Advancing understanding of influences on cervical screening (non)‐participation among younger and older women: A qualitative study using the theoretical domains framework and the COM‐B model

Abstract Background Effective screening can prevent cervical cancer, but many women choose not to attend their screening tests. Objective This study aimed to investigate behavioural influences on cervical screening participation using the Theoretical Domains Framework (TDF) and COM‐B models of behaviour change. Design A qualitative study and semistructured phone interviews were conducted with women invited for routine screening tests within the national cervical screening programme in Ireland. Setting and Participants Forty‐eight women aged 25–65 years were recruited from the national screening register. Results Seven core themes were identified that mapped to three COM‐B components and 11 TDF domains: (1) knowledge of cervical cancer and screening, (2) coping with smear tests, (3) competing motivational processes—automatic and reflective, (4) cognitive resources, (5) role of social support, (6) environmental influences and (7) perceputal and practical influences. A range of knowledge about screening, perceived risk of cervical cancer and human papillomavirus infection was evident. Factors that influenced screening behaviours may be hierarchical—some were assigned greater importance than others. Positive screening behaviours were linked to autonomous motivation. Deficits in physical and psychological capability (inadequate coping skills) were barriers to screening, while physical and social opportunity (e.g. healthcare professional ‘champions’) could facilitate participation. Older women raised age‐related issues (e.g. screening no longer necessary) and had more negative attitudes to screening, while younger women identified practical barriers. Conclusions This study provides insight into screening participation and will aid development of theoretically informed interventions to increase uptake. Patient or Public Contribution Women invited for screening tests through the national screening programme were interviewed. A Public & Patient Involvement (PPI) Panel, established to provide input into all CERVIVA research projects, advised the research team on recruitment materials and were given the opportunity to review and comment on the interview topic guide. This panel is made up of six women with various cervical screening histories and experiences.


| INTRODUCTION
Cervical cancer is a global public health issue with an estimated worldwide incidence of approximately 570,000 new cases in 2018, of which over 116,000 occurred in Europe. 1 Cervical screening programmes operate in many countries and are effective in reducing the incidence of and mortality due to cervical cancer. 2 However, data from the screening registers of 19 European states indicates that screening coverage can vary considerably: from 10% to 79%. 3 Over the past decade, a pattern of falling uptake has been reported in several countries. Initially, there were concerns in some developed countries about uptake in younger women; 4 however, in recent years, a new pattern of lower uptake in older women has emerged in England and Ireland (https://www.cancerresearchuk.org/ health-professional/cervical-cancer-screening-and-diagnosis-statistics# heading-Five). 5 It is important to encourage older women to attend screening as the incidence of and mortality due to cervical cancer remain high in this age group. 2 Screening older women can reduce their cancer risk-women who are screened in their early 50s have a 75% lower risk of developing cervical cancer between 55 and 59 years. 6 Considerable research has investigated the factors that are associated with women's screening participation. Demographic factors such as ethnicity 7,8 ; practical and environmental factors such as accessible appointments and female smear takers 6,9,10 ; and psychosocial influences such as trust, 10,11 embarrassment, anxiety 7,12,13 and concerns about pain/discomfort 6,13 have been identified as being related to screening participation. Age differences also exist, with younger women reporting practical barriers, embarrassment and the perception that they are at low risk of developing cancer. 9,10,12,13 In contrast, older women may make active decisions not to participate; in one study, older women reported low levels of worry about cervical cancer and also perceived themselves to be at low risk of developing cancer. 13 However, the research that has generated these findings has frequently lacked a theoretical grounding.
Assessments of the strategies and interventions that have been tested to increase cervical screening uptake have found mixed evidence of increased participation. [14][15][16] In part, this may be because intervention development has not always been informed by key requirements, that is, empirical data linked to an appropriate theoretical underpinning. 17 The Theoretical Domains Framework (TDF) is a comprehensive integrated theoretical framework-synthesized from 128 theoretical constructs from 33 theories-that can guide the identification of theoretical constructs that influence behaviour. 18,19 The TDF model can be condensed into an overarching behavioural model-the COM-B model-with three central components, capability, opportunity and motivation, that interact in behavioural processes. 18 The TDF and COM-B models have been used to inform intervention design in various healthcare settings, [20][21][22] but have not previously been used to examine influences on cervical screening behaviours.
In the current study, we aimed to identify factors that influence women's decisions on cervical screening (non)-participation using the TDF and COM-B models, with a secondary objective of comparing and contrasting factors relevant for younger and older women.

| Design and setting
This study took place in Ireland. In-depth semistructured telephone interviews were conducted among women who had been invited to attend for routine/follow-up cervical screening tests. The  Each invitation included an information sheet, a reply slip, a consent form and a prepaid envelope. One reminder letter was sent to nonresponders. Recruitment ended when saturation was achieved. 25

| Interviews
A semistructured topic guide (Appendix A) was developed from literature review and informed by the TDF model. The original 12-domain version of the TDF model (v1) was selected as a comprehensive tool that would identify a broader spectrum of potential influences on screening behaviours. 26 Signed consent forms were returned by all participants before interviews. Interviews were conducted by a CERVIVA researcher between August and December 2019 and lasted 45-75 min (mean 60 min). Verbal consent to record the interviews was obtained, and recordings were transcribed verbatim and anonymized. A 'One4All' gift voucher (€25) was sent to participants once the interviews were completed to thank them for their time. All personal data were handled in accordance with the General Data Protection Regulation (GDPR), 2018.

| Coding and analysis
Transcripts were imported into NVivo 10 and an iterative analysis was performed concurrently with data collection. A combination of inductive thematic and deductive framework analyses using the TDF model was conducted. The transcripts were reviewed through familiarization, construction of a thematic framework (TDF domains), indexing, sorting of data and interpretation. 27  Where appropriate, illustrative anonymized quotes from the study participants are included in Section 3.

| Characteristics of the study participants
Interviews were conducted with 48 women-34 were adequately screened (AS) and 14 women were inadequately screened (IS); 17 women were younger than 50 years of age, while 31 women were 50 years of age or older. Detailed information about the participants is shown in Table 1.    Table S3 in the Supporting information.  domain. The factors that linked to physical opportunity included: smear takers; HPV self-sampling kits; and use of information sources.
Irrespective of age or screening histories, women expressed their preference for female and/or experienced smear takers. Some younger IS women described problems with their local smear takers that inhibited their attendance, but they had not made any plans to access different GPs: If there were a clinic or a different patient section in a hospital or whatever, I would have no problems, but it was just the nursing staff at my GP's that I don't like (DM550495, IS, 41 years) Most of the IS women, younger and older, reported that they would use HPV self-sampling kits if they received one in the post.
They felt that kits would be useful and more convenient for them, but highlighted the importance of clear, 'step-by-step' instructions. In contrast, many AS older women said that they would not as they lacked confidence that they would do it correctly.
Similar information sources-such as healthcare professionals, family/friends and the internet-were available to most of the women. AS women spoke about using these resources in the   Consistent with previous research, 12,13,29 many IS women reported limited knowledge about cervical cancer risk factors and believed that they were at low risk of developing it. In contrast, many older women who considered screening unnecessary because of their age or current sexual activity positively assessed cervical screening.
These findings suggest that a hierarchy of influences could exist in determining screening behaviours among older women-the perceived personal relevance of screening was more important than their knowledge of its overall health benefits. Improving knowledge of cervical cancer and screening alone is likely to be insufficient for effective behavioural change strategies. 30 This suggests that modelling and environmental restructuring interventions that increase psychological capability would be useful. Strategies tailored for older women that target their beliefs about salience of screening, which can change over time as they age, would be worth testing. Relevant advertising campaigns with high-profile 'older' women modelling positive screening behaviours or on-screen prompts for GPs to ask older women about screening could also prove effective.
Women highlighted the invasiveness of undergoing a screening test and, as with other studies, described high levels of stress and anxiety before their appointments. 12,13 AS women were confident in their abilities to cope with negative aspects of screening such as finding the test intrusive. Coping strategies such as breathing techniques were commonly used, suggesting that prescreening information that offered women tips and advice on such coping strategies could be useful. As with previous research, many older women reported increasing problems with the smear test procedure-linked to menopausal changes-as they aged. 13 These findings highlight the necessity of supporting older AS women and building physical capability to maintain their adherence. This can be achieved with physical skill development through training or enabling interventions. 20 Practical advice on lubrication or pain medication in information leaflets or as part of standard verbal instructions from smear takers could increase individuals' skills and be beneficial for these older women. The current study also found that promoting physical opportunity with HPV self-sampling kits could prove effective as older women who were IS indicated that they would use HPV kits if given the opportunity to do so. Such kits could be enabling interventions, provided they were accompancied with additional support to develop women's confidence. This additional support could include online video tutorials of self-sampling; information sheets with step-by-step instructions (both clear explanatory text and a visual guide); or 'practise' sessions across healthcare settings, where women are supported by experienced smear takers, receiving guidance from HCPs during their self-collection/while using the kits themselves.
A key finding of this study was that both motivational processes, automatic and reflective, were important in influencing screening behaviours. In contrast to IS women, those who were AS were autonomously motivated to attend screening-salient events such as previous abnormal results played a role in their prioritization of screening. The selfdetermination theory (SDT) on health behaviours suggests that intrinsic motivation will increase when psychological needs of autonomy, competence and relatedness are satisfied. 31 While previous research has identified the central role of healthcare providers in supporting screening, 35-37 the current study offers information specific to older women. The theory of planned behaviour model identifies a gap between intention and behaviour. 38 Our findings suggest that family support may act as a moderating factor on this gap for younger IS women, while health professional 'champions' who encourage women to attend screening may possibly have a similar moderating effect for older women who are IS. However, it should be noted that controlled motivation may not be a reliable influence on screening behaviours over time. 39 SDT suggests that women who are motivated to attend screening for controlled reasons-perceived approval of others-feel a sense of obligation and will only persist if the external pressure is maintained. 33,39 Our findings suggest that persuasion and education interventions that link to environmental/social planning and communication/marketing policies could prove effective in maximizing the positive effects of social opportunity on screening behaviours.
Psychological capability was found to be influential in screening be-

| Strengths and limitations
A major strength of the current study is the use of the COM-B and TDF models in examining influences on screening behaviours. The COM-B and TDF models provided a framework to explore and identify the complex factors that influence these behaviours. A comprehensive behavioural analysis of influential factors could be undertaken, which is an essential step in developing theoretically informed interventions. 40 The links between the TDF and COM-B models to the BCW provide a systematic process to fully understand the nature of screening participation, characterize interventions and link these to specific policy categories. The TDF model also informed the topic guide, which enabled information to be gathered on women's typical screening behaviours. Another strength is the inclusion of women with a range of screening histories. This proved insight into an understudied cohort-those who do not (consistently) engage with screening. We also distinguished younger and older women, which is important, given the indications here of how influences on behaviour may differ in these groups. However, as in any qualitative study, women who were interviewed (both AS and IS) were motivated to participate and may have distinctive views/opinions about screening.
Moreover, it is worth remembering that the interviews took place against the background of a high-profile controversy around CervicalCheck. It is possible that this may have impacted on women's views on screening and, hence, on our findings.

| CONCLUSION
This study has identified that women's screening decisions were influenced by a variety of factors, some of which can evolve over time.
Establishing positive screening behaviours that persist will require tailored strategies that support autonomous motivational processes and increase capability (physical and psychological) and opportunity (physical and social). The study findings can be mapped to specific intervention functions, thereby taking a step towards the development of evidence-based and theoretically informed interventions to improve screening uptake. The changes, if implemented, will mean that all women (aged 25-60 years) will attend for screening every 5 years. This will be different from current screening protocols, where women aged 24-44 years attend for screening every 3 years.' 'Do you have any thoughts on the proposed changes?' 'Do you think you will be more or less likely to participate in cervical screening in the future if these changes are made?' • Future improvements to cervical screening