“I don't mean to be rude, but could you put a mask on while I'm here?” A qualitative study of risks experienced by domiciliary care workers in Wales during the COVID‐19 pandemic

Abstract Domiciliary care workers (DCWs) continued to provide care to adults in their own homes throughout the COVID‐19 pandemic. The evidence of the impact of COVID‐19 on health outcomes of DCWs is currently mixed. The OSCAR study will quantify the impact of COVID‐19 upon health outcomes of DCWs in Wales, explore causes of variation and extrapolate to the rest of the UK DCW population. An embedded qualitative study aimed to explore DCW experiences during the pandemic, including factors that may have varied risk of exposure to COVID‐19 and adverse health and wellbeing outcomes. Registered DCWs working throughout Wales were invited to participate in a semi‐structured telephone interview. 24 DCWs were interviewed between February and July 2021. Themes were identified through inductive analysis using thematic coding. Several themes emerged relating to risk of exposure to COVID‐19. First, general changes to the role of the DCW during the pandemic were identified. Second, practical challenges for DCWs in the workplace were reported, including staff shortages, clients and families not following safety procedures, initial shortages of personal protective equipment (PPE), DCW criticism of standard use PPE, client difficulty with PPE and management of rapid antigen testing. Third, lack of government/employer preparation for a pandemic was described, including the reorganisation of staff clients and services, inadequate or confusing information for many DCWs, COVID‐19 training and the need for improved practical instruction and limited official standard risk assessments for DCWs. Pressure to attend work and perceptions of COVID‐19 risk and vaccination was also reported. In summary, this paper describes the risk factors associated with working during the pandemic. We have mapped recommendations for each problem using these qualitative findings including tailored training and better support for isolated team members and identified the required changes at several socio‐ecological levels.


| BACKG ROU N D
The adult social care sector employed 1.5 million people across 17,700 organisations in England (2019), including 685,000 staff as domiciliary care workers (DCWs; Hayes et al., 2020;Skills for Care, 2021). In Wales, where this study took place, nearly 22,000 people work as DCWs (Social Care Wales, 2021a, 2021b and all are registered with Social Care Wales (a Welsh-Government-sponsored body). DCWs in Wales provide a wide range of support including preventative services, support for independent living and personal care (Social Care Wales [SCW], 2018).
These activities require close contact with clients, who may be at high risk from COVID-19, with some within the UK Government's 'clinically extremely vulnerable' category (Hayes et al., 2020).
Limited and mixed evidence exists on the impact of COVID-19 on this crucial workforce. The Office for National Statistics (ONS) reported mortality rates for 'Care workers and home carers' (SOC2010 code 6145) of 71.1 (males) and 25.9 (females) per 100,000 between 9 March and 25 May 2020 as compared with rates in the general population of 19.1 (males) and 9.7 (females) per 100,000. It is important to note that the occupational classification was absent in many ONS records (17.5% and 37.5% missing for males and females respectively; ONS, 2020). In addition, the impact on other health and well-being outcomes were not reported (or indeed collected). Rates of testing positive for COVID-19 (via a Polymerase Chain Reaction [PCR] test) in June 2020 found rates of infection amongst the DCW workforce to be similar to that of the general population (Public Health England, 2020).
The OSCAR study (Outcomes for Social Carers: an Analysis using Routine data)~is using a mixed-methods design to quantify the impact of COVID-19 upon health outcomes of DCWs in Wales, to explore causes of variation, and to extrapolate to the rest of the UK DCW population (Lugg-Widger et al., 2021). Through data provided on the DCW workforce combined with electronic health record (EHR) outcome data within the Secure Anonymised Information Linkage (SAIL) Databank, the study will be the first to comprehensively quantify the risk of COVID-19 on the UK DCW workforce. This paper describes the (OSCAR) study's qualitative component, which aimed to explore DCW experiences during the pandemic, including factors that may have varied risk of exposure to COVID-19 and adverse health and wellbeing outcomes. The paper reports on risks associated with DCWs working through the pandemic.

| Study design
Within a mixed-methods study, in-depth qualitative telephone interviews were conducted to explore DCW pandemic working experiences and inform anonymised individual-level population-scale routine data linkage (Lugg-Widger et al., 2021).

| Eligibility
Participants were eligible if they were registered or working towards registration as a DCW with Social Care Wales (SCW), had actively worked for at least 4 months during one or both initial waves of the COVID-19 pandemic in Wales and were able to provide consent.

| Identification and approach
The proposed sample size for the qualitative study was 30. DCWs were informed and invited to participate through networks such as care-providing organisations, unions and Social Care Wales.
Informal methods of contacting DCWs included websites, social media (Twitter) and 'snowballing' where participants informed work colleagues of the study. Recruitment and interviews were carried out between February and July 2021. Interested DCWs contacted the qualitative researcher and had the opportunity to discuss any questions they had about the study. DCWs were offered a £20 Amazon voucher to participate. On agreement to interview, the researcher requested basic information to establish eligibility and ensure maximum variation in the sample. Screening questions included the number of years working as DCW, sex, ethnicity, region of work and employer classification. DCWs were immediately notified following screening if the interview could proceed and a suitable time was agreed. domiciliary care workers,qualitative,risk,social care What is known about this topic • Domiciliary care workers provided care in the community throughout the COVID-19 pandemic • Evidence to date is mixed regarding the impact of the pandemic on these workers • Specific risks for care workers posed by working during the pandemic remain under-explored.

What this paper adds
• Team working and care workers' roles changed during the pandemic identifying a training gap for those working in different ways and those redeployed.
• Factors contributing to risk to DCWs health and wellbeing included practical challenges such as shortages of staff and PPE and clients' understanding/following protective measures.
• Lack of preparedness resulted in inadequate and confusing information and training not meeting care worker needs.

| The qualitative interviews and informed consent
Before the interview commenced, the researcher took informed consent remotely via telephone using Skype for Business and recorded the discussion using a research ethics committee (REC) approved consent script (Appendix S2.). The audio-recorded interviews were carried out remotely, with DCWs given the option of using either telephone or online video recording software. The interviews were carried out at a place of choice for the DCWs, (e.g. the office, at home). The interviews were conducted by a clinically trained female, experienced researcher (HP), although her clinical background was not disclosed. The interview guide topics included DCWs' backgrounds, experiences of working practices before and during the COVID-19 pandemic, feelings about how the COVID-19 pandemic affected DCWs personally and an exploration of vaccinations, including the COVID-19 vaccinations. Participants were also given the opportunity to discuss any other topic that they considered relevant. The interviewer made notes and regularly discussed anonymised content with the other qualitative researcher and study team. The interviewer grouped together similar content to make a list of initial codes, when interviewees were no longer raising or responding with new content that could not fit into main list of codes it was deemed that no more interviews were needed and data saturation had been reached.

| Qualitative data analysis
Audio recordings were transcribed verbatim, anonymised and uploaded to NVivo 12 qualitative analysis software (QSR International, 2018). The researcher coded and developed themes from the data using inductive thematic analysis (Braun & Clarke, 2006). Thematic analysis used the six-step process to analyse data as per Braun and Clarke (2006).

| RE SULTS
Recruitment: Following initial contact with the researcher, six DCWs did not reply to the researcher following receipt of the study information letter. Twenty-four DCWs were recruited and interviewed (Table 1), with data saturation being reached at this point. The interviews lasted between 28 and 76 min (mean duration of 50 min).
Characteristics of the participants: Although all interviewed individuals were registered with SCW as DCWs (SCW, 2021b), they comprised two groups; those that visited several client homes (referred to here as Care Workers [CWs]) and those that worked mostly in single supported living homes which were home to several clients (referred to here as Support Workers [SWs]

Staff shortages
Clients still requiring care in spite of the pandemic 'Everyone felt quite the same, you know, we don't really want to be putting ourselves in that situation, but we understand we've got to, because there's no-one else there to do it. We have to, that's our job. The care doesn't stop because they've got Covid' (id 2 CW) 'I never once thought you know, I can't be in work, because I wanted to be there for the people I support, because they needed, you know, me and my colleagues more than ever you know.' (id 16 SW) Clients still requiring care in spite of a reduced workforce 'so suddenly, suddenly you know, you've got five staff members who're experiencing symptoms and they go off and you need to cover their shifts…. It was disastrous.' (id 1 CW) Waiting for results of PCR tests causing lengthier staff absences 'so that's at least 72 hours we need to cover their work' and if the test is positive, the staff member has 'to stay off for 10 to 14 days' (id 2 CW) Fear of catching COVID-19 possibly causing reduced applications for DCWs 'I think because people are scared because of the pandemic so that has definitely had a massive impact in recruiting new staff' (id 1 CW) 'it's recruitment, um it just seems really, really difficult to, to retain and, and employ staff, that are wanting to work frontline'. (id 13 CW) New DCW recruits not suitable for the role 'we're finding that we are employing some people and their just heart, their heart isn't in it, they just want a job and it's like ….you know you've got a job, but then caring is another step up, you know you need to have those personal qualities … you know your empathy and everything' (id 5 CW) Media coverage not helping recruitment 'throughout COVID, I don't believe or I don't feel that kind of care homes, nursing homes or anybody in the care industry, had the best of media coverage. … because you were continually hearing oh, well, care homes are this, and now it's spreading like wildfire in care homes, you know, when actually' (id 24 SW) Lack of childcare leading to staff shortages 'some parents are needing the time off to look after children, because they're at home all day and there's no other childcare provision…… Um, you have to be a keyworker to, to access the childcare provisions in school'. (id 13 CW)

Clients/families and private workers not following safety procedures
Families not wearing face masks, a particular problem 'But it is mostly the families bringing it in, you see, and not wearing their masks. And we've had a couple, we had one little bit of an issue with one client, um, because, bless her, she's got six kids which all come in and out and visit her. Um, some do her shopping and things like that, um, ….again, all coming in not wearing masks'. (id 2 CW) Clients did not follow safety precautions [the clients] 'just could not get the hang of it' (id 7 CW) Private carers not wearing face masks 'So, I do find that they're complacent, and a couple of times I went into the one, the husband and wife and I went in and the private carer was in and she didn't have a mask on and I did have to say to her "Look, I don't mean to be rude, but could you put a mask on while I'm here?"' (id 23 CW)

Initial Shortages of PPE
Shortages of PPE 'Because it was an absolute nightmare getting PPE from your suppliers and of course, suddenly their prices shot up because everyone wanted PPE' (id 1 CW)

DCW criticism of standard issue PPE
Deficiencies of standard issue face masks 'whereas the ones that we were given free from the company …kept steaming up and steaming up, and I think they spent more time playing and cleaning these visors, that that's how they, I think that's how they got infected' (id 21 SW) 'They don't really work, so you've got to lift it up to try to keep … you can't be cleaning someone's teeth, or trying to bath them, if you can't see what you're doing. You need to see' (id 14 SW) Preference for alternative face masks 'I was here in the thick of it, working every day that week because everybody was going down one by one … and I was the only one that didn't get it. … But I had already prepared. I had bought myself this [mask]. I came in here and I did not take that off' (id 21 SW) Time consuming to put on PPE 'it's definitely been more time consuming because of putting PPE on and donning and doffing, taking it off at the right times' (id 24 SW)

Staff shortages
Staff shortages increased significantly during the pandemic (Box 1).
Staff had been advised to shield if considered vulnerable, while others had to isolate, stop working and wait for the results of PCR tests. Some staff remained at home to look after their children since usual carers such as grandparents needed to shield. Some staff were fearful of coming to work during the pandemic, and participants suggested that fear was also the reason for fewer DCW job applications. Participants discussed that media coverage hindered recruitment through anxiety-provoking messages that were considered misleading. Participants indicated that new recruits did not have the required personal qualities for the role. Staff in post had to take on the extra work, including those DCWs with an advanced role, they were busier as hospitals discharged many clients back into the community for their safety.

Clients/families and private workers not following safety procedures
Several DCWs (mostly care workers) stated that some families of clients did not follow the COVID-19 precautionary rules, especially wearing face masks. In these instances, DCWs were required to report this behaviour to social workers who would ring the families to remind them that the client was vulnerable. DCWs would also need to increase cleaning and ask families not to visit their relatives when the DCWs were present. Some clients also appeared not to follow safety precautions, for example, vulnerable clients visiting friends' houses or talking to service employees such as gardeners. Some clients became cross when challenged, with some DCWs not having had any guidance on handling such a situation resulting in uncertainty of the appropriate response. It was also reported that some private care workers did not wear face masks around the clients (one reportedly stating that there was no need as they had previously been infected with COVID-19).

Initial shortages of PPE
Several participants reported shortages of PPE relating to face masks and PPE required for working sleepover night shifts such as plastic mattress covers. One carer reported that a friend even started making masks for them.

DCW criticism of standard-issue PPE
The efficacy and comfort of standard PPE issued to DCWs were questioned with reports of visors steaming up when bathing clients resulting in frequent cleaning or raising of visors to ensure visibility.
Some DCWs bought PPE which they considered were of a superior standard claiming it stopped them from acquiring COVID-19. Some visors were described as looking like a welding mask. Another criticism of PPE was the time required to put these on. Transitioning from a support worker role to a care worker role 'from my point of view of trying to learn a new job, erm, and trying to learn everybody's needs and where they lived and … like you know medication and everything … it was very stressful, it felt, I mean I did feel physically ill, erm … after the shifts because I hadn't had enough to eat. I mean I felt really shaky and … dehydrated and sick you know' (id 9 CW)

COVID-19 training and the need for improved practical instruction
More specific training required 'maybe it could be, maybe a little bit more specific, to you know, workplace…. it was enough information and obviously, we had you know, information from, other resources, but, maybe just, maybe a bit more specific' (id 16 SW) Financial pressure to attend work if ill 'I was panicking a bit, if it did happen, but any situation, obviously, because you've still got to pay your rent and your car and all those sorts of things' (id 3 CW) 'we've got bills to pay haven't we' (id 8 CW)

BOX 3 Pressure to attend work during the pandemic
Potential dishonesty of reporting COVID-19 symptoms to avoid losing pay 'I know there are people that work in places that have gone to work, potentially, with COVID, because they just can't afford to be off' (id 23 CW) 'the minimum wage, how many people that might be minimum wage, can afford to have time off? … it's a conversation that came up once … how many people might be stuck, how many people might tell the truth'? (SW 14) 'usually what you find with staff is, erm, that they can't afford to take that deduction, so they'll come back to work as soon as they can' (id 17 SW) Queries regarding composition of the vaccinations 'I never quite trust, erm, the ingredients in them, they always say there's chemicals … heavy metals or something in them, mercury or something (id 12 SW) Fear that the vaccination may not be safe 'they were saying, right, well we're going to invent this vaccine, and I was thinking, well, well it's all very well saying you're going to invent it, how do we know a, that it's going to work and b, that you're not going to kill us all off, you know?' (id 23 CW) Lack of long term studies regarding the vaccine safety profile 'they are trying their best to assure people that it's gone through the same tests as the flu jab. But because this has been going on for just a year, people are very sceptical …very, very sceptical' (id 2 CW) 'But they don't really know what the long-term effects are. You've also got, erm, people with blood clots at the moment. It's only a tiny, tiny amount of people that's got it, and … but there's this been this panic, hasn't there, about people getting them' (id 19 SW) Fertility issues concerning in view of the 'short span of time it's taken them to make the vaccine'. (id 2 CW)

Futility of vaccinations in view of the virus mutating
'like Covid is, it's a virus that mutates, you never get the same injections, do you, because they mutate… but this is going to be a situation where you have to have the injection because it mutates, you've seen that all over the world, you know, getting their own versions all over the place now' (id 19 SW) Reasons for ultimately having the vaccination despite being initially sceptical 'the vaccination would have had to have gone through you know, rigorous testing and stuff, so you know, I was I, I had them done with you know, no issues really. I, obviously, did a bit of, you know, um, research stuff, and I, I, and I spoke to a few people like um, my GP and stuff, about vaccinations, to get some reassurance'. (id 16 SW) 'I've just sort of accepted it, because you know it's, it's the best thing to do, and you've been told it's the best thing to do, whether or not it is, you just don't know, you know, I don't think anybody can give an accurate answer yet' (id 13 CW) DCWs said they would take time to explain to the clients why they were wearing PPE while showing others wearing them on television, which they said helped. One client was annoyed with the DCW wearing gloves and told them to take all the PPE off.

Management of lateral flow testing
Although LFT kits had been supplied to many DCWs, some were still waiting for these at the time of the interview (February 2021). Many DCWs reported that staff used these twice weekly for asymptomatic testing. The majority of DCWs were positive about using LFTs, which provided peace of mind that they did not have a COVID-19 infection to transfer to clients or family members. Nevertheless, some

Limited official standard risk assessments
Participants described that staff returning to work after being advised to shield underwent a risk assessment. However, only one DCW (in an advanced role), was aware of the Welsh Government Risk Assessment Tool (Welsh Government, 2021). DCWs stated that risk assessment tools were usually used by the employer managers and had little knowledge of these tools themselves. Examples of assessments included those used by the health and safety team and a template that was sent to them from the central management team.
External companies were also used to assess staff who were deemed to be at a higher risk. Many risks assessments were, however, informal with participants reporting that they assessed themselves or each other. Indeed, one participant reported that the company did not know that they had mild asthma. Some DCWs reported that they had to read guidance on risk assessment and sign that they had done so, others stated that they were unaware of any risk assessments for DCWs.
Pressure to attend work during the pandemic Many DCWs were thankful for the statutory sickness payment enhancement scheme made available by the Wales Government (Welsh Government, 2020). However, this was perceived by some as being too low. Several DCWs reported that they had casual contracts or zero-hour contracts where an employer is not obliged to ensure a minimum number of working hours. Consequently, participants reported they would not get paid if they missed work due to sickness or self-isolating. Others reported that they did not earn enough money to qualify for sickness payment which would not include hours accrued working overtime while others still stated that some contracts stipulated that sickness payment would not start for 3 days. It was proposed that some DCWs who worked zero-hour contracts may not be able to afford to take a reduction in pay and may therefore be reluctant to report symptoms of COVID-19 or may return to work too soon. One DCW said that they actually knew people who had gone into work with symptoms. It was also stated that it was unethical that DCWs should lose money for being honest by taking a test for COVID-19. Some DCWs were given the option of taking annual leave instead to avoid losing pay. Others were told that if they wanted to continue their role as a DCW, they must give up any additional employment that they had elsewhere, and this also caused a reduction in pay. As well as financial pressures to attend work, the DCWs mentioned feelings of guilt for taking sick leave due to COVID-19 because they knew that their work colleagues were understaffed.
One DCW reported that it was considered unacceptable amongst their work colleagues not to work a scheduled weekend and an example was given of a DCW completing their weekend shift despite feeling ill. A minority of DCWs felt that management had not been adequately supportive during the pandemic causing them to feel pressure to attend work. One stated that it was management who decided if they had COVID-19 symptoms, rather than sending them for a test. Conversely, a DCW in an advanced role believed that some staff were being untruthful about having COVID-19 symptoms making it difficult for them to alter rotas (Box 3).

Perceptions of becoming infected with the COVID-19 Virus
Whilst some DCWs stated that they were still being careful to wear full PPE after being vaccinated, a minority stated that they were not worried about getting the virus, for example if living in areas with low COVID-19 case rates. It was proposed too that client families were becoming complacent with some no longer wearing face masks, with one example being because it was unnecessary as they had previously been infected with the virus. Post-vaccination complacency was also proposed amongst the DCWs with reports that some were not adhering to the 2-m distancing rule when taking off PPE outside of a clients' house. One DCW held the view that the virus had been manufactured to kill the older generation but had failed in doing so as many young people had also died (Box 4).
This is the first paper to focus on risks relating to the DCW role.
A lack of pandemic preparedness was reported in our study and by Nyashanu et al. (2020b) whose care workers reported no clear strategic policy regarding the pandemic. The practice of discharging people from hospital back to the community for their safety is problematic because of the risk of hospital-acquired COVID-19 (Healthcare Safety Investigation Branch, 2020;Hodgson et al., 2020;Oliver, 2021). As in our data, Nyashanu et al. (2020a)  This issue, reported in other studies, for example, that fear of infection caused sickness rates to increase with some staff taking annual leave entitlement (Nyashanu et al., 2020a) and others still reporting that self-isolation and delay in testing resulted in lack of staff (Nyashanu et al., 2020b). In spite of the fear of being infected with COVID-19 whilst at work, financial pressures of those on contracts with no sickness benefits may have pressured DCWs to work.
These contracts have been noted as problematic in the UK before the pandemic (Ravalier et al., 2019) but have still not been resolved.
Furthermore, Sterling et al. (2020)  and a common approach to addressing that would seem efficient.
Similarly, the pandemic placed stress on a sector with existing workforce challenges and some common responses such as models to enhance teamworking, particularly when working in relative isolation from other members of the workforce would seem appropriate.
We aimed to set the challenges and recommendations (Appendix S1.) within a broader socio-ecological framework (Ingram et al., 2021)  The qualitative study adds strength to the overall OSCAR study by providing contemporary carer-reported insights into risk factors for COVID-19 exposure that can inform the quantitative modelling (Lugg-Widger et al., 2021). Of even greater value perhaps is the detailed nuance achieved about how some risk factors and outcomes are experienced in practice and how interventions to address them may be introduced or improved. While the quantitative analyses will explore mental health outcomes for care workers before and during the pandemic, this qualitative study details a wide range of contributing factors such as workload, financial pressures and uncertainties about protective measures such as PPE availability, use and the role of vaccines.
Efforts were made to maximise the diversity of care workers interviewed by monitoring the socio-demographic profile of the emerging interviewee sample and exploring a range of advertisement and recruitment routes. Although recruitment was concluded at the point where no new major themes were emerging from the interviews, the sample is still likely to reflect bias and it is plausible that different issues may have been reported from a larger and even more diverse group of DCWs. However, the study was not expected to enable fully generalisable findings and it is likely that most broad concerns have been identified, something that appears

CON CLUS IONS
DCWs described a range of experiences that may have served to increase the risk of exposure to COVID-19. These risks operate at varying levels of the socio-ecological context within which care workers operate and recommended courses of action to address them may similarly be required at different levels and by different actors. In-depth understanding of the nature of risk factors can both inform health outcome analyses and interpretation, as well as guide responses to reduce risk including through optimisation of existing approaches such as increasingly tailored training.

AUTH O R CO NTR I B UTI O N S
MR and RCJ are the chief investigators. HP conducted the interviews, LBH was the second coder. FLW contributed to the recruitment and ethical approval of study documents. HP & FLW drafted the paper. All authors contributed to the design and analysis and advised on the drafts of the manuscript. All authors read and approved the final manuscript.