What should a skilled nursing facility dashboard track between star rating updates?

A useful SNF dashboard tracks the inputs to your CMS star rating every day from your own systems: nursing hours per resident day, the MDS items behind quality measures, census, and whether charting is current. That way you see problems before Care Compare does.

A skilled nursing facility dashboard should track the things that drive your CMS star rating, every day, from your own systems: nursing hours per resident day, the MDS items behind the quality measures, census, and whether charting is current. The public “SNF score” on Care Compare tells you where you stood months ago. Your EHR and timekeeping data can tell you where you stand this week.

This is where DashPoint started. Before lakehouses and ML models, we built quality measure, staffing and census dashboards for life plan communities and skilled nursing facilities, fed straight from the EHR (case study). I spent 18 years in healthcare IT before that. Here is what we have learned about what belongs on the screen.

What people mean by “SNF score”

Usually the CMS Five-Star Quality Rating shown on Care Compare. Each nursing home gets an overall rating plus three domain ratings:

  • Health inspections, from state survey results.
  • Staffing, from the Payroll-Based Journal (PBJ) data the facility submits each quarter, including nurse hours per resident day, turnover and weekend staffing.
  • Quality measures, calculated from MDS assessments and Medicare claims.

CMS publishes the methods and cut points in its Five-Star Technical Users’ Guide and revises them regularly. In 2026 alone it respecified the long-stay antipsychotic measure and raised the quality measure rating thresholds. Medicare’s SNF Value-Based Purchasing program also produces a score, which affects payment; that is a separate calculation.

The problem for an operator is timing. PBJ is quarterly, claims-based measures use long look-back windows, and survey results arrive when the surveyors do. By the time a number moves on Care Compare, the staffing gap or missed assessment behind it is old news. Learning about your spring staffing from a public website in the fall is a humbling experience, and a common one.

Care Compare shows where you stood months ago; your EHR shows this week

What to put on the dashboard

Four areas cover most of what administrators and directors of nursing ask for.

Staffing

Nursing hours per resident day by discipline (RN, LPN, CNA), by day, using the same definitions PBJ uses. Weekend versus weekday. Agency hours. Turnover. If you calculate these daily from timekeeping and census, the quarterly PBJ submission should hold no surprises.

Quality measures

The MDS items behind the measures you care about most: falls with major injury, pressure ulcers, antipsychotic use, decline in daily living activities. Track them per resident and per unit as assessments are completed.

Your internal number will not match CMS exactly. CMS applies exclusions, risk adjustment and claims data you do not have. The internal version exists to show direction early. Compare it with Care Compare each time CMS refreshes and write down the gap, so everyone knows how far to trust it.

Census and occupancy

Census by payer and unit, admissions, discharges and length of stay. Census is the denominator for every staffing ratio, so get it right first.

Charting and assessment currency

Whether MDS assessments are on schedule and whether daily charting is complete. Missing documentation hurts twice: it weakens the quality measure data, and surveyors notice it.

How we would build it on Fabric today

Our first dashboards read the EHR directly. That works for one building. With several facilities, more source systems, and finance wanting the same census number as nursing, it is worth putting a data platform underneath. The pattern:

  1. Bronze: land EHR extracts (MDS assessments, census, charting events) and timekeeping and scheduling exports in a OneLake lakehouse, unchanged, as they arrive.
  2. Silver: one row per resident per day and one row per staff member per shift, with facility, unit and payer resolved. Most of the work happens here: matching employee IDs between timekeeping and scheduling, and agreeing on what counts as a census day.
  3. Gold: daily tables for hours per resident day, assessment status, quality measure flags and census, with each definition written down once.
  4. Reporting: a Direct Lake semantic model on gold, with row-level security so each administrator sees their own building.
  5. Checks: a nightly Data Factory pipeline that tests silver before it touches gold, so a bad extract stops the load before it can report a census of zero. A census of zero is great for the staffing ratio and bad for everything else. We used the same circuit-breaker pattern on a K-12 education platform.

Everything lives in Git with separate dev and prod workspaces, and the operator owns the tenant and the code.

Trade-offs to know before you start

  • EHR access sets the schedule. Some vendors offer APIs or data feeds, some offer scheduled exports, and some charge for either. Find out what you can get, and how fresh it is, before designing anything.
  • Resident data changes the setup. Plan workspace permissions, row-level security and who can see resident-level detail from day one.
  • Internal quality measures are estimates. Label them as early warning and show them next to the official CMS figures. If leadership starts treating the estimate as the rating, expect an awkward quarter.
  • A single building may not need a lakehouse. One facility with one EHR and a handful of reports can run on Power BI against a direct feed. The platform pays off with multiple buildings, multiple systems, or when finance and clinical leaders need to agree on the same numbers.

Where to start

Pick one measure that hurts. For many operators that is nursing hours per resident day: the data is internal, the PBJ definitions are published, and you can check your number against your last PBJ submission. Build that one end to end and reconcile it. Then add quality measures and charting.

Start with one measure: nursing hours per resident day

If you want help scoping it, our Fabric assessment covers your sources, the capacity you would need, and a build plan with costs. Get in touch.

  • Post-acute care
  • Skilled nursing
  • CMS star ratings
  • Microsoft Fabric
  • Healthcare analytics

Have a question like this?

Start a conversation
↑