How this is built, what it gets right, and where it is honest about being an estimate

The question the tool answers

If your hospital treated the exact same Medicare patients in FY2027 that it treated in 2024, what would CMS pay you, and how does that compare with FY2026 rates? Volume and coding are deliberately held constant, so every dollar of difference is attributable to what CMS changed: relative weights, standardized amounts, your wage index, and policy adjustments.

Sources, all public

• Caseload: CMS Medicare Inpatient Hospitals by Provider and Service, calendar 2024 (April 2026 refresh). Discharges and actual payments by hospital and MS-DRG.
• Weights and titles: IPPS Final Rule Table 5, FY2026 (CMS-1833-F) and FY2027 (CMS-1849-F, 91 FR 49570, published August 4, 2026), 10%-cap-applied weights.
• Per-hospital factors: FY2026 and FY2027 IPPS Final Rule Impact Files: wage index, teaching and DSH adjustment factors, capital geographic adjustment, uncompensated care per-claim amounts, hospital-specific rates, cost-of-living adjustments, and proxy quality factors.
• Rates: National adjusted operating standardized amounts (Tables 1A/1B, full update column) and the capital standard federal rate (Table 1D), both years.

The payment model

Everything is recomputed in your browser from these inputs, which is why the scenario levers (volume, deductible, sequestration) are exact recalculations rather than approximations. Nothing you type leaves your machine; the tool is static files.

Totals are scaled to the whole book, in two halves

CMS suppresses every hospital-and-MS-DRG cell under 11 discharges, so the published rows are only part of a hospital's Medicare book, and at a hospital with a small Medicare book spread across many codes they are a small part. Earlier versions of this tool reported the published slice and warned about the rest. That was defensible and not useful: a CFO cannot compare a partial figure to a budget. The revenue totals are now estimated up to the whole book, in two halves, because CMS pays in two halves. The payments that follow relative weight (operating base, DSH, teaching, capital, quality, hospital-specific rate) are scaled from the published rows to CMS's own case-mix-weighted volume, taken from the CASETA and CMIV columns of the final rule impact file. The uncompensated care payment does not follow weight, so it is not scaled at all: it is set to the whole-year amount CMS publishes for the hospital in the Medicare DSH Supplemental Data File, including the supplemental amount paid to Indian Health Service and Puerto Rico hospitals. Wherever a total is estimated, the tile says so and gives the published share. The per-DRG detail is never scaled; it is CMS's data exactly as published, which is why it will not sum to the headline. The scaled portion assumes the suppressed volume prices like the published mix; the modeled-versus-paid check deliberately stays on the published rows alone, so it remains like for like. Scale factors are floored at 1.0: at the roughly 1% of hospitals where the published rows exceed CMS's own count, the two files are different vintages, and we do not scale down.

FY2027 is priced ahead of the CMS pricer

CMS has not released the FY2027 IPPS pricer. Every FY2027 figure here is priced from the final rule itself, CMS-1849-F, by our own engine: Table 5 relative weights, Table 1A and 1B standardized amounts, the Table 1D capital rate, and the per-provider factors in the FY2027 final rule impact file. When CMS releases the pricer we will reconcile against it hospital by hospital and republish. Two inputs do not exist yet for anybody. CMS announces the calendar 2027 Part A deductible in November 2026, so the figure here is an estimate carried forward at the FY2026 rate of increase. CMS posts final FY2027 readmission adjustment factors in the autumn after hospitals review their calculations, so we use the proxy factor CMS publishes in the impact file. Both are scenario inputs you can overwrite.

Validation

The operating base engine reproduces CMS pricer output to the dollar at the hospitals we tested: operating base totals, DSH, capital and uncompensated care match published pricer-derived figures exactly; FY2026 teaching add-ons match exactly. Nationally the model prices $82.6B (FY2026) against $70.5B actually paid on the same cases; the gap is almost entirely the Part A deductible (which the caseload file excludes by definition) plus 2% sequestration, and the Scenarios section converts gross to net receipts for exactly that reason. Year-over-year change is unaffected because the same offset sits in both years.

Where we deliberately differ from other tools

1. Retired MS-DRGs are crosswalked and shown, not dropped

CMS deleted 18 MS-DRGs for FY2027 and created new ones in their place. Cases sitting on a deleted code have no FY2027 price. Rather than drop them silently, each deleted code is put onto a blend of the codes its cases now group to, and the result is shown separately in Coverage and held out of the headline, because an allocation we chose should not move your number.

Three of the four families carry no estimate. The five cardiac pacemaker codes collapse into MS-DRGs 210 and 211 along the existing MCC split, one destination each. The six uterine and adnexa codes merge into MS-DRGs 731, 732 and 733 with severity tiers matching one for one, and every new weight falls between the two old branches at its tier, which is what a clean merge should produce. MS-DRG 264 splits into 361 and 362, and the MCC share is solved from the weights themselves at 73%.

The hip and knee family is an estimate and is labelled as one everywhere it appears. CMS deleted MS-DRGs 466, 467 and 468 and 485, 486 and 487, created 449, 400, 403 and 404, and rewrote the grouping logic so that cases coded with both a removal and a replacement code now fall to the wound debridement codes 463, 464 and 465. CMS has not published where the volume went and has said a redistribution report is coming in a future proposed rule. Our shares are set from published revision indications, where infection accounts for roughly 15% to 25% of revision arthroplasty and a larger share at higher severity. MS-DRG 468 is the most sensitive of these, because MS-DRG 449 carries no severity split and pays well above the old without-CC tier.

1a. What we do not estimate, and what that costs you

CMS created MS-DRGs 523, 524 and 525 for extensive and complex spinal fusion by pulling cases out of ten MS-DRGs that all still exist in FY2027. No code was retired to make room, so no retired code maps to them and the 2024 caseload carries no volume under them. We assign nothing. Those three codes carry much higher relative weights than their sources, so at a hospital with substantial extensive fusion work the FY2027 figure here understates what CMS would pay. Sizing that gap needs the Version 44 grouper run against your own claims, which is exactly what a validation engagement does.

1b. The prior cycle: retired spinal fusion DRGs

Effective FY2025, CMS retired MS-DRGs 453, 454, 455, 459 and 460 and split the work into ten successor codes. Roughly 42,000 reported 2024 cases nationwide sit in the retired codes and cannot be repriced one-to-one. Most tools silently drop them. We estimate them: each retired code maps to its severity-matched successor set, and we blend successor weights using national successor-code volumes observed in the same caseload file (three successors are suppressed nationally; their shares are imputed from the observed multi-to-single-level ratio). The estimate is labeled wherever it appears and is never added to your headline numbers.

2. The Medicare-Dependent cliff is named, not buried

The Medicare-Dependent Hospital program lapsed September 30, 2025 and was reinstated only through December 31, 2026 by the Consolidated Appropriations Act, 2026. A hospital paid on a hospital-specific rate therefore faces a cliff part-way through FY2027, and the tool says so on the hospital's own page rather than leaving it to the footnotes. The low-volume adjustment faces the same January 1, 2027 expiry. It is worth up to 25% at a small rural hospital, CMS does not price it in these files, and where a hospital qualifies for one the tool flags that its real position is better than the figures shown.

3. A volume-weighted market average

Ranking hospitals by unweighted percent change lets a 67-case specialty hospital outrank a 6,000-case academic center. We show the unweighted rank because you will be compared on it, and the volume-weighted market average because it is the fairer number.

Which hospitals are covered, and which are not

This tool prices the 2,805 short-term acute-care hospitals paid under IPPS that appear in both the 2024 caseload file and the FY2027 impact file. Medicare pays several other kinds of hospital under entirely separate systems, so they are outside the model rather than missing from it. Facility type is determined from characters 3 to 6 of the CMS Certification Number, using CMS's own published ranges; CMS labels several of these "excluded from IPPS" in that same document.

• Critical access hospitals (1,385) are paid roughly 101% of allowable cost. There is no prospective rate to reprice, so a rate-impact model does not apply to them at all.
• Psychiatric hospitals (650) are paid under IPF PPS on a per-diem basis rather than per discharge.
• Inpatient rehabilitation facilities (415) are paid under IRF PPS using case-mix groups, not MS-DRGs.
• Long-term care hospitals (308) are paid under LTCH PPS using MS-LTC-DRGs.
• Children's hospitals (93) are excluded from IPPS and paid on a reasonable-cost basis. Rural emergency hospitals (52) receive a monthly facility payment.
• A further 289 short-term acute hospitals are IPPS-paid but cannot be priced here, either because they sit under Maryland's all-payer model or because every one of their 2024 DRG cells fell below the CMS suppression threshold.

Search for any of these by name and the tool will name the facility type and the payment system that covers it, rather than returning no result. Counts are from the CMS Hospital Enrollments file and move slightly as hospitals open, close and convert.

Known limits, read before you rely on anything

• Modeled gross amounts. Outlier payments, transfer and short-stay adjustments, new-technology add-ons, and direct medical education / organ acquisition pass-throughs are not modeled. The low-volume adjustment (up to 25% at qualifying small rural hospitals, also restored through 2026) is not priced, because CMS does not carry it in these files. The tool now flags on your own page when you qualify for it and by how much, so you can read these figures as understating you by roughly that amount.
• Every hospital is priced at the full update standardized amount. Quality program penalties to the update are not applied. The teaching factor in the impact file reflects a different resident-count vintage than final pricer files (about 1% of IME at teaching hospitals). Quality adjustments use the proxy value-based purchasing and readmission factors CMS publishes in the impact file; the final FY2027 readmission factors are not published until the autumn.
• The caseload file suppresses any hospital-DRG cell under 11 discharges, so low-volume, high-weight work is underrepresented. Expect counts below your internal numbers.
• Maryland is excluded (all-payer model outside IPPS). Traditional Medicare fee-for-service only; no Medicare Advantage, Medicaid or commercial.
• FY2027 rates take effect October 1, 2026. A hospital whose budget year is the calendar year sees nine months of FY2027 rates and three months of FY2026 rates, so weight this figure accordingly before it goes into a budget.

Terms of use

Estimates, not advice. This tool produces analytical estimates from public CMS data. It is not accounting, reimbursement, billing, legal, actuarial or investment advice, and it is not a substitute for your own analysis. Verify every figure against the primary sources, the CMS pricer, and your own claims and remittance data before using it in any budget, filing, contract or board material. Where this tool and an official CMS publication disagree, CMS governs.

No warranty; no guarantee of payment. Provided as-is, without warranty of any kind. What Medicare actually pays you is determined by CMS and your MAC under the law in force on the date of service, and it will differ from these estimates.

Use and attribution. You are welcome to use this tool for your own hospital's internal analysis. If you circulate output, keep the attribution to A3HCS intact so downstream readers see the same limits you did. The presentation, code and compilation are the property of Advanced A3 Health Consulting; the underlying CMS data is public and not claimed.

Privacy. Calculations run in your browser. We do not use third-party analytics or advertising trackers on this page.

Advanced A3 Health Consulting (A3HCS) is not affiliated with, endorsed by, or acting on behalf of the Centers for Medicare & Medicaid Services or any other government agency. CMS data is used under its public-use terms. Hospital names and CCNs appear for identification only. Analysis prepared by A3HCS. Questions about methodology are welcome at a3hcs.org.