found that the timing of smoking cessation intervention significantly affects smoking abstinence. Participants who received smoking cessation counseling simultaneously with the LDCT screening results by the same physician were up to 3.2 times more likely to quit smoking than participants who received counseling separately from LDCT screening results. Our results may also partially explain the indifferent (between the control and treatment groups) results in the DLCST where a brief one-to-one smoking counseling is provided separately by trained nurses. In the DLCST, no significant impact of lung cancer screening on smoking cessation was found [13].

A simultaneous provision allowed physicians to provide smoking cessation counseling effectively with images of LDCT scan including pulmonary nodules or emphysema or fibrosis which of smoking related findings could astound current smoking participants. As a consequence of their long history of smoking exposure, it is likely that those participating in the lung cancer screening program would also have smoking-related diseases such as emphysema or coronary artery calcification even though no significant nodules were detected [22,23]. Smoking counseling then might have become more personalized when provided with the screening results or may have provoked greater perception and awareness of risk from smoking as in self-regulation model [24]. Though more personalized smoking cessation incorporated with LDCT results seem to be promising to affect smoking behavior, further study is required to provide data on different approaches of personalized smoking cessation programs incorporated with LDCT results [25].

One potential problem raised in this study is that participants with negative LDCT screening results had significantly lower abstinence rates than participants with positive LDCT screening results, consistent with previous findings [13,26]. Negative screening results could signal false reassurance to smokers to continue smoking. Our results imply that physicians’ counselling with simultaneous provision of LDCT screening results in smoking cessation intervention could prevent negative lung screening results becoming a license to smoke because physicians discussed other findings related with smoking. Also, physicians motivated participants to quit smoking to prevent lung cancer. In our study, about 17.3% of currently smoking participants with negative (no significant nodule detection) screening results had quit smoking when they received smoking cessation counseling and screening results separately, but 23.7% of currently smoking participants with negative results had quit smoking when smoking counseling was performed with LDCT scans (p < 0.01).

We also observed that of the 4,136 current smokers who participated in screening and surveyed, 3,115 (75.3%) still had not quit smoking after smoking cessation intervention (Fig. 1). There could be other confounding factors affecting an individual’s decision to continue smoking other than variations in the smoking cessation interventions and demographics tested in this study. For example, participant’s current stress status (e.g., temporary job loss), willingness to quit, or counselling skills of physicians could also affect smoking cessation, but this was beyond the scope of this study [27,28]. Currently no systematic approach exists for participants who continue smoking after participating in lung cancer screening program, thus, identifying other factors associated with the participant’s decision to continue smoking would help in designing more effective smoking cessation interventions to incorporate in lung cancer screening programs.

This study has several limitations. First, the primary limitation is due to a single-arm cohort study; thus, no control group assessed the between-group effectiveness of smoking cessation intervention. The effectiveness of smoking cessation intervention in the K-LUCAS was evaluated by comparing the abstinence rates in the K-LUCAS with the general abstinence rates in Korea [18] and previously reported abstinence rates in lung cancer screening trials [13-17]. Second, this study did not include missing data in the logistic regression model, which could decrease statistical power to investigate the factors associated with success of smoking cessation. It could be possible that continuous smokers after lung cancer screening are more in non-responders than survey responders. Thus, the current study reported the smoking abstinence rate in 6 months after lung cancer screening both by intention to treat analysis of total current smokers participating in K-LUCAS and by per-protocol analysis of survey responders. We did not give any information on survey purpose before phone calling after 6 months of screening, so it is unlikely that the participants of K-LUCAS did not respond to the telephone survey because they continued to smoke. Third, the study has analyzed the effect of different approaches to smoking cessation intervention at individual-level by logistic regression but hospital-level analysis was not included. This is mainly because no data was available to control for possibly associated factors at hospital-level such as physician experience or hospital accessibility which may weaken the results obtained. Thus, we only included a nominal comparison of smoking cessation rates between participating hospitals. Fourth, all measures of smoking cessation are self-reported from telephone surveys without any biochemical verification such as a cotinine test, mostly because of limited financial resources and difficulties obtaining the verification by revisiting hospital 6 months after screening. Hence, the possibility of self-reported bias of smoking abstinence rates or interviewer bias still remains. However, previous study has shown that self-reported measure is still a valid measure of smoking abstinence and may be more feasible with a large sample size as in this study [29]. Fifth, the efficacy of different kinds of smoking cessation interventions include pharmacotherapy at individual level has not been evaluated. All current smokers in the K-LUCAS received face-to-face smoking cessation counseling by a physician and, if they agreed, were provided pharmacotherapy for quitting smoking.

Although no data was available at individual-level on how long the participants received pharmacotherapy or compliance of medication, a recent systematic review has shown the efficacy of different types of smoking cessation intervention. According to the review, in-person smoking cessation counselling could increase smoking abstinence more significantly than web-based or telephone counselling. Therefore, in-person counselling by physician and actively providing pharmacotherapy could increase smoking cessation more effectively targeting to the high-risk smokers participating in lung cancer screening [30]. Indeed, the smoking cessation rates in K-LUCAS (in-person counseling) were 1.7 times higher than that of the UKLS (brochures with information of smoking cessation) which also targeted high-risk smoking individuals [14].

Both the NELSON trial and UKLS trial did not provide in-person smoking cessation counseling after LDCT screening. Though a direct comparison was not available, the smoking cessation rates of 11.9% at 1 year after baseline screening in the LUSI trial and that of 10.9% at 1 year after baseline screening in the DLCST were lower than 24.3% at 6-months after baseline screening in the K-LUCAS. The DLCST trial provided in-person counselling which was less than 5 minutes by nurses at each annual visit [13]. The LUSI trial provided smoking cessation counselling by trained psychologists but showed no statistical significance in lowering smoking cessation rates [17]. An active smoking cessation intervention by physicians and pharmacotherapy in the K-LUCAS would possibly be a reason for higher smoking cessation rates than the NELSON and UKLS that did not provide any in-person counselling and LUSI trial and DLCST which did not provide physician’s counselling and pharmacotherapy [14,15].

Future studies should investigate the differences in the effectiveness of different kinds of smoking cessation intervention in high-risk population eligible for lung cancer screening so that the most efficient intervention can be provided when lung cancer screening is implemented.

In-person mandatory smoking cessation intervention was effective in motivating lung cancer screening participants to quit smoking. When smoking cessation counseling is provided simultaneously with LDCT screening results and lung images obtained by CT scan, the effectiveness of smoking cessation intervention in lung cancer screening participants could increase. Participating in lung cancer screening could be the best opportunity to motivate smokers to quit smoking. It will be very important to strengthen the motivation of lung cancer screening participants to quit smoking and to continuously help them quit smoking to reduce the incidence and death of lung cancer in the future.