Best Practices to Prevent COVID-19 Illness in Staff and People With Serious Mental Illness and Developmental Disabilities in Congregate Living Settings

Persons with Serious Mental Illness (SMI) and Intellectual Disabilities and Developmental
Disabilities (ID/DD) are disproportionately vulnerable to COVID-19 for three reasons: (1)
Medical vulnerability. Smoking, chronic obstructive pulmonary disease, cardiovascular
disease, and diabetes all increase COVID-19 mortality and are about 2-3 times more prevalent
among persons with SMI. People with ID/DD suffer higher rates of COVID-19 risk factors,
including pre-existing chronic conditions, heart defects, obesity, chronic respiratory
problems or lung disease, lower immune function, cancer, and diabetes. (2) Residential
vulnerability. The congregate care settings in which many people with SMI and ID/DD live
carry many of the same higher risks of COVID-19 transmission currently affecting
assisted-living settings and nursing homes across the nation. (3) Health behavior
vulnerability. Some people with SMI and ID/DD have cognitive, behavioral, and physical
challenges that heighten COVID-19 risk by hampering personal protective practices (PPP)
(i.e., hand hygiene, physical distancing, use of face masks). Moreover, the staff who work in
congregate care settings are often subject to high rates of exposure, have low socioeconomic
status, use public transportation, and lack personal protective equipment. This collection of
factors contributes to an extraordinarily high risk of COVID-19 morbidity, and mortality.
Despite payment reforms and mandated best practices for COVID-19 for congregate care by the
MA Department of Public Health, rates of coronavirus illness for residents with SMI and ID/DD
are 8 times higher (12%), and for staff 2 times higher (3.0%), compared to the general
population in the surrounding "hot spot" communities (1.5%) selected for this study. This
tragic health disparity confirms that key decision-makers lack the knowledge of how to
optimally tailor best practices for this highly vulnerable population and the staff who
provide their care to effectively reduce their high risk of COVID-19 and COVID-19 related
mortality.

The investigators' overall goal is to reduce COVID-19 and other infectious-disease incidence,
hospitalizations, and mortality among staff and adults with Serious Mental Illness (SMI) and
Intellectual and Developmental Disabilities (ID/DD) in congregate-living settings in
Massachusetts. The investigators address 2 comparative-effectiveness questions:

With the goal of prioritizing and resourcing actionable best practices: What is the
comparative effectiveness of different types and intensities of five basic preventive
interventions-screening, isolation, contact tracing, personal protective practices (PPP)
(i.e., hand hygiene, physical distancing, use of face masks), and vaccination-in reducing
rates of COVID-19, hospitalizations, and mortality in staff and adult residents with SMI and
ID/DD in congregate care settings? Effectiveness Hypothesis: Of the 5 preventive
interventions, tailored screening, effective use of isolation, and increased vaccination
acceptance will be associated with the greatest reduction of staff and resident COVID-19
rates and related hospitalizations.

With the goal of effectively implementing best practices: What is the most effective
implementation strategy to reduce rates of COVID-19 in congregate care settings for persons
with SMI and ID/DD: (1) Tailored Best Practices (TBP) specifically adapted for staff and
residents with SMI and ID/DD in congregate living settings or (2) Generic Best-Practices
(GBP) consisting of state and federal standard guidelines for all congregate care settings?
Implementation Hypothesis: TBP will be associated with greater implementation fidelity and
lower staff and resident rates of COVID-19 and hospitalization than GBP.

To test these hypotheses, the investigators will pursue three Aims:

Aim 1: The investigators will synthesize existing data collected by the six provider
organizations on (1) rates of COVID-19, hospitalization, and mortality and (2) use of
screening, isolation, contact tracing, PPP, and vaccination in 400 group homes for SMI and
ID/DD. The investigators will also collect qualitative data through surveys and virtual focus
groups on the experience of staff and residents, and on barriers and facilitators to
implementing recommended practices.

Aim 2: The investigators will determine the comparative effectiveness of different COVID-19
preventive practices (screening, isolation, contact tracing, use of PPP, vaccination) by
populating a validated simulation model and engage decision makers and stakeholders in
selecting priorities for best practices. The investigators will apply the Clinical and
Economic Analysis of COVID-19 Interventions (CEACOV), a COVID-19 simulation model that has
already been developed and validated, to simulate the 2,050 residents and 3,300 staff of the
400 group homes for persons with SMI and ID/DD in the study. The investigators will compare
the effectiveness of different types, intensities, and combinations of the five identified
interventions: screening, isolation, contact tracing, PPP, and vaccination.

On completion of the simulation modeling at month 3, the investigators will convene a
COVID-19 Quality Improvement Collaborative (CQIC) Virtual Summit including consumers,
providers, advocates, and key decision makers, during which the investigators will summarize
the results of the comparative effectiveness simulation model and present alternative
scenarios demonstrating the impact of increasing or decreasing amounts or combinations of
various practices. The CQIC Virtual Summit and 1-2 additional brief virtual meetings will
result in recommendations for a final prioritized set of actionable and feasible Tailored
Best Practices (TBP) for implementation. The CQIC will review and adapt relevant COVID-19
training materials and finalize a TBP implementation package.

Aim 3: The investigators will employ a cluster randomized trial design with 200 group homes
randomized to implementation of "Tailored Best Practices" (TBP) compared to 200 group homes
randomized to "Generic Best Practices" (GBP). The investigators will engage in a three-month
observational pre-randomization period to assess the baseline use of preventive practices and
rates of COVID in each group home to better isolate the effect of the introduction of the TBP
intervention. The investigators have also incorporated a 3-month implementation phase in
order to accommodate implementing the TBP intervention with fidelity across 200 group homes
during the same time period. The best practice implementation fidelity and COVID-19 incidence
are co-primary outcomes with group home as the unit of analysis. Within each site, the
investigators will engage in repeated measurement of these outcomes across 6 time points
(Baseline, 3, 6, 9, 12, and 15 months) so that time effects (observed and latent) can be
modeled precisely in the presence of any fluctuations in incidence over time.

By month 16 of the project, the investigators will know the effectiveness of the Tailored
Best Practices (TBP) implementation for a diverse array of state-supported group homes for
SMI and ID/DD with broad generalizability to similar settings across the nation that will be
broadly distributed through dissemination materials at the end of the project.