How Does AI Relate to Healthcare Interoperability?
In a panel on the future of interoperable intelligence, Sarang Deshpande, vice president and chief data and analytics officer at Franciscan Health, said that for the past 10 years interoperability has been about regulatory data-sharing requirements, such as HL7 and Fast Healthcare Interoperability Resources (FHIR). From his perspective, the next phase for interoperability is getting internal systems to connect and talk to each other, improving workflows.
“We have a lot of work to do, but my main thing is moving from data sharing to a well-connected ecosystem,” said Deshpande.
Jawad Khan, chief data and analytics officer at Akron Children’s Hospital, said that interoperability is often limited to discrete data shared system to system; in the age of AI, it needs to be more dynamic and able to change its format on the fly.
Panelists expressed that many organizations’ data isn’t ready for AI and the interoperability required to make AI initiatives successful on a larger scale.
Tamara Wegner, senior director of enterprise platform service at Nordic, said that some organizations are rushing too fast into the AI space.
“You see lofty AI thoughts, but when you peel the onion, the core is still struggling and the foundation isn’t there to get organizations to where they want to go in the AI workstream,” she said.
“Everything and everyone is ready for AI except your own data,” agreed Deshpande, adding that there is a lack of data literacy and stewardship in the industry and many people don’t want to put in the hard, tedious work to prepare data for the future.
EXPLORE: How are health systems transforming their data management?
Preparing Data for AI Implementation in Healthcare
Khan said that it’s important for IT teams to understand full-stack engineering and how it works in the data space. It can help to unify data in ways that data lakes cannot, he explained. AI can play a critical role in making full-stack engineering easier. Governance is also crucial.
“To improve data quality, you have to push as close to data generation as possible. I encourage my team to understand how data is entered into the system,” said Khan. “If you can tackle data quality as close to where the data is generated as possible, then you won’t have to be as concerned on the back end. If you’re doing data quality on the back end, you’ll end up in a vicious cycle.”
To achieve this, data producers need to be a part of data governance committees. He recommended starting on the front end of the data and working toward the back end. Leveraging technology to report on data quality is also helpful, according to Khan.
Scott McEachern, CIO at Southern Coos Hospital and Health Center in Bandon, Ore., emphasized the need for clear steps and understanding the “why” for end users who are generating data.
Khan encouraged healthcare IT leaders to take stock of their internal resources before implementing new tools. Working with partners can help to maximize investments already made, especially electronic health record and enterprise resource planning platforms.
While data stewardship can seem daunting, Deshpande said it’s a good time to ride the AI wave and take a center-of-excellence approach to data.
“Focus on governance, process and the right tools, and then push that work into the business units,” he said. “If you don’t leverage the current excitement, you might miss your opportunity.”
McEachern said that metrics are helpful in getting buy-in from clinicians. He recommended presenting the data in a calm and professional manner to show how long it’s taking them to do certain tasks, then explaining how new processes or tools can help give them time back.
Click the banner below to sign up for HealthTech’s weekly newsletter.
