JMIR Publications

ABSTRACT

Background:

Geospatial data science can be a powerful tool to aid the design, reach, efficiency, and impact of community-based intervention trials. The project titled Take Care Texas aims to develop and test an adaptive, multilevel, community-based intervention to increase COVID-19 testing and vaccination uptake among vulnerable populations in three Texas regions: Harris County, Cameron County, and Northeast Texas (NETX).

Objective:

We developed a protocol for adaptive selections of census block groups (CBGs) to include in this randomized trial from the three regions in a 17-month period from May 2021 to October 2022. The protocol integrated real-time SARS-CoV-2 infection data and contextual social determinants of health (SDOH), known to contribute to health inequities among underserved populations. We aimed to identify and select priority CBGs (PBGs) for inclusion in the community-based randomized trials. Due to the dynamic nature of infections and vaccinations over time, the selection of the priority CBGs was conducted in 10 rounds.

Methods:

We developed the CBG selection criteria based on the COVID-19 burden quantified from SARS-CoV-2 population-level surveillance data, and the community disparity index developed in this study using 12 SDOH measures from US Census data. In each round, PBGs with high COVID-19 burden and the highest community disparity index were identified, from which we further selected geographically non-adjacent sets of CBGs with suitable separation to mitigate potential intervention “spillover”. Candidate CBGs were further chosen based on the input from community partners and community health workers. Selected CBGs were then randomly sampled and assigned equally to two intervention arms of MILI (multi-level intervention) and JITAI (just-in-time adaptive intervention), and one control arm using covariate adaptive randomization at 1:1:1 ratio. These CBGs were mapped on dashboards with population-level data to inform intervention delivery and referral resources.

Results:

A total of 120 CBGs were selected and followed the stepped planning and interventions, with 60 in Harris, 30 in Cameron, and 30 in Northeast Texas (NETX) counties, respectively. COVID-19 burden presented substantial temporal changes, as well as local variations at CBGs. At the CBG level, COVID-19 burden and the community disparity shared similar geographic patterns, but the patterns changed over time during the study period.

Conclusions:

The proposed novel protocol integrated real-time surveillance data and geospatial data science to support the community-based randomized trial design and inform intervention adaptation and delivery within the CBGs. Adaptive selection effectively prioritized the most-in-need communities and allowed for a rigorous evaluation of community-based interventions in a multilevel trial. The proposed protocol can be adapted to other communicable disease control and prevention programs to improve population health and reduce disease burden in areas with disparities.