The utility of Geographical Information Systems (GIS) in systems-oriented obesity intervention projects: the selection of comparable study sites for a quasi-experimental intervention design--TX CORD

Abstract

Background: The Texas Childhood Obesity Research Demonstration project (TX CORD) uses a systems-oriented approach to address obesity that includes individual and family interventions, community-level action, as well as environmental and policy initiatives. Given that randomization is seldom possible in community-level intervention studies, TX CORD uses a quasi-experimental design. Comparable intervention and comparison study sites are needed to address internal validity bias.

Methods: TX CORD was designed to be implemented in low-income, ethnically diverse communities in Austin and Houston, Texas. A three-stage Geographical Information System (GIS) methodology was used to establish and ascertain the comparability of the intervention and comparison study sites. Census tract (stage 1) and school (stage 2) data were used to identify spatially exclusive geographic areas that were comparable. In stage 3, study sites were compared on demographic characteristics, socioeconomic status (SES), food assets, and physical activity (PA) assets. Student's t-test was used to examine significant differences between the selected sites.

Results: The methodology that was used resulted in the selection of catchment areas with demographic and socioeconomic characteristics that fit the target population: ethnically diverse population; lower-median household income; and lower home ownership rates. Additionally, the intervention and comparison sites were statistically comparable on demographic and SES variables, as well as food assets and PA assets.

Conclusions: This GIS approach can provide researchers, program evaluators, and policy makers with useful tools for both research and practice. Area-level information that allows for robust understanding of communities can enhance analytical procedures in community health research and offer significant contributions in terms of community assessment and engagement.

Figure 1.

Maps of TX CORD cities (Austin, Texas, and Houston, Texas), showing the boundaries of the intervention and comparison catchment areas—drawn as a continuous line (green=intervention and blue=comparison) around contiguous census tracts that fall in the higher quintiles of the composite index.

Figure 2.

Maps of TX CORD cities (Austin, Texas, and Houston, Texas), showing the attendance zones of the candidate schools. All the candidate schools, and their attendance zones, were selected from inside the catchment areas that were demarcated in stage I.