Vizient is the nation’s largest provider-driven healthcare performance improvement company, serving more than 65% of the nation’s acute-care hospitals. The company provides analytics-enabled solutions to improve care quality, as well as spend management.
To ensure efficient control of how data is managed and accessed, while also complying with legal, industry, and contractual requirements, Vizient shifted to a decentralized data architecture with distributed data ownership.
We sat down with Charles “Chas” Hatfield, Senior Director, Data Governance at Vizient, to discuss why they made the change, as well as to understand the importance of data ownership from Vizient’s perspective.
Immuta: What makes data ownership important to operations at Vizient?
Chas Hatfield: Vizient has multiple analytics platforms supporting data submitted by healthcare providers throughout the United States, and given the sheer volume of data we collect, we take the role of data ownership very seriously. This includes maintaining data security and secure customer access — it’s a top priority.
We provide various layers of ownership, including at the aggregate level, where we combine data for analytics and benchmarking. This way, our customers can evaluate themselves against their peers. Additionally, we house data consisting of Vizient intellectual property and third-party data to enhance our solutions and analytics. All this data has different levels of ownership and needs to follow proper approvals before it can be consumed by an individual or leveraged for an analytics-enabled solution.
I: Were there any specific pain points you were experiencing that led to the shift toward delegating data ownership? If so, what were they?
CH: One of the main points that led us to a more delegated, or decentralized, approach to data ownership stems from the subject matter expertise needed to be a good mentor of each data set. Ownership of the data is not centralized to one person, but delegated to the people who work with the data the most. Those delegated users come from different business units and access different sets of data.
For example, our supply chain data doesn’t contain the federally mandated protected health information (PHI) that our clinical data does, but it has a high level of confidential information, which requires a different level of security. Our clinical databases each have many data sets, but with slight variations in how PHI data is communicated – each must be managed differently with different parameters. This delegated approach leads to better management of the data by the people who work with it every day.
I: What questions do you need to ask in order to effectively delegate data ownership?
CH: One of the first questions we ask is, “Who do you talk with if you have questions about the data?” We find that those people who are treated as subject matter experts by the everyday users of the data tend to have a good historical knowledge of the data. We translate this collective knowledge into properly managed data, and build a data governance strategy for that data through those findings.
I: What opportunities/business impacts can effective data ownership open up for you?
CH: We find that establishing data ownership leads to more effective use of the data. This happens in two ways primarily:
First, when individuals, such as an analyst, begin utilizing the data, they gain a comprehensive understanding of the data processes and lifecycles.
The second positive outcome is that the knowledge of a data set becomes a part of the user’s data knowledge, and thus the data and its concepts spread wider across the organization. This leads to more data centralization and expansion of use cases where that data can contribute to our data-enabled initiatives.
I: If you could give one piece of advice to other data governance leaders who are aiming to secure and streamline data use, what would you tell them?
CH: Continue fostering discussions on data ownership. Not every data set will initially have an identifiable owner. It will take time to establish who the business owner is. Often, a technology team or an operations team becomes the owner of the data. But it is the analytical users that are the actual drivers of what and how the data is used.
