Best practices to reduce COVID-19 in group homes for individuals with serious mental illness and intellectual and developmental disabilities: Protocol for a hybrid type 1 effectiveness-implementation cluster randomized trial

Abstract

Background

People with serious mental illness (SMI) and intellectual disabilities and/or developmental disabilities (ID/DD) living in group homes (GHs) and residential staff are at higher risk for COVID-19 infection, hospitalization, and death compared with the general population.

Methods

We describe a hybrid type 1 effectiveness-implementation cluster randomized trial to assess evidence-based infection prevention practices to prevent COVID-19 for residents with SMI or ID/DD and the staff in GHs. The trial will use a cluster randomized design in 400 state-funded GHs in Massachusetts for adults with SMI or ID/DD to compare effectiveness and implementation of “Tailored Best Practices” (TBP) consisting of evidence-based COVID-19 infection prevention practices adapted for residents with SMI and ID/DD and GH staff; to “General Best Practices” (GBP), consisting of required standard of care reflecting state and federal standard general guidelines for COVID-19 prevention in GHs. External (i.e., community-based research staff) and internal (i.e., GH staff leadership) personnel will facilitate implementation of TBP. The primary effectiveness outcome is incident SARS-CoV-2 infection and secondary effectiveness outcomes include COVID-19-related hospitalizations and mortality in GHs. The primary implementation outcomes are fidelity to TBP and rates of COVID-19 vaccination. Secondary implementation outcomes are adoption, adaptation, reach, and maintenance. Outcomes will be assessed at baseline, 3-, 6-, 9-, 12-, and 15-months post-randomization.

Conclusions

This study will advance knowledge on comparative effectiveness and implementation of two different strategies to prevent COVID-19-related infection, morbidity, and mortality and promote fidelity and adoption of these interventions in high-risk GHs for residents with SMI or ID/DD and staff.

Clinical Trial Registration Number: NCT04726371.

1. Introduction

There are approximately 14.2 million adults with serious mental illness2 (SMI) and nearly 7.4 million individuals with an intellectual and/or developmental disability (ID/DD) in the US [[1], [2], [3]]. Factors that disproportionately increase the risk of COVID-19 among adults with SMI or intellectual and ID/DD include high rates of comorbid medical conditions, exposures to others in congregate living, and behavioral and cognitive risk factors [[4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21]]. In persons with SMI, chronic obstructive pulmonary disease, cardiovascular disease, and diabetes, are two to three times more prevalent than the general population, increasing the risk of hospitalization and death from COVID-19 [[4], [5], [6], [7], [8]]. In persons with ID/DD, higher rates of congenital heart disease, obesity, chronic lung disease, lower immune function, cancer, and diabetes, predispose to more severe outcomes from respiratory infections [[9], [10], [11], [12]]. Cognitive, behavioral, and/or physical disabilities also create challenges with adherence to hand hygiene, physical distancing, and use of face masks mitigating the risk of COVID-19 [[17], [18], [19]]. Residing in group homes (GHs) and other congregate living environments presents additional risk factors for developing COVID-19.

Staff in GHs risk exposure to SARS-CoV-2 infection associated with direct care of residents, working in multiple GHs or jobs, use of public transportation, and inadequate access to facemasks [22,23]. Front-line workers with limited English language proficiency may not have access to linguistically appropriate health information [24,25]. This combination of factors may elevate COVID-19 risk in staff of GHs.

Despite payment reforms and state mandated best practices for COVID-19 prevention in GHs [26,27], as of May 2020 during the early phase of the SARS-CoV-2 pandemic, rates of COVID-19 for residents with SMI or ID/DD were eight times higher (12%), and for staff two times higher (3%), compared to the general population in the surrounding “hot spot” communities (1.5%) selected for this study [[28], [29], [30]]. A lack of knowledge on how to best tailor COVID-19 prevention practices for vulnerable and underserved residents with special needs, and residential staff who are at risk for COVID-19 and severe disease, motivates this clinical trial. We describe a protocol for a hybrid type 1 effectiveness-implementation cluster randomized trial to test evidence-based COVID-19 prevention practices and the implementation of this intervention in GHs caring for individuals with SMI or ID/DD and residential staff.

2. Methods

2.1. Overview

This is a protocol for a hybrid type 1 effectiveness-implementation cluster randomized trial, to assess the effectiveness of evidence-based COVID-19 prevention practices tailored for adults with SMI and ID/DD and staff in GHs called “Tailored Best Practices” or TBP compared to “Generic Best Practices” (GBP) consisting of standard of care for adults with SMI and ID/DD and staff in GHs. Consistent with a hybrid type 1 effectiveness-implementation trial, we will simultaneously collect data on the process and outcomes of implementation [31]. Implementation and effectiveness outcomes at the GH-level will be assessed at baseline and every three months post-randomization up to a 15-month follow-up. This protocol was approved by the Institutional Review Boards of Mass General Brigham and Massachusetts Departments of Developmental Services (DDS) and Mental Health (DMH).

2.1.1. Study population and setting

Eligibility criteria are adults 18 years and older with serious mental illness (SMI) including schizophrenia-spectrum and affective disorders with persistent functional impairment [32] or intellectual and/or developmental disabilities (ID/DD) living in GHs. The GHs in this study are operated by six human service provider organizations in Massachusetts with 400 GHs with approximately 3300 staff and 2600 residents. GHs refer to group assisted living environments designed to provide both housing and support services 24 h a day to individuals with mental health conditions, or ID/DD. These homes are regulated by Massachusetts DDS and DMH. Residential care staff support residents in activities of daily living, teach skills needed for greater independence, promote community integration and inclusion, and facilitate access to needed behavioral and medical health care.

2.1.2. Site recruitment

The chief executive officers of each of the six provider organizations provided a letter of commitment to participate in the trial and written consent for their sites to be included in the study. To facilitate execution, each provider organization will hire a full-time project manager dedicated to fulfilling the objectives of the study.

2.2. Trial design

This study will employ a cluster randomized trial design randomized at the level of the GH. To ensure equal distribution of important factors between the two study arms, we will utilize a stratified block randomization scheme generated by a study investigator (HY) and stratified based on key variables. Randomization will be blinded and clusters randomized after recruitment. We will randomize approximately n = 200 GHs to the intervention, TBP, and approximately n = 200 GHs to the control comparison, GBP, equally divided by GHs for persons with SMI and ID/DD. Stratification factors will include race and ethnicity of GH staff and residents and a COVID-19 infection risk score based on ranking GHs (i.e., low, medium, high) by prior COVID-19 infections and vaccination of residents and staff (Appendix A). In a three-month observation period pre-randomization, we will assess the baseline use of preventive practices and rates of COVID-19 in each GH to distinguish the effect of the introduction of the TBP intervention. Fig. 1 (below) provides a schematic of the trial design.

Fig. 1.

Hybrid type 1 effectiveness-implementation cluster randomized trial design.

Abbreviations: ID/DD: Intellectual and/or developmental disabilities; SMI: Serious mental illness.

Within each site, we will conduct repeated measurement across six time points (baseline, 3-, 6-, 9-, 12-, and 15-months post-randomization) so that observed and latent time effects can be modeled in the presence of fluctuations in incidence over time.

2.2.1. Logic model

We developed a logic model as the theoretical framework to inform the hybrid trial design consistent with the format by Smith and colleagues (Fig. 2 ) [33]. This framework will support collection of data on the process and outcomes of the implementation and analysis plan and determinants of implementation: intervention characteristics, inner setting, outer setting, individuals involved, and the implementation process [33,34]. The determinants consist of factors that may facilitate or impede implementation based on characteristics of GHs for adults with SMI or ID/DD and key stakeholder input. The intervention tested in this randomized trial, called “Tailored Best Practices”, consists of a multi-component intervention including motivational interviewing, interactive education, trusted messengers, and measurement and feedback to improve uptake of screening, testing, vaccination, face mask use, and hand washing. The intervention, strategies, and outcomes are described below.

Fig. 2.

Logic Model for the Assessment of COVID-19 Effectiveness and Implementation Outcomes.

2.2.2. Community engagement and stakeholder feedback

The project will be guided by principles of community-engaged research and will employ a multi-stakeholder COVID-19 Quality Improvement Collaborative that we developed for this proposal to provide ongoing feedback [35,36]. Members of this collaborative include residents with SMI and ID/DD, caregivers/guardians, provider organization staff, clinicians, and public health and government officials (Appendix A).

2.3. Tailored best practices intervention

TBP consists of a multi-component intervention tailored to residents with SMI or ID/DD and the staff who work in GHs to encourage adherence to current best practices for prevention of COVID-19. The components of TBP were determined by the following data-informed, stakeholder-engaged approach. First, we conducted a review of the literature on interventions to promote adherence to prevention of COVID-19 and other respiratory infections relevant to individuals with SMI or ID/DD (Appendix B). Second, we solicited feedback from employees within participating provider organizations on their experiences with COVID-19 in GHs. Third, we conducted 36 key informant interviews with staff and 24 key-informant interviews with residents, to identify barriers and facilitators to implementing recommended best practices for COVID-19 prevention. Fourth, we used a validated simulation model [37] to determine the comparative effectiveness of different COVID-19 preventive practices (screening, isolation, contact tracing, use of personal protective practices (PPP), vaccination) based on previously collected GH data. Finally, we convened the COVID-19 Quality Improvement Collaborative, the multi-constituency stakeholder working group described above, to identify priorities for TBP.

Through this staged process, we selected the following four core components of TBP:

(1) Motivational Interviewing, (2) Interactive Education, (3) Trusted Messengers, and (4) Measurement, Feedback, and a facilitated House Plan for COVID-19 prevention practices including screening, testing, use of face masks by staff, hand hygiene, and vaccination. The four components of the TBP intervention are described in detail below.

2.3.1. Motivational interviewing

Motivational interviewing (MI) is effective in promoting health behavior change in individuals with SMI [38,39]. MI also has the potential to reduce vaccine hesitancy [40,41]. We adapted principles of MI to focus on COVID-19 prevention practices. The interventionist delivering MI is a staff member, referred to here as a coach, who is hired and primarily supervised by the community-based provider organization. The coaches will be trained in MI and subsequently supervised in biweekly sessions by a board certified adult psychiatrist (SB), with extensive experience providing training in MI in the context of multi-component interventions. MI training will include eliciting personal goals and discussing wellness behavior practices, COVID-19 prevention, tradeoffs of different prevention approaches, and aligning goals for COVID-19 prevention with overall personal goals. Coaches will deliver MI sessions primarily in small groups with separate sessions for residents and staff. GH Program Directors will encourage residents with SMI and staff (serving residents with SMI or ID/DD) to participate in an MI session within the first eight months post-randomization. A follow-up session may occur in the latter seven months of the intervention period if necessary. Residents with ID/DD will not receive the MI component of TBP due to a lack of supporting evidence of effectiveness for individuals with ID/DD requiring the need for adaptation and validation in this population [42].

2.3.2. Interactive education

Interactive education can serve as an important adjunct in health behavior change [[43], [44], [45]]. The Interactive Education component of our TBP will be available to residents and staff in multiple formats to accommodate the needs and preferences of each GH. The research team will create a COVID-19 Educational Toolkit — a collection of materials (e.g., fact sheets, videos, social stories) curated by COVID-19 physician experts. The Educational Toolkit will contain materials that are tailored to people with SMI or ID/DD, non-English language speakers, and people identifying as racial and ethnic minorities. Materials will use simple messaging, low text, and imagery reflecting a multicultural audience, with multilingual versions available [46]. Throughout the study, the toolkit will be amended with evidence-based emergent information on COVID-19 prevention. Materials will be vetted by stakeholder working groups to provide input on appropriateness and accuracy. Residents with SMI and staff will receive Interactive Education during MI sessions. Because residents with ID/DD do not participate in MI sessions, Program Directors will present materials from the toolkit to the residents as Interactive Education. Residents and staff will also have the opportunity to have small group Interactive Education sessions with a COVID-19 physician expert specifically devoted to answering health questions related to COVID-19. All residents and staff will receive Interactive Education within the first eight months of the intervention and an optional “booster session” thereafter based on need indicated by intervention fidelity measures.

2.3.3. Trusted messengers

Trusted messengers, including peers with lived experience, have been demonstrated to be an effective source of health promotion, since personal narratives can support health behavior change when the audience identifies with the message delivered [47,48]. The Trusted Messenger component of TBP will be available in a peer testimonial video or an in-person peer testimonial during a group MI session. Since video testimonials from peers provide an easy to access and scalable format, we will create five peer testimonial videos featuring frontline staff, Certified Peer Specialists, residents, and legal guardians conveying their personal stories with COVID-19 and emphasizing why they chose to receive a COVID-19 vaccine. Residents and staff will view these videos during group MI sessions. In-person trusted messenger sessions consist of the staff, managers, and residents—who have on-going and familiar personal relationships in the GH—sharing their personal stories and experiences related to receiving vaccinations. Coaches will leverage the role of peers to provide trusted information on the potential benefits of TBP.

For individuals with ID/DD or SMI who have legal guardians, coaches will offer a “peer testimonial” video of a family member's perspective on COVID-19 vaccination as a “Trusted Messenger” component. In addition to Trusted Messenger videos, coaches will collaborate with GH Program Directors to identify resident and staff peers who may be willing to share their personal experiences during MI sessions and thus function as supporting “trusted messengers.” All residents and staff will receive a form of Trusted Messenger within the first eight months of the intervention and an optional “booster session” thereafter based on need indicated by intervention fidelity measures.

2.3.4. Measurement, feedback, and facilitated house plan to improve Fidelity to COVID-19 prevention practices

Audit and feedback is an effective approach to facilitate chronic disease management and uptake of infectious disease testing [49,50]. At baseline, 3-, 6-, 9, 12-, and 15-months post-randomization all Program Directors will complete a survey to assess the results of the GH's COVID-19 Best Practices Fidelity, a primary implementation outcome of this study. The survey consists of questions on the frequency and number of staff and residents participating in COVID-19 prevention practices, including screening, testing, use of face masks, physical distancing, handwashing, and vaccination. GHs follow U.S. Centers for Disease Control guidelines for physical distance of six feet between individuals for the prevention of COVID-19 [51]. Mechanisms for maintaining this physical distance in GHs may include floor markers indicating physical distance standards during routines (e.g., waiting on line for medication administration or food or using bathroom or laundry facilities). Following completion of the survey, coaches will provide Program Directors a numeric dashboard (Fig. 3 ) to provide data-based feedback with respect to accomplishments and areas in need of improvement to optimize fidelity to TBP. Each visual display will include a numeric summary of the GH's fidelity performance in relation to the average performance of other GHs in their organization. After receiving the statistical dashboard, each GH will receive assistance from the coaches in creating a GH-specific “House Plan” with actionable steps to improve fidelity performance. Coaches will work with the Program Directors to review areas that need improvement, discuss barriers to increasing the GH's success in adherence to COVID-19 prevention measures, set attainable goals, and identify target completion dates.

Fig. 3.

COVID-19 Best Practices Fidelity Dashboard.

2.3.5. Implementation strategies

As shown in Fig. 2, we will implement TBP using established implementation strategies including community and stakeholder engagement, external facilitation by “facilitation coaches,” and internal facilitation by GH program directors based on the Expert Recommendations for Implementing Change (ERIC) taxonomy [52]. Community and stakeholder engagement will facilitate organization buy-in to eliminate barriers and maximize facilitators. External facilitators will be the “Implementation Coaches” from the community-based research team. As described by Ritchie et al. (2017), facilitation is “a multi-faceted process” that uses “implementation science knowledge and interventions to help and enable others to understand what they need to change, plan and execute changes, and address barriers to change efforts” [53]. External facilitation will motivate behavioral change at the level of staff and residents in GHs so that the intervention is delivered within the local context of the organization and GH. Internal facilitation will be provided by GH Program Directors. Coaches and Program Directors will meet regularly to discuss updated statistical dashboards and review progress made towards reaching pre-determined goals established in the home-specific “House Plan.”

2.4. Generic best practices comparator

The GBP control condition will receive standard recommended best practices for preventing COVID-19 based on recommendations by the CDC and policies of the Massachusetts Executive Office of Health and Human Services. GBP will be implemented by (1) distribution of standard guidelines from the Massachusetts Executive Office of Health and Human Services for public health prevention and management of COVID-19 [27], and (2) standard virtual training of the staff of the GHs in COVID-19 prevention practices outlined by the Office of Health and Human Services. Over the course of the study, the GBP condition will reflect up-to-date recommendations consistent with CDC and mandates from Massachusetts state policy.

2.5. Outcomes and measures

Effectiveness and implementation outcomes are summarized in Table 1 , evaluated at baseline, 3-, 6-, 9-, 12-, and 15-months post-randomization and analyzed at the level of the GH.

Table 1.

Summary of primary and secondary effectiveness and implementation outcomes for a hybrid type 1 effectiveness-implementation cluster randomized control trial of a multi-component intervention to prevent COVID-19 in residents with serious mental illness and intellectual and developmental disabilities and staff based in GHs in Massachusetts.

Outcomes	Description	Source	Frequency
Effectiveness Outcomes
Primary effectiveness outcome: COVID-19 Infection	New incidence of COVID-19 from laboratory testing of residents and staff	Secondary data routinely collected by the 6 participating provider organizations	Every 3 months beginning at baseline
Secondary effectiveness outcome: Hospitalization	New hospitalizations of GH residents due to COVID-19	Secondary data routinely collected by the 6 participating provider organizations	Every 3 months beginning at baseline
Secondary effectiveness outcome: Mortality	New deaths of GH residents due to COVID-19	Secondary data routinely collected by the 6 participating provider organizations	Every 3 months beginning at baseline
Implementation Outcomes
Primary implementation outcome: COVID-19 Best Practices Fidelity	A score calculated based on GH engagement in up to 9 COVID-19 prevention activities (described below)	Surveys of GH Program Directors	Every 3 months beginning at baseline
Primary implementation outcome: COVID-19 Vaccination	Receipt of COVID-19 vaccination	Secondary data routinely collected by the 6 participating provider organizations	Every 3 months beginning from initial introduction of the COVID-19 vaccine
Secondary implementation outcome: Adoption	Score calculated using Acceptability, Appropriateness, and Feasibility measures informed by the RE-AIM implementation framework [54]	Surveys of GH Program Directors	Every 3 months beginning at baseline
Secondary implementation outcome: Adaptation	Open-ended responses on changes made to the TBP or GBP and the reasons for adaptation	Surveys of GH Program Directors	Every 3 months beginning at baseline
Secondary implementation outcome: Reach	The percentage of GHs with at least 80% COVID-19 Best Practices Fidelity	Surveys of GHs Program Directors	Every 3 months beginning at baseline
Secondary implementation outcome: Maintenance	The percentage of GHs maintaining at least 80% COVID-19 Best Practices Fidelity at 15-month follow-up	Surveys of GH Program Directors	Once at the end of the 15-month trial

2.5.1. Effectiveness outcomes and measures

The primary effectiveness outcome will be COVID-19 incidence defined as new laboratory-confirmed SARS-CoV-2 cases among residents and staff in GHs. Secondary effectiveness outcomes will be GH rates of COVID-19-related hospitalization and mortality in residents. These data will be collected from standardized, routine reporting of these indices by each of the provider organizations.

2.5.2. Implementation outcomes and measures

Implementation outcomes will be informed by the RE-AIM Framework (Reach Effectiveness, Adoption, Implementation, and Maintenance) [55,56]. The primary implementation outcomes measure uptake of Best Practices for COVID-19 prevention as dictated by the Massachusetts Executive Office of Health and Human Services. The nine core elements of Best Practices for COVID-19 prevention are: 1) Screening; 2) Testing; 3) Use of Face Masks; 4) Physical Distancing; 5) Hand Hygiene; 6) Cleaning of the Physical Environment; 7) Staff Use of Gowns and Gloves; 8) Vaccination; and 9) Education and Training in Screening and Testing, PPP, and Vaccination. Fidelity will be measured by the COVID-19 Best Practices Fidelity Scale developed for this project. We followed a priori methodology for constructing fidelity scales for evidence-based practices, and the final instrument will be refined with input from key stakeholders [57]. The fidelity scale will assess adherence to the delivery of nine core elements of Best Practices for COVID-19 prevention. In addition to overall fidelity to the nine core practices rated by surveying the program directors, we will also evaluate implementation effectiveness for the primary prevention target of vaccination of GH residents and staff as a co-primary implementation outcome based on electronically collected data from the participating six agencies. The co-primary implementation outcome of COVID-19 vaccination rates reflects findings from our simulation model and CDC recommendations recognizing vaccination as the most important measure in preventing COVID-19, as well as the added value of collecting validated numerical rates of vaccination documented by the provider organizations in addition to survey-based data rated by providers.

A secondary implementation outcome will be Best Practices Adoption, evaluated using Acceptability, Appropriateness, and Feasibility measures of Best Practices for COVID-19 prevention (See Appendix B) [54]. Program directors of GHs will be assessed on fidelity, adoption, and adaptation of the Best Practices for COVID-19 prevention. Evaluation of Adaptation will consist of open-ended responses by program directors on changes they made to the TBP or GBP and the reasons for adaptation. Reach is defined as the percentage of GHs with a score of at least 80% on the Best Practices Fidelity scale. Maintenance is the percentage of GHs maintaining a score of at least 80% on the Best Practices Fidelity at 15-month follow-up. At each data collection point, Reach will be assessed using the Best Practices Fidelity ratings from the Program Director surveys. Maintenance will be calculated using these same ratings following the 15-month follow-up period.

2.6. Sample size, power, and analytic plan

A generalized linear mixed-effects model (GLMM) will be used for analyses. Group (TBP vs. GBP), time (baseline, 3-, 6-, 9-, 12- and 15-months), and group*time interaction will be specified as fixed effects. Intercept and slope at GH level will be specified as random effects to take the intraclass correlation coefficient (ICC) due to repeated measures into account. The treatment effect (i.e., effectiveness of TBP compared to GBP) will be reflected by group*time interaction. To ensure causal relationship between treatment models (TBP vs. GBP) and outcomes, any baseline characteristics associated with both treatment and outcomes will be included in the model as covariate. A stratification factor (baseline COVID-19 incidence) will be included as a covariate. For continuous measures (e.g., Best Practices Fidelity and COVID-19 vaccination rates), identity link and normal distribution will be specified. For COVID-19 incidence, hospitalization, and mortality rates, log link and Poisson or negative binomial distribution (e.g., log-linear model with offset term) will be specified. An additional evaluation of fidelity will include a dichotomous outcome of a threshold of achieving at least 80% Best Practices Fidelity, logit link and binomial distribution will be specified. Subgroup analyses will examine outcomes by type of facility (SMI vs. ID/DD). We will explore outcome differences related to the proportion of ethnic/racial minorities by facility. We will test group*time*facility and group*time*facility*race/ethnicity interactions, followed by subgroup stratified analyses.

For power calculations (Table 2 ) we assume a 2-tailed test with significance level 0.05, intraclass correlation (ICC) of 0.02 to account for the nested effect of the stratification factors of highest and lowest prevalence of COVID-19 infection within GBP and TBP within four administrative groupings of group homes, and ICC of 0.50 for repeated measures at the GH level [58]. For the primary implementation outcome, the continuous Best Practices Fidelity score, we will have 80% power to detect minimum 0.21 standardized mean difference between intervention and control (Cohen's d = 0.21). For a dichotomous fidelity threshold of 80% Best Practices Fidelity, we will have 80% power to detect 7.2% difference (Cohen's d = 0.33). Given COVID-19 incidence as of May 2020 rates based on available COVID-19 prevalence data for GHs residents and staff, we will have at least 80% power to detect 6–10% percentage point difference of COVID-19 infection between GBP and TBP (Table 2). We use conservative assumptions, given the possibility of undetected COVID-19 in GHs. The primary outcome will be COVID-infections in both staff and residents and correction for type I error will be controlled by Bonferroni's method [59]. We plan to examine outcomes among staff versus residents and SMI versus ID/DD homes in subgroup analyses.

Table 2.

Power calculations.

Population	Estimated COVID-19 Prevalence		Difference	Cohen's da	Power
	Control (GBP)	Intervention (TBP)
Residents with SMI & Residents with ID/DD combined	26.5%	17.8%	8.7%	0.28	80%
Group home staff	13.1%	6.8%	6.3%	0.38	80%

^aSmall effect = 0.2; Medium effect = 0.5 [60].

3. Discussion

Individuals with ID/DD or SMI and front-line staff in GHs are disproportionately at-risk of COVID-19 infection and compromised health outcomes. To address this prevention challenge, we will evaluate the effectiveness and implementation of a COVID-19 prevention intervention tailored to individuals with ID/DD and SMI in GHs. We anticipate a variety of challenges in conducting and evaluating the outcomes of this study including the special characteristics of the study sample, the GH setting, and the evolving nature of the COVID-19 pandemic. Our study sample includes residents of GHs with SMI, residents with ID/DD, and the staff of the GHs. In aggregate, our study sample presents the advantage of a generalizable array of providers and residents in public sector behavioral health residential care. However, we anticipate heterogeneity in outcomes that will require assessment in subgroup analyses. Including group home staff in the study sample allows for observation of a critical component of prevention of COVID-19 as potential vectors and as providers of key prevention measures. However, workforce shortages and high rates of staff turnover present a substantial challenge to delivering the TBP intervention and to assessment [61]. High staff turnover in the GHs may impede the ability to deliver the intervention longitudinally and track outcomes as staff move in and out of homes.

The evolving nature of the COVID-19 pandemic with respect to incidence, viral variants, recommended preventive interventions, and the introduction of vaccines during the study timeframe also provides novel methodological challenges, while also providing an opportunity for a dynamic, real-world study in infectious disease prevention. Accordingly, we have selected a pragmatic design to allow for adaptations in trial procedures and intervention content that is responsive to a dynamic pandemic environment and includes the changing epidemiology of SARS-CoV-2, knowledge of COVID-19, state and national policies, and the environmental context [62]. Accordingly, we defined the primary implementation outcome, COVID-19 Best Practices Fidelity, to be a dynamic measure with a flexible denominator, including only the COVID-19 prevention practices that are relevant at the measurement time points. In addition, during the planning phase of the intervention development, we anticipated the introduction of COVID-19 vaccination by adding this strategy as soon as Phase II/III data were available for COVID-19 vaccinations in 2021 in the United States. Based on our validated simulation model determining that vaccination is the most effective strategy for preventing COVID-19, we designated COVID-19 vaccination as a co-primary implementation outcome (in addition to the aggregate fidelity measure of the 9 prevention practices). This presented the added advantage of data based on actual rates of vaccination documented by the provider organizations in addition to survey-based data rated by providers [63,64]. At the same time, with increased availability and uptake of COVID-19 vaccinations during implementation of this protocol, we may find limited ability to detect a difference in COVID-19 effectiveness outcomes.

The design of this study evaluates outcomes for Tailored Best Practices (TBP) delivered with a specific implementation strategy (community and stakeholder engagement with internal and external facilitation) compared to Generic Best Practices (GBP) including usual implementation. As such we will not be able to distinguish the effect of the intervention and implementation strategies. We recognize that a study design with a third arm potentially could provide the capacity to differentiate the effect of the intervention from the implementation strategy. However, a three-armed randomized trial was not feasible due to the urgency and imperative expressed by our community partnering organizations to implement the most robust implementation and intervention aimed at reducing incidence and mortality from COVID-19 in as many GHs as possible; combined with the additional required number of GHs that would be required to achieve statistical power beyond a practical capacity to conduct the study.

Major strengths of this study include the hybrid type 1 effectiveness-implementation design evaluating TBP effectiveness compared to GBP, in conjunction with detailed collection of implementation process and outcome measures for the intervention. The methodology is informed by community-based participatory research and as such we will incorporate key stakeholders including residents, staff, caregivers, and public health and government officials to optimize intervention acceptability and sustainability.

4. Summary

People with serious mental illness (SMI) and intellectual disabilities or developmental disabilities (ID/DD) living in GHs (group homes) and residential staff are at higher risk for COVID-19 infection, hospitalization, and death compared with the general population.

This two-arm effectiveness-implementation trial will assess evidence-based infection prevention practices to prevent COVID-19 for residents with SMI or ID/DD and the staff in GHs in a hybrid type 1 effectiveness-implementation cluster randomized design consisting of 400 state-funded GHs for adults with SMI or ID/DD by comparing “Tailored Best Practices” (TBP) to “General Best Practices” (GBP). This study will advance knowledge on the effectiveness of two different interventions to prevent COVID-19-related infection, morbidity, and mortality and related implementation strategies to promote fidelity and adoption of these interventions in high-risk GHs for residents with SMI or ID/DD and staff with potential implications for optimal infection preventive practices in GHs for vulnerable populations across the nation.

4.1. Institutional ethics approval and informed consent

This study has been approved by the Mass General Brigham Institutional Review Board, the Massachusetts Department of Mental Health Institutional Review Board, and the Massachusetts Department of Developmental Services Research Review Committee. These entities have also granted a waiver of consent from residents and staff for COVID-monitoring data routinely collected by the participating provider organizations. Provider organization Chief Executive Officers have provided written consent for their sites to be included in the study. We will obtain informed consent from all survey participants or their legal guardians, where applicable, prior to the administration of study surveys.

Author contributions statement

Levison: Conceptualization, Methodology, Investigation, Resources, Writing - Original Draft, Writing - Review & Editing, Supervision, Project administration; Krane: Conceptualization, Methodology, Investigation, Resources, Writing - Original Draft, Writing - Review & Editing, Visualization; Donelan: Methodology, Investigation, Resources, Writing - Review & Editing; Aschbrenner: Conceptualization, Methodology, Resources, Writing - Review & Editing; Trieu: Conceptualization, Methodology, Investigation, Resources, Writing - Review & Editing, Supervision, Project administration; Chau: Conceptualization, Investigation, Resources, Writing - Review & Editing; Wilson: Conceptualization, Methodology, Software, Investigation, Resources, Data Curation, Writing - Review and Editing; Oreskovic: Conceptualization, Writing - Review & Editing; Irwin: Conceptualization, Writing - Review & Editing; Iezzoni: Writing - Review & Editing; Xie: Methodology, Writing - Review & Editing; Samuels: Writing - Review & Editing; Silverman: Conceptualization, Writing - Review & Editing; Batson: Conceptualization, Writing - Review & Editing; Fathi: Conceptualization, Writing - Review & Editing; Gamse: Conceptualization, Writing - Review & Editing; Holland: Conceptualization, Project administration; Wolfe: Conceptualization, Writing - Review & Editing, Supervision, Project administration; Shellenberger: Conceptualization, Writing - Review & Editing, Supervision, Project administration; Cella: Conceptualization, Writing - Review & Editing, Supervision, Project administration; Bird: Conceptualization, Writing - Review & Editing, Supervision, Project administration, Funding acquisition; Skokto: Conceptualization, Writing - Review & Editing, Supervision, Project administration, Funding acquisition; Bartels: Conceptualization, Writing - Review & Editing, Supervision, Project administration, Funding acquisition.

Declaration of Competing Interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

Dr. Skotko occasionally consults on the topic of Down syndrome through Gerson Lehrman Group. He receives remuneration from Down syndrome non-profit organizations for speaking engagements and associated travel expenses. Dr. Skotko receives annual royalties from Woodbine House, Inc., for the publication of his book, Fasten Your Seatbelt: A Crash Course on Down Syndrome for Brothers and Sisters. Within the past two years, he has received research funding from F. Hoffmann-La Roche, Inc., AC Immune, and LuMind Research Down Syndrome Foundation to conduct clinical trials for people with Down syndrome. Dr. Skotko is occasionally asked to serve as an expert witness for legal cases where Down syndrome is discussed. Dr. Skotko serves in a non-paid capacity on the Honorary Board of Directors for the Massachusetts Down Syndrome Congress and the Professional Advisory Committee for the National Center for Prenatal and Postnatal Down Syndrome Resources. Dr. Skotko has a sister with Down syndrome.

Appendix A. Supplementary data

Supplementary data: Appendices A and B

Data availability

No data was used for the research described in the article.