Trust and inclusion during the Covid‐19 pandemic: Perspectives from Black and South Asian people living with dementia and their carers in the UK

Abstract Introduction People from ethnic minority backgrounds living with dementia are more likely to be diagnosed later and have less access to health and social care support than their White counterparts in the United Kingdom (UK). Covid‐19 has exacerbated health inequalities and diminished trust from underserved communities in the government and health services. The wider aim of the study was to explore the impact of covid‐19 on Black and South‐Asian people living with dementia and their carers as well as exploring the experiences of dementia care. The present paper specifically explores their views on trust and mistrust using an ecological model. Method Semi‐structured interviews were conducted with 11 family carers and four people living with dementia from South Asian or Black communities living in the community. Thematic analysis was used to analyse data. Design An exploratory qualiative design was used. Results Four main themes were developed exploring trust at the structural, organisational, community and individual level. At the structural level, participants discussed the inequity of Covid‐19, some lack of trust in the UK Government and confusion in its messaging, and the anxiety sometimes leading to curtailment of media usage. At the organisational level, there was some evidence of a perceived lack of person‐centred and culturally sensitive care from healthcare professionals, as well as concerns around care homes as places of safety. At the neighbourhood community level, participants discussed both a distrust as well as a strengthening of relationships and, at the individual level, factors such as knowledge of services, identity, and faith influenced their experience of the pandemic. Conclusions People living with dementia need support at all levels and this study highlights how the pandemic impacted each level. Ways to improve trust in the Government and health professionals alongside culturally adapted health messaging should be explored. Alongside this, an examination of how cultural values and norms may influence help‐seeking responses to dementia and increase trust in services may be helpful post‐pandemic.

For individuals and their families navigating the dementia trajectory, Covid-19 posed additional challenges and amplified many problems. Not specific to ethnicity, these challenges often led to a worsening of dementia symptoms, isolation, disrupted routines and reduced access to face-to-face dementia services. 13,14 Although many carers showed resilience during the national lockdown periods of severe social distancing at the heights of the pandemic, many experienced increased anxiety and depression. 15 Those who had limited support networks seemed to be at greater risk of feeling an increased sense of burden. 16 The pandemic disproportionally impacted people from ethnic minority groups across the globe, with increased severity and mortality risks. 17,18 Explanations for the inequalities related to  included higher prevalence of pre-existing conditions and the impact of social determinants of health (e.g., poorer living environments, less opportunity to work from home). 19 This is despite evidence (prior to vaccine roll out) that people who were Black British or British South Asian were more likely to engage in Covid-19 preventative behaviours (e.g., hand-washing, facemask wearing, and social distancing) than White British people, which was mediated through a higher fear of Covid-19 and lower political trust. 20 Diminished trust in governments and health services is emerging as an important legacy of the pandemic in many countries 21 ; however, it is likely these trust issues are longstanding amongst these groups and COVID-19 has brought these issues to light whilst exasperating them. Trust in scientists in the UK remains high, particularly among older people and those who are higher educated, 22 but by the end of 2020 there was an increase from 42% to 68% in the proportion of the public who believed that the governmental response to Covid-19 had been confusing and inconsistent. 23 Research has already highlighted how a lack of trust in UK-based authorities has led to a low  vaccine update in vulnerable populations. [24][25][26] To reduce health inequalities, it is integral to explore trust in Governments, health services and communities and how it can be improved.
The impact of the pandemic on those living with dementia in the UK has been previously explored, but few qualitative studies have accessed participants from minority ethnic groups. This present paper reports on data collected as part of (anonymised), aimed at developing an evidence-based decision tool to support family carers and people living with dementia when making difficult healthcare decisions during the pandemic. As part of this project, we interviewed people from South Asian and Black ethnic groups living with dementia in the UK and found that the pandemic had increased their feelings of fear and anxiety, impacted negatively on their diet and eating habits, compounded feelings of isolation and curtailed their support networks and health care. 27 It was evident the pandemic was influencing people's confidence in various institutions and required an in-depth exploration through a socio-ecological model to capture the interplay between individual, relationship, community, and societal factors influencing trust. 28

| AIM
The wider aim of the study was to explore the impact of covid-19 on Black and South-Asian people living with dementia and their carers living in the community as well as exploring the experiences of dementia care. The present paper specifically explores views on trust and mistrust using an ecological model.

| Design
An exploratory qualitative design was used with semi-structured interviews; reporting was guided by the Standards for Reporting Qualitative Research framework. 29

| Participants
Participants were recruited from a range of sources, including those involved in previous research, GP Practices, online dementia research recruitment websites (such as Join Dementia Research), social media (e.g., Twitter), and two local organisations. We also approached two other organisations (Together in Dementia Everyday (TIDE) and a local Alzheimer's society, but no participants were identified from these. None of the organisations were specifically for ethnic minority communities but were based in areas with high proportion of people from ethnic minority groups. � People lacking mental capacity to provide informed consent to be interviewed.

| Recruitment
Local and national dementia and carer networks initially approached possible participants or they were contacted directly by the research team. Those interested in participating were asked to contact the researchers directly or agreed to be sent a copy of the study advertisement, study information leaflet and consent form (via email or post as preferred). Interested participants were then telephone screened against the study criteria and given an opportunity to ask questions. Interpreters were able to be used if needed.
Consent forms were completed electronically and returned to the research team via email, where possible. These were then countersigned by the researcher and returned to the participant. The consent forms were adapted in plain English for people living with dementia. If participants were unable to complete the consent form electronically, consent was taken verbally and audio-recorded prior to the start of the interview, with confirmation of this then posted to participants.
Researchers trained in assessing mental capacity consented participants. We followed the principles of the Mental Capacity Act to assess capacity. Participants were screened for mental capacity over the telephone or using video technology; ascertaining whether they understand the information pertaining to the study, can weigh up the risks/benefits of taking part, can retain the information and can communicate their decision and consent to take part.

| Data collection
Individual semi-structured interviews were conducted remotely over telephone (n = 13) or via secure internet video (Microsoft Teams, n = 2). Interviews were conducted by (Anon; n = 14), an academic GP who is British South Asian) and (Anon; n = 1, a researcher who is White British). Interviews lasted approximately 1 h (range: 35-75 min) and were audio-recorded. The interview topic guide was developed with the input of the research team, including our patient and public involvement (PPI) representatives (which included people from ethnic minority backgrounds), and was modified as interviews progressed (See Supporting Information S1). The topic guide focused on the impact of the Covid-19 pandemic on Black and South Asian carers and ARMSTRONG ET AL.

| Data analysis
Thematic analysis guided by Braun and Clark's framework 30 was used to identify, analyse, and report themes. Line-by-line coding was undertaken in NVIVO 12. 31 An inductive approach was taken, where different members of the team read several transcripts multiple times to identify codes from the data. Codes were then synthesised into the themes and subthemes and discussed between the whole team to explore and agree meanings and interpretation over several meetings. A socio-ecological model 28 was applied to the codes when it became evident that the codes were reflecting individual, relationship, community, and societal factors.

| Participants
We recruited 15 participants in total. As this is an underserved group it can be challenging to recruit large numbers. At 15 participants, we could not identify further participants to take part despite using multiple sources and snowballing methods. However, towards the later interviews little new information was being revealed in the interviews. Of the four participants living with dementia, two were female and two male, ranging in ages from 66 to 88 years, and all had been born outside the UK. Of the 11 carers, five were caring for a spouse and six for a parent. Ten carers were female and one was male, ranging in ages from 29 to 85. Of the 15 participants, seven were of Black ethnicity (all were Caribbean) with English as their first language and eight were of South Asian ethnicity (seven Indian and one Pakistani) with English as their second language. All participants had some further education with eight having degrees or equivalent (see Table 1). In the quotations below, P denotes person living with dementia and C denotes carer.

| Themes
All those interviewed reported feeling some degree of lack of trust and a feeling of exclusion at various levels: individual, the community (neighbourhood and groups), organisational (e.g., NHS) and structural (e.g., Government, social norms). Whilst specific questions were asked regarding Government, media and support, the issue of trust came up organically by the participants throughout the whole of the interviews. These findings are presented using a socio-ecological model (see Figure 1). 32

| Inequality with Covid related care and deaths
Nearly all participants were aware that those from an ethnic minority were more likely to catch the virus and trusted this information. The increased anxiety among some participants: When that started to come out, yes. That did start to worry me when they said that black and ethnic minority could contract it more than others. That did start to worry me (Carer of a parent, black ethnicity C7).
Some perceived this inequality was driven by the fact that people from ethnic minorities are disproportionally from poorer backgrounds, have less access to services, and are more likely to be in frontline jobs and unable to work from home: I should be looking into that statistic more, from what I've read it's more people from ethnic backgrounds have mostly… Specifically they're from poorer backgrounds and then that means less access to certain things, maybe healthcare and stuff like that (Carer of a parent, Black ethnicity C7).
Participants handled this knowledge differently, with some preferring to protect the person they were caring for from knowing by not discussing the greater impact Covid-19 was likely to have to their ethnicity group: …but I do think that didn't help. Because that's going to create anxieties within families that I don't think there needed to be… They [Person with Dementia] just know it's a bad virus and you're not supposed to go out. And you could catch it and it can make you ill. But they didn't realise that it was going to affect one group more than the other. Because I think that could be quite distressing (Carer of a parent, Black ethnicity C1).

| Community
The Covid-19 pandemic limited dementia specific community services and general community resources. This appeared to amplify a sense of distrust among some study participants who described withdrawing further from the community and not accessing potential support; however, conversely, others described a strengthening of relationships within their neighbourhoods.

| Distrust and detaching from volunteers and neighbours
Some participants did not accept support supplied by neighbourhood volunteers as they were concerned about possible criminal intent and did not trust them: There were a lot of people on the road that were willing to do volunteering and whatever. We didn't take any of the offers up, because I didn't know these people.

| Individual
Trust was explored differently at the individual level by exploring how likely individual factors, such as knowledge and culture, were likely to influence accessing services and support that they trusted.
Additionally, as explored in detail elsewhere [Authors], we briefly explore here how trust in faith may have supported some of this population.

| Knowledge of accessing trusted services and support
How much support people received appeared to largely depend on whether they knew how to access various services and whether they found them acceptable. Below, a participant describes trying to get hold of a letter to allow her to shop at quieter times and before general opening at the start of the pandemic, and how offers of support kept needing to be repeated: Another participant discussed a system that allowed them to contact their relative living with dementia online and was particularly helpful in linking up with family members in another continent.
The participant had come across this system by chance in conversation with another carer and seemed to have trusted this recommendation: We to pull together and support those more vulnerable, as is often seen in times of crisis. 42 Social and religious support can improve wellbeing during challenging and uncertain times, 43 although these community behaviours tend to decline as emotional and physical resources become depleted. 44 Some individual factors also seemed to play a role in influencing participants' level of trust and the care received, such as their knowledge of how to access services and support; amplifying the trend for ethnic minority groups being less likely to access dementia services. 8,12 Another individual factor that emerged is how the carer identifies with their culture. Some participants felt it was their responsibility and part of their culture to care for their family or elders, although values and beliefs about the role of the family in caring for the person living with dementia will vary between and within different cultures. The impact this has on carer wellbeing and trust is unclear 45 and warrants further exploration.

| STRENGTHS AND LIMITATIONS
Ethnic minority groups are often underserved in research, so although numbers are small, this qualitative study is one of the few to have explored how COVID-19 affected a sample of ethnic minority participants living with dementia and carers in the UK. All but one interview were conducted by a researcher of South Asian ethnicity, which may have helped with 'insider' knowledge and trust among participants from that group. 46 Discussions with the wider study team have enabled a range of perspectives to be included.
Applying the Socio-Economic model allowed the study team to consider how trust was impacted at various levels thereby breaking down future suggestions from individual to societal level. However, limitations of our study are that most of our sample were highly educated, able to use the internet, connected to support networks and could speak English. We acknowledge that there are likely to be additional pressures for those experiencing socio-economic deprivation and those who do not speak English Our sample may have a selection bias with those agreeing to take part having a general interest in research. Additionally, we could not explore cultural differences across sub-groups, but only broadly across people who are Black or South Asian. ARMSTRONG ET AL. -11