Main Menu

Login Form

In order for data to fit into an enterprise data model, data must undergo massive transformations. Binding is the process of mapping the data in the enterprise data warehouse from source systems to standardized and central vocabularies and business rules. This is done so the data can be brought together for analysis. A lot of large healthcare organizations have hundreds of analytics vendors supplying data. All that data has to be brought into the enterprise data warehouse (EDW) https://www.healthcatalyst.com/late-binding-data-warehouse/  in order to ensure the possibility reliable reporting and analysis.

Knowing when and how tightly to bind data to rules and vocabularies is critical to the ability and success—or failure— of a data warehouse. In healthcare, the risks of binding data too tightly to rules or vocabularies are particularly high because of the volatility of change in the industry. Business rules and vocabulary standards in healthcare are among the most complex in any industry, and they undergo almost constant change.

Health Catalyst has brought Late-Binding to the forefront of Data Warehouse technology. The Late Binding Warehouse is a revolutionary architectural model for healthcare analytics. When healthcare organizations combine their enterprise data warehouse with late-binding, they quickly progress to advanced stages of registries and reporting population health and clinical and financial risk modeling.

Early binding is the process where traditional data warehouses try to model the perfect database from the outset, determining in advance every possible business rule and vocabulary set that will be needed. This early binding practice is a time-consuming, expensive undertaking. In healthcare, business rules and vocabularies are dynamic and improve rapidly – and so do the use the cases that data linked across different source systems can serve. Mappings must be redone again and again as data models shift. Early binding architectures – like those espoused by Bill Inmon, Ralph Kimball, and others – force early data bindings into proprietary enterprise data models. Time has proven early-binding architectures to be inflexible, one-size-fits-all solutions, enforcing a compromised, least-common-denominator warehouse.

Health Catalyst’s Late-Binding architecture avoids the consequences of linking data with volatile business rules or vocabularies too early. By waiting to bind data until it’s time to solve an actual clinical or business problem, analysts:

  • Don’t have to make lasting decisions about a data model up front when they can’t see what’s coming down the road in two, three, or five years
  • Quickly adapt to new questions and use cases
  • Have the data they need to perform timely, relevant advanced analytics

Health Catalyst’s Late-Binding architecture avoids wasted time and effort by waiting to bind data until a business case drives it, ensuring data retains its original, undiluted value.

Health Catalyst’s Late-Binding Principles

Health Catalyst’s late-binding principles have been used for over 20 years. These principles continue to deliver proven records for military, manufacturing and healthcare organizations that operate under them. These principles summarized are;

  1. Minimize remodeling data in the data warehouse until the analytic use case requires it. Leverage the natural data models of the source systems by reflecting much of the same data modeling in the data warehouse.
  2. Delay binding to rules and vocabulary as long as possible until a clear use case requires it.
  3. Earlier binding is appropriate for business rules or vocabularies that change infrequently or that the organization wants to lock down for consistent analytics.
  4. Late binding in the visualization layer is appropriate for what-if scenario analysis.
  5. Retain a record of the changes to vocabulary and rule bindings in the data models of the data warehouse. This will provide a self-contained configuration control history that can be invaluable for conducting retrospective analysis that feeds forecasting and predictive analytics.