Epic Analytics Intelligence Guide, 2026
The four-layer Epic analytics architecture, Clarity performance engineering, dashboard governance and revenue cycle intelligence.
What this covers
Epic analytics problems are usually architecture problems wearing a reporting costume. This guide starts from the four-layer stack, Chronicles through Clarity and Caboodle to the presentation layer, then works through the specific failure modes: ETL windows saturated by concurrent reporting, dashboards built on the wrong tier for their audience, and revenue cycle questions asked of tables that cannot answer them.
What is inside
- Chapter 1 · The Epic analytics architecture
- The four layers, what each is for, and a side-by-side of which use cases belong in Clarity versus Caboodle.
- Chapter 2 · Clarity performance engineering
- An issue, root cause and fix table covering the highest-volume tables, including PAT_ENC and CLARITY_MEDICATION.
- Chapter 3 · Dashboard strategy and governance
- Tiers dashboards by audience, data source, refresh rate and governance standard, so a board dashboard and a unit dashboard are not built the same way.
- Chapter 4 · Revenue cycle intelligence
- The dashboards most organizations do not have right, the Epic tables behind each, including ORDER_PROC and COVERAGE, and why they are usually wrong.
- Chapter 5 · Data quality
- Detection and remediation for the corruption that propagates silently into downstream reporting.
- Chapter 6 · The Cogito Cloud and Microsoft Fabric decision
- What the platform choice actually changes, and what it does not.
The point most teams miss
A frequent misdiagnosis: every report running on the Clarity server consumes SQL Server resources, so heavy concurrent reporting during the ETL window presents as "slow ETL" when the real problem is too many reports running at once.
Common questions
When should analytics use Clarity, and when Caboodle?
The guide treats this as the question that generates the most confusion in Epic analytics teams, and argues the answer is specific rather than situational. Using the wrong layer is expensive in both directions, and Chapter 1 gives a side-by-side of which use cases belong in each.
Why do Clarity reports slow down over time?
Epic's default index strategy is optimized for transactional writes, not analytical reads. High-volume tables such as PAT_ENC, CLARITY_MEDICATION and ORDER_PROC are queried constantly by reports but indexed for a different workload, so missing and stale indexes are the most frequently occurring performance issue.
Our ETL is slow. Is that always an ETL problem?
Often not. Every report running on the Clarity server consumes SQL Server resources, so heavy concurrent reporting during the ETL window presents as slow ETL when the real problem is too many reports running at the same time.
How should dashboards be governed?
By tier. Chapter 3 assigns different ownership models by audience: Tier 1 strategic dashboards for the board and C-suite run on Caboodle fact tables, refresh daily or weekly, and permit certified metrics only, where every metric has one approved definition and any change requires governance committee sign-off.
Get the full document
This page summarizes the guide. The full 10-page PDF includes the complete tables, checklists and worked detail.