About the data
Everything here comes from public CMS files, loaded as published and kept with a record of the file and URL each row came from.
Data status
Last checked November 18, 2026 at 6:27 AM (after a scheduled refresh): some warnings · 109 passed · 3 warnings. The site checks for new CMS files on the 3rd and 18th of each month, then checks the data: freshness against CMS's schedule, totals against CMS's official counts, month-to-month jumps, file layouts, and agreement between sources.
All 112 checks
| Result | Area | Check | Detail |
|---|---|---|---|
| OK | Loads | File loads | 13 succeeded, none failed |
| OK | Freshness | Monthly enrollment | latest Nov 2026 |
| OK | Freshness | SNP report | latest Nov 2026 |
| OK | Freshness | CMS monthly totals | latest Nov 2026 |
| OK | Freshness | Contract directory | latest Nov 2026 |
| OK | Freshness | CMS compliance report | newest report September 2026 |
| OK | Freshness | Newest warning letters | sent 2026-05-15 (187 days ago; CMS usually posts them 4–5 months after sending) |
| OK | Freshness | Newest corrective action plans | sent 2026-05-26 (176 days ago; CMS usually posts them 4–5 months after sending) |
| OK | Freshness | CMS enforcement listing | 72 actions listed (newest 2026-11-09); 72 loaded, all tied to a contract or parent |
| OK | Freshness | Display measures | latest 2026; Star Ratings 2027 (CMS posts display measures a few weeks after the ratings) |
| OK | Freshness | Star Ratings | latest 2027; expected 2027 by now (each October for the next year) |
| OK | Freshness | Landscape premiums | latest 2027; expected 2027 by now (each September for the next year) |
| OK | Freshness | Plan payment data | latest 2024; expected 2024 by now (about 18 months after the year) |
| OK | Freshness | Medical loss ratio | latest 2023; expected 2023 by now (about 3 years after the year) |
| OK | Freshness | LIS enrollment | latest 2026; expected 2026 by now (each spring for that year) |
| OK | Freshness | ACO lists | latest 2026; expected 2026 by now (each spring for that year) |
| OK | Freshness | Shared Savings results | latest 2024; expected 2024 by now (each late summer or fall for the year before) |
| OK | Freshness | ACO REACH results | latest 2024; expected 2024 by now (about 18 months after the year) |
| OK | Freshness | ACO members by county | latest 2024; expected 2024 by now (each November for the year before) |
| OK | Freshness | Traditional Medicare by county | latest 2025; expected 2025 by now (each July for the year before) |
| OK | Freshness | Traditional Medicare costs (ratebook FFS data) | latest 2024; expected 2024 by now (each April, two years behind) |
| OK | Totals | MA members, Nov 2026 | ours 36,052,896 vs CMS official 36,053,156 (-0.0%) |
| OK | Totals | PACE members, Nov 2026 | ours 82,913 vs CMS official 83,226 (-0.4%) |
| OK | Totals | PDP members, Nov 2026 | ours 25,456,659 vs CMS official 25,552,335 (-0.4%) |
| OK | Totals | SNP members, Nov 2026 | SNP report 8,502,994 vs plan enrollment 8,502,994 (0.0%) |
| OK | Month to month | Drug plans, employer, Oct → Nov 2026 | 6,322,997 → 6,319,562 members (-0.1%); 285 → 285 plans |
| OK | Month to month | Drug plans, individual, Oct → Nov 2026 | 19,063,024 → 19,137,097 members (0.4%); 369 → 369 plans |
| OK | Month to month | MA employer group, Oct → Nov 2026 | 5,846,990 → 5,847,843 members (0.0%); 1377 → 1377 plans |
| OK | Month to month | MA individual, Oct → Nov 2026 | 21,693,742 → 21,702,059 members (0.0%); 3392 → 3392 plans |
| OK | Month to month | MA special needs, Oct → Nov 2026 | 8,483,873 → 8,502,994 members (0.2%); 1713 → 1713 plans |
| OK | Month to month | Plan IDs are 3 digits | 0 rows that aren't |
| OK | File layout | SNP report, Nov 2026 | 1,800 plans; enrollment filled 96.2%, SNP type 100.0%, contract IDs valid 100.0% |
| OK | File layout | Landscape 2025 columns | all 8 columns we use are present |
| OK | File layout | Landscape 2026 columns | all 8 columns we use are present |
| OK | File layout | Landscape 2027 columns | all 8 columns we use are present |
| OK | File layout | Landscape 2027 values parse | premium read for 100.0% of plans, MOOP for 100.0% |
| OK | File layout | Plan payment data 2024 | 5,539 plans (prior year 5,492); risk score filled 100.0%, A/B payment 100.0% |
| OK | File layout | Enrollment months for payment year 2022 | 12 of 12 monthly plan files loaded (used to weight payment data) |
| OK | File layout | Enrollment months for payment year 2023 | 12 of 12 monthly plan files loaded (used to weight payment data) |
| OK | File layout | Enrollment months for payment year 2024 | 12 of 12 monthly plan files loaded (used to weight payment data) |
| OK | File layout | Medical loss ratio 2023 | 635 contracts; MLR filled 100.0% |
| OK | File layout | Star Ratings 2027 | 605 rated contracts (prior year 621); 45 measures (prior year 45) |
| OK | File layout | MA contracts with a parent organization | 0 MA contracts with members have no parent organization |
| OK | Reconciliation | Rebuilt overall ratings, 2027 | 97.4% of 508 contracts match CMS's published rating exactly |
| OK | Reconciliation | Rebuilt part c ratings, 2027 | 100.0% of 515 contracts match CMS's published rating exactly |
| OK | Reconciliation | Member-months, 2023 | enrollment files 371,542,267 vs MLR filings 372,751,348 across 589 contracts (-0.3%) |
| OK | Reconciliation | Spend pools add up, 2023 | 15 of 15 pools match the MLR line totals for all MA |
| OK | Reconciliation | Spend pool peer ranks | 0 rows with a rank out of range or where the 1,000-member / 20-peer rule says otherwise |
| OK | Reconciliation | Admin booked as overhead, 2023 | 54 contracts marked; 0 with the mark on some rows only |
| OK | Reconciliation | Part D risk sharing, 2023 | CMS's settlement and plans' MLR filings agree for 540 of 634 contracts (85.2%); the rest are mostly estimates filed before CMS settled |
| OK | Prior authorization | Values found in their documents, 2025 | 4239 of 4239 matched the text of the plan's document |
| OK | Prior authorization | Contracts marked found have figures, 2025 | 387 of 387 |
| OK | Prior authorization | Standard approved + denied = 100% | 387 of 387 contracts within 0.6 points |
| OK | Inspector General | Listing, download and parsing | no open problems |
| OK | Inspector General | Reports checked by hand | 27 of 27 |
| OK | Inspector General | Audited contracts in the contract directory | all found |
| OK | Inspector General | Figures consistent | unsupported within sample, refund within estimate, every diagnosis audit has an estimate |
| Warning | Vendors | Quotes found in their sources | 1714 of 1739; run `ma vendors verify` or fix 25 |
| OK | Vendors | Rows tie to known parents, vendors and functions | all do |
| OK | Vendors | Contracts tie to known plans and vendors | 144 contracts |
| OK | Vendors | Contract amounts, terms and quotes look right | all do |
| Warning | Vendors | Vendor market price sources' quotes found | 128 of 138; the rest are on sites that refuse automated reads |
| OK | Vendor need | Map ties to known categories and signals | all rows do |
| OK | Vendor need | Every category lists what the data can't show | all do |
| OK | Vendor need | Categories with a strong signal | supporting signals only: specialty_pharmacy, fitness, telehealth |
| OK | Vendor need | Signals that fire for some contract | none fire now: letter:directory, stars:data_issue |
| Warning | Form 990s | Returns match the IRS copy | 0 of 1544 checked by `ma irs990 verify`; 1544 not yet checked |
| OK | Form 990s | Contractors classified | 0.0% of dollars unclassified |
| OK | Form 990s | Vendor payments tie to known vendors | all do |
| OK | ACO owners | Quotes found in their sources | 54 of 54 |
| OK | ACO owners | Rows tie to known parents and ACOs | all do |
| OK | Appeals (IRE) | Lines found in the fact sheets | 2767 of 2767 |
| OK | Appeals (IRE) | Newest quarter | Q2 2026; expected Q1 2026 or later |
| OK | Appeals (IRE) | CMS's counts against its printed totals | 33 lines in CMS's tables don't add up (counts used): 2021 Q1 grand total, 2023 Q3 EKG, 2024 Q2 Practitioner Services, 2024 Q2 Other, 2024 Q2 grand total, 2024 Q3 Outpatient Therapies … |
| OK | Appeals (IRE) | Service types and a grand total each quarter | every quarter (Maximus: 18 types; C2C: types add up to the total) |
| OK | Services (PBP) | Newest contract year | 2027; expected 2027 by now (CMS posts the next year's files in October) |
| OK | Services (PBP) | Medicare Advantage plans per year | 2025: 5,192, 2026: 5,095, 2027: 5,049 |
| OK | Services (PBP) | Service codes named in ref/pbp_service_categories.csv | all codes used by 10 or more plans |
| OK | Drugs (formularies) | Newest month | Oct 2026; expected Oct 2026 or later (CMS posts mid-month) |
| OK | Drugs (formularies) | Plans with a counted formulary | 5495 of 5495 plans in Oct 2026 |
| OK | Drugs (formularies) | Drug table for the newest month | 963,666 rows for Oct 2026 |
| OK | Drugs (formularies) | Formulary sizes | every formulary lists 500 or more drugs |
| OK | Drugs (formularies) | Formulary ZIPs kept on disk | 2 (the newest two are kept) |
| OK | MLR detail | Quality improvement lines add to the total, 2023 | 635 of 635 contracts within $1,000 |
| OK | MLR detail | Admin lines add to the total, 2023 | 635 of 635 contracts within $1,000 |
| OK | MLR detail | Supplemental categories vs total, 2023 | categories add to 100% of the supplemental total; 0 contracts' categories exceed it (shown as reported) |
| OK | MLR detail | PMPM matches CMS's, 2023 | 43180 of 43180 line values within a cent of CMS's PMPM column |
| OK | MLR detail | Worksheet lines mapped, 2023 | all lines mapped |
| OK | MLR detail | Contracts with a peer group, 2023 | 592 of 592 MA contracts |
| OK | HEDIS | MA HEDIS years | 2024, 2025, 2026; expected 2026 by now (CMS posts each October) and 3 years |
| OK | HEDIS | MA HEDIS 2026 layout | 631 contracts (prior year 661); rate filled for 72.1% of results, enrollment for 100.0% of contracts |
| OK | HEDIS | MA HEDIS 2026 indicator names | every indicator named from the dictionary or national rates |
| OK | HEDIS | SNP HEDIS years | 2024, 2025, 2026; expected 2026 by now (CMS posts each October) and 3 years |
| OK | HEDIS | SNP HEDIS 2026 layout | 1,110 plans (prior year 1,054); rate filled for 96.9% of results, enrollment for 100.0% of plans |
| OK | HEDIS | SNP HEDIS 2026 indicator names | every indicator named from the dictionary or national rates |
| OK | HEDIS | National rates match CMS's, MA 2024 | 88 of 88 within 0.05 points of CMS's published rate |
| OK | HEDIS | National rates match CMS's, SNP 2024 | 39 of 39 within 0.05 points of CMS's published rate |
| OK | HEDIS | National rates match CMS's, MA 2025 | 84 of 84 within 0.05 points of CMS's published rate |
| OK | HEDIS | National rates match CMS's, SNP 2025 | 35 of 35 within 0.05 points of CMS's published rate |
| OK | HEDIS | National rates match CMS's, MA 2026 | 71 of 71 within 0.05 points of CMS's published rate |
| OK | HEDIS | National rates match CMS's, SNP 2026 | 2 of 2 within 0.05 points of CMS's published rate |
| OK | HEDIS | HEDIS 2026 contracts in the contract directory | 631 of 631 (contracts not found have no parent page figures) |
| OK | LIS enrollment | LIS within enrollment, 2024 | all 6162 plans |
| OK | LIS enrollment | LIS within enrollment, 2025 | all 5799 plans |
| OK | LIS enrollment | LIS within enrollment, 2026 | all 5644 plans |
| OK | LIS enrollment | Contract totals, 2024 | 905 of 919 MA-PD contracts within 5% of our Apr 2024 enrollment (98.5%) |
| OK | LIS enrollment | Contract totals, 2025 | 878 of 885 MA-PD contracts within 5% of our May 2025 enrollment (99.2%) |
| OK | LIS enrollment | Contract totals, 2026 | 872 of 875 MA-PD contracts within 5% of our Apr 2026 enrollment (99.7%) |
| OK | Medicaid | Newest enrollment year | July 2024 |
| OK | Medicaid | Plans matched to a parent, 2024 | 100.0% of 67,854,946 comprehensive plan members; not matched: Washington: Multiple Sites |
| OK | Medicaid | Members against the year before, 2024 | -11.1% (76,360,620 to 67,854,946) |
| OK | Medicaid | Parent rules in use | every rule matches a plan in the latest three years |
When each source last had a new file
| CMS source | Published | Files loaded | Last new file |
|---|---|---|---|
| Medicare Durable Medical Equipment, Devices & Supplies - by Supplier and Service | annual | 3 | Nov 19, 2026 |
| Medicare Durable Medical Equipment, Devices & Supplies - by Geography and Service | annual | 3 | Nov 19, 2026 |
| Medicare Physician & Other Practitioners - by Geography and Service | annual | 3 | Nov 19, 2026 |
| Medicaid Managed Care Program Annual Report (MCPAR) public use file | annual | 2 | Nov 18, 2026 |
| Part C Reconsideration Appeals fact sheets (Independent Review Entity) | quarterly | 2 | Nov 18, 2026 |
| County-level Aggregate Expenditure and Risk Score Data on Assignable Beneficiaries | annual | 6 | Nov 18, 2026 |
| MA State/County Penetration | monthly | 37 | Nov 18, 2026 |
| MA and PDP Landscape files | annual | 5 | Nov 18, 2026 |
| Special Needs Plan (SNP) Comprehensive Report | monthly | 37 | Nov 18, 2026 |
| Part C and D Enforcement Actions (CMPs, sanctions, terminations) | ad_hoc | 254 | Nov 18, 2026 |
| Part C and D Star Ratings Technical Notes | annual | 5 | Nov 18, 2026 |
| Monthly Enrollment by Contract/Plan/State/County | monthly | 36 | Nov 18, 2026 |
| Monthly Enrollment by Plan | monthly | 59 | Nov 18, 2026 |
| Monthly Contract and Enrollment Summary Report | monthly | 37 | Nov 18, 2026 |
| MA Contract Directory | monthly | 2 | Nov 18, 2026 |
| Public RFP documents (RFPs, questionnaires, bidder Q&A, proposals) | ad_hoc | 74 | Nov 17, 2026 |
| Form 990 e-file XML (bulk downloads) | monthly | 8 | Nov 14, 2026 |
| Medicaid Managed Care Enrollment by Program and Plan | annual | 1 | Nov 10, 2026 |
| PBP benefits (tab-delimited files and data dictionary) | quarterly | 3 | Nov 3, 2026 |
| HHS OIG audit reports on Medicare Advantage organizations | ad_hoc | 27 | Nov 3, 2026 |
| Medicare FFS data for the MA ratebooks | annual | 3 | Nov 3, 2026 |
| Shared Savings Program and ACO REACH public use files | annual | 24 | Nov 3, 2026 |
| Monthly Prescription Drug Plan Formulary and Pharmacy Network Information | monthly | 13 | Nov 3, 2026 |
| PBP benefits (per-plan JSON) | annual | 3 | Oct 25, 2026 |
| LIS Contract Enrollment by County | annual | 3 | Oct 25, 2026 |
| LIS Enrollment by Plan | annual | 3 | Oct 25, 2026 |
| SNP HEDIS Public Use Files | annual | 3 | Oct 25, 2026 |
| MA HEDIS Public Use Files | annual | 3 | Oct 25, 2026 |
| Medicare Advantage Ratebooks | annual | 3 | Oct 24, 2026 |
| Part C and D Medical Loss Ratio Public Use Files | annual | 3 | Oct 20, 2026 |
| Part C and D Plan Payment Data | annual | 5 | Oct 20, 2026 |
| Part C and D Display Measures | annual | 3 | Oct 20, 2026 |
| Part C and D Compliance Actions (warning letters, ad hoc CAPs) | monthly | 3 | Oct 20, 2026 |
| Part C and D Star Ratings Data Tables | annual | 13 | Oct 20, 2026 |
Sources
- Star Ratings data tables: measure scores, measure stars, domain and summary ratings, cut points, low/high-performing icons, CAI. Where CMS re-issued a table (e.g. the July 2024 and July 2026 recalculations) the latest release is used.
- Display measures (Quality › Display measures): about 60 measures a year CMS publishes with the Star Ratings but doesn't score, 2015 onward, with CMS's national averages from the Display Measures Technical Notes. CMS renumbers them most years, so they're matched across years by name; which direction is better is set by this site.
- HEDIS (Quality › HEDIS): CMS's MA and special needs plan HEDIS public use files, the latest 3 years, posted each October for care the year before. Every contract's audited rate with its numerator, denominator and 95% confidence interval; the Health Outcomes Survey measures (falls, bladder control, physical activity); and risk-adjusted readmissions and preventable hospital stays as observed ÷ expected. National rates weight each contract's rate by its December enrollment, CMS's own formula, and reproduce the rates CMS publishes. A measure links to the Star measure with the same HEDIS code. CMS leaves out contracts under 500 members, and rates on too few members show as NA.
- Star Ratings Technical Notes: each measure's description, calculation, data source, measurement period, weight, star-assignment method, case-mix and CAI use, exclusions.
- Monthly enrollment by contract/plan and by contract/plan/state/county. CMS replaces counts of 10 or fewer with "*"; those are treated as unknown, not zero, and flagged.
- MA contract directory: legal entity, parent organization, tax status, region.
- Monthly Contract and Enrollment Summary Report: CMS's official, unsuppressed totals by contract type (MA, PACE, cost, PDP), used for headline totals and to check plan-level sums.
- Prior authorization (Prior Auth): the figures each organization posts on its own website for the previous year; the service categories each plan's benefit package says need approval in network (CMS's plan benefit package files, 2025 on); the share of each Part D formulary's drugs needing prior authorization, step therapy or a quantity limit (CMS's monthly formulary files); and the national outcomes of appeals sent to the Independent Review Entity (CMS's quarterly fact sheets, 2021 on).
- Traditional Medicare services by state (Vendors › Market size › Chronic condition management): CMS's Medicare Physician & Other Practitioners by Geography and Service file on data.cms.gov, the latest 3 years: Part B services, beneficiaries and dollars Medicare paid clinicians, by state and billing code. Each year comes out about 18 months after it ends. Codes billed for 10 or fewer beneficiaries in a state are left out of that state. We describe each code in our own words. MA plans' payments for the same services aren't public; where a page shows the same use at MA's size, it is our estimate at traditional Medicare rates.
- Traditional Medicare equipment and supplies (Vendors › Market size › Chronic condition management): CMS's Medicare Durable Medical Equipment, Devices & Supplies files on data.cms.gov, by Geography and Service (by the state of the clinician who ordered them) and by Supplier and Service, the latest 3 years. We keep suppliers that are organizations, with city and state only. Suppliers are grouped under their owner only where a public source states it, cited on the page; others appear under the name they enrolled with.
- ACOs and traditional Medicare (Organizations › ACOs): CMS's Shared Savings Program and ACO REACH public use files on data.cms.gov, the latest 3 years of each: the ACO lists, each ACO's financial and quality results, its members by county, and traditional Medicare spending and risk scores by county for beneficiaries who had primary care there. Results for a year come out the following late summer or fall (REACH later). Counties are matched to the penetration file by CMS's SSA codes. Each ACO's participants and skilled nursing facility affiliates come from CMS's current participant lists. Which ACOs a Medicare Advantage parent organization owns is researched ACO by ACO, each tie cited to a public source (the parent's or its subsidiary's own pages, its press releases and filings, or CMS's participant lists) and marked stated or inferred.
- Medicaid managed care (Organizations › Medicaid and Markets › Medicaid): CMS's yearly Medicaid managed care enrollment report on data.medicaid.gov, members on July 1 by state, program and plan, as each state reports them. It comes out about 19 months after the July it counts. Totals count comprehensive managed care plans only; behavioral health, dental, transportation, PACE and case-management programs are listed but not added in, since one person can be in more than one. Plans are matched to parent organizations by this site, using the plan's brand and the parent the state reports, so the Medicare Advantage parent and its Medicaid plans line up. Medicaid is shown beside a parent's Medicare Advantage figures, never added to them.
- Traditional Medicare spending per member (Money › ACOs and the Overview): CMS's county data behind the MA ratebooks: enrollment and Part A and Part B payments for every aged and disabled traditional Medicare beneficiary, and county risk scores normalized to 1.0. It comes out each spring, about two years after the year it covers. It is set beside MA plans' own Part A and B claims from their medical loss ratio filings and CMS's plan payment risk scores.
- Contract types: "MA" here means local and regional coordinated care plans, PFFS and MSA. PACE, cost plans (1876/1833) and demonstrations also use H contract IDs but are counted separately; PACE has its own page.
How the Star outlook works
For each rated contract, the outlook answers three questions about the next Star Ratings release: how close it is to its next half star and how far from losing one, the most likely way up and down, and the chance of each.
- Rebuilding CMS's rating. CMS publishes measure stars and the rounded rating, not the unrounded score in between. The site rebuilds it with CMS's own arithmetic from each year's Technical Notes: the weighted mean of measure stars, the reward factor for high and consistent performance (CMS's published thresholds), the low-income/disability adjustment (CAI), the improvement- and new-measure hold-harmless rules, and CMS's rounding table. It reproduces CMS's published ratings exactly: overall 2024 95.8%; overall 2025 99.2%; overall 2026 90.5%; overall 2027 97.4%; Part C 2024 98.9%; Part C 2025 99.8%; Part C 2026 100.0%; Part C 2027 100.0%. The 2026 overall misses are CMS's July 2026 recalculation, which raised some ratings without changing any measure stars.
- Distances and paths. Each measure's distance to its next cut point and its cushion above its current one, against next year's projected cut points (each boundary moves by its typical yearly move since 2019, capped at CMS's 5-point guardrail for non-survey measures), also expressed in "typical yearly moves" for that measure. The paths are the fewest one-star gains or slips, easiest first, that change the rounded rating, each recomputed exactly.
- Chances. 400 simulated years per contract. Each draws a real past year: cut points move as they did that year, each measure score moves the way scores of contracts at the same star level moved that year, plus a contract-wide good- or bad-year shift. The whole market is simulated together, so CMS's reward thresholds are recomputed each time. The share of simulated years landing on each rating is then calibrated against what actually happened in past years, taking contract size and whether the contract is at 4★ or more today into account.
Tested on past years. Run as if it were 2024, 2025, 2026 (using only data available then; the calibration for each year fitted on the other years) and compared with what CMS published a year later, across 1446 contract-years: Brier score 0.5904 (lower is better) against 1.1342 for "no change" and 0.649 for each year's actual rates of moving, and the exact next rating called 45% of the time against 43% for "no change". When it said 30–40% chance of dropping, 39% dropped.
Limits. Ratings are noisy: no method calls the exact next rating much more than half the time. The outlook assumes this year's weights and rules carry over, with one exception: measures CMS has already removed from the next release are left out, and the reward-factor thresholds are adjusted for the smaller set. For the 2028 Star Ratings that's Statin Therapy for Patients with Cardiovascular Disease and the Call Center – Foreign Language Interpreter and TTY measures (Part C and D), removed by CMS's 2027 Medicare Advantage rule. That rule also drops more measures from the 2029 ratings (appeals, complaints, customer service, members choosing to leave and others) and adds a depression screening measure; the outlook covers only the next release, so those aren't modeled yet. CMS changes other details every year. Scores are CMS's displayed (rounded) values. Improvement measures are held at today's stars. "Quality bonus at risk" is rough: about 5% of a contract's risk-adjusted A/B payments, ignoring double-bonus counties and how plans bid.
Themes (Appeals, MTM, Drug Plan, Medication Adherence, Member Experience…) group measures by pattern so a theme follows a measure even when CMS renumbers or renames it. A contract's theme score is its weight-averaged stars on the measures in that theme.
Compliance and enforcement coverage
- Warning letters from September 2024 (CMS's current report starts there; older reports are no longer published).
- Ad hoc CAPs from 2012. CAPs on the topic "Star Ratings" are CMS's response to low ratings and are counted separately from operational CAPs.
- Civil money penalties, sanctions and terminations: CMS's current list (2021 onward) plus archived copies of earlier lists from the Internet Archive. Coverage of earlier years may have gaps. Contracts come from the contract numbers printed in the CMS letter. When a letter prints none, the action is linked to the contracts whose legal entity CMS names; if CMS posted no letter, to the contracts the same organization's other letters name (a sanction and its release, say). When CMS names only a parent organization, the action is shown on the parent's page and the CMS actions list as organization-wide, and isn't attributed to any one contract.
- A civil money penalty amount is for the whole letter, which can cover many contracts.
- Letters: every action links to the CMS letter where one exists. Warning letter and CAP letters come from the attachments in CMS's monthly report files; enforcement letters were downloaded from CMS (or the Internet Archive) when loaded, because CMS has since moved or removed many older ones. "CMS source" links to the CMS report file or the page (or archived copy) that listed the action.
Money: payments and medical loss ratio
- CMS plan payment data, the latest 3 payment years. CMS publishes each year about 18 months after it ends, once Part D is reconciled. Per plan: average Part C risk score, A/B payment per member per month (standardized to a 1.0 risk score), rebate; Part D risk score, direct subsidy, reinsurance and low-income cost-sharing payments. Per contract: the Part D year-end reconciliation (risk sharing, reinsurance, low-income cost sharing).
- CMS gives no member counts, so plans are weighted by member-months from CMS's monthly plan enrollment files (loaded back to January of the earliest payment year for this). Plans CMS suppresses for 10 or fewer members weigh 0.
- Estimated Part C revenue = A/B payment × risk score + rebate, per plan, times member-months. Against the revenue plans reported in their 2023 MLR filings it comes within about 5%, usually a little under. Member-months agree with the MLR filings within about 1%.
- Part D risk sharing: positive = CMS paid the plan because drug costs ran over its bid; negative = the plan paid CMS back. Plans also report it as revenue ("Part D Risk Corridor Payments") in their MLR filings. For 2023 the two agree for about 85% of contracts; most of the rest filed an estimate before CMS settled (Centene's Wellcare drug plan S4802 reported $300M, CMS paid $157M), so this site uses CMS's settled figure.
- Medical loss ratio: CMS's public use files, the latest 3 contract years, about 3 years behind. Each year gives the adjusted MLR and any remittance owed; from 2023 CMS also publishes the full MLR form (revenue, claims by type, supplemental benefit spending by category, admin costs, broker commissions, member-months).
- Parent figures add up contracts under their current parent, as everywhere on this site.
How the financial pressure score works
A 0–100 score of signs that a Medicare Advantage contract's business is under financial strain. Higher means more strain. It uses only CMS's public data and adds up six parts, each capped so no single sign dominates:
| Part | Up to | What counts | Source |
|---|---|---|---|
| Part D losses | 30 | Part D risk sharing CMS paid the contract because drug costs ran over its bid, per member per year. Nothing up to $50; full points at $200. For a parent, its stand-alone drug plans count too (see below). | CMS plan payment data, latest payment year |
| Star bonus at risk | 20 | All 20 if the contract fell below 4★ this year and lost the quality bonus. At 4★ or more, 20 × the Star outlook's chance of losing 4★ at the next release. Below 4★ in both years, nothing: there's no bonus left to lose. | Star Ratings and the Star outlook |
| Thin margin | 15 | Medical loss ratio above the 75th percentile of the scored contracts' ratios (94.7% for 2023); full points at 100%, where claims and quality spending use up all revenue. Most MA contracts run above 90% (the 2023 median was 90.8%), so a fixed 90% line would flag half of them; only the highest quarter counts. | CMS medical loss ratio files, latest year |
| Rebate falling | 15 | Rebate per member down from the first to the latest payment year loaded; full points at a 25% fall. | CMS plan payment data |
| Losing members | 10 | Enrollment change over the last 12 months; full points at a 15% loss. | CMS monthly enrollment |
| Premium or MOOP up | 10 | 5 each for the premium and the in-network out-of-pocket maximum: the share of the contract's members in plans raising it next year, counting only the part above the market's 75th percentile (most plans raise the MOOP in a typical year, so only unusual increases score), scaled so 100% of members gets the full 5. | CMS Landscape files, next year against this year |
- Between the thresholds, points rise in a straight line: a rebate down 12.5% scores 7.5 of the 15, for example.
- Not scored: a medical loss ratio below 85% is shown as a flag instead. It means the contract owed CMS a remittance, which usually signals high profit, not strain.
- Parents get the member-weighted average of their MA contracts' points for each part, except Part D losses: those pool the Part D risk sharing CMS paid on the parent's MA contracts and its stand-alone drug plans (PDPs), divided by all their Part D members, so a parent losing money on drug plans sold apart from MA shows it. The parts add up to the total.
- Which contracts: MA contracts with members in the latest month (not PACE, cost plans or stand-alone drug plans, though a parent's drug plans count in its Part D part). A part with no data, such as a contract with no medical loss ratio filing yet, scores 0.
- Timing. The parts come from different years. Medical loss ratios are about three years behind and payment data about 18 months; enrollment, Star Ratings and next year's premiums are current. Insurers see their own costs much sooner, so the score can miss strain that began in the last year or two.
- What it isn't: a measure of profit or solvency. It counts public warning signs: a contract can score low and still lose money, or score high after one bad year and recover.
Plans: special needs plans and premiums
- Kinds of plan come from CMS's monthly enrollment by plan: contract type (MA, PDP, PACE, cost), plan type (HMO, PPO, PFFS, MSA), employer group (plan IDs 800 and up) and the SNP flag.
- Special Needs Plan report (CMS, monthly, same 36 months as enrollment; CMS published no January 2024 report): each SNP's type (dual-eligible, chronic condition, institutional), D-SNP integration with Medicaid (CO, HIDE, FIDE, AIP) and C-SNP conditions. Employer plans are omitted, and plans with 1–10 members are blank.
- Landscape files (CMS, the latest 3 years, released each September for the next year): every approved plan by state and county with monthly premium (Part C + D) and in-network maximum out-of-pocket (MOOP). Employer plans and PACE are not included. Where a plan's segments differ, pages show the lowest value with a +.
- Low-income subsidy (LIS) enrollment (CMS, once a year in the spring, the latest 3 years): each plan's Part D members and how many get Extra Help. LIS share = LIS members ÷ Part D members; plans where CMS hides the LIS count (10 or fewer) are left out of the share. CMS doesn't say which month it counts; its counts line up most closely with the April or May monthly enrollment report, and each page names the month. Parent shares cover MA contracts under their current parent.
- Next-year changes match today's plans to the newest Landscape year by contract and plan ID. A plan ID missing next year shows as "ending"; members are usually moved to another plan by the organization, and a renumbered plan also shows as ending.
Caveats
- CMS assigns stars using unrounded scores; displayed scores are rounded, so a score shown exactly at a threshold may sit on either side.
- Enrollment figures on the leaderboard are the latest loaded month, whatever star year is selected.
Sources, rights and corrections
- Public sources only. CMS and other federal data are loaded as published. Other public documents (plans' own websites and booklets, public agencies' board papers and RFPs, company filings and press releases) are read for facts, quoted in at most 25 words, and linked to the source. The site doesn't republish them. The only documents it serves itself are CMS letters, which are federal documents, unaltered. Where a public agency's RFP document has moved and we hold a copy, members can ask for it and we send it on request.
- How we collect. Automated collection names this site, follows each site's robots.txt, and stops at logins and blocks. It never gets past them.
- News shows the headline, publisher and date, linked to the publisher. Read the article for what it says.
- Estimates. Scores, the Star outlook, market sizes and vendor links marked "inferred" are this site's estimates from public data. Each one's method is in its ⓘ note. Reported figures come from the file named beside them.
- No affiliation. This site isn't affiliated with or endorsed by CMS, HHS, NCQA, AHRQ, or any plan, vendor or other organization it names. Names are used to identify organizations, and trademarks belong to their owners. HEDIS® is a registered trademark of the National Committee for Quality Assurance (NCQA). CAHPS® is a registered trademark of the Agency for Healthcare Research and Quality (AHRQ).
- Not advice. The site is for research. It isn't legal, financial, medical or enrollment advice. People choosing a Medicare plan should use Medicare Plan Finder.
- Corrections and removal. If something about your organization is wrong, or you'd like a vendor link or quote looked at, use "Report an error or ask for removal" on the page, or email [email protected]. We look within 5 business days, and a disputed vendor link is hidden while we check it.
Access and cookies
- The Overview, Plans, Markets, Quality, Prior Auth and Money, with their contract and parent pages, are open to everyone. Oversight, Vendors, news and the members' detail sections need a verified email: enter it on the sign-in page and use the 6-digit code (or link) we email you. New sign-ups get one week of access.
- When access ends, email [email protected] about your use case to discuss whether continued access makes sense.
- Cookies: signing in sets one cookie,
ma_session, which keeps you signed in until your access ends (at most 30 days) or you sign out. It's needed for the site to work, so there's no cookie banner. There are no analytics, tracking or advertising cookies. - Your email is used only to give you access and to see which pages are used. It's never shared or sold. See the privacy policy and terms of use.
- AI agents and scripts can use the site on a member's behalf: connect Claude (or another MCP app) by signing in with your email, or make an agent key. Their requests count as that member's: same end date, and they show in the member's activity. Automated access through a signed-in browser isn't allowed: accounts that request pages faster than a person reads them, or that use crawlers or scripts without a key, are blocked.
Version
| Environment | Production (payeralmanac.com) |
| Version | v2026.11.21-1 |
| Code | release @ 564c17f, committed Nov 21, 2026 10:30 AM |
| Deployed | November 21, 2026 at 10:42 AM |
| Database | ma, 74 migrations (latest 2026_11_20_2001_visits.sql) |