Data sharing at the state and local levels requires more resources

At Digital Frontiers in Public Health, panelists from California and Maryland spoke of the difficulties they've faced in garnering data to improve public health outcomes.
By Jessica Hagen
02:36 pm
Share

 Photo: Brian Kramer Photography/Ellison Institute

Joshua Sharfstein, vice dean at Johns Hopkins Bloomberg and former Secretary of the Maryland Department of Health from 2011 to 2014, said that, at the start of his term in Maryland, half or two-thirds of the hospitals were connected for data sharing, and the data would often only provide information for clinical lookups, such as whether the patient underwent a CT scan at another hospital.

"The leaders of that system came and said, 'We envision this as a clinical program, but we don't see where the money can come from to build it out for the whole state,'" Sharfstein stated during Tuesday's event at the Ellison Institute.  

Sharfstein and his partners gave the leaders an ultimatum – he and his team would get hospital systems on board with the governor's help, funding would be provided through Medicaid or other ways, and a provision would be signed into law that designates the leaders as the state's health information exchange.

In turn, the leaders would have to double the size of their board and add members of the public to it, add a provision in the bylaws stating the health secretary can wipe out the board if it's not acting in the best interest of Maryland residents, and the service would be dubbed a public health utility, not just a clinical program.

"It was really a deal between us and the health information exchange (HIE)," Sharfstein said. 

Streamlined data connectivity has been more challenging in Los Angeles County, especially during the COVID-19 pandemic.

Barbara Ferrer, director of the Los Angeles County Department of Public Health, said there were four basic needs for data to analyze what was happening during the COVID-19 pandemic within the county, which houses over 10 million people:

  1. Who is getting sick, and where are we having outbreaks, by sector and geography?
  2. How well do the interventions work, including vaccines and other measures?
  3. What differences were our interventions making, including via telehealth and mobile teams?
  4. What do workers and residents need and think?

"I think the place we do the worst is in that last area, almost as if we don't care. But we do. But when you hear people talk about data, even all of us, we will talk about interoperability. We will talk about the datasets that we have. But the data that we really need to become trusted people in our communities is data about how people feel, and what is it that they think they need, and how can we be helpful?" Ferrer stated. 

Anthony Iton, senior vice president of programs and partnerships at the California Endowment, says when he served as Alameda County's health officer, he was registrar of all births and deaths, which allowed him to take a more focused approach to utilizing data for community health.

"Every death certificate tells you what somebody died of; their race, ethnicity; where they live; and their age. So we could actually paint a story of death in Alameda County about how it occurred. We published maps at the census tract level showing how long somebody could expect to live," Iton said. 

Still, Iton and his team wanted more specific data, but hospital systems refused to share it due to a lack of trust that public health would protect the data. 

"It was sort of indicative of how little leverage public health has," Iton said. 

Claudia Williams, former CEO of Manifest MedEx and one-time senior advisor for health technology and innovation for President Obama's Precision Medicine Initiative, says a critical component to improving health outcomes is access to personally identifiable and de-identified data that is not in the jurisdiction of public health.

"We're never going to be able to reach our bigger goals of health improvement unless there's access to data that allows us to have insights before things happen," Williams said. 

All panelists agreed that change within the health-data ecosystem across the states is necessary for the benefit of public health, but more resources are needed to ensure data interoperability becomes a reality. 

"This work cannot happen without resources, and it's going to need to be sustained over a long period of time," Ferrer said. "We need long-term financial funding at the federal, state and local level. We've got to have the money."

Iton said, "We don't have a national healthcare system. We have a federal healthcare system. So we have a federation of 50 states. And that's baked into sort of the relative distribution of power, the ability to make decisions. The national government can't make decisions that the feds have to follow, and that's problematic. So, the leadership for this has to come at the state level."

Sharfstein argues for the need to prioritize public health outcomes as it's not just about the patient encounter, but rather empowering people who can work with their communities. He stressed the need to change the environments that are driving bad health.

"States need to build a data infrastructure usable, not just by public health, but also by research by clinical folks," Williams said. "This is a public good. We need to invest in it as a public good. We need to align all of our policy levers, and public health needs to benefit from that."

Share