Peter Embi

IU researchers using data to help provide a better understanding of how the addiction crisis is affecting Indiana

Understanding how the opioid addiction crisis is affecting Indiana requires many different types of data that are gathered and held by separate organizations.

For example, when a person overdoses and is brought to a private hospital in an ambulance, the hospital collects data about that hospital visit, but information about the person’s treatment in the ambulance is held separately by the state government. Additionally, relevant information about opioid users' circumstances and living conditions – such as income levels, educational background, crime statistics, and access to transportation – exist in largely separate datasets.

Connecting this information would allow researchers to see patterns in who is becoming addicted and what works for recovery but gaining access to these resources currently takes months or years and requires much additional work to organize distinct datasets in a way that allows them to be used together.

To overcome this challenge and accelerate research that can save lives, Peter Embi and Shaun Grannis are leading the development of the Indiana Addictions Data Commons, part of Indiana University's Responding to the Addictions Crisis Grand Challenge.

The Indiana Addictions Data Commons provides rapid access to critical information by establishing data-sharing agreements among several leading Indiana organizations and preparing this information to be analyzed through powerful computing methods. This work involves making many diverse and distinct datasets compatible with one another.

Participating partners include IU's Regenstrief Institute, which maintains information on more than 13 million patients across more than 100 hospitals; the Indiana state government’s Management Performance Hub, which collects information on public services and their recipients and outcomes, and IU’s Polis Center, which gathers a range of data including location-based information about the characteristics of specific Hoosier communities.

The data commons work will allow individuals' health data to be linked to information about their communities and the public services they receive while protecting patient privacy by ensuring that no names or personal identifying information are connected to the data.

This data will be made freely available through self-service online data dashboards, which will allow researchers as well as healthcare providers, policymakers, and the public to directly access the information. For example, users could select categories from a drop-down menu (such as high school graduation rates, opioid prescriptions, and opioid deaths) for display across a map of the state. These tools can be used to help identify people at risk of addiction, direct resources where they are most needed, and evaluate the effectiveness of Indiana’s response to the addiction crisis.

Currently, while the data compatibility work is underway, the Indiana Addictions Data Commons is offering a virtual help desk to make data available on a case-by-case basis.

This "data concierge" service will allow researchers to request data addressing specific questions, such as how employment levels relate to addiction risk across different communities or how the distance traveled in an ambulance relates to health outcomes after an overdose.

The data commons team will assemble the requested data, make any necessary changes to how the information is organized so that the different datasets can be combined for analysis, and, once the results have been analyzed, assist with data visualization, which creates graphics that make the results easier to understand. A top priority in the current phase of the project is helping other Grand Challenge research teams access information about addiction in Indiana quickly and easily.