In my previous blog post, we analyzed how much people from various countries spend on healthcare. And I posed the questions: "If the US spends more, do we get more? If other countries spend less, do they get worse healthcare?" In this post, I try to answer those questions by using longevity as the measure of healthcare quality.
Before we get started, here's a picture to get you into the mood for a longevity blog post. This is my friend John's Granny, celebrating her 99th birthday. I sure hope I'm still blowing out candles at 99!
And now, on to the data analysis! I started by looking for analyses that have already done on this topic. There are quite a few, but the one that caught my attention was the following graph that Paul pointed me to in his comment on my previous blog (from the ourworldindata.org website).
It was a very interesting graph, and thankfully they had links to the two data sources they used. I downloaded the data (an Excel spreadsheet of life expectancy from worldbank.org, and a csv file of healthcare spending data form oecd.org), put together some code to import it into SAS, and set out to create my own graph.I liked their original graph, so I only made a few small enhancements. I used light blue lines rather than gray (so they were more easily distinguishable from the reference line grid), I added small markers on the red USA line at the locations of the 5-year labels, and I repositioned the country labels so you could better tell which line they went with. Also, if you click my graph image below, you can see the full-size version with html hover-text for the lines, so you can easily tell which line represents which country.
One difference I notice in my graph and the original is that my graph shows more years of data for Germany, and it dips below the USA line in the 70s and early 80s (perhaps that data was not available when they generated the original graph?)
While creating the graph, I noticed there was also data available for several other countries, and I thought it would be interesting to see all the data plotted. So I created a second version of the graph, with all the countries. There wasn't enough room to label each line, but you can use the html hover-text to determine the country names for the unlabeled lines (click the image below to see the interactive version).
Did you find these graphs interesting? Do you like the first graph, or the second graph (with more countries) better? What other data might we use to measure the quality and effectiveness of healthcare?
This content was reposted from the SAS Learning Post. Go there to view the original.