British Journal of General Practice Socioeconomic differences in help-seeking experiences in primary care for symptoms related to colorectal cancer during COVID-19: A UK-wide qualitative interview study

Background COVID-19 has led to rapid changes in health care delivery, raising concern that these changes may exacerbate existing inequalities in patient outcomes. Aim To understand how patients’ help-seeking experiences in primary care for colorectal cancer symptoms during COVID-19 were affected by socioeconomic status (SES). Qualitative men across the recruited using purposive by SES. management, with other methods of getting in touch (e.g., email, 111) taken up more readily by higher SES patients. Our findings suggest that COVID-19 exacerbated disparities between higher and lower SES participants. This study raises awareness around challenges in help-seeking in the context of the pandemic which are likely to persist (post-COVID) as healthcare systems settle on new models (e.g., digital) of care. Recommendations are provided to reduce inequalities of care. how and their perception of the interaction with healthcare professionals. Disparities between higher and lower SES participants were found in how people appraised their symptoms during the pandemic and how they perceived changes in primary care in terms of considerations around the need to visit the GP during the pandemic, understanding of how to access primary care, views about remote technology, safety-netting during COVID-19 and attitudes towards accessing care. Recommendations are made for minimising negative impacts on patient care during and post-pandemic.

1 Introduction Earlier diagnosis of cancer is considered a potential casualty of the COVID-19 pandemic (1) leading to avoidable deaths and significant economic impacts. (2) There have been a number of calls to action to ensure people seek prompt medical help, receive appropriate referrals and follow-up care. (3,4) This is particularly important in the UK context, given the lower cancer survival compared with countries with similar healthcare systems(5) and the likely influence of later diagnosis on these differences. (6) For colorectal cancer, 98% of patients diagnosed at Stage 1 survived their disease for at least one year, compared to 44% of patients diagnosed at Stage 4. (7) The impact of COVID-19 on colorectal cancer care is of particular concern, because it risks perpetuating persistent social/health inequalities in outcomes, (8,9) including late presentation, (10) less access to treatment,(11) delayed treatment, (12) higher mortality (13,14) and poorer survival for more deprived groups. (11) Evidence from population-based and community studies during the pandemic suggest that people were less likely to seek medical help for their symptoms, (15,16) were fearful of catching or transmitting COVID-19 by contacting their GP practice (16) and were less likely to be referred, diagnosed or treated for colorectal cancer. (17,18) It is not yet clear whether these impacts are socially patterned.
Recent data from Public Health Scotland suggests that significantly lower numbers of people were diagnosed with colorectal cancer (25% less) than before the pandemic and there was a trend 3 suggesting that the fall in numbers of people being diagnosed with Stage 1 disease was higher in the most deprived areas compared with less deprived areas. (19) Primary care underwent rapid transformation as a result of the pandemic, (20) with 90% of consultations delivered remotely in April 2020. (21) These changes present trade-offs for patients (22) and clinicians, (20) particularly in the context of assessing potential cancer symptoms. (23) Any positive consequences (e.g. improved flexibility/reducing COVID-19 infection risk), have to be balanced with ensuring adequate diagnostic assessment, (23) and avoiding exacerbation of health inequalities. (24) This study addressed a gap in the literature by providing qualitative evidence about patients' experiences of accessing primary care during the pandemic, and whether these experiences were shaped by socioeconomic differences. We draw on the concept of candidacy as an underpinning framework because it describes the way in which equity in access can differ because of the way patients and health services determine eligibility for healthcare. (25) This evidence can be used to make recommendations to mitigate against exacerbating existing inequalities in care.

Method Approach
Semi-structured interviews were carried out to provide an in depth understanding of patient experiences of the healthcare system when contacting about possible symptoms of colorectal cancer during the COVID-19 pandemic and how this varied by socioeconomic status.

Participant selection and recruitment
Participants were recruited through a research recruitment company (SAROS) who have a database of over 60,000 potential respondents across the UK and can screen based on different variables to gain populations of interest. Screening questions were developed by the research team to identify people from higher and lower socioeconomic (SES) groups across the UK who contacted primary care about symptoms related to colorectal cancer during the COVID-19 pandemic (since March 2020) and administered by the recruitment company. After screening, the research company arranged the interview between the participant and the researcher (AI) and provided the researcher with the participants contact details. Participants were given £50 for taking part in the interview. Symptoms included in the screening questions were those listed on the Bowel Cancer Awareness Measure toolkit (26) and Bowel Cancer UK website. (27) The main index for SES used in this study was education as this has been used in previous research exploring impact of SES on response to cancer symptoms.(28) People with 'O' levels/GCSEs or equivalent or with no formal qualification were in the lower SES group and those with A level or above were in the higher SES group.

Data collection
Interviews were carried out by an experienced qualitative researcher (AI) from October 2020 to November 2020 via phone or Zoom (duration=57 minutes; range: 31-86 minutes). Participants gave verbal consent to take part in the study. Interviews were digitally recorded and transcribed verbatim. The topic guide (Supplementary Box 1) focused on aspects of the Candidacy framework to understand how people decided to seek medical help and how they navigated services. (25) Analysis Transcripts were repeatedly read by one qualitative researcher (AI) to ensure familiarity with the data before coding using framework analysis. We used the Candidacy Framework to help organise the data, but themes were developed using an inductive approach. This framework was useful for exploring individual's identification of their 'candidacy' for accessing and negotiating healthcare services which highlighted barriers and facilitators to accessing care during the pandemic. Comparative thematic analysis (28) was then carried out to explore the differences between higher and lower SES groups. This involved initially analysing interviews by socio-economic group, before moving to analysing differences between groups. This two-step approach to analysis was useful for identifying themes that were running across the whole sample before looking at the differences between individual groups and comparing them with one other. AI had multiple data analysis meetings with three members of the team (KW, GB, CV) to further refine the findings and ensure that the final themes reflected the data. These were then further discussed with the wider team.

Results
A total of 39 participants were recruited (mean age= 50 years, range from 25-78 years) from higher (n=20) and lower (n-19) SES backgrounds across the UK. This sample size was to ensure that we had as close to equal number of participants in both groups to be able to conduct a comparative analysis. For information about participant characteristics including employment status, living arrangements etc., see Supplementary Table 1. The most reported symptoms in both groups was a combination of extreme tiredness for no obvious reason and a persistent and unexplained change in bowel habits (n=6). Telephone consultations (higher SES=17; Lower SES=16) were the most reported initial mode of consultation followed by face-to-face (higher SES=2; lower SES=2), e-consult (lower SES=1) and video (higher SES=1). Three main themes were identified across the data exploring how people decided to seek help, how people navigated services and ease of accessing these, and impact of COVID-19 on how patients interacted with healthcare professionals (Table 1).
[ In contrast, people in the higher SES group were more confident in the likelihood that their symptoms were related to COVID-19, which influenced their assessment of going to the GP practice: "Obviously with a stomach thing it's less likely that I've got COVID, so I suppose they're more amenable to you coming in." (P21, 61, Higher SES, pain in abdomen, persistent and unexplained change in bowel habits)

Relationship between body vigilance and lifestyle modifications
Some people in the higher SES group reported how lifestyle modifications during the pandemic including working from home or being furloughed made them more vigilant about changes in their body. They reported how this increased awareness was a driver for seeking help from the GP: Changes in diet and exercise as well as heightened worry and stress, were described by higher SES as contributing to their symptoms and may have made symptom discrimination more difficult: "I thought, maybe the weight loss was because we are at home, we all went for exercise, healthier eating, cooking more at home, you know." (P16, 50, Higher SES, unexplained weight loss, pain in abdomen)

"Well I wondered at first whether it might be, you get a bit anxious in lockdown and so on and I'm a vulnerable age, so I thought it might be that, but then I wasn't so sure." (P28, 70, Higher SES, persistent and unexplained change in bowel habits)
In contrast, people in the lower SES group did not mention lifestyle changes, working from home or being furloughed as influences on whether they noticed symptoms or sought help from their GP.

How people navigated services 2.1 Accessing healthcare in the face of a pandemic
Both groups weighed up the balance between their own needs against potential risks and NHS resources. However, people in the higher SES group appeared to be more determined to contact their GP for advice, despite the pandemic, in case their condition was serious or worsened: This indicates that having confident expectations about making contact with health services and avoiding contagion could affect help-seeking behaviours, particularly disadvantaging lower SES participants.

Impact of COVID-19 on how patients interacted with healthcare professionals 3.1 Views about utility of remote technology
7 The majority of people in the higher SES group and some in the lower SES group described advantages to remote consultations such as convenience (higher and lower SES), not competing with work commitments (lower SES) and recognising that the use of technology was a positive outcome of the pandemic and the way forward for the NHS (higher SES):

Knowing how and when to seek further help
The increase in remote contact methods for primary care affected how lower and higher SES groups felt their safety was being monitored. For example, people in the higher SES group reported more active forms of safety netting, including a timescale for re-contact when symptoms had not resolved, medication to ease symptoms whilst waiting for their referral appointment, or other remote options of getting back in touch such as emailing or calling 111. This made participants feel reassured:

Summary
Interviews with the public who contacted primary care during the pandemic highlighted disparities between SES groups in how people decided to seek medical help, how they navigated services and how they perceived interactions with healthcare professionals. The overshadowing of COVID-19 and uncertainty about its symptoms had differential effects on both groups, as compared to the lower SES group, the higher SES group appeared more certain about signs of COVID-19 and were more likely to seek a face-to-face appointment and follow-up consultations for their colorectal symptoms, rather than worrying about viral transmission. Furthermore, lifestyle changes because of COVID-19 restrictions allowed people in the higher SES group to be more body vigilant, which appeared to speed up their decision to consult. Awareness of the NHS being under strain and the increased needs of the public may have had a differential effect on both groups in terms of prioritising their own health needs, as the lower SES group appeared more concerned about burdening the NHS. Perceptions around accessibility and safety measures during the pandemic were divided according to SES, with the higher SES group knowing more about how to navigate access compared to the lower SES group. We also found differences in attitudes towards remote consultations, with more reservations reported by people in the lower SES group which may have had an impact on participants' ability to articulate their symptoms and may have implications for future help-seeking behaviours. Finally, with the introduction of remote consultations there was an increased importance of transparent safety-netting and care planning during the pandemic, and differences in the use/perceptions of these may have exacerbated inequalities between higher and lower SES participants.

Strengths and limitations
To the authors' knowledge, this is the first study to explore views and experiences of patients accessing primary care during the pandemic for symptoms of colorectal cancer, and whether there were socioeconomic differences. Participants were recruited from across the UK making the findings generalisable to UK nations.
Participants were identified through a research company and this method of recruitment can result in a highly self-selecting group of participants who will have a degree of technical skills that may increase their competence in engaging with telemedicine during covid restrictions. However, it also increases access to people from lower SES groups who may otherwise be hard to reach and recruit.
The means by which we recruited participants meant that member checking following study completion was not possible, although previous research in the field has found this a helpful strategy for augmenting and interpreting interview data with participants. (30) Although we sampled into higher and lower SES groups to draw out differences, the binary nature of our analyses is a limitation, and does not necessarily capture the nuances across the SES gradient. We also focused on people who sought help during the pandemic as we particularly interested in understanding how changes due to the pandemic influenced accessing primary care. However, it would have been useful to capture views of those who did not seek help too. Finally, we did not capture healthcare professionals' views on how delivery of care changed and the challenges communicating this to patients, however, we are exploring this in another study.
The mean age of our sample was reflective of when the risk of colorectal cancer starts to rise in the general population,(31) although those in the higher SES group were on average slightly older than those in the lower SES group which may have impacted aspects of the pathway (e.g. decision to refer). However, it remains important to ensure safety-netting strategies and communication about next steps are the same for all patient sub-groups, particularly as the incidence of colorectal cancer is rising in younger age groups.(32)

Comparisons with existing literature
These findings support previous research, which applied the candidacy framework to understand how aspects of the doctor-patient interaction influence perceived eligibility for help-seeking based on challenges with recognising that symptoms need medical attention and subsequently how to navigate services.(25) However, they extend them by showing differences between lower and higher SES groups. This study found that the higher SES group reported being more body vigilant during the pandemic and were more likely to seek help, linking this to working from home or being furloughed because of COVID-19. This is in line with a previous study exploring the impact of body vigilance on cancer 'alarm' symptoms, (33) which found that paying more attention to bodily changes was significantly associated with seeking help for cancer symptoms. This was less apparent in the lower SES group.
People in the lower SES group were less certain when discussing the cause of their symptoms and were more likely to worry that their symptoms were COVID-related (which would make them more likely to consult their GP about COVID and risk delaying care for potential symptoms of colorectal cancer). Previous studies have also found that non-recognition of symptom seriousness and lower cancer symptoms knowledge in lower educated groups, (34)(35)(36)(37) is linked to delays in presentation for lower SES groups. (38,39) The finding that people in the lower SES group were more likely to express concern about burdening the healthcare system may be explained within the context of candidacy because people in more deprived areas may witness greater burden of ill health in their communities, and therefore higher levels of frequent attendance in primary care. (40) Safety-netting during the pandemic is also highlighted as important in a previous study as with the use of remote consultations and in situations of uncertain clinical presentations, it can be an effective way of monitoring people to reduce clinical risk. (41) However, patients need to know how to seek further help/re-present if their symptoms persist. This is of particular significance since we know COVID-19 has led to delayed presentations suggesting people require additional support when navigating through the pathway due to new methods of contacting the GP. (3) Many of the participants in this study experienced remote consultations by telephone as opposed to video despite some participants reporting that they would have liked a video consultation if given the choice. However, the rarity of video consultations is usual in general practice as there is a perception that telephone or face-to-face consultations are more useful for majority of circumstances. (21) The importance of considering patient preferences when offering remote consultations and recognising how it may be difficult for certain groups was echoed in a recent study exploring experiences of healthcare professionals during the pandemic. (20) Other studies have also mentioned how remote consultations could potentially widen inequalities in healthcare particularly for older and vulnerable groups.(24, [42][43][44] We found that participants navigated services during the pandemic depending on their perceptions of safety and knowledge of processes (e.g., online appointment systems). This is consistent with a recent survey study that found there was low public awareness of changes to face-to-face consultations. (45) Reluctance to access primary care due to fears of catching COVID-19 is reported by other studies conducted around the pandemic. (3,46,47) Addressing barriers to accessing care may help to reduce this gap, (48) which was already a concern pre-pandemic.

Implications for practice
A summary of recommendations for practice is provided in Box 1. With the uncertainty around COVID-19 symptoms and colorectal cancer symptoms, there is the potential for future impact of long COVID on colorectal symptom appraisal and confusion.(49) Furthermore, the changing nature of guidance about COVID-19 symptoms and differences between NHS and other public bodies in their public-facing communication (e.g., where bowel symptoms are present in some but not others) makes raising awareness of symptoms a challenging avenue. There needs to be consensus on advice that is accessible to the public to avoid ongoing confusion and disseminating these through campaigns and more sustainable information (e.g., on GP practice websites and NHS letters).
There is an urgent need for campaigns to encourage particularly lower SES groups to notice changes in their bodies, such as offering practical action plans (e.g., campaigns to encourage people to set aside time to notice any changes and provide flexible and feasible options to make appointments), as external influences (such as type/place of employment) may make it harder for them to notice and interpret potential cancer symptoms. For people in the higher SES group this will also be important as the easing of the pandemic will result in more people going back to work and being less likely to notice bodily changes.
Reassuring people that safety measures are in place through reliable sources and using different methods to reach people (email, text, letter, practice website) will help, particularly for people from lower SES groups who showed more hesitancy around accessing primary care due to uncertainty about the process and subsequent safety measures.
It is important that people are signposted to appropriate and reliable information about methods of consultation. Proactive efforts to address primary care access and ensuring that patients have an option for face-to-face or remote consultation, will support those with digital access challenges.
Providing people with active safety-netting through different methods (e.g. providing people with a timeframe of when to reconsult and through different methods such as phone, email or app) will help going forwards, especially given the backlog in cancer referrals (17) and the added uncertainty due to the pandemic.
Further research is needed to look at how inequalities may be generated across the care pathway, gathering perceptions from healthcare professionals on the main changes to the cancer pathway and their impact on inequalities, and link changes in care with data similar to epidemiological studies conducted during COVID(1, 8) to see whether the gap is sustained. For example, concerns have been raised that the introduction of quantitative faecal immunochemical test (qFIT) to stratify risk and prioritise patients for limited endoscopic services has led to missed cancers, (50) despite showing promise pre-pandemic. (51) In conclusion, this study provides important insights into barriers accessing primary care during the pandemic and highlights the disparities between higher and lower SES groups including appraisal of symptoms, perceptions around safety and availability of services during the pandemic, use of remote consultations and safety-netting. Recommendations are provided on how we can reduce these inequalities and potentially decrease barriers which may lead to delayed cancer diagnosis. These findings may also be applicable to help-seeking for other symptoms, not just those related to colorectal cancer.