Intro to Data: Using Geographic Location to Enhance Data Visualizations of Mental Health Prevalence in the Military Health System

Two image of maps and graph with data
PHCoE graphic
By Jennifer Greenberg, MPH
January 29, 2018

If you’re following our “Intro to Data” blog series, you know we’ve covered understanding “big data,” examples of big data in the Military Health System (MHS), and moving from data to knowledge and wisdom. For this blog, I’d like to shift directions and dive into how we visualize data through maps.

Every year, we update our Psychological Health by the Numbers section on the PHCoE website with the most recently completed year’s data on mental health condition prevalence, as well as utilization of mental health services, within the MHS. This year however, we’ve totally revamped our prevalence page to not only include 2016 data, but to also allow the user (you) the ability to interact with our visualizations (in this case, graphs and maps). Take a look at the updated Mental Health Disorder Prevalence page (but first read the rest of this blog).

In our previous prevalence reports, you would have seen static images of graphs that were inserted in a downloadable report…not that exciting or user-friendly. But now we have embedded these graphs within the webpage itself and have included handy buttons at the bottom of the embedded portion so that you can flip through different mental health conditions and their prevalence by service and component (active duty, Guard/reserves) from 2005 to 2016.

We typically think of graphs and charts as the main methods by which we can display our data to an audience. However, if your data includes location, you’ve got the makings of a map and an entirely new way to visualize patterns within data. This brings me to the most noticeable update to the prevalence page: the addition of a prevalence map that incorporates some exciting interactive features, including a drop down menu with different mental health conditions and buttons for you to click through the three most recent years.

Stepping back from the shameless self-promotion, let’s talk about maps more generally (and why they’re so cool).

Introduced in 1969 by Waldo Tobler, the first law of geography is that everything is related to everything else, but near things are more related than distant things. It’s a concept at the core of interpreting data driven maps because, as you might imagine, location matters quite a lot to geographers. For example, where someone lives, works, goes to school, or receives health care has the potential to impact some outcomes of interest whether that be prescription drug distributions and subsequent opioid use patterns, population engagement with established networks of health care, or even the cost of your typical grocery store basket of eggs, milk, and cheese. Through the use of Geographic Information Systems (GIS) technology, the application of this overarching concept is more accessible than ever and as a result, allows us the ability to display and analyze data for trends in location and proximity – show me a graph that can do that!

Probably the most familiar type of map made using GIS technology is called a choropleth map. You can find examples on (you guessed it) our prevalence page. These types of maps use graduated color to display differences in the average values being mapped. This way, you can easily display and highlight which geographic divisions (like states) have, on average, higher or lower rates of some variable (such as prevalence of depression) compared to another division. If you see a “cluster” of very dark or very light color in a defined region on your map, that’s interesting and may merit further research and analysis using analytical and statistical tools to determine exactly what’s going on.

Choropleths are just one option though. GIS can also be used to create maps that visualize and evaluate things like:

  • Accessibility and network linkage (Is the military treatment facility (MTF) distribution in region X sufficient for the active-duty service member and dependent population it is meant to support? On average, how long would it take for a service member with an emergency to get to the nearest MTF?)

  • Resource allocation (Are specialty mental health facilities located in convenient areas so that service member populations and their families may be adequately served?)

  • Environmental risk (Are certain groups of military personnel disproportionately more likely to be living near areas high in noise pollution resulting in higher prevalences of insomnia and other sleep disorders?)

Ultimately, the number and types of maps you can design using GIS technology is bound only by your own creativity and, of course, access to appropriate data. If you have ideas of how and why geographic divisions may impact mental health disorders, care delivery, or utilization, please share in the comments below! We are also open to suggestions or comments on our updated prevalence page, so please share those thoughts as well.

Ms. Jennifer Greenberg is a contracted data analyst at the Psychological Health Center of Excellence. She has a bachelor's degree in environmental studies and a Master of Public Health in epidemiology.

The views expressed in Clinician's Corner blogs are solely those of the author and do not necessarily reflect the opinion of the Psychological Health Center of Excellence or Department of Defense.

Add new comment

PHCoE welcomes your comments.

Please do not include personally identifiable information, such as Social Security numbers, phone numbers, addresses, or e-mail addresses in the body of your comment. Comments that include profanity, personal attacks, or any other material deemed inappropriate by site administrators will be removed. Your comments should be in accordance with our full comment policy regulations. Your participation indicates acceptance of these terms.

Please read our full Comment Policy.

You must have Javascript enabled to use this form.
You must have Javascript enabled to use this form.