Nick Stepro, 2017
D3.js SVG, with Illustrator
Data from CMS Public Use File, US Census American Community Survey, and CMS NPPES NPI Registry
More about the Data Visualization
In an age of expanding data and increasingly powerful tools to process and visualize it, it’s increasingly difficult to filter out the noise and find meaning in a sea of numbers. With this piece, Arcadia vice president of product development Nick Stepro has fun with this concept by running over 1,000 combinations of metrics through a correlation matrix. The links he visualizes are both obvious and completely unexpected.
Of the obvious findings – risk scores, healthcare expenditures and inpatient utilization are all positively correlated (though costs and inpatient utilization are most strongly correlated). Less obvious – counties with a higher rate of women physicians are more densely populated, and have a lower per-capita Medicare spend and fewer knee replacements per 1,000. Nurse practitioners are more common in states with less college attendance, and are also associated to higher rates of colonoscopies.
To view a high-resolution image of this data visualization, check out the Arcadia Data Gallery.
Explore the Interactive Version
Nick employs a series of brightly colored partially transparent choropleth maps to visualize geographic variation, and a correlation matrix to show where factors have a negative (orange) or positive (blue) correlation.
You can explore the interactive version via our Data Gallery, and try your hand at comparing different variables.
Want a high-resolution version of this data visualization?
Simply complete the short form below to get a high-resolution copy of this data visualization, suitable for printing. If you do not see our email, please check your clutter or junk folder.