Education increases COVID-19 vaccine uptake among people in Canadian federal prisons in a prospective randomized controlled trial: The EDUCATE study

Education is key to behavioural adoption and acceptability of health interventions. We evaluated the impact of an educational intervention administered 1:1 to individuals incarcerated in four Canadian federal prisons on COVID-19 vaccine uptake. Eligible individuals (those who had refused all COVID-19 vaccines) were randomized 2:1 to receive the educational intervention or not (control group); those who received the intervention completed questionnaires assessing COVID-19 vaccine-related knowledge, attitudes, and beliefs pre- and post-educational intervention. The primary and secondary outcome measures were COVID-19 vaccine uptake and vaccine confidence, respectively. Between May 3 and September 9, 2022, 202 participants were randomized to receive the intervention, of whom 127 (63 %) agreed to participate. Participants who were randomized to the intervention had higher COVID-19 vaccine uptake vs. the control group (5 % vs 1 %, p = 0.046). COVID-19 vaccine-related knowledge, attitudes, and beliefs improved post-intervention. Education increases COVID-19 vaccine uptake and confidence among people in Canadian federal correctional facilities.


Introduction
Canadian correctional facilities have been the epicentre of several large severe acute respiratory syndrome coronavirus (SARS-CoV-2) outbreaks since the beginning of the pandemic [1][2][3]. As incarcerated individuals disproportionately experience social determinants of health associated with increased risk of COVID-19 due to a higher prevalence of chronic diseases such as hypertension and diabetes and, additionally, live in congregate settings with a high risk of transmission, efforts to maximize COVID-19 vaccine uptake, one of the most effective tools for reducing COVID-19 morbidity and mortality, are paramount. However, vaccine hesitancy is not uncommon among people in prison [4][5][6][7][8][9][10][11][12]. Studies investigating reasons for vaccine refusal among people incarcerated in Cana-dian federal prisons found that a lack of information and misinformation were important barriers to COVID-19 vaccine acceptability [9,12], underscoring the potential for educational interventions to improve COVID-19 vaccine uptake.
Education is key to behavioural adoption and acceptability of health interventions. While the evaluation of educational interventions on COVID-19 vaccine uptake among incarcerated individuals has never been studied, various educational interventions have led to increased influenza [13,14], human papillomavirus [15], and varicella-zoster virus vaccine uptake [16], and improvements in overall vaccine knowledge among adults in the general population [17]. That said, studies have confirmed that prison vaccination programs have the potential to increase vaccine uptake if combined with education [10,11]. We thus aimed to evaluate the impact of an educational intervention on COVID-19 vaccine uptake among individuals incarcerated in Canadian federal prisons who had previously refused COVID-19 vaccination.

Study design and setting
We conducted a prospective, randomized controlled study in four Canadian federal prisons. Prisons with the highest proportion of individuals who were never vaccinated against SARS-CoV-2 (25-31 %) prior to study inception were chosen as study sites from among the 43 federal prisons in Canada overseen by Correctional Service Canada (CSC) where incarcerated adults serve sentences of two years or more [18]. This included four all-male institutions: Collins Bay Institution (CBI), Millhaven Institution (MI), and Joyceville Institution (JI) in Ontario and Kent Institution (KI) in British Columbia. Study site characteristics are shown in Supplementary  Table 1. Study participation was voluntary and, as per CSC regulations [19], participants were not eligible to receive an honorarium or any incentive for their participation.

Participants
As of April 20, 2021, incarcerated individuals across all 43 federal sites had access to an mRNA COVID-19 vaccine. We included individuals aged 18 years or older incarcerated at one of the four study sites who refused to be vaccinated against SARS-CoV-2 and who were able to consent to study participation in either English or French. Individuals who posed a security risk to the research team as determined by facility staff and anyone whose remaining sentence was less than one month were excluded from the study.

Educational intervention
Building on our qualitative work identifying reasons for vaccine refusal among people incarcerated in Canadian federal prison [9,12], we consulted with an educational consultant to design an educational intervention tailored to addressing four primary reasons: 1. Risk perception; 2. Lack of knowledge (i.e. mRNA technology, efficacy, booster immunity); 3. Lack of trust; and 4. Perceived lack of benefits (personal and societal). The intervention was approved by CSC and the Indigenous Initiatives Directorate to ensure cultural sensitivity. It was designed to be administered 1:1 with visual aids and to be 15 minutes in duration (Supplementary Appendix A). To address issues of mistrust with CSC, a research assistant (unaffiliated with CSC) was trained to deliver the intervention. A facilitator guide was developed to answer potential questions and was validated by an Infectious Diseases specialist (NK) (Supplementary Appendix B).

Data collection
The research assistant was provided with a de-identified list of potential participants at each study site. This was used to randomize individuals 2:1 into two groups, those who would be randomized to receive the intervention (Groups A and B; Group A being those who agreed to participate and Group B being those who declined) versus those who would not be contacted and who would serve as the control group (Group C). Eligible participants were approached by the research assistant near their units/ranges and were provided with a detailed description of the study. Participants who agreed to participate were escorted to a nearby interview room where written consent was documented. Participants were given two self-administered paper questionnaires to complete and those who required assistance with reading and writing could request support from the research assistant. The first questionnaire (the pre-intervention questionnaire (Supplementary Appendix C); $20 minutes) assessed socio-demographic and clini-cal characteristics, as well as knowledge, attitudes, and beliefs towards COVID-19 vaccines. We developed the survey questions by adapting questions from the World Health Organization Strategic Advisory Group of Experts Working Group on Vaccine Hesitancy [20], Leger's North American Tracker -Concerns about COVID-19 [21], Statistics Canada's Canadian Perspectives Survey Series 3 [8], the National Advisory Committee on Immunization Acceptability Matrix [22], an online survey of vaccine acceptability among the Canadian general public [23], and the 5C vaccine confidence scale [24]. We pilot-tested the questionnaire with 15 people in prison to ensure clarity before study inception. Immediately following the pre-intervention questionnaire, the research assistant delivered the educational intervention to the participant 1:1. Thereafter, participants were given 10 minutes to ask questions. When questions were beyond the scope of the facilitator guide, participants were advised to consult their respective health care teams. A shorter post-intervention questionnaire ($5 minutes; Supplementary Appendix D) was given immediately after the educational intervention. This questionnaire focused on knowledge, attitudes, and beliefs towards COVID-19 vaccines and participant willingness to accept vaccination. Following the postintervention questionnaire, the research assistant offered an mRNA COVID-19 vaccine (Moderna Spikevax or Pfizer-BioNTech Comirnaty) to each participant and documented acceptance or refusal. Participants who accepted COVID-19 vaccination immediately following the intervention (''immediate uptake") were then vaccinated by a CSC nurse as soon as possible. Participants who refused vaccination were approached 30 days later by the research assistant and re-offered vaccination (''delayed uptake" vs. ''delayed refusal"). Individuals who declined participation (Group B) and incarcerated people in Group C who were not invited to participate were not exposed to the intervention. All individuals who declined vaccination in Group A and who remained unvaccinated in Groups B and C were offered a COVID-19 vaccine by CSC Infectious Disease nurses on a weekly basis. Anonymized baseline characteristics (age, ethnicity, results of SARS-CoV-2 testing during incarceration, and receipt of 2020-2021 influenza vaccine) and COVID-19 vaccination status were provided to the research team by CSC for those in Groups B and C at the end of study recruitment. This data was self-reported and collected in the pre-intervention questionnaire for Group A.
The recruitment period spanned five months (May 3 -September 9, 2022). Assuming a 2 % vaccine uptake rate among Group C without any intervention, a total sample size of 118 consented participants (Group A) and 59 participants in Group C was necessary to detect a 10 % increase in vaccination rate after the intervention with a level of significance of 0.05 and 80 % power versus [25]. Given anticipated recruitment challenges and the likely degree of refusal to participate (estimated at 40 %), we increased the total sample size for Groups A and B to 200 participants total.

Outcomes
The primary outcome measure was vaccine uptake, measured as the proportion of participants who accepted a COVID-19 vaccine following the educational intervention (''immediate" or ''delayed" uptake). The secondary outcome measure was COVID-19 vaccine confidence, measured by changes to knowledge, attitudes, and beliefs post-educational intervention.

Data analysis
Summary statistics were calculated to describe the study sample; continuous variables were reported in median and interquartile range (IQR), and categorical variables were reported in counts and proportions. The Mann-Whitney U and Pearson's Chi-square tests were used to compare continuous and categorical variables, respectively. A p-value < 0.05 was considered significant. For the primary outcome, an intention-to-treat (ITT) analysis with a twosample Z-test of proportions was used to compare vaccine uptake between those randomized to receive the intervention (Groups A and B) vs. the control group (Group C), with a null hypothesis of no difference in population proportions against a one-sided alternative that vaccine uptake is greater among those who were randomized to receive the intervention compared to controls. A perprotocol (PP) sub-analysis with a two-sample Z-test of proportions was used to compare vaccine uptake between Groups A and C. For the secondary outcome, a paired sample t-test was used to compare the mean change in pre-and post-intervention vaccine confidence among participants who received the intervention. Questions measured using a Likert scale were scored out of 5: 1 = ''strongly disagree," 2 = ''mostly disagree," 3 = ''uncertain," 4 = ''mostly agree," and 5 = ''strongly agree." All analyses were performed using IBM SPSS (version 29.0).

Results
Overall, 321 (of $ 1,120) individuals across the four study sites refused all prior COVID-19 vaccines. Of these, 303 met all eligibility criteria and were randomized 2:1 to the intervention or control groups, with 202 invited to participate in the study (Fig. 1). A total of 61 (32 %) declined study participation (Group B) and 14 (7 %) were excluded, leaving 127 participants who agreed to receive the intervention (Group A).
Overall, the median age of participants was 31 years (Table 1). Over one-third (35 %) of participants self-identified as Black, 27 % as white, 11 % as Indigenous, and 24 % as another visible minority.
The majority (75 %) had received a prior positive SARS-CoV-2 test and a minority (8 %) had received the 2020-2021 influenza vaccine during incarceration. Groups A and B were similar to Group C in terms of age, race/ethnicity, proportion who received the 2020-2021 influenza vaccine, and prison facility.
All participants in Group A completed both questionnaires. The distribution of knowledge, attitudes, and beliefs towards COVID-19 vaccination and the change in means between pre-and postintervention questionnaire responses are presented in Fig. 2 and Table 2, respectively. Following the educational intervention, there were significant improvements in knowledge and more favourable attitudes and beliefs towards COVID-19 vaccines. There were no improvements in participants' perception of the COVID-19 pandemic as a real threat and the risk of COVID-19 in prisons.

Discussion
This prospective, randomized controlled study evaluated the impact of an educational intervention -tailored to the reasons for vaccine refusal identified by people in federal prison -on COVID-19 vaccine uptake and confidence among individuals incarcerated in four Canadian federal prisons that had the highest proportion of COVID-19 vaccine refusal at a time when vaccines were freely and readily available in prisons. We found that those who received the educational intervention experienced a small but statistically significant increase in vaccine uptake, improvements in COVID-19 vaccine-related knowledge, and more favourable attitudes and beliefs towards COVID-19 vaccines. Our results are in line with other studies that have demonstrated modest increases in vaccine uptake [26] and vaccine acceptability [27] following educational interventions, underscoring that while education is inarguably important, a multimodal approach that incorporates education alongside or in addition to other interventions will be essential for increasing vaccination among vaccine-hesitant incarcerated individuals.
Interestingly, the educational intervention had an immediate effect on vaccine uptake but failed to influence COVID-19 vaccine uptake 30 days post-intervention. Sun et al. [28] showed that intermittent long-term education is needed to maintain motivation and encourage behaviour adoption for vaccination, suggesting that additional educational sessions (or other interventions) may have been necessary to address reasons for ongoing hesitancy over time. Furthermore, we previously found that incarcerated individuals' attitudes towards vaccination are often influenced by their support systems [12], implying that educational approaches such as ones that expand the provision of information to outside support networks may be one novel approach to improving vaccine uptake in prison settings [29].
Addressing distrust, the primary reason we identified for immediate vaccine refusal after the educational intervention, will be an essential component of improving COVID-19 vaccine uptake among people in prison. Cognizant of the challenges in building trusting relationships between correctional staff and people in prison, the intervention was intentionally delivered by an individual unaffiliated with the prison system to provide participants with an independent source of information. Nonetheless, the majority (93 %) of participants refused COVID-19 vaccination postintervention, with one-third declining vaccination due to a lack of trust. Providing information, educational content, and messaging about vaccines from a trusted source has been identified as a key component of successful interventions. For example, to further foster provider-patient trust, studies have shown that providers who adopt dialogue-based approaches or campaigns that involve leaders and peers result in higher vaccine uptake [30,31]. In federal corrections, trusted individuals could include Indigenous Elders or prison committee members. Consideration should be given to the implementation of ''inmate volunteer" models of care where incarcerated individuals take more active roles in prison health care delivery. These models have had positive effects on trustbuilding and the empowerment of people in prison [32], and may improve vaccine uptake going forward.
Our study has limitations. First, our study was conducted in four of 43 Canadian federal correctional facilities where all incarcerated people were male, thereby decreasing the generalizability of our results to other correctional institutions or to those who house women. That said, gender has not been found to predict vaccine uptake in prior studies utilizing educational interventions [13,33,34]. Second, we made the assumption that the reasons for vaccine refusal identified during prior qualitative research conducted among this population [9,12] were generalizable to the entire federal incarcerated population. Ideally, we would have conducted qualitative interviews in each of the selected sites prior to recruitment and tailored the education intervention accordingly. Third, baseline characteristics of study participants presented in Table 1 were self-reported for Group A only. Despite these limitations, this is the first study to evaluate the effect of an educational intervention on COVID-19 vaccine uptake among incarcerated people who had refused COVID-19 vaccination and adds to the dearth of literature regarding the role of education in improving vaccine uptake and confidence.

Conclusions
Education increases COVID-19 vaccine uptake and confidence among people in Canadian federal correctional facilities. Future efforts should seek to involve people in prison in both the codesign and delivery of prison-based educational tools to address multiple barriers to vaccination and ultimately improve COVID-19 vaccine uptake and confidence.

Statement of ethics
The study was approved by the McGill University Health Centre Research Ethics Board (REB #2022 -7868).

Funding
This work was supported by a CIHR Operating Grant: COVID-19 Vaccine Confidence awarded to NK. NEB is supported by a Canada Research Chair Tier 2 in Infectious Disease Prevention. NK is supported by a career award from the Fonds de Recherche Québec -Santé (FRQ-S Junior 1).

Data availability
Data will be made available on request.

Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: [TZ, HP, CD, OV, NEB, LW, PR, SD, MM, ST, MB, MM and RK have no conflicts of interest to declare. NK reports research funding from Gilead Sciences, advisory fees from Gilead Sciences, ViiV Healthcare, Merck and Abbvie, and speaker fees from Gilead Sciences and Merck.]. Questionnaire item: Average score out of 5 (1 = ''strongly disagree", 2 = ''mostly disagree", 3 = ''uncertain", 4 = ''mostly agree", 5 = ''strongly agree"