The phrase Enterprise Data Governance is often meant to encompass governance of everything data related in an organization, from data entry to storage to analysis to reporting and advanced analytics. However, the term Analytics Governance has emerged to address the use of data to answer questions. In this context, Data Governance applies only to the part of the data life cycle from data entry through storage/data management.
Because Enterprise Data Governance is such a large and often overwhelming undertaking, breaking down the effort into smaller chunks is helpful and necessary as long as we don’t lose sight of the big picture. For example, analytics is much more difficult and requires extensive manual scrubbing and mapping when the underlying data isn’t clean and standardized. So if Analytics Governance is attempted in isolation of Data Governance, it won’t get very far very fast.
Current EDG Subdivision
In the diagram above, there’s something missing that is vitally important when it comes to data, at least in the healthcare context we operate in: metrics. Since metrics are a use of data, it could be argued that they should fall under Analytics Governance. But we believe that metrics are worthy of being called out as a separate category for several reasons:
- Metrics garner a high amount of attention in healthcare organizations and are important to all levels of administration, as well as to caregivers.
- Healthcare metrics are complex; and the more clinical the subject matter, the more complex metric inclusions and exclusions become.
- Performance against externally defined metrics is often tied to reimbursement by both government and commercial payers.
- Performance against externally and internally-defined metrics is often tied to individual compensation (primarily of physicians and executives).
- Multiple departments within a healthcare organization develop metrics, and those metrics can be very similar (for example, both Finance and Performance Improvement are often interested in “average length of stay”; however, they are usually calculated differently). This seeming duplication can cause enormous confusion.
- An incredible amount of waste and rework is involved in the development, validation, and maintenance of metrics across the organization.
- “Analytics” tends to refer to answering questions that haven’t already been answered, performing data discovery, and looking out to the unknown future. Metrics, however, are generally products of “settled science” (so to speak). For example, the best practice for treating patients presenting at the emergency department with stroke-like symptoms is known; metrics are used to measure and monitor adherence to those best practices.
Therefore, we propose adding a third subdivision of the discipline of Enterprise Data Governance: Metric Governance.
Proposed EDW Subdivision
Like Analytics Governance, attempting Metric Governance in isolation is very challenging and inefficient, so the underlying data (Data Governance) must be taken into consideration as well. But addressing metrics as a way to divide up an overwhelming and complex body of work into smaller focus areas just makes sense. Plus the tasks involved with Metric Governance are unique in comparison to Data and Analytics Governance.
For each metric, the governance process involves:
- Identifying an owner responsible for the definition and performance of the metric
- Gathering input from those with “skin in the game” related to the metric, as well as any external definitions that need to be considered, and then facilitating an agreed-upon enterprise metric definition
- Creating the detailed metric definition with inclusions and exclusions fully spelled out
- Creating and monitoring goals related to the metric
- Working with the metric developer to ensure the actual metric matches the agreed-upon definition
- Monitoring and validating the metric on an ongoing basis to ensure it continues to measure what was intended (especially since underlying data can change without anybody knowing about it!)
- Working with BI developers to make sure metrics are being pulled from the metric repository and not re-created in each report and dashboard, and that metrics are being displayed in a helpful and informative way
- Creating metric documentation and ensuring that information is widely and readily available (and understandable) to those throughout the organization who require it
Adding Metric Governance as a subdivision under Enterprise Data Governance may be an urgent need only in the healthcare industry, due to the incredible complexity of and focus on metrics. We’ll leave it to others to determine if their industries would benefit from this nuance. But regardless of the world outside healthcare, we’re convinced that healthcare organizations would benefit from a much more formal discipline of Metric Governance.
Interested in exploring this idea further? Contact us to discuss the Compendium Metric Hub, the intelligent metric hub that stops “metric chaos” before it starts. The Metric Hub comes complimentary with the Compendium data catalog solution.
