Disease maps as context for community mapping: a methodological approach for linking confidential health information with local geographical knowledge for community health research

Disease maps as context for community mapping: a methodological approach for linking confidential health information with local geographical knowledge for community health research

by Kirsten Beyer, Sara Comstock, Renea Seagren

Biological, social, and even economic reasons can be linked to certain health issues and concerns. Recently, people have started to approach health care and analysis through these various lenses, rather than only looking at the biological and physical aspects of human health. Beyer et al. used a multidimensional approach to gather both health data and community understanding about colorectal cancer in Storm Lake, an city in Iowa that has greater incidence and rates of colorectal cancer than the majority of Iowa.

GIS has been used to provide a deeper understanding of the spatial patterns for health hazards, exposures, and outcomes at different scales. This is contributed greatly to health researchers because they are beginning to analyze how patterns could be used for disease surveillance and identifying patterns between health outcomes and different causal factors. Researchers have run into some problems when trying to analyze health patterns because detailed health data, such as disease records, are extremely confidential.

The researchers chose to utilize community-based participatory research (CBPR) in their study because they wanted to involve their subjects in all phases of the research to better understand the community’s perspective. This approach would allow for the local knowledge of the community to be observed, which would better help with the development of health and education programs.

This particular study was aiming to connect confidential, geocoded health data with community generated information using GIS to advance research on colorectal cancer. Health data was obtained through the Iowa Cancer Registry and the Bureau of Vital Statistics. They had to be careful in their representation of the disease maps because they did not want any personal information to potentially be linked to particular households. The avoid this issue, they used adaptive spatial filtering that calculated a disease rate for each grid point on a map. A circular filter was used that expanded until enough data was include to calculate a stable rate.The filters varied in diameter according to the number of disease cases in the area and the areas in between the grid points were smoothed over using an inverse distance weighting function in order to produce a continuous representation of the spatial pattern. Since there were overlapping filters, the data was being repeatedly aggregated, which helped protect data confidentiality because geographic origin was made unclear.

Separate maps were created to display colorectal cancer incidence, late-stage, and mortality. These maps are shown below.

                                        

Once these maps were created, the participatory part of the research was then conducted. Seven focus groups were established and 12 interviews were conducted for a total of 60 participants. All the participants lived within a 20 mile radius of Storm Lake, a rural, northwestern Iowa in Buena Vista county. This county is characterized by poverty, low education, and designated as a medically under-served area.



The groups were participated in a presentation on colorectal cancer; a presentation about how researchers calculate disease rates and map them; a brainstorming discussion about why colorectal cancer is a greater problem in Storm Lake; and a mapping exercise where participants added features to a map that were thought to increase or decrease risk for colorectal cancer. The map to the right displays the community-generated features that the participants indicated as having a positive, negative, or unsure influence.








Overall, they found that that participants were concerned with environmental contamination, especially water quality, as a negative influence for cancer risks. Other features that were believed to be a negative influence were workplaces, fast food restaurants, poor quality housing, etc. Some of the features they believed to have a positive influence were associated with nutrition, physical activity, and healthcare. The map below was created displaying the community-generated features to visualize what/where people were perceiving influence to be located.






The community-based approach of the study allowed the researchers to gain a greater understanding of how the community members understand this particular health issue, allowing for more effective approaches to solving the problem. GIS was initially useful in determining areas that had the highest rates of colorectal cancer. Additionally, GIS proved to be a great resource of displaying information gathered from the participants. This spatial representation allowed both the participants and researchers to explain and analyze community understanding.

-Lisa Morse

source: Beyer, K., S. Comstock, and R. Seagren. 2010. Disease maps as context for context for community mapping: a methodological approach for linking confidential health information with local geographical knowledge for community health research. Journal of Community Health 35:635-644.