Open Data Fabric Coming to CMS, But Culture Remains a Barrier

The Centers for Medicare and Medicaid Services (CMS) will soon release an open data fabric, officials said Wednesday at the Health IT Summit. CMS CIO Patrick Newbold said the agency is pushing hard for open standards to improve communication and integration across federal health agencies and partners.

“We’re challenging those who work with us to adopt that open architecture, and I am proud to say that by the end of this month we will have a true open data fabric,” said Newbold. “Just moving things to the cloud doesn’t solve the silo issue, but when you can integrate the cloud, the security controls, the processes together, that’s what’s going to allow partners like CMS and NIH to exchange their best valuables.”

Newbold added that adopting the right infrastructure for CMS has helped unlock data that otherwise would still remain in siloes. Newbold said CMS needed a hybrid, multi-cloud environment rather than locking into one vendor solution.

“Instead of forcing everyone to a particular cloud service provider, we’re going to have multiple so that the best workloads, the best needs, is the one we can use to do what we need,” said Newbold.

Christine Pacheco, federal civilian sales director at Equinix, echoed Newbold and added that conversations with agencies about data storage are shifting to a hybrid multi-cloud approach. Pacheco said as officials look to pilot more AI use cases, hybrid multi-cloud environments become the more realistic option.

“With a hybrid multi-cloud view, they can maintain what they need to on prem, and then have what is shareable in an interconnected, secure environment,” said Pacheco. “That collaboration can be more efficient and also a little bit more strategic, in that you’re creating pockets of ‘here’s what we can’t share, here’s what we absolutely will share’.”

Navigating the Open Data Cultural Challenges

Amid the technical challenges, the culture of encouraging researchers to lean toward open data is also a barrier. National Institutes of Health’s National Heart, Lung and Blood Institute Scientific Solutions Delivery Branch Chief Sweta Ladwa said enforcing data sharing and management policies has helped soften concerns around the cultural shift to open data.

“The culture of research has always been it’s mine, mine, mine, and it’s very difficult to open up that data beyond the consortia,” said Ladwa. “That culture has persisted, but I think it’s getting better, and I think some of it is due to recognizing that you have to enforce things.”

Chris Kinsinger, assistant director for catalytic data resources at NIH, said the recent BioGenesis Mission — a sub-initiative of the Energy Department’s Genesis Mission to use AI to accelerate health research outcomes — relies heavily on interoperable and findable data.

“I think good data infrastructure is findable, and also really interoperable with the other resources out there. We don’t have to consolidate, but we need to use common standards and data models,” said Kinsinger. “At NIH we have 27 institutes and centers, and unbelievable data resources across them. But because it’s so federated and so spread out, it’s really hard to find.”

He cautioned against top-down interoperability efforts that are led with too heavy of a hand. Kinsinger said that asking people to give up control of their data is difficult and often isn’t responded to well.

“We’ve seen a lot of bottom-up driven approaches that provide enough coordination that people get excited about it,” said Kinsinger. “Then people are willing to put energy into it, and that’s worked pretty well in a number of environments.”