- MS-DRG
- CMS relative weight
- MS-DRG CMI
- APR-DRG
- SOI subclass
- ROM subclass
01 · Working principle
Decompose the average
Begin with eligible discharge counts and matched-weight coverage. Then compare the distribution of MS-DRGs and identify which groups contribute most to the weighted difference. Stratify only where definitions and sample sizes support the review.
- Confirm the discharge-date fiscal year and weight file.
- Show eligible, matched, unmatched, and excluded discharges.
- Identify high-contribution DRG shifts rather than treating CMI as a black box.
- Review coding or documentation as possible lenses, not presumed causes.
02 · Working principle
What not to conclude
A higher CMI does not automatically mean patients were clinically sicker, documentation improved, revenue increased, or performance improved. Those propositions require their own evidence. CMI is useful precisely because it is bounded: an aggregate of applicable weights, not an all-purpose hospital score.
Executive checklist
What leaders should ask next
- Did the eligible denominator or match rate move?
- Which DRGs explain most of the weighted change?
- Are weights aligned to discharge-date fiscal year?
- Which interpretation remains a hypothesis?
Interpretation boundary
Limits to keep visible
- CMI is an aggregate of applicable DRG relative weights, not a complete measure of patient severity or hospital performance.
- MS-DRG and APR-DRG should not be combined into one synthetic metric.
- Classification and payment references change; use the appropriate effective version rather than automatically using the newest file.
Authoritative references
Sources and method notes
References reviewed September 28, 2026. They support the definitions and governance context; the operating-review guidance is sanalytics’ editorial interpretation, not an endorsed industry standard.
- Centers for Medicare & Medicaid Services: Case Mix Index
CMS definition of hospital CMI using Medicare discharges; the page distinguishes transfer-adjusted and unadjusted calculations.
- Centers for Medicare & Medicaid Services: MS-DRG Classifications and Software
Official MS-DRG classifications, software, and effective periods, including midyear grouper updates. Use the version applicable to the discharge date.
- Centers for Medicare & Medicaid Services: FY 2026 IPPS Final Rule Home Page
FY 2026 Table 5 includes MS-DRG relative weights and geometric and arithmetic mean length of stay. It is a named reference year, not a recommendation to use FY 2026 for every cohort.
- Agency for Healthcare Research and Quality: APR-DRG Severity of Illness
APR-DRG severity-of-illness context and four-level subclass structure.
- Solventum: APR-DRG Classification System
The methodology owner describes APR-DRG relative weights, separate four-level SOI and ROM subclasses, and the need for weights appropriate to the population.