Community-engaged randomised controlled trial to disseminate COVID-19 vaccine-related information and increase uptake among Black individuals in two US cities with rheumatic conditions

Abstract

Introduction

Inequities in COVID-19 infection and vaccine uptake among historically marginalised racial and ethnic groups in the USA persist. Individuals with rheumatic conditions, especially those who are immunocompromised, are especially vulnerable to severe infection, with significant racialised inequities in infection outcomes and in vaccine uptake. Structural racism, historical injustices and misinformation engender racial and ethnic inequities in vaccine uptake. The Popular Opinion Lleader (POL) model, a community-based intervention that trains trusted community leaders to disseminate health information to their social network members (eg, friends, family and neighbours), has been shown to reduce stigma and improve care-seeking behaviours.

Methods and analysis

This is a community-based cluster randomised controlled trial led by a team of community and academic partners to compare the efficacy of training POLs with rheumatic or musculoskeletal conditions using a curriculum embedded with a racial justice vs a biomedical framework to increase COVID-19 vaccine uptake and reduce vaccine hesitancy. This trial began recruitment in February 2024 in Boston, Massachusetts and Chicago, Illinois, USA. Eligible POLs are English-speaking adults who identify as Black and/or of African descent, have a diagnosis of a rheumatic or musculoskeletal condition and have received >=1 COVID-19 vaccine after 31 August 2022. POLs will be randomised to a 6-module virtual educational training; the COVID-19 and vaccine-related content will be the same for both groups however the framing for arm 1 will be with a racial justice lens and for arm 2, a biomedical preventative care-focused lens. Following the training, POLs will disseminate the information they learned to 12–16 social network members who have not received the most recent COVID-19 vaccine, over 4 weeks. The trial’s primary outcome is social network member COVID-19 vaccine uptake, which will be compared between intervention arms.

Ethics and dissemination

This trial has ethical approval in the USA. This has been approved by the Mass General Brigham Institutional Review Board (IRB, 2023P000686), the Northwestern University IRB (STU00219053), the Boston University/Boston Medical Center IRB (H-43857) and the Boston Children’s Hospital IRB (P00045404). Results will be published in a publicly accessible peer-reviewed journal.

Trial registration number

NCT05822219.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The unique nature of the community-engaged study design and popular opinion leader (POL) model has the potential to reach a diverse population often hesitant to engage in traditional research, with the aim of reducing racial inequities in COVID-19 vaccine uptake and hesitancy.

• The potential to assess POL engagement among social network members provides a unique opportunity to study the effectiveness of community-led health information dissemination and whether this differs from the framework used for the POL training.

• POL participation and social network member outreach may be limited by health and technological literacy, cultural differences, vaccine access issues (eg, for undocumented or uninsured individuals) and primary language other than those included in this study (English, Spanish and Haitian Creole).

• Interactions with a traditional biomedical institutional review board with limited experience with community-engaged research resulted in delays and modifications to the original plan. However, the minimal risk nature of this study allowed the approved protocol to be consistent with the project’s original goals to maximise community engagement.

Introduction

While significant public health measures exist to address COVID-19 vaccine hesitancy and promote uptake among historically marginalised racial and ethnic groups,¹ inequities in vaccine uptake persist. Marginalised racial and ethnic groups, specifically Black and African American communities, Hispanic/Latinx communities and Indigenous communities who have been made vulnerable to adverse health outcomes by systemic and structural racism, discrimination or oppression of other kinds,2,4 continue to experience poorer outcomes from COVID-19 infection and lower rates of vaccine uptake. On 12 September 2023, the Centers for Disease Control and Prevention (CDC) released an updated COVID-19 vaccine (monovalent, targeted against the XBB family variant).^{5 6} As of 20 January 2024, the percentage of Black (17.9%) and Hispanic/Latinx (13.1%) individuals who have received the updated COVID-19 vaccine remains below that of White individuals (25.7%).⁷ COVID-19 vaccine hesitancy among marginalised racial and ethnic groups may be engendered by historical and current experiences of racism, access barriers and concerns regarding the science of vaccines and their development.^{8 9} Studies of individuals who identify as belonging to these racial and ethnic groups with rheumatic conditions have similarly demonstrated a disproportionate burden of COVID-19 adverse outcomes¹⁰ and persistent vaccine hesitancy.¹¹ This hesitancy manifests as overall low vaccination rates among individuals with systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA), despite high rates of serious infection, with inequities by race, ethnicity and insurance status.12,14

The updated vaccine presents an opportunity to reduce racial and ethnic inequities in COVID-19 vaccine roll-out, hesitancy and uptake and to diminish health misinformation15,17 with innovative public health measures. Community-engaged research, a collaborative research approach to integrate community members’ unique social and cultural understanding of the community,^{18 19} has been shown to lessen barriers to COVID-19 vaccination among marginalised racial and ethnic groups.^{20 21} A promising strategy to promote vaccine acceptance is the Popular Opinion Leader (POL) model, which trains trusted community members to disseminate information about health-related behaviours through their social networks.^{22 23} The POL model has been shown to increase community knowledge, reduce stigma and prompt behavioural changes^{22 24} and has been used by our study team to improve knowledge and awareness of clinical trial participation²⁵ and lupus²⁶ among individuals of Black and/or African descent with rheumatic conditions.

Our team of academic and community partners combines expertise from multiple academic medical institutions in Boston (Brigham and Women’s Hospital (BWH), Boston Children’s Hospital (BCH), Boston University/Boston Medical Center (BU/BMC, a safety net hospital) and Chicago (Northwestern University (NU)) and community leaders and organisations (Boston—Vital CxNs, True Alliance Center, Chicago—Alliance Chicago), a neighbourhood association (Boston—Mission Hill Health Movement, Inc.), support groups (Boston—Women of Courage, Inc., Boston Arthritis and Lupus Support Group), patient advocates, community activists, the Lupus Society of Illinois, and community and public health education experts. Three community partners were previously trained as POLs through SLE-related projects.^{25 26} Seven physician academic partners and four community partners participated in preliminary interviews to understand stakeholder’s perspectives to increase COVID-19 vaccine uptake among Black individuals with rheumatic conditions.⁸

We aim to conduct a randomised controlled trial (RCT) to test the efficacy of a community-based POL intervention with a racial justice or biomedical framework to increase COVID-19 vaccine uptake and reduce hesitancy among the social networks of individuals who identify as Black and/or African descent with rheumatic conditions. The intervention arm (arm 1, racial justice framework) curriculum is based on “Public Health Critical Race Praxis” (PHCRP),²⁷ which recognises race as a social construct, emphasises the ways social systems and institutions reinforce racism, and calls on researchers to critically examine a priori assumptions related to issues of race at every stage of a biomedical study.^{27 28} As defined by Bailey et al, structural racism ‘refers to the totality of ways in which societies foster racial discrimination through mutually reinforcing systems of housing, education, employment, earnings, benefits, credit, media, healthcare and criminal justice. These patterns and practices in turn reinforce discriminatory beliefs, values and distribution of resources’.^{3 29} With this framework, intervention arm (arm 1) modules will acknowledge and address the role of structural racism in vaccine deliberation³⁰ and rheumatic disease health outcome inequities,³¹ with an emphasis on community-level COVID-19 vaccine uptake as a strategy to advance racial justice. Arm 2 uses a biomedical framework,³² emphasising the physical and biological roots of health outcomes, COVID-19 vaccine mechanisms, and the importance of preventative care. As individuals who are immunocompromised often have a less robust vaccine response, it is especially important for their social network members to be vaccinated for a ‘cocooning effect’, to minimise transmission.³³ Therefore, this model aims to improve an individual’s vaccination uptake and uptake of their close contacts to improve protection for the immunocompromised individual. Arm 2’s strategies are more focused on improving and protecting the health of the individual and their close contacts, compared with arm 1’s strategies that emphasise the health of the community.

We hypothesise that social network members of POLs receiving vaccine safety and efficacy training with racial justice framing (arm 1) will exhibit greater COVID-19 vaccine uptake when compared with social network members of POLs trained with a biomedical framework (arm 2). We also hypothesise that POLs trained with racial justice framing (arm 1) will reach out to larger social networks with higher percentages of individuals of Black and/or African descent compared with arm 2.

Methods

Study design

This study is a community-based cluster RCT to compare the efficacy of two POL training frameworks to increase COVID-19 vaccine uptake and reduce vaccine hesitancy among social network members of individuals identifying as Black and/or African descent with rheumatic or musculoskeletal conditions (figure 1). Recruitment began in February 2024, and we plan to complete data collection over the course of 1.5 years. The initial 18 months of this study grant (June 2022–January 2024) were dedicated to developing study materials and receiving institutional review board (IRB) approval. The study design described in this manuscript reflects protocol version 8 (date: 16 February 2024).

Figure 1. Study overview. POLs, popular opinion leaders.

Material development

Module development

Informed by interviews with community and academic stakeholders in the greater Boston and Chicago areas,⁸we engaged stakeholders to develop iterations of culturally competent and scientifically accurate modules for POL training. We combined stakeholder interview findings⁸ and a literature review to develop the first iteration of the POL training modules and module pre- and post-tests. A base curriculum with parallel information on COVID-19 vaccine safety and efficacy was developed, with the enrichment of racial justice-related information (arm 1) or biomedical information (arm 2). All modules and pre- and post-tests were reviewed by two or more academic and community partners across sites and subsequently reviewed by a study team member and physician to confirm an eighth-grade literacy level. After the intervention period, an optional seventh module will provide details regarding the two intervention arms and share initial findings.

Recruitment material development

Community and academic stakeholders were continuously engaged in study recruitment material development. This review process resulted in significant changes to materials to reach a diverse POL and social network population. The study title—ACTIVATE: Academic Community parTnerships to Increase Vaccination in Black and/or African descent identifying communiTiEs—was revised multiple times to respond to feedback received. Additionally, to include the 14.5% (~25 000) Haitian-identifying and 13.5% (~23 000) Black Latinx-identifying individuals in Boston,³⁴ as well as the Haitian-identifying population (~4700 individuals) in Chicago,³⁵ social network member enrolment criteria was expanded from English-speaking individuals to Haitian Creole-speaking and Spanish-speaking individuals. This required the translation of all social network materials, including the translation of the Vaccine Hesitancy Scale (VHS) survey.36,38

Evolving enrolment criteria

To reflect evolving changes in CDC COVID-19 vaccine recommendations, study eligibility requirements and study materials were changed several times between study initiation (June 2022) and first recruitment (February 2024). POLs will be required to have received a COVID-19 vaccine after September 2022 (bivalent Omicron booster)³⁹ or the most recent COVID-19 vaccine (approved on 12 September 2023)⁶; social network members will be required to have not received the most recent COVID-19 vaccine (approved on 12 September 2023).⁶ Based on the input from a community partner, social network members are eligible if they did not receive the most recent vaccine due to a COVID-19 infection within the prior 3 months so long as they intend to receive it following this window.⁴⁰

IRB approval process

The community-engaged nature of the POL model posed a challenge to IRB approval for two main reasons. The first was the role of POLs as both study subjects and extensions of the study team. The second was concern regarding the consent process. Both factors differed from traditional biomedical studies. The original proposal relied on consent implied by the willingness of the community member to engage with the POL in a minimal risk conversation, rather than requiring written informed consent. However, IRB members raised concern about the indiscriminate boundary between POLs as study subjects participating in the RCT, and as study team members delivering the intervention to their social network members and collecting their data. Initially, IRB members suggested POLs be trained as Individual Investigators, undergoing general and study-specific training relating to human research protection (eg, Health Insurance Portability and Accountability Act (HIPAA), Belmont Principles (respect for persons, beneficence and justice)), consent procedures, data collection and data privacy, and sign an extensive Individual Investigator Agreement.⁴¹ However, we received input from our academic and community partners that this strategy would be problematic for several reasons. First, there was a concern regarding a potential conflict of interest by asking individuals to be research subjects and paid data collectors for the same overarching study. Second, the proposed approach seemed antithetical to the goal of the POL model where POLs were trusted, trained community members/leaders and not extensions of academic research teams. Third, one community partner expressed concern that by being asked to follow specific rules as a ‘researcher’, they would be going against the social norms and dynamics of their social network, which would limit their impact and reach. We provided this feedback to the IRB in a letter and in response, they decided that this would be an exempt study since it is a minimal risk behavioural intervention, under the condition that POLs receive training (module 5) on human research protection. The training was designed by our study team and reviewed, edited and approved by community and academic partners and by the IRB. Additionally, the IRB suggested that POLs share a study flyer with their social network member that provides a brief study description, specifying that the POL is working with community organisations and hospitals in Boston and Chicago and has been randomised to discuss health information from a certain perspective. The flyer further describes what study participation will look like, risks, benefits, study staff contact information and Mass General Brigham/NU IRB contact information. The first study survey that social network members complete will include a secondary brief study description (online supplemental material 1), and survey completion will serve as implied consent.

While the initial intention was for a single IRB, the exempt determination meant that this was not possible. Therefore, each site (NU, BCH, BU/BMC) had to obtain individual approval from their respective IRBs. All sites eventually deemed this study exempt; however, this need for site-specific approval led to significant study delays.

Study setting, participants and eligibility criteria

We will enrol POLs from the greater Boston and Chicago areas who are 18–85 years, of any gender, English-speaking, self-identify as Black and/or African descent, of any ethnicity, and self-report a physician diagnosis of a rheumatic or musculoskeletal condition including systemic rheumatic diseases (connective tissue diseases (eg, SLE), inflammatory arthritis (eg, RA, spondylarthritis), crystalline arthritis (eg, gout) or a musculoskeletal condition (eg, osteoarthritis (OA), osteoporosis). We broadly defined rheumatic conditions to include individuals at high risk for COVID-19 infection due to immunosuppression and the nature of the systemic rheumatic condition (eg, SLE, RA),⁴² gout itself, which has been associated with a higher risk of COVID-19 infection and adverse outcomes^{43 44} or due to comorbidities (obesity and cardiovascular disease in gout⁴⁵ and OA,⁴⁶ and older age in osteoporosis).^{47 48} Based on community partner feedback, to represent the heterogeneity of Black communities in Boston and Chicago, we expanded our original study enrolment criteria from Black and/or African American to include individuals identifying as ‘of African descent’. Additionally, we will encourage the participation of POLs who, in addition to English, speak Haitian Creole or Spanish. We will require that POLs have received at least one COVID-19 vaccine after 31 August 2022 (bivalent booster or updated COVID-19 vaccine (2023–2024 formula)) to enable POLs to share their vaccine experiences and rationale. For inclusion as a POL, we will also require suggestion of social connectedness, with a score of one or more on an adapted Berkman Syme Social Network Index survey⁴⁹ (online supplemental material 2). POLs will be excluded if they are incarcerated or if they are currently pregnant.

POLs will be asked to disseminate information to 12–16 of their social network members who are ≥18 years old, any gender, and English-speaking, or Haitian Creole or Spanish-speaking if the POL speaks the same language and have not received the updated COVID-19 vaccine (2023–2024 formula) at the time of the first POL conversation. We will not require social network members to be Black and/or African descent; however, POLs will be encouraged to prioritise individuals who identify as belonging to these groups.

Recruitment and screening

We will leverage our community and academic partnerships to recruit POLs across the greater Boston and Chicago areas. We aim to recruit POLs who represent the heterogeneous and diverse Black and/or African descent-identifying communities in these areas. At the academic clinical centres (BWH, BCH, BU/BMC, NU), rheumatologists will identify patients who meet eligibility criteria, specifically recruiting charismatic patients who participate in rheumatic disease-advocacy efforts and community groups. Several vignettes will be given to providers highlighting examples of POL characteristics (eg, a hairdresser who is a member of a lupus support group). Rheumatologists will be asked to introduce the study to their patients using a flyer with a brief description of the study and a QR code that links to a more detailed fact sheet. To capture individuals who receive care outside of participating academic medical institutions, well-connected community partners, patient/community advocates and support group leaders will be asked to identify individuals meeting the same criteria. Community leaders will provide potential POLs with the same flyer and QR code for the fact sheet. Individuals who may be interested in receiving POL training will be instructed to contact the study team via phone or email. POLs will be prescreened for eligibility criteria and verbally consented after review of the study fact sheet (online supplemental material 3). POLs will be told that they will be randomised to one of two training arms, but the differences in the curricular frameworks will not be disclosed. POLs will complete a demographic survey prior to randomisation (figure 2).

Figure 2. POL prescreen and randomisation flow. MGB, Mass General Brigham; POL, popular opinion leader; REDCap, Research Electronic Data Capture.

Randomisation

Our randomisation scheme will use stratification and permuted blocks with block sizes alternating between 2, 4 and 6. POL randomisation will be stratified by site (Chicago/Boston), age (18–45/46–85) and gender (men/women).

Interventions

This trial will have two intervention arms. Both arms will include a six-part training with identical scientific content addressing COVID-19 vaccine hesitancy, safety and efficacy, delivered by the same team of multidisciplinary, racially and ethnically diverse individuals. Arm 1 will receive their training enriched with a PHCRP²⁷ framework embedded in every module (except in module 5: Human Subjects Training, Collecting Research Data education, which is standardised across both arms). Training in arm 1 will acknowledge and address the role of racism in vaccine deliberation and emphasise vaccine uptake as a form of racial justice. Training in arm 2 will focus on a biomedical approach to discussing COVID-19 vaccine hesitancy and uptake, emphasising preventative care like vaccinations and a healthy lifestyle (eg, nutrition, exercise).

Intervention timeline, surveys, data collection

POLs will be asked to complete surveys at the time of enrolment to capture demographic information, COVID-19 vaccination history, previous engagement as a POL and community-based organisation participation (figure 3). Following enrolment, the virtual group training sessions will begin. Virtual education sessions will be recorded and sent to those unable to attend the live session. Each module will include a pre-and post-test. To confirm POL understanding of modules, all post-tests will be reviewed. At the subsequent module training, common errors and key slides from the previous training will be reviewed.

Figure 3. POL engagement timeline. POL, popular opinion leader.

Within 1 week of module 1 instruction, POLs will be contacted by a study team member and guided through an adaptation of the validated, HIPAA compliant, REDCap (Research Electronic Data Capture, a secure web-based software platform)^{50 51}-based Personal Network Survey (PERSNET),⁵² a 15–20 min evaluation that captures and maps an individual’s deidentified social ties and network.⁵² The first administration of the PERSNET aims to capture the POL’s personal social network and to help the POLs identify who in their network they might target for intervention. POLs will be asked to identify (1) people they discuss important personal matters with, (2) people who they socialise with and (3) people who they engage with about health matters (not mutually exclusive), and the interconnectedness and demographic and health-related information of those individuals, which will result in a map of the POL’s personal social network.

After the training, POLs will be directed to disseminate information to their social network members. POLs will be asked to reach out to each social network member at least twice and to record the time, date of each contact, the platform used and the zip code where the person was when reached. POLs will have worksheets for each contact to keep track of these data using a secure system.

Following their first conversation with social network members, POLs will share a demographic information survey, a COVID-19 vaccination status survey and the VHS survey³⁶ with the social network member. POLs will share a unique REDCap^{50 51} code and link to each social network member for survey completion. For social network members with lower digital literacy, paper survey copies will be provided. These surveys will be available in English, Spanish and Haitian Creole. Four weeks after their first contact with the social network member, POLs will distribute a second VHS and COVID-19 vaccination status survey link and code to the member to assess changes in the member’s vaccination status and hesitancy.

Following survey distribution to 12–16 POL social network members, POLs will complete an adaptation of the HIPAA compliant, secure, REDCap^{50 51}-based PERSNET,⁵² to capture the POL’s ‘outreach’ network. In the outreach network, POLs will be asked to list the people they discussed COVID-19 vaccination with over the last 3 months, the interconnectedness of those individuals, and demographic and health-related information of those individuals, which will result in a map of the POL’s outreach social network.

Following completion of this quantitative component, the network research team will engage the POLs in a brief (approximately 20 min) semistructured interview to understand the quality of network ties (connections between people) and the mechanisms of information spread. Interviews will specifically focus on the place of connections, ease of access and barriers to tie formation and will be recorded, transcribed verbatim.

Outcomes

The primary outcome will be COVID-19 vaccine uptake among social network members, which will be compared by intervention arm. The secondary outcome will be changes in social network member vaccine hesitancy, assessed using the VHS.³⁶ In addition, we will assess both structural (quantitative description of ties in a POL’s network) and compositional (proportion of characteristics across all members of the POL’s networks) features of the POL’s personal and outreach networks by arm and map the interconnected outreach networks of the POLs.

Power estimates and statistical analysis

We estimate that if 30 POLs per arm reach 14 members each (n=420 per arm), we will have 80% power to detect a 12% difference in vaccine uptake (primary outcome) and a clinically meaningful difference of 0.53 of vaccine hesitancy (secondary outcome) between arms. We estimate that this sample size will have $\ge$80% power to detect an effect size of 0.75 or greater for network characteristics. We plan to enrol 39 POLs per training arm to account for retention rates of 70% for POLs and 65% for social network members.

A two-sample proportions test will be used to compare social network vaccine uptake and change in total score from the 10-item VHS survey between groups. Unadjusted and adjusted generalising estimating equation models will be used to account for social network members by POL. The primary predictor variable will be the intervention arm, and covariates will include the city (Boston or Chicago) and other factors that are imbalanced between the intervention arms.

Quantitative metrics of the personal and outreach networks in terms of structural and compositional traits will be generated and compared across the two study arms using multivariable methods that adjust for baseline demographic characteristics. Qualitative analyses will delineate the quality of network ties and mechanisms of vaccine information spread. POL’s semistructured interviews will be recorded, transcribed and analysed using the constant comparative method.⁵³

Patient and public involvement

Community representatives, including patients with rheumatic conditions, were engaged in every stage of this research project, which is reflected in coauthorship and in the acknowledgements section based on the nature of their contributions.

Monitoring

For data monitoring and management, reporting of unintended effects, and auditing procedures, see the Data Safety Monitoring Plan (online supplemental material 4).

Discussion

The POL model is a promising community-engaged public health strategy employed by this trial to improve COVID-19 vaccine uptake and reduce hesitancy. This study aims to assess the impact of a POL education curriculum with a racial justice framework compared with a more conventional biomedical preventative care framework to increase COVID-19 uptake and hesitancy. The study will also investigate how training content impacts information dissemination among POL’s social networks.

There are several limitations to this study. The first is the potential for technological barriers that may impact both POL and social network engagement and survey completion. Second, we are only able to provide POL training in English, although we will be able to enrol POLs who are bilingual in English and Spanish or Haitian Creole to allow for dissemination to non-English speaking social network members. However, we were unable to translate social network materials into additional languages spoken by the diverse Black population in Boston (eg, Cape Verdean Creole, Portuguese)^{34 54} and, therefore, some social network members may not be represented unless they are bilingual. Third, our study materials were informed by stakeholder interviews⁸ and community partner and academic partner feedback from two US cities and may not be generalisable to other US cities. Fourth, there is the potential for bias to the null. In our previous studies, participants have, unprompted, raised issues related to racism and this may occur in conversations with POL trained in arm 2 (biomedical framework). We suspect that even if racism is raised in conversations with arm 2 trained POLs, it will not bias our study given the emphasis of related topics in arm 1 modules, which will be further strengthened in role-play scenarios. If both arms increase vaccine uptake equally, we do not feel this would suggest that racial justice training is not essential, but rather that other issues may also contribute to vaccine hesitancy. It is possible that there will be overlap of POLs’ social network members. To address this, we will train the POLs to first ask their network members whether they had a recent conversation with someone else trained in this study; if so, they will not include them.

In conclusion, this trial will evaluate the impact of a community-based intervention with a racial justice or biomedical framework to increase COVID-19 vaccine uptake and reduce hesitancy among social networks of individuals who identify as Black and/or African descent with rheumatic conditions. If the intervention is effective, this model could be broadened to other vaccines, diseases and geographical locations to promote equity in vaccine uptake.

Ethics and dissemination

This protocol has been approved by the Institutional Review Boards at Boston Children’s Hospital (recruitment), Boston University (recruitment), Mass General Brigham (BWH within MGB is co-primary site), and Northwestern University (co-primary site). Verbal informed consent will be obtained from POLs; implied consent will be obtained following fact sheet receipt and survey completion by social network members. Subjects are blinded to intervention arm; the study team will not be as they will be administering the training. Data analysts will be blinded to arm assignment. Results will be shared with participants in a meeting following the project and publications that result will be sent to all POLs. The findings from this work will be published in a peer-reviewed journal and made publicly accessible.

This content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. The funder did not play a role in the study design, collection, management, analysis or interpretation of data, in the writing of the manuscript, or in the decision to submit the manuscript for publication.