Purdue IE’s Published Dashboard

(Preview, use link above^ to see full dashboard)



This was my 3 minute presentation script (to explain the dashboard):

Hi, my name is Pamela Yuan and I am presenting on behalf of my team, Purdue IE, in the Data Visualization track of the 2017 IN Medicaid Challenge.

In this track, our team was tasked to use innovative technologies to visualize and bring a story alive

Purdue IE created and published a dashboard enables us to effectively diagnose the health care system and shed light on areas to further research, once the data is inputted.

To demonstrate the effectiveness of our dashboard, let’s focus on the emergency claims submitted from the database:

Here we used our dashboard to filter for emergency claims related with, “OBS”

As you can see at the top of the dashboard, the location and density of the population affected by this type of disease is displayed

To the bottom left, the demographic such as the age and race, along with the respective densities are shown

Finally, to the bottom right we find the most interesting trend in our story.

As our team was constructing this graph, assigning red to the number of the recipients/ year and blue to the total dollar value of the claims/year, we played with the data by trying to display the different types of diseases and viewing the respective

We then made a discovery that all the emergency disease types could be classified into 3 types: Ideal, Expected, and Distressed.

In Ideal we would see an increase of recipients/year and a decrease in total dollar amount of claims/year

In Expected, we would see a parallel trend between the two factors

In Distressed, We can see (in our example with OBS) that the total dollar amount increasing, while the number of recipients decreasing.

While, intuitively we would expect the total dollar amount should decrease with the reduced number of recipients, this logical discrepancy highlights an area opportunity for researchers to find out why the claims/recipients are increasing.

Since the information about the affected population is already displayed, simultaneously when the trend is identified, researchers can see different epidemiology spread of what areas are most affected, to begin.

Furthermore, if this type of disease presents a problematic trend, researchers can use the filters of our dashboard to find different type of diseases that have healthier implications and share best practices from the ideal cases to the expected and distressed cases, to improve the overall public health.

At a high level, the significance of using our dashboard to reveal interesting stories such as the one demonstrated and displaying the insights on the recipients affected will catalyze a more efficient approach to problem solve in the healthcare system.