Cancer Mortality Incidence Ratio in the US from 1999 to 2018

Authors

• Sam Lindsay
• Peter Xu

Motivation

• A previous study showed disparities in MIRs between White and African American population in South Carolina.
• Growing recognition of socioeconomic factors in individual health outcomes.
• Provide some future study topic

What it does

• Generate Altair interactive charts for changes in MIR by states and cancer types from 1999 to 2018.
• Generate Altair interactive map showing race differences in MIR across the US.

How we built it

• We used Pandas DataFrame and GeoPandas GeoDataFrame to clean and join datasets.
• Altair is utilized to create the interactive charts.

Challenges we ran into

• The datasets are worse in quality than we originally expected. Many regions are lacking essential data for the calculation of MIR.
• The datasets are inconsistent on symbols of None values.
• Learning Altair using their documentation.

Accomplishments that we're proud of

• We correctly cleaned and joined the datasets, including removing rows with empty entries, calculating the MIRs, and joining DataFrame with GeoDataFrame.
• We successfully produced functional interactive charts.

What we learned

• The datasets are imperfect, and vast amount of regions in the US is lacking proper data coverage.
• Different states perform differently on decreasing MIR, and most types of cancer remain relatively constant on their MIRs.

What's next for Cancer Mortality Incidence Ratio in the US from 1999 to 2018

• The insufficient data coverage in obtained datasets from CDC points at possible racial structural violence. Further investigations are needed from medical anthropologists and policy makers.

Built With

• altair
• geopandas
• github
• html
• pandas
• python