Data Privacy in the World of Data Democratization and Sharing

Data Privacy in the World of Data Democratization and Sharing

Hosted by CDO Magazine and ComSpark, the CDO Midwest Summit 2022 marks its success through immensely engaging sessions on numerous topics related to data and technology. The session “Data Privacy in the World of Data Democratization and Sharing” brings together the following speakers to discuss the democratization of data and the challenges inherent in data sharing, internally and externally:

  • Chris Higgins, Sr Manager - Data Engineering & Architecture, GE Aviation
  • Matt Snider, Chief Information Officer, US Urology Partners
  • Krishna Tummalapalli, Director of Data Engineering and BI, OhioHealth

Drew Stark, Sales VP at Ataccama, moderates the session.

Snider begins the discussion by mentioning how the democratization of data adds to the organizational challenges of disparate data sets. To solve disparate data problems, he suggests using Dashboard MD, a third-party product that integrates data from multiple EMRs into a single reporting platform.

Tummalapalli emphasizes that organizations must ask three questions when it comes to data privacy and sharing:

  • Why share the data?
  • What is the minimum sharable data set to meet objectives?
  • How do we know if the data is right, and what is happening with the data?

He believes these questions help develop a better understanding of data privacy and move the enterprise needle.

For Higgins, the three key components are people, processes, and technology. He affirms that the people aspect is crucial, especially in the initial classification phases. Higgins adds that communicating well with people who understand the systems and data makes it easier to track where data comes from and what to do with it.

According to Higgins, his organization has a natively built solution for tracking data within the lakehouse. “Within that process, we classify things like PII, but what is equally important to us are the products and parts and the data on those parts. That is really where our value lies,” he shares.

According to Higgins, when the organization migrated from on-premises to the cloud, technology was the first thing they focused on. Technology helps the organization democratize data and aids the need for security and governance, he notes.

Higgins maintains that data retention should be done for the sake of use, not just for retention. He adds that while it is necessary to retain important data and back it up, it is equally essential to get rid of unused data.

Sharing his point of view, Snider says that keeping 14 years of patient data in EMR is questionable from a technological standpoint. However, from the business value proposition standpoint, having this amount of data is a boon. “It is a tough balance to determine what proper retention is, and it goes back to appropriately classifying your data and determining what the underlying value of that data is,” he adds.

According to Tummalapally, the best data retention policy is to maintain data for seven years and use it anonymously for analytic purposes after that. He urges IT leaders to update or recatalog enterprise inventory.

Higgins shares that inventorying is a value-add while deploying enterprise architectural solutions. He recalls creating a maturity curve and inventorying at the asset and product level, which helped the organization transform.

Data sharing in health care revolves around interoperability, says Tummalapally. He adds, “We are calling it the integration platform. That would be your point to the external world, and then try to integrate everything internally with the integration platform.”

Higgins further pinpoints that the data owners are external to the organization. GE Aviation has built external relationships on trust, which positively impacts organizational data sharing.

On a similar note, Snider mentions partnering with different external partners and sharing patient data through interfaces. He adds, “Technical integration of those things is not very difficult. The difficult piece is the negotiation work that happens up front.” He encourages organizations to share data cautiously.

For Tummalapally, identifying the right ways of sharing and monitoring them are essential.

In conclusion, Higgins states that organizations must enable data product managers with the mechanism to share data in a controlled and productive way. Snider suggests organizations start with one classification and build off of that.

Watch other CDO Midwest Summit 2022 sessions HERE

Related Stories

No stories found.
CDO Magazine
www.cdomagazine.tech