Project 3: Understanding the Distribution of Racial Disparities for Breast Cancer Mortality in Florida from 2011-2015

Background:

This course presented students in the field of Geographic Information Systems (GIS) and was critical in the development of GIS knowledge, competence in geographic databases, and familiarity with computer software and hardware. For my final project I chose to explore the distribution of breast cancer mortality in Florida in addition to the racial disparities associated with it. Despite advances in health care, black women continue to have the highest breast cancer mortality rate out of all races in the United States. Even more, this deviation is not due to biological differences, but rather because of lack of health care and preventative measures mixed in with racial discrimination. Breast cancer mortality is approximately 40% higher among black women compared to white women, with faster decreases in mortality among white women. Therefore, the purpose of this project was to understand the distribution of racial disparities existing at the county level in Florida for breast cancer mortality rates through spatial autocorrelation.

​

Geographic Data:

Data for this analysis were collected from ArcGIS online under the search ‘breast cancer US’. Using some GIS techniques (further explained below), the data was extracted to only reflect attributes within the state of Florida. 

​

Geographic Methods/Models:

The Select By Attribute tool was used firsthand to extract data to be only that of Florida instead of the entire country. Next, since the data was set to break naturally through Jenks as default and given that the data for Black mortality and White mortality differed, a more uniform classification method was needed for better comparison, which ended up being Quantile. Lastly, the Spatial Autocorrelation (Moran’s I) tool was administered three times (one for the total mortality rate, one for the White mortality rate, and one for the Black mortality rate). Each run of this spatial analysis generated a report that was used for analyzing the results.

​

Key Results:

Based on the Spatial Autocorrelation reports generated by ArcGIS, all three maps have a clustered distribution (Figure 1). For the total mortality rate, the Moran’s Index is 0.204, z-score of 2.18, and p-value of 0.03. For the white mortality rate, the Moran’s Index is 0.208, z-score of 2.20, and p-value of 0.028. For the black mortality rate, the Moran’s Index is 0.314, z-score of 3.25, and p-value of 0.001. These statistics indicated that there is an extremely low likelihood that these clustered patterns could be the result of random chance.

​

Conclusion:

In conclusion, the Spatial Autocorrelation analysis proved that black women are experiencing much higher mortality rates in breast cancer than white women. It has also been illustrated that these deaths are clustered. Further implications would be to understand the underlying causes that contribute to this clustering pattering, specifically for black women. 

​

Key Maps:

​

​