Objective: To explore the joint effect of distance and neighborhood-level demographics on pet adoptions from an animal shelter by means of client segmentation, geospatial tools, and epidemiological techniques.
Sample population: Geocoded locations of 1,563 adoptions from an animal shelter in eastern Massachusetts and 4,700 geospatial control locations.
Procedures: Pet adoptions were geocoded to addresses by use of spatial analysis software for area-based analysis. A population of neighborhood control locations (proportional to human population at the level of the block group) was randomly generated. Adoption and control locations were spatially joined to census data and to defined lifestyle segments. The association of geodemographic variables with adoption was assessed by use of multivariate logistic regression.
Results: The shelter service area was not well described by individual geopolitical boundaries such as county. Locations that were < 9.7 km (6 miles) from the shelter, had higher block group median income, or had a greater proportion of households that were composed of married couples with children in the block group were associated with increased odds of being an adoption location. Neighborhood lifestyle characteristics affected the association between affluence and adoption. Controlling for distance of a location to the shelter markedly affected the strength of association with adoption.
Conclusions and clinical relevance: Results suggested that distance as well as neighborhood lifestyle characteristics profoundly influences shelter adoption patterns. Client segmentation and geodemographic analysis can reveal communities with increased likelihood of pet adoption from a shelter and help clarify adoption patterns.