R in a Nutshell, 2nd Edition

A More Complex Example: Medicare Data

To help show how to use ggplot2 to solve problems, and to better understand the grammar of graphics, I’ll use a real, complicated data example: U.S. Medicare cost and outcome data. See Medicare Data for more information.

Medicare Data

To demonstrate ggplot2, I tried to find a rich and complicated real-world data set. You can download the data from the website Medicare; it’s straightforward to load the raw data into R.

I have included several R data frames based on this data in the nutshell package:

outcome.of.care.measures.national

A small data set that shows the national average mortality and readmission rates for heart attacks, heart failure, and pneumonia.

medicare.payments

A data set that shows the average payment to each hospital for 70 common conditions. Average payments are available only for hospitals that treated a sufficient number of patients with each condition; otherwise, HIPAA makes it illegal to disclose this information.

medicare.payments.by.state

Similar to medicare.payments, but summarized at a state level.

For more details on these data sets, use the online help.

Let’s start with a simple example: average mortality and readmission rates for three common medical conditions. We’d like to compare national treatment effectiveness statistics for three common diseases. This is a fairly simple data set: there is one dimension (the readmission rate), three conditions (Heart Attack, Heart Failure, and Pneumonia), and one factor variable (Measure) ...