Big Data: The Approach – Pt 1

This is one in a series of blog posts discussing Divurgent’s 2014 IT Organization Best Practice Study.  To view earlier blogs from this series, click here.

In addressing Big Data, all but the smallest healthcare organizations are exploring the potential value of the data collected by their information systems, and how to manage/disseminate the data.  The potential uses for the data collected through the EHR (conditions, treatments, outcomes, etc.) when combined with the financial data available in revenue cycle applications is limited only by the resourcefulness of the organization. Unfortunately, at this time, there is no solution available, requiring custom and commercial systems to be cobbled together to meet this growing need.

Organizations are approaching these issues in one of two ways.

The First Approach. To build a custom “data warehouse” based on the data models contained within their existing applications (i.e. EHR, revenue cycle). This “data warehouse” is not usually a single, monolithic structure, but a series of custom data bases linked together in a federated model.

The Second Approach. In contrast to the “build your own from scratch” approach, the second method for creating a “data warehouse” is to use a single vendor’s data model as a starting point and expand that data model to include additional, relevant data.

Besides the need to build an effective infrastructure to hold and disseminate big data, data governance is the next biggest challenge. Data Governance is the integration of people, processes, and technology that enables an enterprise to realize and leverage the value of its data and information. Recognizing the need for strong data governance, all study participants either had a data governance process in place, under development, or planned. 



“There is no one solution for big data. There is not one vendor that can handle multiple sets of data without tying to fit that data into a predefined data model with its own context for the data.”

“The concept of a huge data warehouse that is everything to everybody is not going to happen.”

“We bought a data model that proved inadequate and we are now adding additional data elements. Building out the new glossary is all part of the data governance process.”

“I proposed a solution, a single source of the truth program. We got the requirements and got agreement on the key data items needed to run the business. We then reached consensus on the definitions.”

The topic of the next blog entry will be Big Data: The Strategy – Pt 2

To hear more about our study, please contact Divurgent at

Leave a Reply

Your email address will not be published. Required fields are marked *