Using geospatial analysis to monitor health disparities in people and cats across urban neighborhoods

Background: Premature mortality is a strong marker of health disparities in the human population. Area-based analyses have been used to identify neighborhood-level disparities in human health, but have not yet been used to assess whether similar disparities exist in companion animal populations.

Purpose: To assess shelter-associated mortality in the cat population using animal shelter data from Boston, MA and to compare this with human premature mortality and socioeconomic characteristics, by neighborhood.

Methodology: Geocoded addresses from 15,285 cat intake records (2004–2008) were spatially joined to neighborhood and census tract polygons using Arc-GIS software. Shelter-associated deaths were calculated per capita, and compared with previously published data on human premature mortality. Poisson mixed-effects models were used to smooth mortality rates and calculate relative risks.

Results: 14 to 40–fold gradients in cat mortality were detected across 16 neighborhoods (0.27–3.9 cats/1,000 persons) and 156 census tracts (0.15–6.47 cats/1,000 persons) of Boston. Shelter-associated cat deaths were significantly correlated with human premature mortality at the neighborhood level (R2=0.50, P=0.002), and with a variety of quantitative indicators reflecting socioeconomic deprivation. One neighborhood with cat mortality 50% lower than expected was served by a low-cost clinic providing subsidized preventive veterinary care and pet sterilization; the clinic also supported trap-neuter-release programs.

Conclusions: Health disparities were detected in the feline population, by neighborhood, and the geographic distribution was similar to that reported for people in an urban setting. Geospatial techniques can precisely identify areas of greatest need to better target intervention programs to improve cat welfare.