Five-Star Health Inspections: Rebuild the Weighted Score Before Explaining a Rating Change

Reconstruct the score before discussing the stars

A citation count loses information about severity, survey date, repeat visits and scoring exclusions. The September 2026 CMS Five-Star Technical Users' Guide, pages 4–7, describes the health-inspection algorithm. Lower weighted scores represent fewer scored deficiencies and revisit points; the score is then translated into a rating using state thresholds.

Since July 2025, the method uses the two most recent standard surveys. The most recent receives a three-quarter weight and the second a one-quarter weight. Complaint and focused infection-control findings retain a three-year lookback: findings in the most recent 12 months receive three-quarter weight, while findings from months 13–36 receive one-quarter weight. A third standard survey may remain visible without contributing to the rating.

Build a citation ledger with a scoring disposition

For each finding, retain survey type and date, tag, scope/severity, status, base points, applicable period and whether the finding is included. The guide excludes Life Safety findings from this calculation. It also describes exclusions for specified tags and waivers. Citations pending IDR or IIDR remain visible but do not enter the score until the dispute process finishes.

Do not deduplicate solely by tag. The guide applies specific 15-day rules to overlapping standard, complaint and focused infection-control findings, including different handling of repeated infection-control citations. Preserve the original survey identities and apply the rule before calculating the cycle total. The same tag on two dates is not automatically one event.

Use the scope/severity table rather than an informal seriousness label. For example, ordinary D-level findings receive four points and E-level findings eight. Substandard-quality-of-care and past-noncompliance rules can change the points associated with a finding; the official classification must be preserved.

A worked score reconstruction

This fictional facility has no special scoring exclusions, duplicate findings or substandard-quality adjustments in the included findings. The review uses CMS's current weights and assumes all survey classifications have been verified.

The weighted score is 36 × 0.75 + 16 × 0.25 = 31. A first revisit adds no points under the guide; the example uses a second revisit, which adds 50 percent of the applicable cycle score. Do not add every listed revisit percentage cumulatively without applying the CMS method.

Suppose the older complaint finding ages beyond the 36-month lookback while the other facts remain unchanged. The older total becomes 12, and the weighted score becomes 36 × 0.75 + 12 × 0.25 = 30. That one-point improvement is an aging effect, not evidence of a new corrective action during the month.

Translate the score using the correct state context

Retrieve the state cut-point table for the relevant refresh and retain it with the calculation. CMS recalibrates state thresholds monthly, but generally holds a facility's health-inspection rating until its own weighted score changes. It is therefore insufficient to compare today's cut points with an unchanged historical score and announce that the rating must move.

The example supports a verified score of 31, followed by 30 after the assumed aging event. It does not establish a star rating without the applicable state data and any special rules. Facilities meeting the abuse-icon criteria have a health-inspection rating cap; facilities with only one standard survey have insufficient information for the rating. Check those conditions before signing an explanation.

Publish a change explanation that can be checked

A useful management note states the prior and current score, which findings entered or left, changes in revisit or dispute status, the applicable state thresholds and any cap. Keep a separate line for the overall facility rating, which also uses staffing and quality-measure ratings under its own combination rules.

For a multi-facility software decision, test whether an analyst can reproduce the score from dated source records and preserve the cut-point version. A dashboard that reports only the star and citation count cannot distinguish improvement in care processes from aging, reweighting or a resolved data issue. That distinction matters when leaders choose the next quality intervention.