How integration of SDOH data is helping improve opioid response, pop health

As hospitals and health systems have become more aware of the key value of social determinants of health, the questions have shifted somewhat. Now, many providers and public health agencies are focused on more nuts and bolts questions: related to data aggregation and integration into electronic health records and clinical workflows.

At HIMSS20 in March, one public health expert from Regenstrief Institute Center for Biomedical Informatics will explain how health information exchanges can help – and show the state of Indiana is beginning to make headway with some key social determinant use cases.

Regenstrief's creation of the Indiana Network for Population Health aims to enable new applications using social, genetic, environmental and other data for individuals and populations, says Brian E. Dixon, director of public health informatics at the Indianapolis-based Regenstrief, an NGO affiliated with Indiana University.

At Regenstrief, "my role is to lead our portfolio projects that are relevant to population health," Dixon explains. "My background is in computer science and information science and I kind of came into healthcare through that – and have been working to develop information systems that focus on population health, especially those that meet the needs of large health systems."

The goal of the INPH, he says, is to build out an information infrastructure – a data platform, a governance mechanism – to enable management, sharing and use of pop health data among different stakeholders.

The immediate plan for the platform is to "help with the opioid epidemic in the state of Indiana," Dixon explained, and to "lead the development of a novel data infrastructure to help meet the needs of healthcare systems and health organizations trying to work together to address it."

But he and others hope the initiative, "an integrated repository for social determinant data along with healthcare data and claims data," will eventually be able to accommodate other high priority population health use cases – allowing for key SDOH data to move more freely among hospitals, clinics, payers, laboratories and public healthcare organization.

'Upstream issues'

"In the past we’ve largely thought about bringing together different clinical information from Hospital A’s EHR and Hospital B’s EMR to share data on the patients they take care of," said Dixon.

"But in the context of the opioid crisis we realized we need to bring in non-healthcare data sources along with healthcare data sources in order to get the full picture of what’s going on – whether that be data from the census bureau, or data with collections organizations, for example, to understand the people who’ve died from the opioid epidemic putting their healthcare record together with the toxicology report and the coroner’s report as well as, for example, we’re working to integrate emergency medical system data."

"It’s one thing to get ahold of data. It’s a whole 'nother ball of wax to analyze it and then put it back into operations."

Brian E. Dixon, Regenstrief Institute

Oftentimes, he said, "we’re able to treat someone on scene for example with Narcan, but then that person isn’t necessarily taken to an emergency department. So what we’ve discovered is that we need to bring all these data sources together so we can really get a picture of the opioid burden and the different ways people interact with the healthcare system in order to address all the upstream causes."

Public health professionals "spend a lot of time talking about prevention of disease," said Dixon. "For example, if we can identify people who are pre-diabetic, maybe we can work with behavior to help them keep from becoming diabetic.

"So if we know for example that people turn to opioids because of mental health conditions or because of life stress or economic situations or because they get addicted to it because they were on pain meds from a surgery, or what they really need is substance abuse treatment but they don’t have the insurance to pay for substance abuse treatment – if we can identify those sorts of things earlier and get people into social services or other agencies that can help them deal with those, what we call upstream issues, we can prevent opioid abuse disorder from setting in the first place," he said.

But that's easier said than done, of course.

"The real question is how do you tap into those outside data sources," said Dixon. "So I really want to share the model we’ve come up with in Indiana to do that, and I’ll be talking about the infrastructure that we’ve set up, the methods that we use to gather the data, the techniques we use to integrate the data together, and then the methods we’re using to make that data available to our partners.

"It’s one thing to get ahold of data," he added. "It’s a whole 'nother ball of wax to analyze it and then put it back into operations. So we’ve been working with our partners for two years now to set up our approach and we’re at the stage where we’re ready to share our model with others – and to hear from others, because maybe there’s other things we haven’t thought about that we could do. So I’m also looking for input from the audience on maybe other things that we should be thinking about or maybe potential partners we should be thinking about."

Brian E. Dixon will describe Regrenstrief's efforts in his HIMSS20 session "Integrating Social Determinants Into an HIE Network." It's scheduled for Wednesday, March 11, from 1-2 p.m. in room W209C.

Dixon was interviewed for this story by Jonah Comstock.