The statements that accurately describe what Medicare measures when assigning star ratings are those identifying clinical quality measures, patient experience and access surveys, and administrative/operational data such as complaints and plan responsiveness. Medicare Advantage star ratings are not based on premiums, profits, or marketing spend. They are calculated from dozens of specific measures grouped into weighted categories, then converted into a single 1-to-5 score. Two true statements must name actual measure domains, not financial or enrollment outcomes. This explainer details the exact architecture behind those ratings.
How the star rating calculation actually works
The Centers for Medicare & Medicaid Services builds star ratings from a fixed menu of measures, then weights and transforms them mathematically. Understanding the pipeline explains why certain statements about the ratings are true and others are false.
Each contract year, CMS collects data across roughly 40 to 50 measures for Medicare Advantage plans with prescription drug coverage. Each measure falls into one of several domains: staying healthy, managing chronic conditions, member experience, member complaints, and customer service. Drug-plan measures form an additional set. Measures are not equally important; CMS assigns each a weight of 1, 2, or 3, and the triple-weighted measures drive most of the final score.
The transformation matters. CMS converts each measure's raw performance into a 1-to-5 star score using clustering algorithms, not simple percentage cutoffs. A plan at 85 percent on a measure might earn 4 stars in one year and 3 stars in another if the national distribution shifts. Those measure-level stars are then weighted and averaged into domain scores, and domain scores roll up into the summary rating. A plan's overall rating is therefore a weighted composite, not an average of averages.
The critical mechanism is the weighting and the domain rollup. A statement claiming ratings reflect "overall plan value" misdescribes the process, because value depends on premiums and cost-sharing, which are not star measures. A statement claiming ratings reflect "member satisfaction alone" is also wrong, because clinical quality measures carry substantial weight.
Plain language: the stars are a weighted report card built from checklists, not a popularity contest.
The measures, quantified
The measure set is finite and published. CMS releases the technical specifications annually, and the measures fall into recognizable buckets that map directly to the accurate statements.
Clinical quality measures include screening and treatment metrics. Examples include colorectal cancer screening, blood sugar control for members with diabetes, blood pressure control for members with hypertension, and medication adherence for statins, diabetes drugs, and blood pressure drugs. Adherence measures are typically triple-weighted because they predict outcomes and costs.
Patient experience measures come from the Consumer Assessment of Healthcare Providers and Systems survey, known as CAHPS. These measures capture access to care, communication with providers, and overall plan ratings. CAHPS measures are weighted heavily because CMS treats member experience as a distinct quality dimension.
Administrative and operational measures include complaint rates, call-center performance, and the timeliness of appeals decisions. CMS also includes measures on special needs plan care coordination and, in recent years, statistical adjustments for social risk factors.
The scale runs 1 to 5 in half-star increments for the summary rating. Plans at 4 stars or higher receive quality bonus payments, which is why the thresholds matter commercially. Approximately 5 percent of the summary rating can come from a plan's performance on CMS's own administrative review, depending on the year and contract type.
The takeaway: accurate statements must map to clinical measures, CAHPS measures, operational measures, or the weighting and cut-point mechanics.
Where accurate statements diverge from inaccurate ones
Statements about star ratings fail in predictable ways. The table below contrasts accurate and inaccurate framings across the domains people most often confuse.
| Statement theme | Accurate description | Why inaccurate versions fail |
|---|---|---|
| Clinical quality | Ratings include screening, chronic-condition, and adherence measures | Quality is not measured by plan premiums or network size |
| Patient experience | Ratings include CAHPS survey results on access and communication | Satisfaction is one domain among several, not the whole score |
| Complaints and appeals | Ratings include complaint rates and appeal timeliness | Complaint volume is a measure, not a marketing signal |
| Weighting | Measures are weighted 1 to 3 and rolled into domains | Simple averaging ignores documented weighting |
| Cut-points | Measure stars use clustering, not fixed percentages | Fixed-threshold claims contradict CMS methodology |
| Bonus payments | 4-star plans earn quality bonus payments | Bonus eligibility is not itself a measure |
| Enrollment | Ratings do not directly measure enrollment growth | Enrollment is an outcome, not an input to the score |
The pattern is clear. Statements that name a measure domain and its mechanism are accurate. Statements that name a business outcome, a premium level, or a single survey question are not.
A counterintuitive detail: a plan can improve its clinical scores and still lose a half-star if its CAHPS performance slips, because domain weights and clustering cut-points interact. Professionals watch the measure-level stars, not just the summary, because the summary hides which domain moved.
Edge cases that change what the ratings measure
Several structural features mean the ratings measure different things for different plan types.
Medicare Advantage plans with drug coverage are rated on the combined measure set. Standalone prescription drug plans are rated separately on drug-specific measures. Special needs plans for dual-eligible or chronically ill members have their own measure sets and weighting. Cost plans and demonstrations may be rated differently or not at all.
Contracts, not brands, are the unit of analysis. A single insurer may operate multiple contracts with different ratings, so a statement about "the company's rating" is imprecise. Ratings also apply to the contract's service area, so the same contract can perform differently across counties.
New plans and low-enrollment contracts may receive no rating, displayed as not applicable. A statement claiming every plan has a rating is false. The star rating also reflects a multi-year data window; some measures use one year of data, others use two or three, which smooths volatility but delays improvement.
Plain language: the rating belongs to a contract in a place, not to a logo.
Timeline: when the measures and ratings take shape
Timing explains why a statement about "current" ratings can be stale.
| Stage | Timing | What happens |
|---|---|---|
| Measurement year | Calendar year | Plans deliver care; data accrue |
| Data submission | Following spring | Plans submit HEDIS and administrative data |
| CAHPS survey | Following spring | Member experience data collected |
| CMS calculation | Following summer | Measures scored and weighted |
| Ratings release | Following October | Star ratings published for next year |
| Enrollment use | Following fall | Ratings displayed in plan comparison tools |
| Bonus payment | Following year | 4-star plans receive quality bonuses |
The lag is roughly 18 months from measurement to public display. A statement describing ratings as a real-time snapshot is inaccurate; they are a backward-looking composite.
Misconceptions and red-flag warnings
Misconceptions cluster around what the score represents and who controls it. Watch for these red flags in any claim about star ratings.
Claims that ratings measure "value for money" or premium competitiveness; cost-sharing and premiums are not star measures. Claims that a single survey question or a single clinical metric determines the rating; no single measure dominates. Claims that ratings are set by fixed percentage cutoffs; CMS uses clustering and weighting. Claims that a 5-star rating guarantees better outcomes for every member; ratings are population-level composites. Claims that plan ratings and drug-plan ratings are interchangeable; they use different measure sets. Claims that ratings update continuously; they are annual and lagged.
The professional caveat: because cut-points shift with national performance, year-over-year star changes can reflect the distribution rather than the plan. Analysts compare measure-level stars and the underlying rates, not just the summary.
Practical actions: how to verify a statement about star ratings
You can test any statement against the published methodology in a few steps.
First, identify the named measure or domain. If the statement cites a clinical measure, a CAHPS measure, a complaint measure, or the weighting system, it is likely accurate. Second, check whether the statement refers to a contract and service area rather than a brand. Third, check the measurement year, since ratings lag by roughly 18 months. Fourth, confirm the plan type, because Medicare Advantage, standalone drug plans, and special needs plans use different measure sets. Fifth, reconcile thresholds explicitly: the summary rating uses half-star increments from 1 to 5, the bonus threshold sits at 4 stars, and measure-level stars use clustering rather than fixed percentages.
Useful thresholds to remember: 4 stars triggers quality bonus payments; measures are weighted 1, 2, or 3; the measure count runs roughly 40 to 50 for combined contracts. If a statement conflicts with any of these, treat it as inaccurate until verified against the CMS technical specifications.
A practical verification habit: pull the plan's measure-level star table, not the summary, and confirm which domain moved.
Limitations of what the ratings can tell you
Star ratings measure documented performance on a fixed measure set, within a contract and service area, over a lagged window. They do not measure every dimension of care quality, and they do not measure premium value or network breadth directly. Measures change between years, so long-run comparisons require care. Plans with small enrollment or new contracts may lack ratings entirely. The ratings are a composite signal, useful for comparison but not a guarantee of individual outcomes.
For insurance and coverage decisions, this article is general educational information, not a recommendation or plan-selection advice. It does not promise any outcome or rating. Consult a qualified local insurance advisor or CMS plan-comparison tools for decisions about your specific coverage.