Report: Data interoperability key to unlocking AI’s public health potential
J Studios via Getty Images
New survey findings suggest public health organizations are keen to adopt the technology to boost operational efficiency, but improving data interoperability is a crucial first step.
Data interoperability is top of mind for public health leaders as they grapple with tightening budgets, changes in reimbursements rates, evolving policy mandates and persistent workforce shortages, according to a new report.
The drive to improve data services comes as agencies nationwide are increasingly turning to artificial intelligence as a tool in government operations to enhance efficiency and streamline workflows, according to a report released Tuesday by data platform company Snowflake. The findings are based on a survey of 183 health care decision-makers, including public health agency officials.
Indeed, 85% of respondents said data interoperability was a “higher or much higher priority” to their organization than in previous years, while 57% of respondents identified AI implementation as a reason for that, the report found.
“Interoperability is not a top industry priority simply because regulators have said it should be,” Shahran Haider, deputy chief data officer of NYC Health and Hospitals, said in the report. “It is the connective tissue that enables AI-driven value across the full spectrum of operational efficiency and clinical improvement opportunities, benefiting not just individual organizations, but our health care system and society as a whole.”
Data interoperability is the necessary infrastructure foundation to deploy AI tools that enable health care providers to more efficiently track and monitor health information across sectors and departments. The report says it can help identify outbreaks or population trends to inform public health responses. Such data coordination can help reduce duplicative testing efforts, lower overall spending and waste, and streamline administrative processes.
In fact, more than 70% of respondents said in the survey that improving patient experience and operational efficiency and decision-making also motivated their desire to improve data interoperability in health systems.
Many data systems at the state and local level are still siloed, however, which impedes public health agencies’ ability to manage and prevent public health threats, said Dr. Georges Benjamin, executive director of the American Public Health Association. Additionally, public health technologies across the U.S. are at various stages of maturity, with many agencies still relying on fax machines that may not be able to communicate with more advanced computer and data systems.
“If we had better data modernization and interoperability, we could do a much better job picking those cases up early and [leveraging] AI, because [the tech] can be very helpful to do tracking, forecasting and hypothesizing what to do to try to address public health threats,” said Benjamin.
Indeed, improving population health was one of respondents’ top drivers for improving data interoperability to ultimately adopt AI solutions, like agentic AI, according to the report. The technology, for instance, can assist with monitoring electronic health records, analyzing historical data or public health reporting capabilities.
Nearly half of respondents also cited AI’s potential for combating fraud, waste and abuse by, for example, enhancing the screening of billing, claims and prior authorizations. The report pointed to those efforts as a reason to prioritize data interoperability.
States must continue their efforts to facilitate data improvements to ensure public health systems can stay abreast of data and tech advancements, like AI-enabled health care, Benjamin said.
The California Department of Health Care Services, for example, has implemented a statewide data exchange framework since 2024 that aims to require health care entities in California to participate in a health data and information sharing network, under the state’s interoperability project.
“The real secret here is getting information in a timely way,” Benjamin said. “You can’t begin to talk about AI until you put those good data systems in.”