December 18, 2019

Within cities there can be stark differences in health across neighborhoods. Often, surveys of city residents are conducted to provide information related to health and then reported to the public at the city level. However, understanding health characteristics within smaller areas of a city, such as by census tract, zip code or neighborhood allows researchers, city officials, and public health practitioners to identify areas of high need and allocate resources. However, health data available for small areas is often limited because few events occur or because the number of survey respondents  living in the area can be small. Small area estimation techniques are a way to overcome problems from smaller sample sizes to provide more reliable estimates.

The Drexel Urban Health Collaborative (UHC) utilizes a method of small area estimation called Bayesian hierarchical models that employs both spatial and temporal smoothing to borrow strength from nearby neighborhoods and also from the same neighborhood across time. This method was developed by Harrison Quick, assistant professor of Epidemiology and Biostatistics at the Dornsife School of Public Health and recently published in a paper “Trends in Tract-Level Prevalence of Obesity in Philadelphia by Race-Ethnicity, Space, and Time” in the journal Epidemiology.

Researchers at the Urban Health Collaborative developed a companion methods brief to provide examples of the small area estimates method and their utility. The UHC publishes methods briefs to provide leadership to the broader community of scholars looking to create a rigorous urban health evidence base. These briefs highlight best practices and emerging directions for collecting, analyzing, and presenting urban health data.

UHC has produced small area estimates for Philadelphia on over 25 health outcomes, health behaviors, and social determinants to describe and understand health disparities. For more information, please contact the Urban Health Collaborative

To learn more, view the methods brief here.