Improving COVID-19 Vaccine Uptake in Nursing Homes

SARS-CoV-2 vaccine, now being administered to skilled nursing facility (SNF) residents and staff, has highly variable acceptance between facilities. The investigators need to develop and disseminate effective strategies to increase vaccination immediately. For SNF residents and staff, the investigators will develop and implement a scalable multi-pronged intervention that educates, builds trust and supports the informed consent process aimed to increase SARS-CoV-2 vaccination. The investigators will conduct a cluster randomized trial to compare the effect of electronic messaging and education (i.e., usual care) versus a multi-pronged 'high touch' intervention to reduce vaccine hesitancy in skilled nursing facility staff and residents among a random sample of facilities across four SNF chains. As part of the 'high touch' intervention, the investigators will identify and train local opinion leaders. The investigators will offer these leaders assistance through real-time support for questions and provide consenting specialists. During the second wave of vaccination, the investigators will provide the intervention facilities with positive reinforcement for staff and will identify local champions to garner support and empowerment of staff. Finally, in the intervention facilities, the investigators will provide additional funds to support COVID-19 testing, in order that facilities have access to enough testing kits for patient or staff who develops symptoms following vaccination. This trial will be randomized within four SNF chains in order to evaluate the effect of a multi-pronged strategy to improve SARS-CoV-2 vaccine acceptance among direct care staff and long-stay nursing home residents. In four chains, eligible facilities will undergo randomization between usual care versus adding the 'high touch' intervention, implemented in two waves. Randomization and roll out of the intervention will occur at the facility level. The investigators hypothesize that: (1) the intervention will increase vaccination of SNF residents by at least 10 percentage points versus facilities usual care alone; (2) staff of SNFs with the intervention will have at least a 10 percentage point greater vaccine uptake of vaccine than staff in SNFs that do not participate in the high touch intervention; and (3) within intervention SNFs, improvements in vaccine uptake will be similar across staff and resident race/ethnicities.

An IRB waiver of consent was obtained for this study. The investigators will conduct a cluster randomized trial to compare the effect of electronic messaging and education (i.e., usual care) versus a multi-pronged "high touch" intervention to reduce vaccine hesitancy in skilled nursing facility (SNF) staff and residents among a random sample of facilities across four SNF chains. This will be a cluster randomized trial where the intervention is applied at the facility level. The primary interest is the effect of this intervention on SNFs that are characterized by having a relatively high proportion of residents who are Black or Latinx. The four SNF chains that have been selected have already given assent to participate in this trial. The facilities will then be stratified into three categories based on racial composition of residents: 1. \< 25% Black and Latinx residents 2. 25-40% Black and Latinx residents 3. \> 40% Black and Latinx residents Facilities will undergo constrained randomization within each chain and stratum to ensure that the proportion of Black and Latinx residents is balanced across the intervention arms. A total of 60 SNFs will be randomized to the intervention, allocated proportionally across the strata. The SNFs that were removed due to institutional instability will be compared separately to the control arm to assess potential bias due to selection into the study. Staff in facilities randomized to intervention arm will be informed by corporate leadership that they will be participating in a program to maximize COVID 19 vaccination among staff and residents. They will not be informed that this is part of a trial. Individual SNFs will be randomized to the intervention or usual care; randomization will be stratified by chain and by the proportion of minority residents based on three groups: \<25%, 25%-40%, \>40%. The research implementation team will not be masked to facility assignment. However, the PIs, the lead statistician, and programmers will be masked. All SNFs in the four chains are prepared to receive electronic messaging and educational material (i.e., usual care) through the American Health Care Association (AHCA). Within the four SNF chains that have agreed to participate in the trial, the investigators will ask the CEOs if there are any facilities that should be excluded due to leadership instability or other inability to participate in the multi-pronged intervention. Remaining facilities will be randomized to additionally receive the multi-pronged intervention versus continuing usual care. The entire trial will take place over 11-15 weeks: each facility in the high touch intervention will be involved in approximately 1-3 week start up activities, 6-8 weeks of vaccine administration (three scheduled deliveries for vaccine approximately 3-4 weeks a part), and 4 weeks of data collection. Facilities in the usual care group will follow the same timeline for enrollment and data collection. During the start-up period in the high touch facilities, the research team works with the leadership and opinion leaders in each SNF to optimize program roll-out within each unique environment. Among four SNF chains, the research team will randomize eligible facilities to receive an additional "high touch" intervention. These high touch facilities will receive the electronic messaging and educational material described above. In addition, these facilities will work with the research team on the following: 1. Facility Opinion Leader. At each intervention facility, the research team will work with the facility administration to identify local opinion leaders among nursing assistants (NA), nursing, dietary, and housekeeping. The opinion leaders will participate in the following activities: 1) Participate in an initial informational meeting with the research team and other facility opinion leaders; 2) Identify a local champion who could help participate in educational materials; 3) Participate in the social media messaging described in the Electronic Messaging section above; 4) Engage the research team for support and problem solving. The research team will invite all of the opinion leaders to participate in a one hour virtual informational meeting with members of the research team and other facility leaders. Meetings will be organized by discipline (e.g., nursing, dietary) and SNF chain. The research team will offer a few make-up sessions for staff who are unable to attend. During these meetings, basic information on vaccine safety and efficacy will be covered, leaving the majority of time for an open question and answer session. These sessions will NOT be recorded. Opinion leaders who participate in these meetings will be given a gift card for their time. Our research team will provide opinion leaders with direct contact information (email and phone number) of the research team so that they may ask questions during implementation. Insight Therapeutics will also work to identify a support team that can offer guidance and problem solve during implementation. 2. Consenting Specialist. Through Insight Therapeutics, the research team will employ external staff members to facilitate the clinical consent for vaccination process. Each facility will make up to ten referrals of residents who were not vaccinated during the first of the three available vaccine dates to the consenting specialists. Consenting specialists will contact each proxy, review risks and benefits of the vaccine, and answer questions. The research team will provide a 1-800 number for proxies who have additional questions/hesitancy and will offer a group Zoom call for interested proxies to review risks and benefits. As indicated, this consenting process will be a clinical consent for the vaccination itself - not a study-specific informed consent process to participate in research. The investigators received a waiver of informed consent for the overall intervention study. 3. Building Trust Locally. The facility opinion leaders will be encouraged to identify well respected persons in the community (e.g., minister, teacher, government leader) who are willing to provide a message promoting trust in the vaccine. Through Insight Therapeutics, the research team will reach out to these leaders and coordinate the video messages and implementation plan. Messages will be distributed widely within a facility by email, website, text and/or social media. Further, the research team will prepare the community leaders to serve as an additional support for the facility opinion leaders during implementation. 4. Positive Reinforcement. The research team will create and distribute buttons, T-shirts, and masks that promote awareness about vaccination (e.g., Ask me about the COVID-19 vaccine! OR Vaccinated for You!). These items will be distributed through facility leadership at each facility, with recommendations to give each staff member these goodies when vaccinated. 5. Testing Supplies. The research team will provide funds to each facility in the high touch intervention arm, that the facility may use to acquire additional COVID-19 testing kits. This will enable frequent testing of any residents and staff that experience symptoms following vaccination. Given the cost of most point-of-care testing kits, these funds will support the cost of approximately 200 test kits. The research team will suggest that facilities follow the CDC recommendations for testing following vaccination. The research team will additionally facilitate kits for facilities that are experiencing difficulty securing the test kits. The high touch intervention will be implemented in two waves. The first cycle of vaccine administration, the research team will focus on identifying opinion leaders and positive reinforcement. During the second round, the research team will add building trust locally, a consenting specialist, and testing supplies. Nursing home data are collected prior to the start of the study for descriptive purposes and to inform the development of a list of eligible facilities for recruitment. These include elements from Nursing Home Compare, including: the number of beds, hospital-based, special care dementia unit, nursing and nursing assistant hours/resident/day, and number of deficiencies on state inspections. Resident data is already being collected for all facilities within the four chains as part of the RADx-UP supplement. Existing data transfer agreements from all four chains have been signed and authorized. The research team plans to use data from the electronic medical record, as well as data from the Minimum Data Set (MDS) for this study. Resident characteristics will be obtained during baseline only (that is during the 3 months before the vaccine is first delivered to the facility) whereas vaccination data will be obtained during the 6-8 weeks of implementation and 4 weeks of follow-up. Resident and outcome data will be aggregated at the level of the facility. The hypothesis that will be tested is whether facilities that receive the high touch multi-pronged intervention will achieve a greater number of staff and residents vaccinated as compared with facilities randomized to usual care. The research team will use an intention-to-treat approach as the primary analytic approach, including all facilities that were randomized to the intervention regardless of implementation of the intervention components. Additional exploratory analyses will estimate a complier average causal effect to assess the effect of the intervention on those SNFs who fully engage in the intervention. A similar approach will be used to determine the effect of the high touch intervention on staff vaccination. The logistic model described for the primary analysis will be extended to include race and ethnicity indicators as well as interaction terms, to better understand if the treatment effect is heterogeneous across different subgroups of residents.