Barriers to colonoscopy in UK colorectal cancer screening programmes: Qualitative interviews with ethnic minority groups

Abstract Objective People from ethnic minority backgrounds are less likely to attend colonoscopy, following faecal immunochemical test screening, and are more likely to be diagnosed with colorectal cancer at an advanced stage as a result. The aim of this research was to explore the barriers and facilitators to attending colonoscopy, perceived by ethnic minority groups living in the United Kingdom. Methods Semi‐structured online and telephone interviews were conducted with thirty men and women of Black‐African (n = 5), Black‐Caribbean (n = 5), South Asian (n = 10) and White British (n = 10) descent. Participants were eligible for screening, but had not necessarily been invited for colonoscopy. All interviews were conducted in the participant's first language and were assessed using Framework‐analysis, in line with a conceptual framework developed from previous interviews with healthcare professionals. Results Five thematic groups of barriers and facilitators were developed: ‘Locus of control’, ‘Cultural attitudes and beliefs’, ‘Individual beliefs, knowledge and personal experiences with colonoscopy and cancer’, ‘Reliance on family and friends’ and ‘Health concerns’. Differences were observed, between ethnic groups, for: ‘Locus of control’, ‘Cultural attitudes and beliefs’ and ‘Reliance on family and friends’. Black and South Asian participants frequently described the decision to attend colonoscopy as lying with ‘God’ (Muslims, specifically), ‘the doctor’, or ‘family’ (Locus of control). Black and South Asian participants also reported relying on friends and family for ‘language, transport and emotional support’ (Reliance on family and friends). Black‐African participants, specifically, described cancer as ‘socially taboo’ (Cultural attitudes and beliefs). Conclusions The results highlight several targets for culturally‐tailored interventions to make colonoscopy more equitable.


| BACKGROUND
Colorectal cancer (CRC) is a leading cause of morbidity and mortality in Europe. 1 Several large randomised controlled trials have shown that regular faecal immunochemical test (FIT) screening, between the ages of 45 and 80, can significantly reduce CRC-mortality among people who complete the test. 2 As a result, most European countries have implemented FIT-based screening programmes for the early detection of CRC. 3 As with all screening, the extent to which the benefits of FIT are realised, and the extent to which they are equitable, is highly dependent on uptake of the test, as well as any necessary follow-up investigations (colonoscopy being the gold standard for FIT screening). 4 However, in a recent international survey of 35 FITbased screening programmes, Selby and colleagues (2021) found that the mean proportion of participants, with a positive FIT result, who attend colonoscopy, was 79%, with completion rates ranging from 39% in the programme with the lowest level of colonoscopy attendance, to 100% in the country with the highest. 5 As with uptake of FIT, evidence suggests that attendance at colonoscopy is lower among those from an ethnic minority background, compared with those of White British ethnicity. 6 Such disparities may contribute toward ethnic inequalities in CRC outcomes seen in the UK. 7 For example, Black and South Asian adults, living in the UK, are more likely to have lymph node involvement at diagnosis (compared with White adults), 7 and are less likely to be diagnosed via screening (the diagnostic route associated with the best outcomes for CRC). 8 To date, the majority of qualitative research exploring nonattendance at colonoscopy has been conducted with White, English-speaking, adults. 9 The little research that has been conducted with non-White populations has been conducted primarily in the USA, with Black African Americans. 9 Key findings from a recent review of the literature indicate that procedural costs, perceived threat to masculinity and lack of insurance are among the most prevalent barriers to colonoscopy. However, the findings from these studies are not necessarily transferrable to the UK and other European countries, because of cultural differences between countries, differences in healthcare delivery, and the fact that colonoscopy is often offered as the primary screening test in the USA. 9 The aim of this research, therefore, was to explore the perceived barriers and facilitators of colonoscopy, among ethnic minority groups (including native speakers of English and patients with limited According to the most recent census (a national survey of UK households, undertaken by the Office of National Statistics), 9.3% of the population in England and Wales identify as 'Asian, Asian British or Asian Welsh' and 2.5% identify as 'Black, Black British, Black Welsh, African or Caribbean' (the census defines ethnicity as "The ethnic group that the person completing the census feels they belong to […] based on their culture, family background, identity or physical appearance"), 10 making these two of the most common ethnic minority groups in the UK (data for and Wales Scotland and Ireland have not yet been published, and so the ethnic composition for these regions is not currently known; however, England and Wales account for 92% of the UK population, and thus reflect the majority of the population). 10 The overall attendance at colonoscopy, within the screening programme, and across the population, is 80%.

Study design and participants. Online and telephone interviews
were conducted with members of the British public, who: (1) were of screening eligible age in England, Wales, Scotland and Northern Ireland (i.e. aged 60-74 years), (2) had been invited to participate in bowel cancer screening at least once, and 3) were of either a South Asian, Black African, Black Caribbean or White British ethnic background (these ethnic groups were selected for inclusion in the research, on the basis that they are the least likely to attend colonoscopy and receive a diagnosis through the bowel cancer screening programme [with the exception of the White British group, which was included to help disentangle 'ethnic group-specific barriers' to colonoscopy from 'universal barriers']). Non-attendance at colonoscopy was not a requirement for inclusion in the study, as previous research has found that those who decline colonoscopy are unwilling to participate in such interviews. 11 Participants were subsequently those who were characteristically unlikely to attend colonoscopy, as opposed to those who have been invited for (and not attended) colonoscopy. Key characteristics of participants (e.g. age, gender, ethnicity first language, etc.) were determined through a short survey, administered at the beginning of the interview (Appendix A).
Sample size and recruitment strategy. To minimise language barriers to participation, targeted recruitment strategies were employed, with White British participants being recruited via Social Media, and Black and South Asian participants being recruited through Agroni: a multi-disciplinary research organisation with extensive specialist knowledge and experience working with Britain's ethnic minority group communities. 12 Participants recruited through Agroni were identified and recruited by a team of professional recruiters, who visited community centres, Mosques, Mandirs, and charities, with which they had existing connections. Individuals who appeared to meet the eligibility criteria were approached by the recruiters, who discussed the study with them on a one-to-one basis (due to the sensitive nature of the topic; i.e. cancer). Those who expressed an interest in taking part in the study were given Agroni's contact details, and asked to call or email for further information about the research. All but one person given Agroni's contact details went on to complete an interview (i.e. 20/21). Twitter and Facebook were used to recruit participants via social media. A digital flyer was used to provide details about the research and invite participants to call or email the lead researcher (RK), if interested in taking part.
A stratified sampling approach was used to ensure equal numbers of male and female participants, as well as participants of different ethnicities (the researchers had no prior relationship with the participants). The recruitment strategy also aimed to sample participants from across the UK, but was not stratified by region or nationality.
On the basis that data saturation is achieved after 9-17 interviews, 13 we aimed to recruit 10 participants from each ethnic group (i.e. Black, South Asian and White British). Data saturation was subsequently assessed by reviewing whether new codes were developed following analysis of the 10th interview of each group.
Data collection. Interviews took place between July 2021 and July 2022. To minimise barriers to disclosing culturally sensitive information, interviews were concordant for gender, ethnicity and the participant's first language. RK (White male) conducted all interviews with White British participants, who identified as male (n = 5). NG (White female), meanwhile, conducted all interviews with White British participants, who identified as female (n = 5).

Finally, Agroni researchers conducted all interviews with Black and
South Asian participants, identifying as either male (n = 10) or female (n = 10). Zero participants identified as any other gender (e.g.

non-binary).
To ensure Agroni researchers understood the aims of the study, and thus conducted the interviews appropriately, RK met with them prior to data collection, to discuss the interview schedule and address any questions about the research. No additional training was given to Agroni researchers, as they were all experienced researchers with prior experience conducting qualitative interviews.
Interviews lasted 36 min on average (range: 24-67 min) and were conducted using a semi-structured interview guide, which was developed by three members of the research team (RK, CvW and KW), all of whom have PhDs in Psychology. The interview schedule was developed using a conceptual framework, which was developed through interviews with specialist screening practitioners in a previous study led by the research team. 14 The interview schedule was pilot tested with two individuals (one by RK, with an individual of White British ethnicity; one by Agroni, with an individual of South Asian ethnicity), prior to data collection, to ensure participant comprehension of the questions (no changes were made to the interview guide). Questions focussed on patients' perceived barriers and facilitators to colonoscopy attendance, following participation in a FIT-based screening programme for CRC (see Appendix B).
An audio recorder was used to record the interviews. The recordings were anonymized, translated and transcribed verbatim by K International: a UK-based translation and transcription services provider. 15 Following transcription, the audio files were deleted by K International and the research team.
Participants were given the option to participate in either an online or telephone interview. Participants who opted to participate in an online interview were given a choice of MS Teams or Zoom. All participants opting for an online interview opted to participate via MS Teams. The interviews were conducted remotely, from the researchers' and participants' homes. No one was present during the interviews, besides the researcher and the participant.
Informed consent. Informed consent was obtained from participants, before the interviews commenced. For White British participants, an information sheet and consent form were provided, via email, to those expressing an interest in taking part. A mutually convenient date for interview was then agreed, for those who returned a completed consent form (via post or email). The purpose of the study, the right to withdraw from the study, and the right to skip any questions, were repeated at the beginning of the interview, along with the participant's decision to proceed with the interview.
For South Asian, Black African and Black Caribbean participants, Agroni researchers explained the purpose of the study to potential participants, in their first language, over the phone. For those interested in taking part, Agroni Researchers subsequently read through the information sheet and consent form (again, over the phone).
Those who were interested in participating, after providing verbal consent, were offered an interview on a mutually agreeable date. As with the White British interviews, the purpose of the study, the right to withdraw from the study, and the right to skip any questions, were repeated at the beginning of the interview, along with the participant's decision to proceed with the interview. Data analysis. Transcripts were analysed using framework analysis. 16 This method is well-suited for generating knowledge relevant for health and clinical practice, 17 allowing for in-depth understanding within individual cases, as well as analyses of key themes across the data-set. While this is a flexible method, not aligned with a particular epistemological, philosophical, or theoretical approach, this study adopted an interpretivist approach, which recognises the importance of situating the researcher in the context of that which is being studied, 18 to offer an interpretive understanding of the meaning participants ascribe to their own experiences. 19 A detailed overview of the analysis process is provided below:   Two researchers (RK and NG) initially independently developed ideas for codes using the same sample of transcripts (n = 6; 20%). Codes were generated inductively (i.e., from the data) and deductively,

| Initial development of codes (stages 1-4)
according to a framework previously developed from interviews with specialist screening practitioners, led by the research team (i.e., codes were used/adapted from a list of previously curated codes, as and when relevant to the data; see Appendix C). 14 Stage 4: Development of a working analytical framework. Researchers (RK and NG) met to compare their independently generated ideas for new and revised codes and, through discussion, consensually agreed on a working analytical framework that could be applied to further transcripts (i.e., a 'revised' set of codes and the meaning of each).

| Application of codes (Stage 5)
Stage 5: Applying the analytical framework. One researcher (NG) then coded the remaining transcripts using the revised codes. Where a new transcript was judged to contain text that could not be satisfactorily coded using the existing codes, new codes were created, or existing codes updated to more accurately reflect the data they represented. Any new codes were discussed and agreed with RK, before NG subsequently revisited previously coded transcripts to apply the new codes (if/where relevant).

| Use of coded material to develop themes (Stages 6 and 7)
Stage 6: Charting data into the framework matrix. The data were charted into a matrix, to provide a summary of the transcript material assigned to each code per participant (some cells were blank where no material existed in a transcript for a particular code; see: https:// osf.io/pvk3w/). The charted framework matrix was then reviewed by three researchers (RK, NG and ET) to develop themes (all stages of coding and data analysis were carried out in Microsoft Excel).

| Participant characteristics and data collection
In total, 15 men and 15 women participated in the study (none of whom dropped-out or withdrew from the research). Participants  Table 1; individual-level characteristics are reported in the appendix (see Appendix E).  Table 2. A summary of the themes, subthemes and codes developed is presented in Table 3. In addition to discussing it as being primarily the doctor's decision, South Asian participants frequently reported that the family held an important role in the decision-making process, and that they would follow the advice of their partners and children. Black participants also highlighted that they would discuss it with their family, but argued it was not the family's decision to make.

| Description of themes
White British participants, meanwhile, predominantly discussed making the decision independently of others, and going online for information and advice to help them make their decision. This process was, for example, discussed by one patient, who had recently attended colonoscopy.

(The influence of) God and religion
South Asian and Black African adults, who were of Muslim faith, also frequently discussed God, saying they would "trust in God" and attend the appointment. Paradoxically, other South Asian and Black African participants, also of Muslim Faith, indicated they would not go, as "No one can change the decision of God, not the doctor, nor any other physician".
White British adults, and those who were not of Muslim faith, meanwhile, did not discuss God, despite many of them being of Christian faith.

Theme two. Cultural attitudes and beliefs
Cultural attitudes and beliefs (religious and non-religious) were also frequently discussed by participants, and included a broad range of cultural taboos, and the importance of having a female healthcare professional perform the examination.

Cultural taboos
Colonoscopy, colons, rectums and cancer were all discussed as culturally taboo topics by participants. For Black African participants, cancer, in particular, was reported to be culturally taboo and feared by the community. Black African participants also discussed the need to undress and have "something in your ass" (referring to the endoscope) as "evil" and "shameful" (these views were not discussed by Black Caribbean participants). South Asian participants, meanwhile, indicated that, while they could not discuss such issues as colonoscopy readily with their peers, they could (and would) discuss them with their family. For White British participants, however, it was "bums" and "bowel habits", specifically, which were discussed as being socially taboo.

Importance of Female HCPs
Only one participant (Black African) highlighted the importance of having a same-sex practitioner, specifically for Muslim women.
Their views on this were very strong, however, with them stating: colonoscopies, even when the test was described as "painful", as they demystified the procedure. Those who previously attended colonoscopy also discussed the peace of mind it provided, and referred to this as an important motivation for attending future colonoscopies.
The same motivator was reported as a reason for going to colonoscopy, by those who had not yet been invited. Those with a family history of bowel cancer, in particular, endorsed attending/wanting to attend colonoscopy, for this reassurance.

Lack of trust in western medicine
One Black Caribbean participant, who had previously had a positive bowel cancer screening test result, reported that they did T A B L E 3 Summary of the themes, subthemes and codes identified. not attend colonoscopy, as they "did not trust Western medicine". As only one participant endorsed this view, and was not discussed in relation it to their ethnic background, or religious beliefs, in any way, it was not possible to attribute this belief to their cultural background.

Valuing health
Participants from all ethnic groups highlighted the value they place on their health, and discussed the importance of doing "everything for the betterment of our health" as a reason for attending colonoscopy and other healthcare appointments.
4. Theme four. Reliance on family and friends.
Many participants reported that they relied on friends and family when attending hospital appointments, and that the same was/would be true for attending colonoscopy. Participants relied on family and friends in a number of ways, including transport (getting to and from the hospital), translation services (interpreting the information materials, the nurses and doctors, etc.) and emotional support. While reliance on family and friends did not appear to be intrinsically linked to ethnicity, it was reported more frequently, and more prominently, by Black and South Asian participants.

Reliance on family and friends as unofficial interpreters
Being unable to read English was described as a barrier to attending colonoscopy (by Black African and South Asian participants), one which could not be overcome without the help of friends and family. Being unable to speak English was also described as a barrier to attending colonoscopy (again, by Black and South Asian participants). Here, too, some participants described relying on the support of family members as interpreters, with some reporting that official interpreters were not always available (where they were available, however, participants stated they would use them to overcome language barriers).

Reliance on family and friends for transport
Several Black and South Asian participants also reported relying on family and friends to take them to the hospital. In one instance, this was because they were "not allowed to go anywhere alone", while, in another, it was because they had health issues, making it difficult for them to attend by themselves. One White British participant, who had previously attended colonoscopy, indicated that they would need support from friends and family to get to and from the hospital.

Reliance on family and friends for emotional support
Black and South Asian participants also reported wanting to bring a friend or family to their appointment for emotional support (while not overtly linked to ethnic group, the need for emotional support from others was not as prominent among the White British participants interviewed).

Theme Health concerns
Finally, participants of all ethnic groups discussed pre-existing health conditions, and the risk of getting COVID, as barriers to attending colonoscopy.

Existing health conditions
Several participants discussed existing health conditions as barriers to colonoscopy, either because they were higher priority, or because they presented possible complications (for example, one participant, who was taking warfarin, was worried about potential bleeding, while another, who had multiple morbidities, was concerned how the doctors would accommodate their physical condition).

COVID
Several participants also expressed concerns about COVID. Such concerns were expressed by participants from all ethnic groups, but were more prevalent and pertinent among those of Black and South Asian ethnic groups. Interestingly, others felt the pandemic was "over", and that COVID was "not an issue".

| Comparison with existing literature
This study is the first to explore barriers and facilitators to colonoscopy among British ethnic minority groups, in their first language. This has been attempted previously in the UK, without success (colonoscopy decliners were unwilling to meet with researchers to discuss their decision not to attend colonoscopy). 11 Innovative approaches to collecting these data, therefore, may be required. One possible approach would be to record the pre-colonoscopy consultation in which patients decide whether to attend colonoscopy (although this would not capture the barriers for those who do not attend the pre-colonoscopy assessment, which is a minority of colonoscopy non-attenders). 6 Such approaches have been employed in other areas of healthcare, and provided valuable insights into patient-doctor decision-making. 24 In addition to the above, quantitative research is needed to understand how barriers and facilitators interact with one another, and which of the perceived barriers and facilitators are predictive of nonattendance at colonoscopy. This could be achieved through national surveys, distributed to both attenders and non-attenders, which may be more acceptable to non-attenders than interviews (a similar approach has been used previously for flexible sigmoidoscopy screening non-attenders, and successfully quantified the barriers for this). 25 This study also has several implications for policy. First, to reduce barriers to screening (and colonoscopy), there is strong public health mandate to update national systems, so that invitations are sent in the first language of recipients (these data are recorded on the GP clinical systems, from which national screening programmes obtain the necessary information to administer invitations [although their current access to these systems does not allow them to access information about a patient's first language and, therefore, need to be amended]). Second, there is a need to address cultural taboos and stigma, surrounding cancer, the colonoscopy procedure, and other sensitive topics, in order to make access to screening and colonoscopy further equitable still (this could be achieved through changes to the invitation materials, which, as discussed, could be tailored to patients' first language).
Finally, there is a wider need to educate the population that bowel cancer can be a treatable disease, particularly when it is diagnosed early, and that participating in screening (and attending follow-up colonoscopy) can improve patient outcomes (this could also be achieved through changes to the invitation letter, as well as social media campaigns, which could be targeted according to the individual's first language).

| Study limitations
This study has several limitations. First, it is subject to hypothetical bias, as several participants had not received an abnormal bowel cancer screening result, but rather, were asked to imagine they had.