As the saying goes, the things that scare you and the things that actually kill you are rarely the same things. So seems to be the case with vaccines which, despite overwhelming evidence proving them to be one of the greatest lifesaving measures in the history of medicine, have been made to justify themselves to small but significant portions of the US population who are electing to not vaccinate their children. As a result, 2019 there have already been the same number of cases of measles in 2019 as there were in all of 2016 and 2017 combined (see chart below from the CDC in my visual examples.)
Finding data
I went in search of data on vaccine rates over time, as I was interested in seeing if vaccination rates varied based on race (or cultural heritage). I also wanted to see if rates for specific vaccines (for example, the MMR vaccine which was falsely linked to autism in 1998) differed from each other.
The dataset I selected was produced by the Centers for Disease Control and Prevention (CDC) and found at data.gov.
The data provided vaccination rates for 9 different vaccines from 1995-2009 broken down into 6 different racial and ethnic groups. The 2011 CDC Report “Health, United States 2011” includes this dataset as part of its analysis and identifies the rate being reported as the percentage of children ages 19-35 months (see table 85).
Examples
The CDC released this bar chart on measles which shows an alarming outbreak in 2014. However, I feel that this would have been more effective as a continuous line graph rather than blocking each year separately. As a continuous line you could see how the momentum of the outbreak built up and then subsided.
The CDC presented this information as a table, which is useful to pinpoint specific dates, but a line graph would more easily show overall movements over time.
In General, I think the New York Times has great data visualizations, including this great series of graphs about climate change. Another NY Times example below shows the movement of trends over time, which is the kind of information I was hoping to glean from my data.
Creating Graphs
I used free online software called Tableau public to create graphs based on my data. Tableau easily recognized and organized the data I imported as a .csv. It understood how to interpret my data in ways that I didnt. Knowing what I do now about how Tableau can ingest and convert data will change how I look for datasets for Tableau in the future.
I found the Tableau design tools to be mostly intuitive (you can do most of your layouts using drag-and-drop methods). I was able to make a quick line chart from my data and start to see some trends.
I focused on two different comparisons: Rates of all 9 vaccines from 1995-2009 to compare against each other
1995 – 2008 MMR vaccines by race or ethnicity
I would have liked to have also created a graph of overall vaccines by race or ethnicity but I wasn’t able to combine the individual vaccine data into a single “all vaccines” category. So I picked MMR because this is the vaccine that has had the most controversy.
https://public.tableau.com/profile/micaela.walker#!/vizhome/Vaccinations1994-2010/Dashboard2
One thing I experimented with in both graphs is integrating the legend into existing spaces. For the MMR graph I chose to forgo a separate legend altogether when I realized that the individual race graphs basically functioned as a legend itself.
Findings
Vaccines by type
Different types of vaccinations had very different rates of use. While some, such as DTAP and Hep B, had a relatively steady increase in use over the entire time period, while the combined series (4:3:1:3:3:1:4) made a dramatic entry in 2007, climbed in popularity in 2008 and then took a relative nosedive. This may be related to the fact that many people are increasingly resistant to “overloading” babies’ systems with too many vaccines at once, preferring to spread them out over time. It’s also possible that this new combination of vaccines was far more expensive combined than as separate vaccines.
Vaccines by race and ethnicity
It is interesting and a bit depressing to see that all races apart from Native Hawaiian and Pacific Islander are in decline at the end. The peak for most races is around 2002 apart from Asian children who dipped down around then and then began to climb in rates until a fall right at the end.
Future prospects
In the future I would like to explore vaccines by religion and education level of the parents. Vaccines, in particular, are a bet on the future, rather than fixing something that is already wrong, and require some amount of faith to be effective.
Medicine is a unique combination of scientific and intensely personal decisions that play out in interesting ways in data.