Geospatial Hot Spots and Cold Spots in US Cancer Disparities and Associated Risk Factors, 2004–2008 to 2014–2018

PEER REVIEWED

This figure shows the results of a hot spot analysis: 1 large cluster on the US mainland, mainly in the Plains states and the Midwest, and a few counties in Hawaii. Cold spots were located in the Southeast, a portion of the Northeast, 2 states in the Mountain West (Utah and Idaho), and portions of Texas, Louisiana, and Alaska.

Figure 1.
Changes in total cancer death rates at the county level between 2004–2008 and 2014–2018 in the US.

Return to Article

This forest plot showed the results of random forest machine-learning analysis for the relative importance of predictors of changes in total cancer mortality rates between 2004 and 2018 in US counties. Less than a high school education and preventable hospital stay are the 2 top variables.

Figure 2.
Relative importance of predictors of changes in total cancer mortality rates between 2004 and 2018 in US counties.

Return to Article

This forest plot shows the results of random forest machine-learning analysis for the relative importance of selected predictors for changes in total cancer death rates among hot spots and cold spots from geospatial analysis of US counties from between 2004–2008 and 2014–2018. For hotspots, preventable hospital stay, being aged 65 years or older are the 2 top variables. For cold spots, no mammography screening and preventable hospital stay are the 2 top variables. For the average area, alcohol consumption and being aged 65 years or older are the 2 top variables.

Figure 3.
Relative importance of predictors of changes in total cancer mortality rates in hot spots and cold spots from geospatial analysis of US counties, 2004–2008 to 2014–2018.

Return to Article

Top