Improving uptake of COVID-19 testing and vaccination in a homeless population: mixed-methods evaluation of community health worker-led education in a shelter Authors: Rebecca A Ziolkowski, Lara Balian, Shreya Sridhar, Natalia M Rodriguez Published: 2024 ABSTRACT AbstractObjectivesAccess to healthcare and accurate health information is a persistent issue for people experiencing homelessness (PEH), one heightened by the COVID-19 pandemic and exacerbated by issues of health literacy and mistrust. Strategies involving community health workers (CHWs) have the potential to improve education and service delivery for PEH. The objective of this study is to assess PEH knowledge, attitudes and barriers to COVID-19 testing and vaccination, and the impact of CHW-led education within a homeless shelter in Indiana.Design/settingBetween November 2021 and April 2022, 15 sessions were held at a homeless shelter that combined focus group discussions with COVID-19-related education. Pre-/post-tests measured knowledge gains and attitude shifts among PEH. Guided by the socioecological model, thematic analysis of focus group transcripts identified key motivators and barriers to COVID-19 vaccination and testing.Participants78 PEH (73% male, 74% white, mean age=48).Outcome measuresCOVID-19-related knowledge gains and willingness to receive testing and vaccination among PEH.ResultsCompletion of an educational session was associated with a significant increase in knowledge scores (mean pre-test score=10.76/15, SD=2.16; mean post-test score=13.58/15, SD=1.72; p<0.001), with most participants (91%) improving by up to 10 points. The proportion of unvaccinated participants (42%) willing to be vaccinated increased from 21% to 39%. Participants willing to get tested for COVID-19 after a known exposure increased from 83% to 92%. Qualitative findings from the sessions provided insights into key knowledge gaps and revealed participant preferences towards health education format and topics.ConclusionCHW-led COVID-19 education increased knowledge gains and positively influenced attitudes towards vaccination and testing among PEH. Findings highlight the importance of tailored, community-driven interventions for pandemic response and emphasise the crucial role of education in building trust and promoting health literacy to improve health outcomes for vulnerable populations. STRENGTHS AND LIMITATIONS OF THIS STUDY Combines qualitative methodology and descriptive quantitative measures to assess knowledge, motivators, demotivators and barriers to COVID-19 testing and vaccination. Combines qualitative methodology and descriptive quantitative measures to assess knowledge, motivators, demotivators and barriers to COVID-19 testing and vaccination. Conducted at one point in time, limiting assessment of sustainability and long-term impact. Conducted at one point in time, limiting assessment of sustainability and long-term impact. Study design was informed by ongoing community-based participatory research efforts with a homelessness services organisation. Study design was informed by ongoing community-based participatory research efforts with a homelessness services organisation. Participants were clients of the homelessness services agency and did not include unserved/unsheltered people experiencing homelessness. Participants were clients of the homelessness services agency and did not include unserved/unsheltered people experiencing homelessness. Introduction Within the context of health determinants, access to quality education and healthcare are recognised by the Healthy People 2030 initiative to have significant impacts on health outcomes, a recognition that demands deeper consideration when addressing the unique challenges faced by people experiencing homelessness (PEH).1 Nationally, the homeless population experiences a mortality risk 3.5 times higher than housed individuals,2 bears a disproportionate burden of disability and disease and requires a higher level of care and access to health services, leading to an average life expectancy of 64–69 years, a sharp contrast to 76–81 years for the general population.25 These health disparities worsened during the COVID-19 pandemic, particularly as congregate shelter settings became increasingly high-risk environments. Unable to adhere to ‘stay at home’ guidelines, PEH were often pushed back to shelters by public officials or the clearing of encampments under claims of public safety.6 This movement meant putting PEH in environments where social distancing, quarantining and other mitigative behaviours were often impossible. As expected, shelter outbreaks were reported across the USA, where up to 66% of shelter users were positive for the virus.7 The heightened risk of COVID-19 infection intensified adverse outcomes for PEH as it compounded with pre-existing barriers to accessing services and their experiences of mistrust, stigma and discrimination within healthcare settings.810 As a result of the challenging landscape of COVID-19, PEH saw an elevated risk of hospitalisation and higher mortality rates, highlighting the urgent need for targeted interventions to address their unique vulnerabilities in the wake of communicable diseases like COVID-19.10 11 Subsequently, early publications called for regular testing and prioritisation of vaccination to protect PEH from adverse health outcomes and to limit transmission.7 12 Yet multiple countries have encountered similar challenges in vaccinating their homeless populations, with studies observing lower vaccination rates in PEH compared with the general population.13 14 To address this, strategies used internationally have included enhancing accessibility, collaborating with community partners, fostering trust and engaging targeted education programmes.15 These interventions had varying levels of success but often met barriers.11 16 The pandemic’s amplification of deep inequities in information and resource access furthered the isolation of the PEH demographic where low health literacy levels have been linked to poor health knowledge and access.7 17 In particular, health literacy was identified as a contributing barrier to COVID-19 testing among PEH.16 For vaccinations, a significant association between low intention to be vaccinated against COVID-19 and low health literacy was drawn, as evidenced by vaccine hesitancy rates among PEH that ranged from 35% to 48% throughout North America and Europe.17 The mitigatable nature of health literacy as a barrier to testing and vaccination acceptance highlights the importance of interventions centred around building trust and increasing access to accurate health information for informed decision-making. One such solution that emerged throughout the pandemic was the call for community health workers (CHWs) as crucial intermediaries between people who are underserved and healthcare systems.1821As an already present and trusted workforce in close proximity to the population they serve, their deep knowledge and relationships within their communities can be leveraged to bridge gaps in healthcare access and information. Their roles often include outreach, education and navigation services, putting them in a unique position to reach and educate through culturally sensitive and accessible care.18 Though minimal, studies explored how CHWs were able to connect to communities hit hardest during the pandemic, how CHWs positively impact under-represented and marginalised populations, and in some instances reach PEH specifically.2225 This study builds on prior community-based participatory research (CBPR) conducted in collaboration with a homelessness services organisation in Indiana, to examine the impacts of COVID-19 on PEH.8 9 Early findings from our initial work revealed how knowledge gaps, misinformation and unreliable sources affected PEH perspectives on COVID-19 mitigative behaviours, highlighting the need for improved information dissemination and education. This work also demonstrated the potential outreach capacity of CHWs9 and supported the deployment of CHWs in homeless shelters as health educators and navigators to health and social services. In March 2021, the homelessness services organisation trained and deployed CHWs to implement and conduct free, voluntary COVID-19 rapid antigen testing, making it available to all guests and tenants of their shelters and supportive housing units. As part of the National Institutes of Health Rapid Acceleration of Diagnostics-Underserved Populations (RADx-UP) initiative, this study aimed to identify motivators, demotivators and multilevel barriers to shelter-based COVID rapid testing and vaccination and to assess the impact of CHW-led education on the COVID-related knowledge, attitudes and beliefs of PEH. Methods Patient and public involvement This study was informed by ongoing CBPR co-led in partnership with a homelessness services agency in Indiana.8 9 26 27 Research questions were of mutual interest to the community partner organisation and the academic team and were informed by the homelessness service providers’ experience working with the community during the COVID-19 pandemic. The community partner organisation was not involved in the recruitment or conduct of the study but did allow the research team to post flyers and make announcements about the study onsite at the organisation. All study findings were shared with the community partner organisation in real time for their subsequent action and dissemination back to the community. Patients were not involved in this study as it was not clinical research. Study aims and rationale What are key motivators, demotivators and barriers to COVID rapid testing and vaccination among PEH? What are key motivators, demotivators and barriers to COVID rapid testing and vaccination among PEH? What is the effect of CHW-led education on the COVID-related knowledge, attitudes and beliefs of PEH? What is the effect of CHW-led education on the COVID-related knowledge, attitudes and beliefs of PEH? We designed a focus group formatted educational intervention to first learn participants’ knowledge, attitudes and beliefs about COVID-19 testing and vaccination while also providing the opportunity to share reliable information and clarify/correct misconceptions. The decision to integrate a focus group and educational session for the intervention was based on crucial insights gained from PEH in ongoing CBPR studies.8 9 27 The structure of an informal and conversational approach to an educational intervention is supported in literature analysing active learning or group discussion styles.28 29 This study used a mixed-methods approach by combining quantitative pre-/post-intervention assessments of participant knowledge and attitudes with additional qualitative insights on motivators and barriers captured during the focus group sessions. A certified CHW (RAZ) trained in research and ethics contributed to all aspects of this study and led key aspects including intervention design, recruitment, data collection, analysis and findings dissemination. Prior to the study, the CHW had experience working with this homeless community in the context of pandemic response and had carried out health programming at the shelter including conducting health assessments, navigation to healthcare appointments, insurance and pandemic relief resources. While this work allowed the CHW to build rapport with the community, they did not share lived experiences of homelessness. Recruitment, consent and compensation In recognition of the marginalisation faced by PEH, and their heightened risk of substance use, mental and physical health challenges, there were essential ethical considerations involved in this study design. Our community–academic partnership provided crucial insights for understanding population-specific risks and benefits and incorporating these into recruitment, eligibility, consenting, compensation and all other aspects of study design. Recruitment was done passively, through flyers and general announcements made onsite at a homelessness engagement centre during meal times; no direct recruitment occurred. Participation eligibility required interested subjects to be 18 years or older, current shelter guests or supportive housing tenants of the homelessness services organisation in Lafayette, Indiana and able to give informed consent. The following four questions were used to assess consent capacity, to ensure participants could comprehend the purpose and potential outcomes of the study, as well as their participation: Can you tell me what you understand to be the purpose of this study? Can you tell me what you understand to be the purpose of this study? Do you have to participate in this study? Do you have to participate in this study? Can you tell me what will happen if you agree to take part in this study? Can you tell me what will happen if you agree to take part in this study? What will happen if you decide not to be in the study? What will happen if you decide not to be in the study? If the participant correctly responded to all four questions, they were eligible to participate in the study. If adequate consent capacity was not found, participants were informed they were not able to participate at this time. Interested participants were able to sign up for a focus group at a later date and would be assessed again for consent capacity. During the consent process, researchers clearly explained that their relationship with the organisation or services received would in no way be affected by their participation in the study. Furthermore, the role of the CHW as a member of the research team was clearly defined to mitigate the potential for role confusion between researcher and service provider and maintain research and ethical integrity.30 All focus groups provided dinner and participants received a US$60 gift card to a local grocery store as compensation for their time. Compensation amounts were agreed on by both community and academic partners to ensure fairness. Study procedures and materials Focus group educational sessions were conducted between November 2021 and April 2022. Each focus group session took up to 2 hours and included the informed consent process, demographic intake questionnaire, pretest, focus group, educational intervention and post-test. Each group had 4–6 participants and adhered to the facility’s COVID-19 safety policies and guidelines at the time. The consent forms, intake and pre-/post-test questionnaires were read aloud to participants privately, and responses were recorded in REDCap’s secure web platform. No one else was present besides the participants and researchers for the entirety of this process. Educational materials The educational materials were adapted from the Centers for Disease Control and Prevention (CDC) guidance made available during the pandemic. All questions and materials used were adjusted to appropriate literacy levels and context, with expressions like ‘stay at home’ altered due to the inappropriateness in a homeless shelter environment. Content for materials was tailored and further informed by the CHW’s experience working with the population as well as findings from the previous CBPR work.9 The moderator guide and education PowerPoint slides covered any information participants were asked during the pre-tests/post-tests. The educational content presented was reviewed by a nurse who worked part time at the shelter and was updated regularly throughout the intervention to include any updates on vaccinations, COVID-19 strains, etc, as needed and per participant feedback (such as to further clarify key points). Intake and pre/post-test questionnaires The intake questionnaire included demographics and questions to understand participants’ experiences with COVID-19 testing and vaccination. The pre-tests and post-tests, composed of identical questions, consisted of 15 true/false knowledge questions and 4–11 multiple choice attitude and belief questions, dependent on skip logic (online supplemental appendix A). Post-tests were administered immediately after intervention completion. Educational focus group interventions Once participants completed the intake process, they were brought into a conference room at the shelter, provided dinner, asked by the CHW to select a pseudonym and given a name tag. The pseudonym served as an additional privacy measure but also became a welcomed icebreaker as the process of choosing a fake name often leading to light conversations with other participants and the moderator. Focus group rules and guidelines were then explained by the CHW, including courtesy measures like asking participants to refrain from sharing any information disclosed by other participants for confidentiality, and that the intent of the group session was not to reach consensus but to respect differing views and opinions. Delivered via PowerPoint presentation (online supplemental appendix B) with a moderator’s guide, the educational format was designed to promote conversation to understand attitudes and beliefs about COVID-19, and also be informative by asking mostly yes or no questions to engage participants actively before providing an answer. Participants were first prompted with a question and the group engaged in initial discussions allowing researchers to gain insights on topics like testing and vaccine awareness, motivators and demotivators. Subsequently, the CHW presented the correct answer prompting additional discourse for clarification and participant questions. The beginning of the session was signified by the start of the audio recording device and the CHW initiating engagement and on-topic discussion by asking for a ‘show of hands’ of whether participants had experienced COVID-19 testing. In addition to knowledge based yes/no questions, open-ended attitudes/beliefs centred questions were included, such as ‘what do you know about COVID-19’ or ‘what have you heard?’ After the focus group and post-test were completed, participants were thanked for their participation and compensated by gift card. Data storage, confidentiality and analysis All audio recordings were transcribed using Otter.ai and stored in a secure electronic database on a duo-protected Box folder. Quantitative analysis Quantitative data were collected via REDCap, cleaned and analysed using Excel and SPSS V.26 software. Descriptive analyses, including means and SDs, were applied to participants’ responses. Paired t-tests were then used to compare the pre-test and post-test group means, assessing for overall knowledge gain. Additionally, the mean scores per item were compared with determine knowledge gains per item from pre-test to post-test questionnaire. Results for knowledge questions were sorted into levels based on significance; high gains in knowledge/evidence of knowledge gaps (p<0.001), moderate gains in knowledge (p<0.05), low gains in knowledge/already high scores/not statistically significant (p>0.05). Findings from free response options for testing and vaccine questions were sorted into topical themes. Qualitative analysis Qualitative data were collected through the educational focus group sessions to provide additional contextual insights into quantitative findings. Recorded discussions were transcribed and reviewed for accuracy by members of the research team. A codebook was developed both deductively from the interview guide and inductively as codes were iterated or added during analysis. Each transcript was analysed thematically in NVivo V.12, a qualitative coding software program, by at least two independent coders who met at regular intervals for a consensus review of the coded material. Coders began with open coding, followed by axial coding. Themes related to multilevel motivators and demotivators for testing and vaccination were identified, guided by the socioecological model, which posits that health behaviour is influenced by multiple levels of factors (ie, individual, interpersonal and structural).31 Focus group sessions concluded when the research team determined thematic saturation was achieved.32 Results A total of 15 focus groups were conducted with a total of 78 PEH. The majority of participants identified as white (74%, n=58), with 18% of participants identifying as black (n=14), 8% (n=6) as Hispanic and 8% (n=6) as other. Most (73%, n=57) identified as men, 24% as women (n=19) and 3% (n=2) as transgendered or non-binary (table 1). Nearly half (49%) were more than 50 years of age, 27% (n=21) between 40 and 49, and 24% between 25 and 49 (n=19). Most (60%, n=47) were staying at a shelter, 22% (n=17) were in permanent supportive housing (PSH) and 12% (n=9) were unsheltered. Table 1Characteristics of focus group participants (n=78)Characteristicn%RaceBlack or African American1418White5874Other, more than one race or American Indian/Alaska Native68EthnicityHispanic, Latino or Spanish origin68Age18–240025–39192440–492127>503849SexMale5874Female2026GenderMan5773Woman1924Transgender or non-binary23HousingUnsheltered912At a shelter4760Permanent supportive housing1722Other56Education6th–8th grade569th–12th grade, no diploma2228High school graduate or GED completed3140Some college level/technical/vocational degree2026Loss of employment income since pandemicYes3140No or don’t Know4760Employment statusWorking now1519Unemployed and/or looking for work2532Unable to work due to disability2329Retired56Other or prefer not to answer1013Insurance statusPublic (Medicare, Medicaid, Tricare)6685Private (purchased directly or through Employment)68I do not have health insurance56Don't know11Loss of insurance due to pandemicYes34No6988Don’t know68Health statusExcellent56Very good1418Good2431Fair2329Poor1215Disability statusDisabled3950Not disabled3950COVID-19 testing practicesn% Knew where to get COVID-19 testingStrongly agree or agree7596 Felt it is easy to get tested for COVID-19Strongly agree or agree7495 Have been tested for COVID-19Yes6583At the shelter (38) No1317COVID-19 vaccine practicesn% Received vaccineFully vaccinated4051Received booster (11)Received 1 dose56No3342 Characteristics of focus group participants (n=78) Almost all knew where to get tested for COVID-19 (96%, n=75) and felt it was easy to get tested (95%, n=74). Of those who had been tested for COVID before (84%, n=65), most (58%, n=38) were tested at least once at the shelter. Roughly half of the participants had been fully vaccinated (51%, n=4), 6% (n=5) were partially vaccinated and 42% (n=33) were not vaccinated. Motivators and demotivators to COVID-19 testing and vaccination Table 2 presents a comprehensive list of COVID testing and vaccine motivators and demotivators identified from both quantitative pre-test data (see online supplemental appendix C for survey results) and qualitative focus group data. Motivators were conceptualised as reasons participants did or would get tested or vaccinated and how they would encourage PEH to get tested or vaccinated. Demotivators were conceptualised as reasons they did not or would not get tested or vaccinated. Per the socioecological model, these were organised into the individual level (knowledge, attitudes, beliefs and experience), interpersonal level (influences from individuals in social network) and structural level (organisational, environmental, political or cultural influences, such as rules, policies and information dissemination). Table 2COVID-19 testing and vaccine motivators and demotivatorsTestingVaccineMotivatorsDemotivatorsMotivatorsDemotivatorsIndividualWant to know their results (eg, for peace of mind or safety)ExposureSymptomsBeing in a high-risk setting with close proximity to others (ie, in shelter)Pain or fear of painDidn’t feel the need to get testedFear of results or isolation (social and physical)Concerns about test accuracyTo prevent severe illness or deathTo keep self, community and/or family safeHas chronic health condition/illnessBelief that things will not go back to normal till everyone is vaccinatedFear/mistrust of vaccineReligious beliefsConcern of vaccine side effectsNot concerned about getting really sick from COVID-19Belief that having COVID is enough protectionDoesn’t like getting shotsInterpersonalFamily, friends, healthcare provider or shelter staff said to get testedTo protect others or prevent spreadSocial isolationDoctor said to get vaccinatedPeer pressureVaccine hearsay (eg, gives you covid, harms your health/kills you, contains a tracking chip/harmful products, is part of a government conspiracy, causes infertility)StructuralFree/available (ie, at shelter)Incentive (eg, gift card, food or refreshments)Option for rapid test and/or self-test vs PCRRequired/mandatory (eg, for work)Dissemination of COVID-19 educationOther people administering testInadequate information disseminationWait timeShelter lockdownsCredibility of test siteTesting personnelVaccine is freeIncentiveRequired/mandatory (eg, by jail, shelter or workplace)Misinformation COVID-19 testing and vaccine motivators and demotivators Testing At the individual level, symptoms, exposure, wanting to know their results (eg, for peace of mind) and being in a high-risk setting such as a shelter were identified as motivators, ‘Because with us being homeless I mean, we go in and out of public places all the time, and so we never know on whether or not we're going to be around somebody else that has it’—FG13. Pain or fear of pain was a prominent discussion point in focus groups as an individual-level demotivator or undesirable aspect of testing, ‘They aint gotta shove that thing all the way up your nose anymore either… A lot of people are scared because of that [P4: I was!] … That’s why I didn’t do it for a long time.’—P3, FG2. Other individual-level demotivators included concerns about the accuracy of testing, faith/religion, not feeling the need to get tested, fear of results and at the interpersonal level, social isolation, ‘I think most people don't get tested because they're scared of what they—[P5: they scared of the results]—and then they feel isolated and people avoid them’—P1, FG10, PSH. In terms of interpersonal-level motivators, many were motivated to test when someone in their social network suggests it (ie, friends, family, shelter staff or healthcare provider), and to protect others and prevent spread of the virus, ‘I’m concerned about, you know, spreading it to other people, you know, giving it to kids and the elderly people. Concern of safety. [P5: I agree with that] [Participants agree].’—P4, FG4. At the structural level, available and free testing (ie, at the shelter) was a motivator, ‘But since it’s here, and it’s available, free… Let’s just get tested.’—P1, FG3. In general, having rapid tests was preferred, because despite some concern about test accuracy, the pain experienced or feared by participants was specific to PCR tests. Having the option to take their own nasal swab was also a structural-level motivator, ‘I’d rather do it myself than have… instead of have someone else does.’—P1, FG2, as some were uncomfortable with other people swabbing their nose, ‘I don't like people touching on me um poking at me.’—P5, FG1. Many who were tested did so because they were required or perceived it as a requirement (eg, by hospital, jail, or work), ‘I had to do it to keep my job.’—P4, FG1. While many did not believe testing should be mandatory, particularly by the shelter, it was generally viewed as an effective way to get people to test, including the suggestion for the shelter to ‘Enforce the rules that we got’—P4, FG11. In one focus group, participants suggested offering food or refreshments as an incentive to get guests to test, ‘Offer them food [other participant: yeah, especially the homeless community] [Participants recommend coffee, donuts, hot chocolate, and apple cider]… Refreshments and stuff, so there’s always food like refreshments or something. Someone will weasel their way over there to see what that’s about.—Unknown Participant, FG7, PSH County health department-mandated lockdowns, which required the shelter to restrict access for individuals not exposed on the identification of a positive case at the shelter, were a noteworthy deterrent to testing. This was criticised by many participants as hindering PEH’s ability to obtain services, reminiscent of the challenges presented by the national lockdown, ‘… maybe a negative is that anytime someone does test positive around here they issue a lockdown and then if someone hadn’t been here that day, if they hadn't been vaccinated or don't want to for religious preferences or whatever the reason being, um they're denied services until lockdown’s over and usually when one person tests positive more people end up testing positive.’—P2, FG12. Other demotivators included faith/religion, and as one person suggested, ‘credibility of the test site’ and ‘testing personnel’. Having inadequate information dissemination was also mentioned, ‘I never needed to [get tested] but to be honest with you, I didn't know you can get tested. No one ever told me that.’—P4, FG14. This translated to the need for COVID-19 education dissemination at the shelter as a method to motivate and encourage testing among guests. Participants made several suggestions on topics and ways to disseminate education, such as emphasising the severity of COVID, highlighting the cost of testing and vaccination, and making more announcements at the shelter to get the word out, ‘P1: Give them some general information like the knowledge you just told us… P3: Be informed about … cost… Let them know how many people have died. Then they might wanna go get a shot.—FG2’ Vaccination At the individual level, preventing severe illness or death and keeping oneself and others safe were important motivators for many participants, ‘… it [the vaccine] reduces your chances of getting it [COVID] and reduces the like, if you do get it, it reduces your chances of getting like severely sick. And it also reduces your chances of spreading it.’—P3, FG3. Focus group discussions (FGDs) revealed many misconceptions about vaccines that explain fear and mistrust of the vaccine as a demotivator, such as the belief that the vaccine was developed too quickly and that the vaccine contains a live virus, ‘It’s just like the flu virus the flu—I don’t want that shot, why would you give me the flu to make me better? It’s just crazy.’—FG9 (PSH). There was also scepticism on how much protection the vaccine could provide, given that COVID-19 infection is still possible, ‘You can still get corona; you can still get corona virus with the shot.’—FG7, PSH. Others reported religious beliefs, concern of vaccine side effects, belief that having COVID-19 was enough protection, and not being concerned about getting really sick from COVID-19 as demotivators. At the interpersonal level, while many reported keeping others safe as a motivator to getting vaccinated, some were motivated by a doctor’s suggestion, or as one person stated ‘peer pressure’. Hearsay from social networks created more misconceptions about the vaccine that served as demotivators, for example, that the vaccine is ‘part of government conspiracy and contains a tracking chip’, ‘A lot of people won’t take the [vaccine] cause they think they put in a chip ya or you get tracked.’—P3, FG5. Other hearsay shared were that vaccines ‘can give you COVID’, harming your health through ‘heart attacks, blood clots, all kinds of things’, and even cause death, ‘Heard you get the shot, you die.’—P1, FG13 At the structural level, while providing free vaccines, receiving an incentive to be vaccinated and requiring vaccination (eg, by one’s workplace) were identified as motivators, misinformation about COVID and the safety of the vaccine contributed to uncertainty of getting vaccinated, ‘… the clearer we are about you know, some of the information, that’s, that’s probably one of the biggest fights that’s been about this is, is misinformation and getting correct and reliable information has probably been one of the biggest things for most Americans.’—P3, FG Post-intervention knowledge gains and attitude shifts The mean pre-test score was 10.76 out of 15 (SD=2.16), while the mean post-test score was 13.58 out of 15 (SD=1.72) (p<0.001). Most participants (n=71, 91%) improved by at least 1 and up to 10 points in their post-test scores. Table 3 shows pre-/post-test knowledge scores per item, accompanied by qualitative insights shared during group sessions to provide additional context for the numerical scorings. The findings are categorised into three sections: high, moderate and low knowledge gains. Table 3Knowledge gains and qualitative findings by pre/post questionnaire questionItem (correct answer) pre %, post %, p valueThemeRepresentative QuoteHigh gains in knowledge/evidence of knowledge gaps (p<0.001)COVID-19 vaccines affect fertility (false)38%, 87%, <0.001Hearsay of vaccines effect on fertility‘I heard they make you sterile’ – P4, FG14Trust in vaccine regulation‘I imagine if they suspected that it could cause problems with your reproductive system, they wouldn't be giving it out.’ – P3, FG10 PSHUncertainty about ability to know possible effects‘That’s up in the air because how they gon’ know til nine months down when somebody got pregnant.’ – P3, FG11Not directly concerned or effected by fertility‘P4: I don’t know I’m not a lady. P2: Don’t plan on having any kids’ – FG9Some of the COVID-19 vaccines can give you COVID-19 (false)42%, 85%, <0.001Belief that vaccines can give you COVID-19‘…them one of my- been one of my biggest deterrents… is the weakened form of the virus’ –P3, FG1Uncertainty if vaccine side effects are from COVID-19 infection‘I know some people who took the vaccine and said they got sick, I don’t know if they got COVID or not, but said they felt bad and everything.’ – P6, FG6Knowledge that vaccine causes symptoms, not COVID-19‘It’s not the virus, it’s your body’s response to fighting off’ – P3, FG 10 (PSH)How much do covid tests cost? (Free)54%, 94%, <0.001Cost uncertainty‘I don't know if it’s free or not’ – P4, FG1It is possible to test negative but still have COVID-19 (true)69%, 94%, <0.001Knowledge of false negative likelihood of rapid vs PCR‘I'd say if it’s a rapid test, it’s more likely that it’s [false negative] possible vs the PCR test.’ – P2, FG12You can get your results from some COVID-19 tests within 15 min (true)73%, 97%, <0.001Knowledge of result time differences of rapid vs PCR‘The rapid you get it back right there and then… the other one is one to four days.’ Unknown Participant, FG7Knowledge of swab location difference of rapid vs PCR‘You’ve got the brain scraper, and then you got the inside of your nose scraper.’ – P5, FG14Knowledge of accuracy differences of rapid vs PCR‘…the only thing I’ve heard is that the one brain tickler is more accurate of the two.’ – P3, FG11COVID-19 vaccines prevent severe illness, hospitalisation and death from COVID-19 (true)60%, 81%, <0.001Knowledge that vaccine protects from severe illness‘…it’s not like 100% effective…it reduces your chances of getting it and…if you do get it, it reduces your chances of getting like severely sick. And it also reduces your chances of spreading it.’ – P3, FG3Belief that vaccines are not protective because you can still get COVID-19‘No and I'm here to tell you I was very upset when I found that out. I thought it was- I said, well, I've been vaccinated, and they're like that don't matter.’ – P4, FG14Moderate gains in knowledge (p<0.05)Everyone who gets COVID-19 will have symptoms (false)62%, 77%, 0.002Aware of possibility to be asymptomatic‘we have different symptoms no tastin or smellin and a lot of people have COVID and don’t know they have it’ – P5, FG10 PSHUnaware of possibility to be asymptomatic‘I guess I'm surprised I could have it and not know it…I’m pretty open minded about it, I guess it could be possible but until I get it, it’s not true [laughs].’ – P2, FG9 PSHCOVID-19 vaccines can help reduce the spread of COVID-19 (true)77%, 91%, 0.007Knowledge that vaccine lowers transmission, but not fullyP2: It’s probably less likely [to spread] but they still can. – FG12Uncertainty of vaccines impact on transmissionYeah, ‘cause I can be a carrier, right. The vaccine aint going to keep you from being a carrier? – P1, FG3Unaware transmission is still possible when vaccinated‘I did not know that you could-you can spread it if you had it and you were vaccinated.’ – P1, FG2If you tested negative for COVID-19 last week it means you are negative now (false)79%, 91%, 0.019The virus that causes COVID-19 spreads easily between people (true)86%, 97%, 0.006Awareness of ease of spread‘[COVID is] very contagious.’ – P1, FG3Knowledge on modes of transmissions (eg, airborne)‘Touch, breathing on somebody. Sneezing on somebody, coughing’ – P2, FG3‘Yeah, I heard some people can like, carry it and pass it without actually getting sick themselves.’ – P3, FG3Belief of fomite transmission‘You know it can be spread that way. By money, by anything.’ – P5, FG4‘Yeah, it can live on surfaces for…14 days’ – Unknown P, FG7 PSHScepticism of how easily COVID spreadsP2: As easy as they said it spreads out, I would’ve thought I’d had it about three times….’ – FG9 PSHYou don't need a COVID-19 vaccine if you've had COVID-19 before (false)81%, 91%, 0.045Belief you should still get vaccine‘P4: Definitely. P3: You can still get it again.’ – FG5Don’t believe vaccine is necessary‘Technically they say yes.’ – P1, FG11‘Shouldn’t it be dealt with, like when you were a kid, somebody’s sick, they threw you in with the sick kid and said ‘go ahead and get it” you get it now, you won’t get it later.’ – P2, FG9 (PSH)Low gains in knowledge/already high scores/not statistically significant (p>0.05)You may experience side effects such as arm soreness,fever or fatigue from the COVID-19 vaccine (True)85%, 92%, 0.109Awareness or experience with side effects‘They have different side effects to different people” – P5, FG10 (PSH)‘You get cold and your legs ache, that’s really the only symptoms after the vaccination.” – P5, FG13You can still get COVID-19 even if you've had the vaccine (True)86%, 92%, 0.199Awareness of COVID-infection risk when vaccinated‘I think you can. Just lowers your risk of getting it. P3: Yeah, lowers your risk of dying.” – P4, FG5Unaware that COVID-infection is possible when vaccinated‘I guess I didn't realize that if you had both shots and the booster, that you could get COVID again” – P2, FG 11You can still get tested for COVID-19 if you've been vaccinated (True)94%, 99%, 0.103Awareness of testing purpose even when vaccinated‘Yeah, you can get it again.” – P2, FG3If your COVID-19 test result is positive it means you likely have COVID-19 (True)90%, 90%, 1.000False positive concerns‘It could be a false positive as well, if you ain’t feeling nothin or have any symptoms… that still shows positive, it could be wrong.” – P1, FG4 Knowledge gains and qualitative findings by pre/post questionnaire question High knowledge gains were observed for specific items, namely misconceptions about vaccine fertility (pre-test: 38%, post-test: 87%, p<0.001) and COVID-19 vaccines giving you the virus (pretest: 42%, post-test: 85%, p<0.001). There was a spectrum of viewpoints on both topics (see table 3), centred around the uncertainty of effects and ranging from complete belief in misconceptions to trust or knowledge of vaccine regulations and purpose. Knowledge gains and varying viewpoints were seen in the testing-related questions as well, including the cost of COVID-19 tests (pretest: 54%, post-test: 94%, p<0.001) with some participants unaware of free test access, the possibility to test negative but still have COVID-19 (pretest: 69%, post-test: 94%, p<0.001) and participants’ awareness of rapid testing options, asked as whether it is possible to have your results in 15 min (pretest: 73%, post-test: 97%, p<0.001). The supporting quotes in table 3 represent the scope of individuals existing knowledge and awareness of the testing topics, demonstrating, for example, how some participants understood the variation between the testing options by using result time, swab location/depth and accuracy as differentiation methods of rapid versus PCR (often representing PCR with comments like brain scraper, the one that touches your brain, etc). Participants displayed mixed feelings about the effectiveness of COVID-19 vaccines in the comments they shared, and topics such as protection against severe outcomes (pre-test: 60%, post-test 81%, p<0.001) and minimising the spread (pre-test: 77%, post-test: 91%, p<0.05) saw high to moderate knowledge gains. There was a general understanding that some degree of protection is offered through vaccination, with one participant stating they were upset when they discovered protection was not 100%. Other areas saw lower or no significant knowledge gains due to high pretest scores, however, the qualitative findings still provide important insights into these questions. For example, whether or not to get a COVID-19 vaccine if you’ve had COVID-19 before, participants generally knew that it is recommended to get it (pretest 81%, post-test 91%, p<0.05), but despite awareness, some didn’t necessarily agree with this recommendation by health agencies (‘THEY’ as depicted in table 3). Attitudes towards testing and vaccination were positively influenced postintervention. Of those not vaccinated (n=33), 21% were willing to get the vaccine initially, and this increased to 39% after the educational intervention, with ‘I don’t know’ responses decreasing from 39% to 18%. Participants’ willingness to get tested after exposure also increased from 83% to 92% after the session (see figure 1). Percentage of COVID-19 testing and vaccine willingness among participants, pre-education and posteducation intervention. PEH perspective on CHW-led education sessions and preferred methods related to information sharing During the FGDs, participants provided valuable feedback on their experience with this style of educational intervention, discussed their preferred information sources and expressed their interests in various topics beyond COVID-19. Participants mainly provided positive feedback and appreciation for the information shared in the groups, which they found to be informative and trustworthy coming from the CHW. In addition, one participant brought up that they not only benefited from what the CHW shared, but found value in a setting with an opportunity to learn from others, ‘I haven't been keeping up on anything lately, and what that man said [other FGD participant] uh sort of informed me that there’s more on the way, testing, better testing and uh I like that…’—P4. The participants also identified specific sources they trust when seeking information, including the news and the CDC. However, scepticism was expressed, and participants contributed concerns of misinformation to changing guidance delivered by health authorities, one stating, ‘A lot of contrary information. When Fauci first came out, masks are not necessary, masks are not necessary. Now two masks. You know, it just changes every day. It’s ridiculous’—P6, FG8. Others mentioned online media’s responsibility for the spread of misleading information, ‘I think that was probably one of the biggest things that hurt this whole situation we're in, is social media and this the era we live in of cell phones and spread of just hearsay and misinformation is so bad nowadays is like wildflower.’—P3, FG12. Overall, participants highlighted the importance of trustworthy and credible sources in gaining accurate information and that they value information from community members and experts, such as local university professors in a relevant field of study, local community officials, including CHWs and health department employees. This expanded to people with lived experience or who work in close proximity to the topics they would educate, deeming them more trustworthy sources because they have ‘seen more’. Discussions also included nurses and medical doctors, however, some preferred community members stating ‘Doctors- they don't explain stuff completely.’—P1, FG3. Discussion This study investigated the impact of CHW-delivered education for PEH in Indiana, aiming to address misinformation and promote COVID-19 testing and vaccination. Overall, considerable knowledge gains and positive attitude shifts were achieved, demonstrating the impact of CHW-led education on COVID-19-related knowledge, attitudes and beliefs. The blended focus group and education session garnered insights towards critical knowledge gaps, motivators, demotivators and barriers to COVID-19 testing and vaccination. Guided by the socioecological model, these insights further our understanding of the multilevel interactions influencing PEH access to and utilisation of testing and vaccination during the COVID-19 pandemic. This revealed key incentives as well as legitimate yet often mitigatable barriers including fear of pain from testing, inadequate dissemination of information and vaccine misinformation. In regard to testing, our findings align with other studies that found barriers such as cost, logistics and the impact of limited culturally sensitive messaging on health literacy to be codependent and compounding.16 Our previous study also identified that prior to testing being available at the shelter, participants’ confusion around testing costs in healthcare settings if uninsured led to not being tested.9 This study found that despite the mitigation of cost and other structural barriers by having testing available onsite at the shelter, critical knowledge gaps and lack of adequate information dissemination regarding free testing availability still presented barriers to uptake. The results of our intervention showed improved knowledge and attitudes towards testing, demonstrating the effectiveness of and highlighting the need for educational interventions, in addition to improving accessibility. A scoping review and meta-analysis of COVID-19 prevalence and vaccine hesitancy in the homeless population found that suspicion of adverse side effects and doubts over vaccine clinical trials were significant contributors to hesitancy, and emphasised factors like low health literacy as well. Our findings align with these concerns, as our FGDs revealed similar misconceptions about vaccines, such as fears regarding the speed of vaccine development and concerns or doubts about vaccine efficacy. Our intervention resulted in improved vaccination attitudes and knowledge, supporting the review’s suggestion for COVID-19 vaccine education among PEH. Additionally, our study also highlights the influence of misinformation from social networks as a barrier to vaccination acceptance. Though online media was identified as a source of information for PEH, our participants expressed a desire for better sources. They expressed interest and need for information that is consistent, accessible and diverse, and they want to be informed about their community by their community. This provides a key opportunity for public health to leverage interventions led by CHWs and other outreach professionals with shared lived experience and underscores the importance of future interventions that consider the crucial perspectives of PEH. Educational interventions remain an important method for promoting preventive health behaviours among PEH and are often called on as a recommendation for homeless shelters.3 9 17 33 However, these recommendations are often impractical for shelter staff to implement. Homeless shelters and other high-risk congregate settings are typically not well-resourced healthcare facilities that can simultaneously implement CDC guidelines such as testing while also upholding their essential mission of providing shelter and food to those in existing crises. Already limited, their ability to provide education and resources to inform and motivate individuals about available resources, such as COVID-19 testing and vaccination, is stretched. Even when staff at homeless shelters have the capacity to disseminate health information, our previous CBPR studies reported findings of PEH populations experiencing poor communication and discrimination with homeless shelter staff, pointing at other community assets like CHWs or the local health department to remediate health literacy barriers.9 The findings from our study align with existing literature in highlighting the effectiveness of CHWs in reaching populations excluded or met with barriers in accessing health information and resources during a pandemic by expanding on community relationships to build trust, establish knowledge, and address hesitancy and misinformation.22 One study used Community Health Outreach Workers to administer incentivised testing to unsheltered PEH and found that along with using mobile testing to improve accessibility, trusted community members can help overcome barriers to accessing and accepting COVID-19 testing and vaccines.25 This study also contributes important insights on CBPR as a valuable tool for centring the priorities and knowledge of communities in research efforts. CBPR can meaningfully inform appropriate educational interventions for PEH, including key content and delivery approaches. This study leveraged PEH and CHW relationships to understand, convey and tend to the unique challenges faced by this population during the COVID-19 pandemic. Combining an educational session with a qualitative focus group discussion led to a comprehensive understanding of the complex factors contributing to the COVID-19-related knowledge and practices of PEH. This style and its administration by CHWs allowed for real-time tailoring of educational material based on participant needs and was supported by the feedback collected from participants during sessions. There were several limitations of this study. First, it focuses on a specific PEH population in Lafayette, Indiana, and recruitment was limited to people using the engagement centre services. This may have led to the exclusion of unsheltered and unserved homeless populations, limiting the generalisability of our findings. Furthermore, participant knowledge gains and attitude shifts were only evaluated immediately following the intervention, limiting our understanding of whether these gains were sustained. Future work should target a larger, more diverse population of PEH, including people experiencing housing instability, and those who are unsheltered, as well as cover a longer period that includes follow-up with participants over time to measure the sustainability of the intervention’s impact. Future work should also examine the holistic impact of using and encouraging shelter-based COVID-19 rapid testing on the shelter environment. supplementary material