The willingness of UK adults with intellectual disabilities to take COVID‐19 vaccines

Abstract Background Given the much greater COVID‐19 mortality risk experienced by people with intellectual disabilities (ID), understanding the willingness of people with ID to take a COVID‐19 vaccine is a major public health issue. Method In December 2020 to February 2021, across the United Kingdom, 621 adults with ID were interviewed remotely and 348 family carers or support workers of adults with ID with greater needs completed an online survey, including a question on willingness to take a COVID‐19 vaccine if offered. Results Eighty‐seven per cent of interviewees with ID were willing to take a COVID‐19 vaccine, with willingness associated with white ethnicity, having already had a flu vaccine, gaining information about COVID‐19 from television but not from social media, and knowing COVID‐19 social restrictions rules. A percentage of 81.7% of surveyed carers of adults with ID with greater needs reported that the person would be willing to take a COVID‐19 vaccine, with willingness associated with white ethnicity, having a health condition of concern in the context of COVID‐19, having had a flu vaccine, being close to someone who had died due to COVID‐19, and having shielded at some point during the pandemic. Conclusions Reported willingness to take the COVID‐19 vaccine is high among adults with ID in the United Kingdom, with factors associated with willingness having clear implications for public health policy and practice.


Introduction
Willingness to take vaccines (or in the negativevaccine 'hesitancy') has long been a topic of research in public health and has become especially significant in the context of the current COVID-19 pandemic, leading to multiple research studies internationally (for reviews, refer to Lin et al. 2020;Sallam 2021). Large within and across-nation general population surveys of adults have been carried out since early in the COVID-19 pandemic (i.e. from as early as March 2020). These research studies have generally found: considerable variability in willingness to accept a COVID-19 vaccine across countries (Kerr et al. 2020;Lazarus et al. 2020); reducing willingness from the beginning of, or early in, the pandemic to the end of 2020 (Kerr et al. 2020;Lin et al. 2020); and that correlates of vaccine willingness/hesitancy vary across populations and countries (Lazarus et al. 2020).
People with intellectual disabilities (ID) have been shown to be at considerably increased risk for COVID-19 infection and mortality following COVID-19 infection compared with adults without ID (Clift et al. 2020;Gleason et al. 2021;Henderson et al. 2021;Williamson et al. 2021). The subpopulation of individuals with Down syndrome may be particularly vulnerable, potentially due to increased risks of immune dysregulation, congenital heart disease and respiratory conditions (Clift et al. 2020;Dard et al. 2020). Given such findings, and the underlying health inequities experienced by people with ID (Emerson and Hatton 2013), there have been calls internationally for people with ID to be prioritised for vaccines (e.g. Hotez et al. 2021). Countries including the United Kingdom (DHSC 2021) have acted in response to such calls, and the emerging data, to include people with ID in vaccine priority groups.
In this context, understanding vaccine willingness among populations of people with ID is crucial to inform national and potentially international public health policy, particularly as there is some evidence that people with ID can be fearful of a range of medical procedures, including needle phobia (Marriott 2017;Kupzyk and Allen 2019). In addition, there is not a tradition of vaccine hesitancy in developmental disabilities research, except notably to understand views about vaccines in parents of autistic children (e.g. Goin-Kochel et al. 2020). As yet, there are also few data available about COVID-19 vaccine willingness and people with ID. Lunsky et al. (2021) gathered data from 3371 workers in developmental disability services in Ontario, Canada, in early 2021 asking them about their own willingness to take the vaccine. The majority (62% very likely, a further 20% likely) reported that they would likely take the vaccine when offered. Given the importance of protecting individuals with ID, especially within residential settings (Landes et al. 2020), the Lunsky et al. study provided useful data to inform educational initiatives to increase vaccine uptake among paid care workers. Factors associated with non-intent to have a COVID-19 vaccine were: being younger, and endorsing beliefs about the lack of vaccine benefits, worry about rapid development of the vaccine, and concerns about side effects (Lunsky et al. 2021).
We could find only one study with any data from people with ID themselves. In a survey of family carers, paid staff, and people with intellectual and developmental disabilities in New York State, USA, Iadarola et al. (2021) reported data on vaccine willingness from 91 people with disabilities (including people with ID). 'Vaccine intent' was defined as definitely or probably being willing to have the vaccine and included a small number of people who had already had a vaccine, with 83.5% reporting positive vaccine intent. Correlates of vaccine intent were not examined separately for the group of disabled people responding directly to the survey.
Given variability in vaccine willingness across countries, variability in correlates of vaccine intent across countries, and the lack of data about vaccine willingness directly from or about individuals with ID, additional country-specific data are needed urgently to inform vaccine policy. In the current study, we report data about willingness to take a COVID-19 vaccine from: 1 adults with ID across the four UK countries (England, Northern Ireland, Scotland and Wales) who could take part in a remotely conducted interview (e.g. via videolink or telephone) 2 carers of adults with ID who were not able to take part in a direct interview with researchers.
For both samples, we report factors associated with willingness to take the COVID-19 vaccine.

Participants
Two samples were recruited: adults with ID who were interviewed by a researcher (Cohort 1), and adults with ID who would not be able to take part in an interview, where family carers and support workers reported via an online survey (Cohort 2). Full details of the survey methods and participants can be found in Flynn et al. (2021). Selected demographic information is presented in Table 1 for both cohorts. For Cohort 1, most of the participants were aged 16-44 years (68.2%), just over half were men (51.3%), the vast majority were white (94.7%), and 11% were people with Down syndrome. Substantial proportions of participants in Cohort 1 were living with their family (41.4%), living alone or with a partner (36.3%) or living with other people with ID in some form of supported housing or residential accommodation (22.3%). For Cohort 2, most people with ID were aged 16-45 (80.8%), a majority were men (56.5%), and the vast majority were white (94.1%). Almost half were people with 'Profound and Multiple Learning (Intellectual) Disabilities' (44.0%), and 18.3% were people with Down syndrome. Most adults with ID in Cohort 2 were living with their family (63.6%), relatively few people were living alone or with a partner (10.7%), and a quarter of people were living with other people with ID in some form of supported housing or residential accommodation (25.7%). Table 1 shows the set of measures used for the analyses in this paper, including how they were coded for the purposes of analysis. Responses reported by small numbers of participants (e.g. participants with a gender identity other than man or woman) were collapsed into broader categories or excluded from the specific analyses involving the relevant variable, but these participants were included in all other analyses (e.g. people identifying as a gender other than male or female were excluded from analyses concerning gender, but were included in all other analyses). An empty cell in Table 1 indicates that the particular measure was not collected with the particular cohort.

Measures
The selection and wording of measures was finalised through extensive consultation with groups of people with ID (particularly for Cohort 1 interviews) and family organisations (particularly for Cohort 2 surveys), to maximise relevance and accessibility. Where possible, survey items were adapted from those used in larger UK population surveys concerning people's experiences through the COVID-19 pandemic, such as the Office for National Statistics Opinions and Lifestyle Survey (ONS 2021) and the Understanding Society COVID-19 survey (Understanding Society 2021). A shortened version of the Pandemic Anxiety Scale (McElroy et al. 2020) was also used, retaining four of the seven items related to worries about COVID-19 infection or leaving the house, and excluding worries about job and financial prospects.
For the analyses included here, measures included demographic factors, items related to health and COVID-19, and (for Cohort 1) items concerning information and knowledge about COVID-19 restrictions and anxiety related to the pandemic.

Procedure
Recruitment of people into the study was facilitated through multiple methods across the United Kingdom, including through collaborating organisations in each country, social media, and wider networks of ID and family organisations across England, Northern Ireland, Scotland and Wales. Potential participants could express interest in the study via telephone, e-mail, social media or clicking a link to the survey (for family carers and support staff only) on the research project website. Contact details of people who had indicated an interest in taking part in Cohort 1 were sent to research teams in each country, who contacted each person to talk through the project and send them the easy read participant information sheet. If people were still interested in taking part, at least 24 h later, the interviewer arranged to go through the consent process and, if the person consented, For Cohort 1, trained research interviewers directly interviewed adults with ID via Zoom, telephone, Microsoft Teams, WhatsApp video call, Messenger video call or FaceTime, depending on the interviewee's preference. All interviewees had the capacity to take part in the interviews and gave their consent to do so before the interview was conducted. Data were entered directly into Qualtrics ™ during the interviews by the interviewers. Three people preferred to self-complete an online version of the survey, and in these cases, this was made available to them at their request. Participants were also able to have a supporter of their choice (e.g. family member and support staff) present at the interview. In all cases, flexibility was paramount to ensure that people with ID were able to participate in their preferred way. Interviews took typically 45 min to complete, and were usually completed in one sitting. Short breaks were offered during interviews when needed. All interviewers had experience of research interviewing and were trained via online training sessions within each country, with additional training sessions across all UK interviewers and regular supervision for 953 interviewers. Of those who consented to be interviewed, 86% (n = 545) answered the question on vaccine hesitancy. For Cohort 2, information was collected via an online Qualtrics ™ survey about adults with ID who were not able to take part in an interview with a researcher. To gather data on this group, we surveyed their family carers or paid support staff. Eighty-three per cent of participants were family carers of an adult with ID, and 15% of participants were paid support staff of an adult with ID. The remaining 3% of participants were other people who knew the adult with ID very well (e.g. a friend).

Data analysis
Cohorts 1 and 2 data sets were analysed separately throughout, using SPSS 26. For categorical variables, potential associations between willingness to take a COVID-19 vaccine were analysed using Fisher's exact test (for 2 × 2 tables) or chi-square (for 3 × 2 tables), and relative risk (RR) (with 95% confidence intervals) for all comparisons. For mean scores on the four items taken from the Pandemic Anxiety Scale (McElroy et al. 2020), a t-test (with Cohen's d to evaluate effect size) was conducted between those willing or unwilling to take the COVID-19 vaccine. All tests were two-sided or two-tailed.

Ethical approval
Research ethics approval was sought and obtained from the Manchester Metropolitan University Faculty of Health, Psychology and Social Care Faculty Research Ethics Committee.

Data availability
A quantitative data set will be archived online in a form that will be available to researchers after all waves of data collection for the project have been completed.

Cohort 1interviews with adults with intellectual disabilities
The vast majority of adults with ID interviewed (87.0%; n = 474) said that they would take a coronavirus vaccine if they were offered one, with 8.4% (n = 46) saying they were not sure, and 4.6% (n = 25) saying they would not take the vaccine. Table 2 first presents potential associations between demographic factors and willingness to take a COVID-19 vaccine. The only statistically significant association concerned ethnicity (Fisher's exact P = 0.033), with people in white ethnic groups 22% more likely (RR 1.22) to be willing to take a COVID-19 vaccine if it was offered, although the 95% confidence interval for RR did contain 1, indicating that the two groups may not be distinct in their vaccine willingness. Table 2 also presents potential associations between health-related factors and willingness to take a COVID-19 vaccine. There was a statistically significant association with people having had a flu vaccine in the 2020/21 winter flu season (Fisher's exact P < 0.001): people who had had a flu vaccine were 19% more likely (RR = 1.19) to be willing to take a COVID-19 vaccine than people who had not or did not want a flu vaccine.
In terms of factors related to information and knowledge about coronavirus and social distancing rules, Table 2 reports that two information sources were associated with willingness to take a COVID-19 vaccine. People who got information about coronavirus from the television were 15% more likely to be willing to take a COVID-19 vaccine (Fisher's exact P = 0.002; RR = 1.15), and people who got information from social media were 12% less likely to be willing to take a COVID-19 vaccine (Fisher's exact P = 0.003; RR = 0.88). People who said they knew the social distancing rules where they lived were also 14% more likely to be willing to take a COVID-19 vaccine (Fisher's exact P = 0.013; RR = 1.14).
Finally, Table 2 shows there was no evidence of a difference in shortened Pandemic Anxiety Scale mean scores between those who would be willing to take a COVID-19 vaccine and those who were not sure or would not be willing to take a vaccine (t = 1.84; df = 533; P = 0.067 (two-tailed); Cohen's d = 0.25).

Cohort 2survey of carers (including paid support staff) of adults with intellectual disabilities
The vast majority of carers of adults with ID surveyed reported that the person they care for would take a coronavirus vaccine if they were offered it (81.7%; n = 259), with 13.2% (n = 42) reporting they were not sure, and 5.0% (n = 16) reporting that the person would not take a vaccine. Table 3 first presents potential associations between demographic factors and carer reports of the willingness of the person with ID to take a COVID-19 vaccine. The only statistically significant association concerned ethnicity (Fisher's exact P = 0.043), with people in white ethnic groups reported to be 30% more likely (RR 1.30) to be willing to take a COVID-19 vaccine if it was offered, although the 95% confidence interval for RR did contain 1. Table 3 also presents potential associations between health-related factors and willingness to take the COVID-19 vaccine. There were four statistically significant associations.
People with ID were 20% more likely (RR 1.20) to be reported by carers as willing to take a COVID-19 vaccine if the person had a health condition of concern if they caught COVID-19, such as epilepsy (Fisher's exact P = 0.004). There was also a statistically significant association with people having had a flu vaccine in the 2020/21 winter flu season (Fisher's exact P < 0.001): people who had had a flu vaccine were 61% more likely (RR 1.61) to be reported to be willing to take a COVID-19 vaccine than people who had not or were reported to not want a flu vaccine. People with ID were 20% more likely (RR 1.20) to be 956

Discussion
Direct comparisons with other UK vaccine willingness data are not possible given different timings of data collection and also variations in phrasing of questions. However, data derived from the Understanding Society UK panel survey reported 82% vaccine willingness in over 12 000 participants (Robertson et al. 2021); vaccine acceptance rates in UK samples repeated from May to October 2020 were consistently above 70% (Kerr et al. 2020); and in a sample of over 5000 UK adults vaccine acceptance was 71.7% in September to October 2020 (Freeman et al. 2020). All these were conducted with general population samples, and it was not reported if any data from people with ID were included. These data suggest that acceptance of COVID-19 vaccination (as reported in the current research) is at least as high in people with ID, if not slightly higher. Not only may this be good news given the additional risks of people with ID to serious consequences of COVID-19 infection, but it indicates that people with ID are likely to play their part in COVID-19 public health programmes. More carers were unsure about whether people with more significant impairments would be willing to have a vaccine (Cohort 2) than people with ID themselves (Cohort 1). However, this was mainly associated with a perception among Cohort 2 respondents that suitable supports and adjustments may not be in place to enable people with more severe/profound ID and multiple health needs to receive a vaccine (Flynn et al. 2021).
We did not observe associations between vaccine willingness and either age or gender; factors that have been examined extensively with inconsistent findings in other COVID-19 vaccine willingness research (Freeman et al. 2020;Kerr et al. 2020;Lazarus et al. 2020;Robertson et al. 2021). We examined a number of other demographic factors that may be associated with vaccine willingness and found the same association in both cohorts: individuals from white ethnic groups were more willing to accept COVID-19 vaccines compared with combined all other ethnic groups. These findings are similar to those in other UK samples (Freeman et al. 2020;Robertson et al. 2021). We did not have sufficient sample size to examine vaccine willingness in specific minority ethnic groups, but vaccine hesitancy is associated in particular with Black and Pakistani/Bangladeshi communities in other UK research (Robertson et al. 2021). More recent analysis of COVID-19 vaccine coverage has found that, among non-shielding adults with ID aged 16-64, vaccination rates are considerably lower across black, mixed, South Asian and other ethnic groups compared with white groups (OpenSafely 2021). Future research, therefore, should attempt to examine more fine-grained analysis of any associations between ethnicity and vaccine willingness in adults with ID.
We found no evidence that adults with Down syndrome were more or less willing to take a COVID-19 vaccine, and this is encouraging given that they have been identified as a particularly at risk group (Clift et al. 2020).
In the current study, health-related factors were consistently associated with vaccine willingness in both cohorts of adults with ID. Vaccine willingness was associated with indicators of health vulnerability in Cohort 2 (having a health condition of concern in relation to COVID-19, and having shielded at some point during the pandemic). Greater willingness in both cohorts associated with having received a flu vaccine in the 2020/21 winter season may indicate experience with vaccination is facilitative for adults with ID, or that again additional health vulnerability (and thus need to have a flu vaccine) is associated with vaccine willingness, or both explanations may be true. In Cohort 2 (but not Cohort 1), vaccine willingness was associated with the person being close to someone who had died of COVID-19.
Finally, for Cohort 1 (questions were not asked for Cohort 2), getting information about COVID-19 from television bulletins was associated with increased willingness, getting information from social media with lower levels of willingness, and reporting knowing about local pandemic related restrictions (i.e. being better informed) was associated with increased willingness.
This study has a number of limitations to be borne in mind when considering the findings. Participants identified themselves or people they were caring for or supporting as people with ID (or 'learning disabilities' in current UK parlance). The representativeness of the samples of participants is unknown (and there are no reliable data across the United Kingdom to make these comparisons). Although both samples were diverse in terms of gender, age, country and living situation, there were small numbers of people from any minority ethnic communities (5% in Cohort 1; 6% in Cohort 2). Reliable data on the ethnicity of adults with ID across the United Kingdom is surprisingly lacking; the 2011 Census for the population as a whole reported 12% of adults aged 18+ from minority ethnic communities in England and Wales (ONS 2018); 4% for adults aged 16+ in Scotland (National Records of Scotland n.d.) and 2% for people of all ages in Northern Ireland (Northern Ireland Statistics and Research Agency 2016).
As the data are cross-sectional, we did not think it appropriate to build regression models examining the independent contribution of different correlates of vaccine willingness. This may be important in future. However, at this stage and to provide as much detail to policy makers and public health practitioners as possible, we focused on individual associations with appropriate effect size measures so that readers and organisations can independently assess the value of different correlates identified in the current study.
We suggest that these first data on vaccine willingness in adults with ID may be useful to inform public health efforts. In particular, subgroups less willing to accept COVID-19 vaccines could be the focus of educational interventions especially if these can be informed by additional research to understand hesitancy in some sub-groups. Given that adults with ID reported using television broadcasts as a main source of information about COVID-19 (Flynn et al. 2021) and that accessing information from the television was also associated with increased vaccine willingness, vaccine information targeted at (or made appropriately accessible for) adults with ID via television broadcasts may be particularly useful, compared with the current focus on producing easier-to-read web resources. In terms of adjustments that should be made in vaccine programmes, our findings also suggest that experience with vaccines (or perhaps other preparation to increase familiarity with/reduce anxiety about the vaccination procedure) may be crucial for some adults with IDmaking reasonable adjustments at all stages to improve the likelihood of people being vaccinated is likely to be crucial in maximising vaccination rates (Rotenberg et al. 2021).
Given the variation in COVID-19 vaccine willingness around the world, variability in predictors of willingness in different countries, and the lack of data on vaccine willingness among people with ID, more research, allied to urgent action, is clearly needed.