Defining neighborhood clusters in a multilevel study of barriers to cervical cancer prevention, detection, and treatment among Latinas in Los Angeles

Studies using multilevel modeling to analyze neighborhood health effect typically define neighborhoods based on contiguous units with arbitrarily drawn boundaries (e.g. census tracts).  This approach may not necessarily reflect the cultural, institutional and physical characteristics of neighborhoods, hence biasing the results.  Based on previous work of the Metamorphosis Project (http://www.metamorph.org/), we developed a novel approach to define neighborhoods in such a way that responds to this challenge, and captures residents’ everyday communication environment where decisions related to health--such as cervical cancer prevention, detection, and treatment--are shaped and acted upon.  From April 2012 to December 2013, 1660 Latina aged 21-50 and living in Los Angeles County participated in a quantitative survey on knowledge, attitudes, and behaviors with respect to cervical cancer.  Using Geographic Information Systems (GIS), we began with identifying the spatial distribution of participants, followed by a review of data and maps from various sources such as Metamorphosis, US Census, Mapping LA (http://maps.latimes.com/neighborhoods/), and Healthy City (http://www.healthycity.org/) as well as land use maps (http://cityplanning.lacity.org/). We then conducted hot spot analysis to examine the clustering of socioeconomic and demographic characteristics within the natural catchment areas of participants.  This multistep process yielded 25 distinctive neighborhood clusters that will be used for multilevel analysis of barriers and facilitators to cervical cancer related behaviors for our study population. This innovative approach improves current methods of defining neighborhoods and limits the bias associated with aggregating dissimilar units in studying not only cervical cancer related outcomes, but also health in general.