Cross-Sectional Study of University students attitudes to on campus delivery of COVID-19 vaccines and future-proofing MenACWY and MMR vaccination rates by adopting COVID-19 vaccine roll-out strategies

Background University students are a critical group for vaccination programmes against COVID-19, meningococcal disease (MenACWY), and measles, mumps and rubella (MMR). We aimed to evaluate risk factors for vaccine hesitancy (refusal or intention to refuse a vaccine) and views of university students about on-campus vaccine delivery. Methods Cross-sectional anonymous online questionnaire study of undergraduate students at a British university in June 2021. Chi-squared, Fishers exact, univariate and multivariate tests were applied to detect associations. Results Complete data were obtained from 827 participants (7.6% response-rate). Two-thirds (64%; 527/827) reported having been vaccinated against COVID-19 and a further 23% (194/827) agreed to be vaccinated. Other responses were either unclear (66) or indicated an intention to refuse vaccination (40). Hesitancy for COVID-19 vaccines was 5% (40/761). COVID-19 vaccine hesitancy was associated with black ethnicity (aOR, 7.01, 95% CI, 1.8-27.3) and concerns about vaccine side-effects (aOR, 1.72; 95% CI, 1.23-2.39). Lower levels of vaccine hesitancy were detected amongst students living in private accommodation (aOR, 0.13; 95% CI, 0.04-0.38) compared to those living at home. Uncertainty about their personal vaccine status was frequently observed for MMR (11%) and MenACWY (26%) vaccines. Campus-associated COVID-19 vaccine campaigns were definitely (45%) or somewhat (16%) favoured by UK-based students and more so among UK-based international students (62% and 12%, respectively). Conclusions Vaccine hesitancy among students of black ethnicity and those living at home requires further exploration because attitudes in these groups may affect COVID-19 vaccine uptake. High levels of uncertainty among students about their MMR and MenACWY vaccine status are also a concern for the effectiveness of these vaccine programmes. This issue could be tackled by extending the capabilities of digital platforms for accessing vaccine information, such as the NHSapp in the UK. Sector-wide implementation of on-campus vaccine delivery may also improve vaccine uptake, especially for international students.


Introduction
Definitions of primary endpoints 110 scores were derived for all students and analysed for differences between ethnic groups and term time residence locations as an alternate measure of vaccine 114 hesitancy. The potential utilisation of on-campus vaccination programmes was defined based on responses to question 15 split between those in favour 115 (definitely increase, somewhat increase) and those who were ambivalent (neither, somewhat decrease, definitely decrease).

124
Univariable analyses were performed on unweighted survey results using chi-squared, Fisher's exact or Wilcoxon's rank sum tests. False discovery rate 125 (FDR) correction was used to adjust p-values for multiple testing. COVID-19 vaccine hesitancy and preference for on-campus vaccinations (COVID-19 and 126 MMR) were dichotomized and used as dependent variables. Multivariable analysis was performed using logistic regression on weighted survey results, with 127 vaccine hesitancy and preference for on-campus vaccinations (COVID-19 and MMR) as dependent variables. Predictors included gender, ethnic group, age 128 group, course studied, year of study, experience of harassment, experience of COVID-19 related death, concern over side-effects from the 129 Oxford/AstraZeneca vaccine, concern over hospitalization from COVID-19, concern over spreading COVID-19, home area (local, national, international), 130 residence while studying (home, halls of residence, private accommodation, other) and a psychometric score on self-determination/fatalism. Home area was 131 . CC-BY 4.0 International license It is made available under a who has granted medRxiv a license to display the preprint in perpetuity.
is the author/funder, (which was not certified by peer review) The copyright holder for this preprint this version posted February 7, 2022. ; https://doi.org/10.1101/2022.02.07.22270394 doi: medRxiv preprint derived using information on post-codes and international status with students being classified as local if they came from either Leicester or the wider county 132 of Leicestershire, 'national' for students from the rest of the UK and 'international' for students ordinarily living overseas. Experiences of COVID-19 related 133 deaths were classified into a Yes or No category according to responses to question 21 (S1 Table) with the Yes responses including family member(s), friend(s) or someone else. Survey data were weighted using the raking method in the survey package 4.0 for R based on national student distributions for ethnic group (White, Asian, Black and Other) and gender (S2, S3 and S4 Tables). Constraints in the national data made it necessary to remove students of 136 unknown gender (n = 6).

138
Statistical differences between the distributions of VAX scores for different groups were determined using a Games-Howell pairwise test with FDR 139 correction.

Response rate, sample characteristics and COVID-19 vaccination uptake 143
In June 2021, all University of Leicester (UoL) undergraduate students were invited to participate in a study of the uptake and attitudes to COVID-19 144 vaccines. Complete answers were provided by 7.6% (827/10,869) of participants. Respondents were young (94% 18-25 year olds), ethnically diverse (25% 145 Asian, 8% Black, 58% White, 9% Other) and included 10% (n=86) international students. Response rates were higher among females (11% above the level 146 for UK universities and 14% above the UoL level) and an ethnicity profile intermediate between UoL and UK university undergraduate populations (S3 and 147 S4 Tables). The distribution among year of study was however strongly representative of the UoL population (S5 Table). Two thirds (64%) of students 148 (527/827) reported having had a COVID-19 vaccine at the time of questionnaire completion [74% (390/527) had Pfizer/BioNTech, 23% (121/527)

149
AstraZeneca, 3% (16/527) another vaccine]. A further 194 students (23%) expressed a willingness to become vaccinated, giving a total of 85% who had been 150 or were willing to be vaccinated. Results for 66 students were excluded from further analysis of vaccine hesitancy due to uncertainty about their intention to 151 vaccinate (the selected response was 'I have not had a vaccination but have been told that I will be offered a vaccination in the near future'). Removing these 152 students from the denominator gave an overall willingness rate of 95% (721/761). There were 40 students (5%) who indicated that they had refused or would 153 refuse a COVID-19 vaccine.

154
. CC-BY 4.0 International license It is made available under a who has granted medRxiv a license to display the preprint in perpetuity.

Univariable analysis of vaccine hesitancy 156
The results of the univariable analysis for vaccine hesitancy are shown in 40 students (5%) who indicated that they had refused or would refuse a COVID-19 157 vaccine.

159
. Ethnicity, course studied, concerns around side-effects (particularly the AstraZeneca vaccine), concerns around spreading COVID-19 to others, place of 160 residence while studying and VAX score were all found to be significantly associated with hesitancy after correcting for multiple testing. There was a weak 161 trend for an association of age with vaccine hesitancy; this could not, however be explored further due to banded collection of age data and the narrow age 162 range of this cohort.

164
Analysis of the individual VAX scale questions (see S2 Figure) showed that only 29% of hesitant students disagreed that natural exposure to a disease was 165 safer than vaccination compared to 80% of vaccine-willing students. By contrast, 70% of hesitant students, but also 54% of willing students, had concerns 166 about the safety of vaccines (the statement was 'Although most vaccines appear to be safe, there may be problems that we have not yet discovered').

167
Approximately half (49%) the hesitant students and 82% of willing students agreed that >95% vaccine coverage was required to prevent the spread of 168 infectious diseases.

170
A surprising observation was that high proportions of both vaccine-willing (42%, 299/721) and vaccine-hesitant (48%, 19/40) students had experienced a 171 COVID-19 death among relatives or other acquaintances (S3 Figure). This outcome was, however, not associated with differences in hesitancy (Table 1). . CC-BY 4.0 International license It is made available under a who has granted medRxiv a license to display the preprint in perpetuity.
is the author/funder, (which was not certified by peer review) The copyright holder for this preprint this version posted February 7, 2022.

Multivariable analysis of vaccine hesitancy 177
The multivariable analysis identified associations between vaccine hesitancy and ethnicity, course of study, side-effects and place of term-time residence, as 178 found in the univariate analysis, and additionally with experiences of death among contacts (Table 2). For course studied, studying medicine and allied 179 professions (e.g. midwifery, nursing and physiotherapy) had a significantly lower likelihood of being vaccine hesitant (OR 0.1, 95% CI 0.02-0.5, adjusted P = 180 0.021) compared to humanities, law and social sciences. For ethnicity, hesitancy among black students had a high odds ratio (OR 7.01, 95% CI 1.81-27.3, 181 adjusted P value = 0.021) as compared to white students (Table 2). Students living in private accommodation (OR 0.13, 95% CI 0.04-0.38, adjusted P=0.004)

182
were less vaccine hesitant than students living at home.

184
Hesitancy was strongly associated with concerns over side-effects from the Astra-Zeneca (Table 2) and Pfizer/BioNTech vaccines (OR 2.1, 95%CI 1.5-3.0, 185 adjusted P<0.001; data not shown). Concerns about side-effects were, however, lower for the Pfizer/BioNTech vaccine (S4 Figure). Surprisingly, the 186 multivariate analysis detected a positive association between experiences of a COVID-19-related death in a family member, friend or other contact with 187 vaccine hesitancy (Table 2). This association remained even when only close contacts (friends; family) were considered (odds ratio 6.4, 95% CI 1.9-21.6, 188 adjusted P = 0.02; data not shown).

189
190 . CC-BY 4.0 International license It is made available under a who has granted medRxiv a license to display the preprint in perpetuity.

Analysis of VAX scores 194
Both Asian and Black ethnic groups had significantly lower VAX scores than White ethnicity, indicating a higher level of vaccine hesitancy in the former 195 groups ( Figure 1). Similarly, we observed that home students had significantly lower VAX scores than students living in private or other accommodation 196 ( Figure 1). The mean VAX score for students living in halls was higher but not significantly different to those living at home, indicating a trend for home 197 students to be more vaccine hesitant than those who lived in other locations during this academic year.
198 199 . CC-BY 4.0 International license It is made available under a who has granted medRxiv a license to display the preprint in perpetuity.

Knowledge of MMR and MenACWY vaccine status among students 207
Views on MMR and MenACWY vaccines are shown in Table 3. Very few students (2-4%) self-reported not having had the MMR or MenACWY vaccines, 208 but an additional 8% did not know if they had had their MMR vaccine and 23% did not know if they had had their MenACWY vaccine (Table 3).

209
International students were more likely not to know their vaccination status compared to local students (Table 3). Additionally, 15% of UK students did not 210 know that the MMR and MenACWY were available free of charge in the UK and 6% reported not knowing that COVID-19 vaccines were also available free CC-BY 4.0 International license It is made available under a who has granted medRxiv a license to display the preprint in perpetuity.
is the author/funder, (which was not certified by peer review) The copyright holder for this preprint this version posted February 7, 2022. ; https://doi.org/10.1101/2022.02.07.22270394 doi: medRxiv preprint of charge. Again, these proportions were significantly higher among international students. More than half (57% and 61%, respectively) of students favoured . CC-BY 4.0 International license It is made available under a who has granted medRxiv a license to display the preprint in perpetuity.
is the author/funder, (which was not certified by peer review) The copyright holder for this preprint this version posted February 7, 2022.

Univariable and multivariable analyses of attitudes to on-campus vaccinations 219
The univariable analysis of on-campus MMR/MenACWY vaccine programmes identified a significant association with MMR vaccine status indicating that 220 those who responded with either a 'Yes' or 'Don't Know' response for their vaccine status were in favour of these programmes (S7 Table). However, these 221 responses were not significant in the multivariate analysis after correction for multiple testing (S8 Table). The univariable analysis of on-campus COVID-19 222 vaccine provision found significant associations with COVID-19 vaccine hesitancy and term-time residence (S7 Table). In the multivariable analysis, vaccine 223 hesitancy was negatively correlated with on-campus provision (S8 Table). Multivariable logistic regression of term-time residence indicated that students 224 studying in halls (OR 3.5 95% CI 1.6-7.6, adjusted P = 0.021) or private accommodation (OR 2.6, 95% CI 1.3-4.99, adjusted P = 0.03) were in favour of this 225 provision.

228
University students are a critical group for illness and spread of infectious diseases and hence are an important target for vaccination programmes. Our survey 229 of University of Leicester students was unique in that we evaluated both attitudes to and mechanisms for uptake of the three major vaccines targeted to this 230 population group in the UK. Our study indicates that ethnicity, concerns over side-effects and place of residence are key determinants of COVID-19 vaccine 231 hesitancy. We also found high levels of uncertainty among students about their MenACWY and MMR vaccine status. As an approach to facilitating vaccine 232 uptake, students were asked about on-campus provision of vaccines and reported to favour this approach. Based on our findings we elaborate key 233 recommendations for improved vaccine delivery to this population sector.

235
Our observed high uptake (64%) of COVID-19 vaccines was surprising given that this age group only became eligible for COVID-19 vaccines during the 236 course of our data collection. These uptake levels were significantly higher than the wider young adult population at that time (P<0.0001 as compared to 237 . CC-BY 4.0 International license It is made available under a who has granted medRxiv a license to display the preprint in perpetuity. is the author/funder, (which was not certified by peer review) The copyright holder for this preprint this version posted February 7, 2022. ; https://doi.org/10.1101/2022.02.07.22270394 doi: medRxiv preprint Leicester 18-24 year olds; S1 Figure), suggesting that these students were more proactive about accessing COVID-19 vaccines than their peers. This high 238 uptake may also have been partially attributable to high participation in the study by pro-vaccine students and/or to surge vaccinations in the Leicester 239 COVID-19 hotspot, including a pop-up vaccination centre in De Montfort Hall, adjacent to the UoL campus, that took place just prior to initiation of the 240 survey. We also observed a high willingness for uptake of COVID-19 vaccines (95%) among the University of Leicester survey respondents. These findings 241 may reflect the effectiveness of both the delivery of vaccines to students and the information campaign on the benefits of these vaccines. The subsequent 242 evidence of uptake rates of >90% in an ONS study of UK university students [32] indicates that, despite the caveats (see below), our study was a reasonable 243 predictor of student attitudes to vaccines. Intriguingly, we also found that 93% (37/40) of the COVID-19 vaccine hesitant individuals reported having had at 244 least one of the MMR and MenACWY vaccines with 50% (20/40) having had both; this would suggest that these individuals are specifically concerned about 245 the COVID-19 vaccines or that vaccine hesitancy has developed during their transition to adulthood.

Determinants of vaccine hesitancy among university students 248
A key concern for vaccination programmes has been to identify groups of individuals with lower levels of vaccine uptake. Our univariable analysis indicated 249 that ethnicity was strongly linked to vaccine hesitancy, with the multivariable analysis further linking this hesitancy to the black ethnic group within the wider 250 university student population. As these findings were based on small numbers of vaccine-hesitant individuals, we examined the VAX scores for all 251 individuals of each ethnic group and found that both the Asian and Black ethnic groups had significantly lower average scores than the White ethnic group 252 indicating a general trend towards hesitancy among the minority ethnic groups (Figure 1). Other studies have also found evidence of vaccine hesitancy 253 associated with ethnicity [33-36] and specifically with students of black ethnicity [36]. Hesitancy in these groups has been linked to discrimination, mistrust 254 of healthcare organisations, misinformation, lower perceived vaccine efficacy/safety. A substantial proportion of vaccine hesitant individuals (37.5%; 15/40) 255 in our study agreed with a statement that COVID-19 vaccines had not been thoroughly tested in different ethnic groups (S9 Table).

257
Our study also identified some evidence that students registered in medical courses were significantly less likely to be vaccine hesitant than students 258 registered for other courses. Other studies have also observed higher uptake of vaccines among medical students [8]. Medical students may be more 259 knowledgeable about vaccines and interested in preventing illnesses, as inculcated in the duty of care for healthcare professionals, and hence are likely to be 260 more in favour of vaccination programmes than other groups.

261
. CC-BY 4.0 International license It is made available under a who has granted medRxiv a license to display the preprint in perpetuity.
is the author/funder, (which was not certified by peer review) The copyright holder for this preprint this version posted February 7, 2022. ; https://doi.org/10.1101/2022.02.07.22270394 doi: medRxiv preprint A novel finding was of an association between vaccine hesitancy and students who lived at home. This association may be part of a general trend as the 263 average VAX scores for all students living at home was significantly lower than those living in private accommodation or other, mainly mixed, 264 accommodation types (Figure 1). Despite lockdowns and extensive online teaching, the home student group was small as only 7% of national and 19% of 265 international students reported remaining at home for the 2019/2020 academic year. Higher vaccine hesitancy in students living at home may be explained by 266 a consideration of the concerns around spreading of COVID-19. While their risk of hospitalization (and hence personal safety) was not a significant predictor 267 of vaccine hesitancy in our multivariate regression (P = 0.6), we found that there were lower levels of vaccine hesitancy among students concerned about 268 spreading COVID-19 (OR 0.5, 95%CI 0.3-0.81, P=0.024). Students living at home were less concerned about spreading COVID-19 than those living away 269 from home (50% and 62%, respectively). We postulate that this attitude may be driven by a lower perceived risk of the potential for spreading the disease due 270 to reduced day-to-day social interactions compared to students living in halls or private accommodation.

272
A concern raised by our study is of an association between potential side effects of COVID-19 vaccines and vaccine hesitancy. Our study was performed a 273 few months after concerns about side-effects of the AstraZeneca vaccine became apparent, leading to recommendations for young people to receive the 274 Pfizer/BioNTech vaccine instead. Concerns over side-effects were associated with vaccine hesitancy in both our univariate and multivariate models,

275
indicating that this important factor could reduce vaccine uptake. Unfortunately, we did not query which side-effects in particular the students were concerned 276 about and hence it is possible that common but mild side-effects may be enough to dissuade ambivalent students from receiving vaccinations. However, we 277 also note that many students still obtained COVID-19 vaccines despite these concerns suggesting that other factors may over-ride the perceived risks of side-278 effects. It will be interesting to observe whether this effect has dissipated over time with more information on the low occurrence of severe reactions and 279 improved strategies for preventing occurrences.  Organisation for preventing measles and mumps outbreaks [19,21,38]. It is likely that many of the students who are uncertain about their vaccine status have 291 not had these vaccines. A particularly concerning feature of our study was that a higher proportion of international students had not had the MMR and 292 MenACWY or did not know their status for these vaccines; these students will be at a higher potential risk of contracting and/or spreading the diseases 293 targeted by these vaccines. In England, Public Health England recommends that anyone who is uncertain about their vaccine status or has missed a vaccine 294 dose should be offered these vaccines [39]. Most students will be unaware of this recommendation, indeed 20% of UoL students did not know that these 295 vaccines are free (Error! Reference source not found.). Addressing these issues is important for maintaining both direct protection and herd protection 296 across the population for measles, mumps and four of the five major serogroups responsible for invasive meningococcal disease in the UK.

On campus provision of vaccines 299
A high proportion of students were in favour of on campus provision of MenACWY, MMR and COVID-19 vaccines. These proportions were even higher 300 among international students who were based in the UK, which may reflect difficulties in understanding how to access the UK medical system. The 301 statistically significant evidence of support for provision of on-campus COVID-19 vaccinations by students who are not studying at home is an indication that 302 students value easy access to vaccinations. The success of COVID-19 vaccine pop-ups and the University of Nottingham's programme for delivery of the 303 MenACWY vaccine in the pre-pandemic era indicate that students value and will access these services [9].

Recommendations 306
Harnessing new approaches developed during the UK COVID-19 vaccine roll-out is a potential positive legacy of the pandemic to build-back-better for future 307 generations. Empowering digitally-aware young people to take responsibility for their own health and to engage in community health policies is an 308 . CC-BY 4.0 International license It is made available under a who has granted medRxiv a license to display the preprint in perpetuity.
is the author/funder, (which was not certified by peer review) The copyright holder for this preprint this version posted February 7, 2022. to vaccines by international students should be a gold-standard for the university sector as the financial gains accrued from these students must be allied to 313 high welfare provision. On-campus vaccination programmes should be widely adopted for all relevant vaccines as the preferred strategy of getting vaccinated 314 prior to attending university has never been optimal. 315 316

Strengths and limitations 317
A strength of this study is that it is the first to simultaneously evaluate uptake, knowledge and attitudes to COVID-19, MMR and MenACWY vaccines among 318 university students. A further strength is the high number and ethnic diversity of the participant population. The use of multivariable regression was a strength 319 that allowed for adjustment for confounders and for identification of significant associations between variables and vaccination parameters with the potential 320 to inform vaccination policies.

322
Findings from this study may, however, be affected by the inherent limitations of cross-sectional studies. The response rate of 8% indicates that there may 323 have been a response bias due to demographics. The strengths and limitations arising from a range of demographics were considered above but it is possible 324 that biases from other unaccounted demographics (e.g. socioeconomic status) may confound generalisability of our data to the wider UK student population.

325
A significant potential limitation of our study, and inherent in many studies of vaccination, is enhanced participation by individuals with pro-vaccine attitudes  The findings from this study indicate that there may be differences in uptake and access to the COVID-19, MenACWY and MMR vaccines among university for hesitancy may be required in order to delivery more effective, 'tailored' vaccine information and to develop methods for enhancing trust and acceptance of 334 vaccines in these groups. High levels of uncertainty about personal vaccine status and availability of the MMR and MenACWY vaccines were observed and 335 are likely to impact on vaccine uptake. On campus vaccination delivery was found to be widely favoured particularly by on-campus and international 336 students. These knowledge gaps and delivery approaches should be considered in future student-focussed vaccination campaigns and explored through 337 additional research. Our findings indicate that adopting 'best-practices' of easy access and digital vaccine information within the university-sector may 338 breakdown barriers and future-proof uptake of all required vaccines among students.    CC-BY 4.0 International license It is made available under a who has granted medRxiv a license to display the preprint in perpetuity.
is the author/funder, (which was not certified by peer review) The copyright holder for this preprint this version posted February 7, 2022. ; https://doi.org/10.1101/2022.02.07.22270394 doi: medRxiv preprint as this cohort has the highest uptake and because a high proportion of University of Leicester students would have been able to take advantage of surge 472 vaccinations in Leicester.