Mapping the spectrum of psychological and behavioural responses to low‐dose CT lung cancer screening offered within a Lung Health Check

Abstract Background Research on the psychological impact of low‐dose computed tomography (LDCT) lung cancer screening has typically been narrow in scope and restricted to the trial setting. Objective To explore the range of psychological and behavioural responses to LDCT screening offered as part of a Lung Heath Check (LHC), including lung cancer risk assessment, spirometry testing, a carbon monoxide reading and smoking cessation advice. Methods Semi‐structured interviews were carried out with 28 current and former smokers (aged 60‐75), who had undergone LDCT screening as part of a LHC appointment and mostly received an incidental or indeterminate result (n = 23). Framework analysis was used to map the spectrum of responses participants had across the LHC appointment and screening pathway, to their LDCT results and to surveillance. Results Interviewees reported a diverse range of both positive and negative psychological responses, beginning at invitation and spanning the entire LHC appointment (including spirometry) and LDCT screening pathway. Similarly, positive behavioural responses extended beyond smoking cessation to include anticipated implications for other cancer prevention and early detection behaviours, such as symptom presentation. Individual differences in responses appeared to be influenced by smoking status and LDCT result, as well as modifiable factors including perceived risk and health status, social support, competing priorities, fatalism and perceived stigma. Conclusions The diverse ways in which participants responded to screening, both psychologically and behaviourally, should direct a broader research agenda to ensure all stages of screening delivery and communication are designed to promote well‐being, motivate positive behaviour change and maximize patient benefit.


| INTRODUC TI ON
Lung cancer is most frequently diagnosed at an advanced stage (49%-53% at stage 4) 1 yet early detection markedly improves prognosis; with five-year survival increasing from 6% at stage 4 to 82% at stage 1A for non-small-cell lung cancer. 2 Implementing low-dose computed tomography (LDCT) lung cancer screening for high-risk groups achieves a stage shift to earlier diagnosis. 3 LDCT screening reduced the relative risk of lung cancer mortality by 20% in the US National Lung Screening Trial (NLST) 3 and by 26% in the Dutch-Belgian trial NELSON. 4 However, there has been concern about the psychological burden of LDCT screening, particularly because the NLST had a high false-positive rate. 3 Trials have since differentiated 'false-positive' results (suspicious for cancer) from 'indeterminate' pulmonary nodules (usually benign but require surveillance). 5 Protocols for incidental findings are also evolving, with some reporting only those conditions for which diagnosis would lead to clinical benefit. Evidence from the trial setting suggests any distress induced by these types of results is relatively short-lived and not clinically significant. [6][7][8] While research has focused on whether abnormal results cause any clinical psychological morbidity, the ways in which individuals respond to screening may be more diverse and encompass positive as well as negative dimensions. For example, for some people, screening may provoke conscious awareness of risk and distress whereas for others, regular screening may offer a reassuring safety net, providing a positive means of managing risk of lung cancer mortality.
The psychosocial component of the Danish Lung Cancer Screening Trial (DLCST) developed a condition-specific measure of more diverse psychosocial consequences, 9 which consists of responses spanning social, cognitive and attitudinal domains (eg focus on airway symptoms, existential values). Interestingly, these responses were observed among both 'screened' intervention participants and 'no screen' control participants, 10 suggesting aspects of the screening pathway other than the LDCT test itself (eg communicating individual risk status, lung function tests) may have psychological consequences. There are also likely to be individual differences in the way people experience screening and respond to abnormal results, with studies beginning to implicate cognitive risk factors such as higher affective risk perceptions and self-blame in adverse psychological outcomes. 10,11 The downstream effects of lung cancer screening on behaviour may be similarly diverse. Ongoing research efforts are directed to understand whether screening undermines or promotes smoking abstinence and how to effectively embed cessation advice and treatment. 12 However, a qualitative study in the United States found lung screening participants also reported improvements in diet and physical activity, 13 suggesting the behavioural impact extends beyond smoking. Indeed, evidence drawn from studies of screening for other cancer types, 14 as well as the diagnostic setting, point to potentially wider-ranging positive and negative effects across prevention and early detection behaviour. For example, false-positive breast screening results have been associated with lower subsequent screening uptake, 15 and all-clear diagnostic test results as well as negative bowel screening results with reduced concern about symptoms and delayed presentation. 16,17 The scope of psychological research on the impact of LDCT screening therefore needs expanding to understand the diverse ways in which individuals respond to screening as well as the potential implications for patient experience, well-being and cancer control. This evidence should inform the design of patient-centred screening communication and delivery to optimize screening benefit. The present study aimed to explore and map the diverse spectrum of positive and negative psychological and behavioural responses among individuals with indeterminate and incidental LDCT screening results across the entire screening pathway.

| Participants
Current and former smokers (aged 60-75) were interviewed four to eight months (Mean: 6 months) after having LDCT screening as part of the Lung Screen Uptake Trial (LSUT). 18 LSUT was a randomized controlled trial aiming to improve uptake of lung cancer screening by high-risk individuals (current or recent (quit < 7 years) former smokers, aged 60-75 years) as part of a 'real-world' demonstration screening service. Individuals were invited to a Lung Health Check (LHC) appointment, which included a medical and smoking history, spirometry and carbon monoxide tests, NCSCTaccredited 'Very Brief Advice' on smoking cessation 19 and a LDCT scan for those who were eligible based on their risk of lung cancer.
Three months after their LDCT scan, purposive sampling was used to recruit a heterogeneous subsample of participants from LSUT who varied in smoking status (current and former), LDCT results (incidental and indeterminate pulmonary nodule) and socio-economic position (SEP; high and low). A sampling matrix was drawn from a preliminary trial database using these three characteristics. LSUT participants, who fit each of the eight different combinations of characteristics and had consented to further contact, were invited to take part by letter. A researcher (SLQ) checked the demographic and smoking characteristics of those who responded over the phone to ensure each combination of characteristics from SMJ is a Wellcome Trust Senior Fellow in Clinical Science (WT107963AIA

| Analysis
Data were analysed in NVivo v.11 by SK using a framework approach 20 to thematic analysis. The primary objective was to explore the spectrum of positive and negative psychological and behavioural responses to lung cancer screening and their potential influences. SK familiarized herself with the data and developed a coding framework inductively using four transcripts to identify important and recurrent themes. The coding framework was reviewed by SLQ using a subset of transcripts and any disagreements were resolved prior to its application to the remaining data. Following coding, data in each transcript were indexed, charted and summarized systematically into themed matrices so that comparisons could be made by smoking status (current, former) and type of LDCT result (indeterminate, incidental). Six randomly selected transcripts were independently coded by SLQ to review the appropriateness of the indexing codes and themes. Any discrepancies in the coding framework were discussed and resolved.

| Sample characteristics
A total of 129 LSUT participants were invited to take part, 55 responded and 28 were selected for interview. Sample characteristics are presented in Table 1  result) and smoking status (CS/FS = current/former). Our coding frame is also available in File S1.

| Psychological responses
Psychological responses varied along the screening pathway and are presented for each stage.

| Prior to the Lung Health Check (LHC)
Most regarded their invitation to the LHC as a positive occurrence

| When told about LDCT eligibility
Some participants expected to be offered a LDCT scan at their LHC, because this possibility was outlined in the invitation materials or because they perceived their lung health as poor. Others had not expected to be offered a scan, which came as 'a bit of surprise' [P13/ Nod/FS]. Apprehension about being scanned was only raised by one participant who was concerned about the potential result. Instead, participants described reacting positively to being eligible for a LDCT scan; seeing it as an opportunity to find out whether something was wrong with their lungs.

| Spirometry test results
The type of spirometry result appeared to have the potential to posi-

| Waiting for LDCT results
Extreme anxiety concerning the possible LDCT result was rare.

| Receiving LDCT results
Some perceived that their risk of lung cancer remained just as high following their LDCT results as it had been before: 'I still think it's

| Waiting for and undergoing follow-up tests
Overall, few concerns were raised by those who underwent surveillance for a pulmonary nodule. Knowing that a follow-up LDCT scan was scheduled or having been psychologically prepared for the possibility of an indeterminate result during the LHC appeared to have provided reassurance. However, for the small minority who did report anxiety, this was significant. Those undergoing diagnostic procedures were understandably more concerned. Of note, one participant reported breathlessness due to anxiety but initially interpreted this as a symptom of lung cancer.

| Behavioural responses
The ways in which participants responded behaviourally also varied.
Participants predominantly talked about anticipated future behaviours, but a variety of actual changes to behaviour were also mentioned. These are further outlined in this section.

| Future anticipated screening participation
The majority intended to take part in any future lung screening

| Future anticipated symptomatic help-seeking
There was little evidence that taking part in LDCT screening un-

| Smoking behaviour
Three participants reported that they had stopped smoking either before they attended their LHC (which they set as a date for them

| Factors influencing psychological and behavioural responses
This section outlines the various factors interpreted as influencing participants' responses to the LHC. On the other hand, a lack of pre-existing concern or symptoms was linked to a corresponding lack of concern about the upcoming LHC 'I didn't really feel concerned… I haven't seen any symptoms or anything A lack of concern due to potential misattribution of symptoms to 'the quality of the cigarettes nowadays' also appeared to adversely affect future help-seeking intentions.

| Existing concerns about lung health and smoking history
Interestingly, prior concern may have been instrumental in whether an abnormal LDCT result motivated behaviour change.
Pre-existing concern about health appeared to motivate positive behaviour change regardless of the type of LDCT result, including motivation to quit or reporting (temporarily) cutting down smoking.
However, for those with no prior concerns, an indeterminate result alone did not appear to be sufficient to motivate change in smoking behaviour with two participants noting how they had started to smoke more after temporarily cutting down and appeared to be more likely to engage in compensatory behaviours.

| Competing Priorities
Two participants were focused on their other existing medical conditions and consequently the LDCT results were considered relatively unimportant: 'I've got so many things wrong with me… just another thing' [P12/Inc/FS], whereas another individual diagnosed with an indeterminate result revealed that they thought 'about it more because I wasn't expecting it to be a cause for concern' [P10/Nod/FS].
Similarly, others mentioned that the invitation to the LHC, although it was readily embraced and acted upon, came when they had been faced with challenging life circumstances such as a daughter's cancer diagnosis.

| D ISCUSS I ON
This study broadens our understanding of the psychological and behavioural impact of LDCT lung cancer screening. We identified a more diverse spectrum of responses experienced throughout all stages of the screening pathway than has been described previously.
As expected, these varied widely encompassing positive responses as well as negative responses, with wider-reaching anticipated implications for future prevention and early detection behaviour.
Individual differences in response appeared to be influenced by concern about future symptoms. Current smokers appeared to be most likely to anticipate seeking help promptly; a group that has been found to be less likely to seek help for lung cancer symptoms compared with non-smokers in previous research. 21 Our participants' accounts were hypothetical, and those who attend screening may be more proactive in seeking help than the general population.
Nevertheless, these findings suggest there may be something about the screening process that fosters symptom awareness, which could provide an opportunity to encourage prompt presentation among a high-risk group.
Participants with indeterminate results reported increases in distress consistent with previous research. 6,8,22 However, there was an apparent lack of concern about incidental findings (eg COPD) among some participants who were understandably pleased their scan showed no sign of lung cancer but in some cases regarded this an 'all-clear' for their respiratory health. This suggests that incidental findings carry low risk of psychological distress which is reassuring, but also implies risk of over-reassurance akin to 'clear' screening results raised by a previous study. 23 Indeed, current smokers with an incidental finding more frequently reported cutting down on smoking rather than making a quit attempt compared with those who had indeterminate results. Communicating incidental results may therefore be an opportunity to capitalize on the initially positive emotional response to motivate positive behaviour change that could ultimately halt disease progression.
Indeed, our findings suggest the LDCT screening pathway could provide multiple opportunities to support cancer prevention. Those participants, who stopped smoking, did so at different points along the pathway including after receiving the initial invitation letter as well as following indeterminate results. There was also evidence of improvements in exercise and diet supporting findings from a previous US qualitative study. 13 Indeed, other studies have proposed cancer screening as a largely acceptable context for providing advice about multiple behavioural risk factors 24 and some degree of behaviour change was observed without intervention by one study. 25 Integrating a broader cancer prevention approach may be especially beneficial for lung cancer screening participants given existing data showing that behavioural risk factors cluster. For example, those who smoke are also less likely to be physically active or drink within recommended alcohol limits. 26,27 Further research is needed to quantify these responses and understand how to best provide and integrate advice and signposting to support within lung screening services.
While this study focused on the psychological and behavioural responses that high-risk individuals had to screening, there are other possible types of consequences that were not studied. These could include physical (eg exposure to radiation) or financial harms, as proposed in Harris and colleagues' taxonomy. 28 Future work will be needed to explore how these other types of harm might affect individuals' psychological and behavioural responses along the screening pathway. We used purposive sampling and recruited from a 'real-world' demonstration pilot of lung cancer screening; strategies intended to increase the ecological validity of our results.
Nevertheless, it should be acknowledged that LSUT participants may differ from the wider lung screening population as the eligibility criteria were purposefully narrower; recruiting a relatively older (aged 60-75) cohort of predominantly current smokers and recently quit former smokers living within the two most deprived quintiles nationally. This means the data reported here may not fully reflect the range of views and experiences of lung screening attenders. Our sample is also likely to be subject to self-selection bias as our participants volunteered to be interviewed. We cannot rule out the possibility that those who did not take part would have reported different responses to lung screening. Furthermore, while a minority of our participants received a clear LDCT result, we focused on individuals with indeterminate and incidental findings, and so further research is needed to better understand responses among those who receive a clear screening result.
Finally, our findings may be subject to recall bias, which may have led participants to recall fewer responses and of a lower intensity.
In conclusion, the ways in which individuals respond to LDCT screening both psychologically and behaviourally are more diverse than have been described by previous studies and span the entire pathway, beginning with the screening invitation. The ways in which screening is delivered by health-care professionals and communicated to participants should therefore be evidence-based and patient-centred at every stage, not just at the LDCT test. Importantly, negative responses may be reduced through psychological preparation for the different types of screening results and there is potential to capitalize on positive responses to support positive behaviour change in cancer prevention, symptom awareness and screening adherence. The present findings should help direct a broader psychological research agenda driven to optimize patient benefit from LDCT lung cancer screening.

ACK N OWLED G EM ENTS
We would like to acknowledge the substantial intellectual contribution made to the study design by Professor Jane Wardle, who passed away prior to its conduct. We would also like to thank all of the study participants who kindly gave up their time to be interviewed.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.