Before asking whether performance improved, establish who belongs in the measure, what counts as success, and which evidence can support the answer.
01
Define the population first
The denominator identifies eligible observations; the numerator identifies those meeting the measure’s criteria. Eligibility windows, exclusions, and qualifying evidence belong to the selected specification. Use the applicable year’s full methodology.
Cut points translate a measure result into a star category. They are not universal percentage targets. Different measures can have different directions: a higher rate may be desirable for one measure and undesirable for another.
The CMS performance-data collection is the starting place for official rating files and supporting documents. For an internal inventory, separate pharmacy, clinical, survey, and CMS administrative inputs so a missing source never silently becomes a zero.
01Give every measure a durable record02Trace the inputs before the result03Separate missing from not applicable04Reconcile with an exception register
COMMON QUESTIONS
A little more clarity.
The roadmap includes licensed Part C validation, X12 834 and NCPDP intake, and Part C research-chat coverage. Production specification fidelity remains dependent on licensing and reconciliation.
Is every CMS measure a Stars measure?
No. CMS uses measures across multiple quality programs. Confirm the program and reporting year before using a definition found in a search result.
Can claims answer every quality question?
No. A data feed can only establish what it contains. A claim may identify a service without the clinical result needed by a particular measure. Keep missing laboratory, clinical, or survey evidence visible.
Is this a licensed measure specification?
No. This is an educational navigation guide. It does not reproduce licensed PQA or NCQA specifications or value sets. Use appropriately licensed materials for production calculation and reporting.