Policy Synthesis

Healthcare Research Dataset Catalog

Searchable, filterable catalog of 20 major healthcare research datasets spanning Medicare claims, Medicaid enrollment, provider files, national surveys, and synthetic data. Click a row to expand the full description. Click column headers to sort. Keyboard shortcut: press 3 to switch to this tab.

Coding System Distinctions

How the five core Medicare coding and payment classification systems relate to each other — what they do, where they apply, and how they interact in bundling and value-based care contexts.

Feature	DRG Diagnosis-Related Group	HCPCS Healthcare Common Procedure Coding System	HCC Hierarchical Condition Category	ICD International Classification of Diseases	APC Ambulatory Payment Classification
Primary use	Inpatient hospital payment under IPPS.	Billing codes for procedures, supplies, drugs, and services (FFS). Level I = CPT; Level II = non-physician supplies/drugs/etc.	Risk adjustment to predict annual expected cost (premiums/capitation and quality models).	Clinical classification of diagnoses (ICD-10-CM) and inpatient procedures (ICD-10-PCS).	Hospital outpatient payment under OPPS.
Care setting	Acute inpatient stays.	All settings (professional, outpatient, ASC, DME, ambulance, etc.).	Population-level risk for Medicare Advantage/ACA and some Medicaid models.	All settings; foundational clinical coding on every claim/record.	Hospital outpatient departments (HOPDs); some ED/clinic/ancillary.
Payment model	Episode-based: one payment per stay, adjusted by DRG weight, wage index, etc.	Fee-for-service: line-item payment per code (subject to edits, bundling).	Risk-adjusted: additive risk score applied to premiums/benchmarks.	Not a payment system by itself; feeds DRG, HCC, quality, epi studies.	Prospective: APC weight × conversion factor; packaging & multiple-procedure rules under OPPS.
Basis of grouping	Groups from ICD-10-CM/PCS diagnoses & procedures, age/sex/discharge status; CC/MCC tiers.	CPT/HCPCS codes describing the specific service/supply/drug; units and modifiers refine payment.	Mapped from ICD-10-CM diagnosis codes recorded over a look-back year; disease hierarchies.	ICD-10-CM (diagnoses) and ICD-10-PCS (inpatient procedures) with clinical rules/edits.	Groups CPT/HCPCS codes into clinical/resource bundles (status indicators determine packaging).
Scope	Single inpatient episode.	Individual services and supplies.	Annual view of chronic disease burden per member.	Complete clinical picture across encounters (conditions & procedures).	An outpatient visit/encounter; multiple APCs may apply.
Additivity / interaction	Hierarchical: one DRG per stay; CC/MCC levels modify payment (not additive).	Generally not additive across bundled edits; each line evaluated (NCCI edits, modifiers).	Additive with hierarchies: most unrelated HCCs add; within a disease family, higher-severity HCC suppresses lower.	Multiple diagnoses/procedures can be listed; downstream systems (DRG/HCC) impose hierarchy.	Mixed: some APCs pay separately; many ancillary items are packaged into a primary APC; multiples discounted.
Typical inputs	Principal diagnosis, OR/non-OR procedures, comorbidities (CC/MCC), demographics, discharge status.	Procedure code + units + modifiers; sometimes drug strength (J-codes), DME details.	All qualifying chronic/severe diagnoses from claims/encounters in the model year.	Provider documentation abstracted to ICD codes; POA, laterality, encounter type, etc.	Outpatient CPT/HCPCS codes, status indicators, units; sometimes device edits, composite logic.
Governing body	CMS (Medicare IPPS; MS-DRGs).	CMS (HCPCS Level II) and AMA (CPT = HCPCS Level I).	CMS (MA/ACA risk models; state variations for Medicaid).	WHO (ICD), with U.S. maintenance by NCHS (ICD-10-CM) and CMS (ICD-10-PCS).	CMS (OPPS/APC program).

DRG

Creation (claim/part): Assigned by grouper to institutional inpatient (UB-04/837I) claims — Type of Bill 11x, paid under Medicare Part A (IPPS). Medicaid IP claims in TAF also use UB-04 (state policies vary). Inputs: principal DX, procedures, CC/MCC, age/sex, discharge status.
Input to bundling/VBC: Payment unit for IPPS (Part A); anchors episodes in DRG-based cost measures and readmissions programs (HRRP); state Medicaid programs may use DRG-based IP payment.
Used with other data: Wage index, case-mix, outlier thresholds, discharge/transfer status, transfer rules.

HCPCS (CPT Level I / Level II)

Creation (claim/part): Reported on professional (CMS-1500/837P) and hospital outpatient (UB-04/837I, TOB 13x) claims in Medicare Part B; also used in MA (C) encounters and Medicaid (TAF). Level I maintained by AMA (CPT); Level II by CMS.
Input to bundling/VBC: Drives OPPS/APC grouping (Part B), MPFS pricing (B), and episode/cost measures in value-based programs; similar logic in MA/Medicaid payment edits.
Used with other data: Status indicators (OPPS B), units, modifiers (25/59/JW/JZ), revenue codes, NDC for drugs (especially Part D contexts).

HCC — CMS Risk Adjustment

Creation (claim/part): Derived annually from ICD-10-CM diagnoses on institutional (UB-04/837I: A/B) and professional (1500/837P: B) claims across settings; computes RAF for Medicare Advantage (Part C); some Medicaid programs adopt HCC-like risk models.
Input to bundling/VBC: Adjusts MA (C) capitation and ACO benchmarks; used for risk stratification and performance attribution.
Used with other data: Demographics (age/sex/dual/disability), model interactions, encounter completeness rules to produce final RAF.

APC

Creation (claim/part): Assigned to hospital outpatient (UB-04/837I) claims — TOB 13x, paid under Medicare Part B (OPPS). State Medicaid programs may use OPPS-like rules. Built from CPT/HCPCS + status indicators (includes packaging and C-APC logic).
Input to bundling/VBC: Payment unit for OPPS (B); packaging creates de facto ambulatory bundles; used in outpatient episode/cost measures in VBP contexts.
Used with other data: OPPS conversion factor, wage index, multiple-procedure discounting, device/drug pass-through status, required modifiers.

Claim Forms & Type of Bill Codes

UB-04 / Form CMS-1450 → 837I Institutional claim form/transaction. Used for Medicare A/B facility billing to MACs and by many Medicaid State Agencies for institutional claims. Setting indicator: Type of Bill (TOB).

CMS-1500 → 837P Professional (non-institutional) claim form/transaction. Used for Medicare Part B physician/supplier claims and Medicaid professional billing. Stewarded by NUCC. Setting indicator: Place of Service (POS).

TOB 11x Hospital Inpatient (e.g., 111 = admit-through-discharge). Claims paid under Medicare Part A — IPPS, DRG-based payment. Medicaid institutional IP claims in TAF also carry TOB.

TOB 13x Hospital Outpatient (e.g., 131 = outpatient claim). Claims paid under Medicare Part B — OPPS; APC-based payment applies.

Place of Service (POS) 2-digit code on professional claims (CMS-1500/837P) identifying the care setting. Required for Medicare Part B; widely used by Medicaid and other payers.

Important Variables & Cross-Program Linkage

Core data elements that recur across Medicare and Medicaid research files, with notes on which Medicare Part or program each variable belongs to and how to use it for linkage, cost, utilization, and disease burden analysis.

Function	Medicare FFS (A/B)	MA Encounters (C)	Professional (B)	Part D (D)	Medicaid (TAF)
Beneficiary key	`BENE_ID`	`BENE_ID`	`BENE_ID`	`BENE_ID`	`BENE_ID` (in CCW-linked TAF)
Claim window	`CLM_FROM_DT` / `CLM_THRU_DT`	same names	same names	fill date / PDE date vars	service dates in TAF
Diagnoses	ICD-10-CM (principal/secondary)	same	fewer DX fields	n/a (drug)	ICD-10-CM
Procedures	ICD-10-PCS (IP A) + HCPCS/CPT (OP B)	HCPCS/CPT	HCPCS/CPT	NDC	ICD-10-PCS/HCPCS/CPT parallels
Site / setting	TOB (11x, 13x) on UB-04/837I	encounter type mirrors	POS on 1500/837P	plan & pharmacy context	UB-04/TOB & professional POS
Payment amounts	`CLM_PMT_AMT` + `REV_CNTR` lines	encounter pricing fields (when populated)	allowed/paid at line/claim	gross drug cost, plan paid, patient pay	TAF paid/allowed fields by claim type

Note: BENE_ID is the CCW link key across Medicare and certain Medicaid data (when available via CCW) — not a general-purpose ID outside the CCW environment.

Elements by Analytic Domain

Cost (payment/pricing)

Institutional (UB-04/837I): claim-level CLM_PMT_AMT; sum of revenue-center payment/charge lines in Medicare A (IP) and B (OP). Medicaid TAF has analogous paid/allowed fields.
Professional (CMS-1500/837P): allowed/paid amounts under Medicare Part B (Carrier/DME), line-level and claim-level rollups; similar constructs in Medicaid TAF non-institutional.
Pharmacy (Part D PDE): gross drug cost, plan paid, patient pay, LIS, TrOOP contributors under Medicare Part D; Medicaid pharmacy has parallel elements in TAF.
Plan context: Part C/D plan characteristics (premiums/service area) for benchmarking; state/plan constructs in Medicaid.

Utilization (counts/intensity)

Units: HCPCS lines, revenue-center units (OP), E/M visits, admissions/stays (IP/SNF via MedPAR roll-ups under Medicare A), days supply (PDE Part D).
Dates: CLM_FROM/THRU, ADMSN/DSCHRG for LOS, readmissions, episode triggers — primarily Medicare A/B/C; fill date for D; analogous in Medicaid TAF.
Setting indicators: TOB 11x (IP/Part A), 13x (OP/Part B) for site-of-service splits; POS for B/C professional/encounter claims; Medicaid uses TOB/POS analogs in TAF.

Disease Burden (clinical severity/risk)

Diagnoses: ICD-10-CM across A/B/C (and Medicaid TAF) feed HCC risk models used for MA (C) and sometimes Medicaid.
Procedures: add clinical complexity — ICD-10-PCS in IP Part A, HCPCS/CPT in B/C.
Part D (D): contributes medication exposure and adherence metrics for chronic disease burden analysis.

Mini-Glossary

BENE_ID Encrypted CCW Beneficiary ID. Stable, linkable key across CCW files for Medicare FFS (A/B), MA encounters (C), Part D (D), and CCW-linked Medicaid (TAF). Use: link claims/encounters/PDEs to enrollment and each other; build denominators.

CLM_ID Claim header identifier. Revenue-center/line segments appear beneath the header. Exists in Medicare FFS (A/B), MA encounters (C), and analogous constructs in Medicaid TAF (institutional/professional).

CLM_FROM_DT / CLM_THRU_DT "Statement covers from/thru" dates. Present in Medicare FFS (A/B) and MA encounters (C). Part D uses fill date variables; TAF carries service dates for Medicaid.

CLM_PMT_AMT Claim-level Medicare payment amount on institutional (UB-04/837I) claims. Paired with revenue-center line amounts for granular cost analysis.

MedPAR Medicare Part A stay-level roll-up from inpatient hospital and SNF claims. Convenient for admissions/LOS analytics; not line-level.

RAF (Risk Adjustment Factor) The numeric risk score produced by the HCC model. Combines HCC scores and demographic factors (age/sex/dual/disability). Applied to MA capitation benchmarks and ACO financial benchmarks.

NDC (National Drug Code) 11-digit drug identifier on Part D Prescription Drug Events (PDEs) and Medicaid pharmacy claims. Identifies the specific drug, strength, and package size.

TAF (T-MSIS Analytic Files) Research-optimized Medicaid claims/encounters (institutional and non-institutional), eligibility, and provider files derived from T-MSIS. Access via ResDAC/CMS.

NPI (National Provider Identifier) 10-digit identifier for individual providers and organizations. Present in Medicare A/B/C/D contexts; Medicaid TAF has parallel provider keys. Use to link claims to POS file, HCRIS, and MIPS performance data.

Quick Starter List — Variables to Pull First

Linkage & scope

BENE_ID — A/B/C/D/TAF
CLM_ID — A/B/C/TAF
CLM_FROM_DT, CLM_THRU_DT — A/B/C/TAF
CLM_ADMSN_DT — IP Part A, some TAF

Clinical

Principal/secondary ICD-10-CM — A/B/C/TAF
ICD-10-PCS — A inpatient, some TAF
HCPCS/CPT — B/C/TAF
NDC — D, Medicaid pharmacy

Payment

CLM_PMT_AMT, REV_CNTR amounts — A/B
Carrier/DME allowed/paid — B
PDE gross/plan/patient pay — D
TAF paid/allowed fields — Medicaid

Context

TOB (11x = Part A, 13x = Part B) — A/B
POS — B/C/TAF
Provider NPI — A/B/C/D/TAF
Plan IDs — C/D; Medicaid plan IDs in TAF

Commonly Conflated Concepts

Payment Models & Episodes

Conflated pair	What each one actually is	Distinctions
Episode window: TEAM vs BPCI-A / CJR	TEAM uses a 30-day post-discharge window. BPCI Advanced and CJR use 90 days.	All three share the same shape — a triggering hospitalization followed by an accountability period — so the window reads as a property of bundled payment generally. It is set per model, and TEAM's is the shortest of the three.
HHVBP: original model vs expanded program	The CMMI model ran in 9 states and ended 2021-12-31. The expanded HHVBP is a separate nationwide program, permanent from 2023.	Same name, same acronym, and the nationwide start falls just after the model's end — so the two read as one continuous program. One is a time-limited demonstration in 9 states; the other is permanent nationwide policy.
ACO REACH's predecessor	REACH replaced Global and Professional Direct Contracting (GPDC). MSSP ran in parallel throughout and still exists.	MSSP and GPDC were both accountable-care programs running at the same time, which makes either a plausible ancestor. REACH's lineage runs through the Direct Contracting family only; MSSP is a parallel track, not a predecessor.
DSH adjustment percentage vs Medicaid inpatient day share	The 11.75% figure is the DSH adjustment percentage — a payment output of the formula. Medicaid inpatient days are an input to it.	Both are percentages attached to the same hospital, and Medicaid volume is what drives the formula — so the output reads as though it were the volume measure. One is a payment adjustment; the other is patient days.
Metropolitan area vs CBSA	TEAM's 188 geographies are Core-Based Statistical Areas, which include micropolitan as well as metropolitan areas.	Most CBSAs are metropolitan and CBSA counts are usually reported in metropolitan contexts, so the term reads as a technical synonym. The set is strictly larger — micropolitan areas are CBSAs and are not metropolitan.

Policy, Coverage & Statute

Conflated pair	What each one actually is	Distinctions
OBBBA §71109 vs §71301 / §71302	§71109 narrows Medicaid eligibility (effective Oct 1, 2026). §71301/§71302 narrow marketplace premium tax credit eligibility.	Both sit in the same statute, both restrict eligibility for overlapping immigrant populations, and both land in the same window — so they read as one provision. They act on different programs, reach different groups, and take effect on different dates.
Rule finalized vs rule in force	The RADV extrapolation rule was finalized in January 2023. It was vacated by the Northern District of Texas on September 25, 2025; appeal pending.	A finalized rule is normally an enforceable one, so the two states are ordinarily the same thing. Vacatur separates them: the rule exists in final form and cannot currently be applied.
Current Medicaid enrollment vs unwinding-era peak	Roughly 73.5M as of May 2026. The widely-quoted ~94M is the continuous-enrollment peak, before the 2023–24 unwinding.	Both are recent national Medicaid totals, and the peak is the more frequently circulated of the two. They describe different periods, on either side of the unwinding — roughly a 20M difference in any denominator built from them.
Georgia Pathways vs full-expansion population	~240,000–250,000 is the Pathways-specific estimate. ~447,000–536,000 is the full-expansion eligible population.	Both answer "how many Georgians could gain coverage," and Pathways is the state's actual vehicle — so its population reads as the state's whole coverage gap. One counts people eligible under Pathways' conditions; the other counts everyone a full expansion would reach.
Waiver-dependent vs regulation-authorized Community Supports	Under CalAIM, 12 of 14 Community Supports are authorized by 42 CFR §438.3(e)(2), independent of the waiver. Only 2 depend on it.	All 14 are administered as one benefit under one program, so the set reads as sharing one legal basis. Authority splits between federal managed-care regulation and the waiver, and that split is what determines which services survive a non-renewal.
Draft vs final guidance	FDA's PCCP draft guidance issued April 2023; the final guidance issued December 3, 2024.	Both are FDA guidance documents with the same title and largely overlapping content, and drafts circulate widely enough to be cited as operative. Only the final document carries the date that governs submissions.
Partnership announced vs guidance issued	Joint Commission and CHAI announced their partnership in June 2025; their first joint guidance followed September 17, 2025.	Both events belong to one collaboration, fall within a single quarter, and coverage of the second usually recaps the first. One is an agreement to work together; the other is a published document organizations can act on.
QHIN designation vs CMS-Aligned Network pledge	TEFCA QHIN designation is a formal ONC/ASTP status. The CMS-Aligned Network pledge is a separate voluntary commitment.	Both are national interoperability credentials, both signal a commitment to data exchange, and many of the same organizations hold both. Neither implies the other — an organization can hold either, both, or neither, which is why they are tracked here as independent fields.

Data & Method

Conflated pair	What each one actually is	Distinctions
ICD-10-CM vs ICD-10-PCS	CM codes diagnoses in every setting. PCS codes procedures on inpatient claims only.	They share the ICD-10 name and sit side by side on the same inpatient claim, so "ICD-10" reads as one code set. CM spans every setting; PCS appears only on inpatient claims, while institutional outpatient (13x) carries HCPCS like professional claims.
MBI vs HICN	The MBI replaced the HICN in 2018. Files spanning the transition may carry either.	Both are the Medicare beneficiary identifier of their era and fill the same slot in a linkage plan. Which one is present is a property of the file's vintage, not a choice the analyst makes.
Wilson vs Clopper-Pearson interval	For 2 of 5, Wilson gives roughly 12%–77%; Clopper-Pearson is wider, near 7%–85%.	Both are standard binomial confidence intervals for the same proportion, and output often labels either one simply "binomial" or "exact." At small denominators they differ materially, so the method and the bounds have to be reported together.
Estimating propensity scores vs matching on them	scikit-learn estimates the scores. econml does doubly-robust and heterogeneous-effect estimation. Neither performs the matching step.	They are consecutive steps in a workflow usually named as a single procedure — "propensity score matching" — and the estimation library is the one most write-ups mention. Matching is a separate step requiring its own implementation.
HCC V28 vs V24	V28 moves from 86 to 115 categories, removes about 2,294 codes, and adds constraining. 2026 is the first 100% V28 year.	Both are CMS HCC models scoring the same conditions for the same purpose, and the transition years blended the two, which makes them look interchangeable. The differences are structural, so a RAF is only comparable within one version.
IRF payment: CMG vs DRG	Inpatient rehabilitation facilities are paid by Case-Mix Group via HIPPS codes. Acute inpatient hospitals are paid by MS-DRG.	Both group an inpatient stay into a payment category from diagnosis and clinical detail, and an IRF is an inpatient facility — so acute-care logic reads as applying. They are different prospective payment systems with different groupers.

Special Topics: Access & Cost under OBBBA

A standalone deep-dive on the One Big Beautiful Bill Act (H.R. 1 / P.L. 119-21, signed July 4, 2025) and its 2026+ coverage and cost implications — national provisions, five state impact categories, individual mandate variation, a state-by-state drill-down, and a topic-organized reading guide. Static reference content, edited in place like the Data Taxonomy reference sections. Keyboard shortcut: press 5 to switch to this tab.

Introduction & National Context: OBBBA's Core Provisions
Categories of Impact: Five State Response Patterns
Individual Mandate Variation: State Penalty Rules for 2026
Disproportionately Affected Populations
The 2026–2030 Landscape: Coverage Loss Amid Accelerating Cost Growth
State Drill-Down: California, Montana, Georgia — plus any-state lookup
1. Any-State Drill-Down (all 50 states + DC)
2. The Bay Area as Bellwether
Works Cited by Topic: A Reading Guide
Static/Dynamic Refresh Protocol
Limitations & Caveats

Throughline across every section below: OBBBA's federal savings are consistently smaller than the state, county, and provider costs it shifts downstream — and no reconciliation event (a follow-on bill, a court injunction that sticks, or a funding backstop) has yet closed that gap.

How estimates are handled on this page. Where a range exists, this page leads with the most conservative credible figure — normally the issuing agency's own central estimate — and names the higher-end scenario separately rather than blending the two or headlining the larger number. Where no state- or organization-specific figure is traceable to a named source, the page says so plainly instead of extrapolating one from a national average. Every headline figure carries the source it came from; if a number here has no attribution, treat it as unverified.

National Context: OBBBA's Core Provisions

P.L. 119-21 (H.R. 1) is a budget reconciliation law, not healthcare-specific legislation — its healthcare provisions are four of many titles. Two of the biggest compounding factors in 2026 (the expired enhanced ACA premium tax credits and the broader MA risk-adjustment/RADV litigation) are not OBBBA provisions; they are separate, coincidentally-timed policy events. This section covers OBBBA's actual text first, then flags what compounds it.

Medicaid Work Requirements

What it requires: 80 hrs/month of work, work/education program, or community service (or a $580+/month income equivalent in some state implementations) as a condition of Medicaid eligibility for the ACA expansion population, ages 19-64.
Effective: By January 1, 2027 nationally, across the 40 expansion states + DC. CMS published the implementing interim final rule (CMS-2454-IFC, "Community Engagement Requirement for Certain Individuals") June 1, 2026. Self-declaration of exemptions is permitted for 2027; supporting documentation is required starting 2028. Applicable enrollees move to 6-month (from annual) redeterminations.
Early movers: Nebraska, Montana, Iowa, and Arkansas are enforcing ahead of the federal deadline on their own state timelines — CMS's own interim final rule names all four (detail in Categories of Impact, below).
CBO estimate: ~10M combined coverage loss by 2034 across OBBBA's Medicaid and marketplace provisions — see Affected Populations for sourcing caveats.

Immigrant Eligibility Changes

§71109 — narrows Medicaid eligibility for lawfully present immigrants: The Medicaid-side companion to §71301/§71302 below — effective October 1, 2026, it restricts federal Medicaid reimbursement to citizens, lawful permanent residents, Cuban/Haitian entrants and COFA migrants — ending federal eligibility for refugees, asylees, parolees and trafficking victims. UC Berkeley Labor Center attributes ~200,000 Californians losing federally-funded full-scope Medi-Cal specifically to the refugee and asylee categories. This is the provision that drives FQHC/safety-net revenue exposure for immigrant-heavy caseloads, distinct from the marketplace-subsidy provisions below.
§71302 — five-year rule (marketplace, not Medicaid): Eliminates ACA premium tax credit (PTC) eligibility for lawfully present immigrants below 100% FPL who are ineligible for Medicaid because of immigration status (including those in their first 5 years in the U.S.). Effective for tax years after December 31, 2025 — already in effect. Affects Covered California marketplace subsidies, not Medi-Cal.
§71301 — narrowed "eligible alien" definition (marketplace, not Medicaid): Restricts PTC-eligible immigration categories to lawful permanent residents, Cuban/Haitian entrants, and COFA migrants. Refugees, asylees, and Temporary Protected Status holders lose PTC eligibility entirely. Effective January 1, 2027.

Premium Tax Credit Repayment Caps Eliminated

§71305: Repeals the repayment caps in §36B(f)(2)(B) that, through 2025, limited how much excess advance PTC a household under 400% FPL had to repay if their actual income came in higher than estimated (caps varied by income tier and filing status, from $375 up to $1,625 single / $3,250 other filers at the highest tier below 400% FPL).
Effective: Tax year 2026 onward. Any marketplace enrollee whose actual income exceeds their estimate — for any reason, at any income level — now owes back the full excess credit, dollar for dollar. This is the "uncapped clawback" risk referenced throughout this tab.

Provider Tax / SDP Caps & Rural Health Transformation Fund

Provider tax freeze, then phase-down: A nationwide freeze effective October 1, 2026 locks all states' provider-tax rates at their July 4, 2025 levels; from FFY2028 (Oct 2027) the allowable ceiling then phases down from 6% to 3.5% in expansion states specifically, through 2032.
State-directed payment (SDP) caps: Limits SDPs to 100% of Medicare rates in expansion states, 110% in non-expansion states, phasing down starting January 1, 2028 — reducing a major supplemental-financing lever hospitals and states have used since 2021. Note the revenue-impact window for SDPs runs mainly 2028-2032, not 2026-2028 — later than the provider-tax freeze/phase-down above. Nursing facilities and ICF/IDD provider taxes are exempt from the 6%-to-3.5% phase-down and keep the 6% safe harbor.
Rural Health Transformation Fund: $50B, distributed FY2026-2030 via CMS-approved state transformation blueprints. Widely expected by rural-health analysts (National Rural Health Association, hospital associations) not to fully offset the combined revenue loss from coverage attrition and the provider-tax/SDP caps above.

What compounds OBBBA (not part of the law itself): The IRA-era enhanced ACA premium tax credits expired December 31, 2025 without congressional renewal — subsidized marketplace enrollees' average annual premium roughly doubled, from $888 to $1,904, and the "subsidy cliff" at 400% FPL returned for everyone. Separately, MA risk-adjustment (HCC V28) and RADV litigation is compressing plan margins for reasons unrelated to OBBBA (see Top Shifts, Ontology tab). Both land in the same 2026-2027 window as OBBBA's provisions and are treated as compounding catalysts throughout this tab, not attributed to OBBBA itself.

Categories of Impact: Five State Response Patterns

States are not implementing OBBBA uniformly. Grouping by exposure type — rather than by region or party — is more predictive of operational outcome. Evidence grade follows each row's sourcing quality: confirmed (official agency data or GAO/auditor report), modeled (advocacy/research-org projection), insufficient (qualitative reporting only, no verified figure as of this pass).

Category	States	Exposure Type	Action Plan	Projected Coverage Loss	Cost Impact	Evidence
Early Adopters	NE, MT, IA, AR	Rapid, state-initiated timeline ahead of the Jan 2027 federal deadline	Mass beneficiary notification campaigns; state-run online resource hubs	NE: ~200 in the first disenrollment round (July 31, 2026) — the state Medicaid director's own figure, used here as the headline; advocacy groups separately project 20,000-40,000 eventually, which this page does not adopt as the expected outcome. MT: ~13,000 adults (~18% of expansion population) by the 2027 enforcement completion. IA: enforcement begins Dec 1, 2026; ~32,000 estimated coverage loss (analyst projection). AR: began July 1, 2026, with enforcement delayed until January 2027 (per CMS's own IFR, which names all four states as early movers) — a non-enforcement grace period matching Montana's.	Rural hospitals, FQHCs absorb uncompensated care first	Confirmed (NE, MT, AR) / Modeled (IA)
Waiver-Only / Non-Expansion	GA (verified); TN, WI (cited in prior tracking, not independently re-verified this pass)	Non-expansion states running (or considering) standalone work-requirement waivers	Vendor-heavy eligibility-system spending over direct care	GA: ~7,400-9,175 enrolled vs. a Pathways-eligible population commonly cited at ~240,000-250,000 (missed its own ~25,000 first-year target, per GAO)	GA: $54.2M administrative vs. $26.1M care spending since 2021 (GAO, Sept 2025)	Confirmed (GA) / Modeled (TN, WI)
Expansion+Waiver Intersection	HI, MA, NY, OR, UT	Dual-track compliance — existing state waiver terms plus the new federal work requirement	State call centers; dual-eligibility-system rebuilds	Qualitative reports of administrative strain and call-center wait times; no state-specific figure independently verified in this pass	Public hospitals in expansion states with high dual-eligible share	Insufficient
Administratively Overwhelmed	ND	Capacity failure — eligibility-system backlog builds before work-requirement enforcement even begins	"Good faith"/self-declaration workarounds; emergency eligibility-office staffing (Missouri improved after a staffing increase)	ND: pending-renewal share rose from 11.2% (Q4 2025) to 24.8% (Q1 2026, its highest on record); 69.3% of applicants waited >30 days. ND HHS estimates only 5,000-7,000 of ~23,000 expansion members will actually be affected once its own work-requirement deadline (Jan 1, 2027) arrives — nearly half pre-qualify for exemption. Nationally: median share of pending Medicaid renewals rose from 4.3% (Q2 2025) to 7.0% (Q1 2026); share of new applications taking >30 days to process rose to 21.9% in Q1 2026.	State budgets absorb overtime/temp staffing first, then rural providers absorb the access impact	Confirmed
Litigants	25 states + DC (23 states via AG + 2 via governor: KY, PA), incl. MA (lead), CA, NY, OR	Compliance-under-protest — must implement while suing	Multi-state coalition litigation: Commonwealth of Massachusetts et al. v. Oz (filed June 2026; MA is the named lead plaintiff, CA a co-plaintiff) challenges the interim final rule's narrowed "medically frail" exemption; a separate 21-AG + 1-governor coalition (filed N.D. Cal.) challenges the 2027 Notice of Benefit and Payment Parameters' expanded catastrophic-plan eligibility (HHS's own estimate: 2M lose coverage in 2027, 5M by 2030)	A federal judge denied the preliminary injunction in Massachusetts v. Oz on July 29, 2026 — the rule stays in effect nationwide while litigation continues (a motion to reconsider was filed Aug 26, 2026; outcome not yet resolved as of this pass) — deferred, not avoided	Same downstream cost pattern as the state's underlying category (e.g., CA is also Expansion+Litigant)	Confirmed

Texas was removed from the Administratively Overwhelmed category above: Texas never adopted ACA Medicaid expansion, so it has no expansion population subject to the work requirement, and its confirmed application backlog (~1,600 → 210,000+ applications, Jan-Aug 2026; on-time processing fell from >90% to 62%) is attributed by advocates primarily to new SNAP accuracy requirements and the shift to 6-month redeterminations, not work-requirement enforcement. It's still a real and useful finding — cross-program eligibility-system strain (SNAP verification changes compounding with Medicaid redetermination cycles) is straining state capacity in both non-expansion states like Texas and expansion states like North Dakota, independent of whether work requirements are the direct cause.

Individual Mandate Variation: State Penalty Rules for 2026

The federal individual mandate penalty was zeroed out by the 2017 Tax Cuts and Jobs Act; five jurisdictions have since enacted their own. As the federal enhanced-subsidy expiration raises marketplace premiums nationally, these state penalties interact directly with affordability in the states that have them — going uninsured now carries both a higher unsubsidized premium risk and, in five places, a direct tax penalty.

State Mandate Type	2026 Penalty	States
Income + flat-fee floor	Greater of 2.5% of income or a per-adult/per-child flat fee: CA ~$950/adult, ~$475/child, max ~$2,850; NJ ~$695/adult + 2.5% income, capped at the statewide average Bronze-plan premium; RI $695/adult (~$57.92/mo), $347.50/child, flat-dollar cap $2,085, also capped at Bronze-plan cost; DC is higher — $795/adult, $397.50/child, $2,385 family cap (DC's 2026 figure was not yet published as of this pass; the 2025 amount is shown)	CA, NJ, RI, DC
Income/age sliding scale	Income-tiered by multiple brackets (% of FPL) rather than one flat maximum; sources conflict on the top-tier figure — one aggregator cites ~$135/mo (~$1,620/yr), this project has also used ~$182/mo (~$2,184/yr). Not independently confirmed against MA's own TIR 26-1 guidance in this pass — verify directly with Mass.gov before citing a specific number.	MA
Reporting only	No financial penalty — coverage status reporting required	VT
No mandate	No penalty, no reporting requirement	Remaining ~44 states

Note: the flat-fee amounts above are 2026 figures (DC's is 2025, pending 2026 publication) and differ meaningfully by state within the "income + flat fee" category — DC's floor runs roughly $100/adult higher than NJ/RI's; CA's is highest of the four. Treat the bracket as directionally consistent, not identical, across jurisdictions.

Disproportionately Affected Populations

Medicaid / Marketplace Coverage Loss

CBO's cumulative scoring across OBBBA's Medicaid and marketplace provisions estimates roughly 10 million more people uninsured by 2034. Component estimates vary by provision and by which CBO score vintage is cited — treat 10M as an order-of-magnitude anchor, not a precise figure, and check for updated CBO scoring each refresh cycle.

Subsidy Clawback Exposure

OBBBA §71305 eliminated repayment caps starting tax year 2026 (see National Context). Anyone whose actual income exceeds their marketplace estimate — a raise, a spouse's new job, a good freelance year — now owes back the full advance premium tax credit with no cap, at any income level. This is a new, uncapped downside risk distinct from the separate 400% FPL subsidy-cliff issue.

ESRD & Disability Dual-Eligibles

CMS's dual-enrollment data brief shows disability-based Medicare eligibility present in roughly 38-43% of dual-eligible (Medicare-Medicaid) beneficiaries (2006-2015 series). A second independent verification pass could not locate a directly comparable, CMS-sourced Medicare-only-beneficiary figure to compute a disparity ratio — treat any specific multiplier (this project has used both "2x" and "4.7x" in different drafts) as unverified until a single CMS source with both figures is located. What is confirmed: dual-eligibles carry a materially higher disability-based eligibility share and are simultaneously exposed to Medicaid work-requirement and eligibility churn on the Medicaid side of their coverage.

Immigrant Eligibility Narrowing

Lawfully present immigrants under 100% FPL in their first 5 years in the U.S. lose PTC eligibility for tax years after 2025 (already in effect). Refugees, asylees, and TPS holders lose PTC-eligible "alien" status entirely starting Jan 1, 2027 — a population previously treated as marketplace-eligible under ACA rules since 2014.

The 2026–2030 Landscape: Coverage Loss Amid Accelerating Cost Growth

OBBBA is not an isolated event — it compounds an existing cost trend. PwC's 2027 medical cost trend survey (published June 11, 2026) puts Group at 9% and Individual at 8.5% — the highest commercial trend in 17 years, independent of anything in OBBBA. Aon's "9.5%" figure needs a year caveat: this project has two differently-dated Aon sources — one from Sept 2025 headlined "9.5% in 2026" and one from Aug 2026 headlined "9.5% in 2027" — and could not resolve in this pass whether that's a genuine coincidence across two separate annual reports or a citation mix-up; verify against Aon's original report before citing a specific year. Mercer's 8.2% and WTW's 11.1% figures circulate widely for this period but were not independently re-verified against each firm's own published report in this pass; cite with that caveat. The chart below plots each catalyst on two axes — its effect on access (horizontal) against its effect on cost (vertical) — while marker shape and colour preserve the catalyst's root: an OBBBA provision, another federal law or policy, a market/structural force, a state action, or litigation. Reading those two dimensions together is the point: it shows not just how much pressure exists but where it originates, and therefore which lever could relieve it.

Each catalyst is placed by its access effect (horizontal) against its cost effect (vertical); marker shape and colour carry its root. Hover any marker for the full detail.

OBBBA provision Other federal law / policy Market / structural State action Litigation

Positions are qualitative judgements about direction and rough magnitude, not measured coordinates — the chart is a way to see which catalysts pull against each other, not a quantitative model. Note that the two quadrants doing most of the work are opposite corners: nearly everything OBBBA does directly sits in access lost, cost up, while the offsets that might pull the other way (its own Rural Health Fund, state value-based-care programs, IRA drug pricing) are fewer, smaller, and in several cases depend on authority that expires or is still in litigation.

Fills a gap in the existing Top Shifts cascade (Ontology tab): MA risk-adjustment litigation and the Rural Health Transformation Fund were previously untracked in this project's OBBBA-adjacent content; both are now cross-referenced here and in the Payer & Company Map (rhtf_prog node).

State Drill-Down: California, Montana, Georgia — plus any-state lookup

Three distinct categories, chosen for contrast: CA (Expansion + Litigant, largest Medicaid program in the U.S.), MT (Early Adopter, rapid state-initiated rollout), GA (Waiver-Only/Non-Expansion, the only state with a multi-year work-requirement track record). Each card covers exposure, action plan, strengths/weaknesses, tipping point, cost-shift cascade, and incentive structure.

California — Expansion + Litigant

Exposure: Largest Medicaid program in the U.S. (Medi-Cal). Three distinct, non-conflicting DHCS/LAO estimates, not competing numbers: DHCS puts 4.8M enrollees subject to the work requirement (2.2M expected exempt), with 1.1M as its own central estimate of coverage loss and up to 3.4M as a higher-end/worst-case scenario (modeled on other states' early work-requirement experience); the LAO separately projects ~2M more Californians uninsured by end of 2030 from OBBBA plus state policy changes combined (a broader net figure spanning Medi-Cal, marketplace, and immigrant-eligibility effects, netting out those who find other coverage — LAO attributes roughly 90% of it to OBBBA — LAO report). Effective Jan 1, 2027; 6-month redeterminations from March 2027; $580/month income accepted as an alternative to 80 hrs/month.
Action plan: DHCS "Tracking Federal Impact" public hub; CA is a co-plaintiff (not lead) in Commonwealth of Massachusetts et al. v. Oz (MA is the named lead plaintiff; 25 states + DC total, filed June 2026) over the medically-frail exemption rule — a federal judge denied the coalition's preliminary injunction on July 29, 2026, so the rule remains in effect while litigation continues — and co-leads a separate 21-AG + 1-governor suit (filed N.D. Cal.) against the 2027 Notice of Benefit and Payment Parameters. CHCF publishes provider/plan implementation guidance.
Strengths vs. weaknesses: Mature statewide MCO infrastructure and existing CalAIM 1115-waiver social-needs screening could double as an exemption-documentation rail (strength) — but CA faces its own concurrent state budget shortfalls even as DHCS's own reporting flags implementation capacity risk (weakness), and CA's sheer population size means even a small procedural-failure rate produces large absolute coverage-loss numbers.
Tipping point: Whether paperwork/documentation failure — rather than actual non-work — ends up driving most disenrollment, as it did in early-adopter states. The figure this page uses is DHCS's own central estimate of ~1.1M; the 3.4M number is a high-end scenario modeled on other states' early experience and should not be treated as the expected outcome. If procedural failure dominates, the realized number moves toward the upper end of that span rather than past it. (CHCF, citing DHCS)
Cost-shift cascade: Federal Medicaid match reductions shift onto the state general fund, county eligibility offices, and public/safety-net hospitals (e.g. Zuckerberg SF General, LAC+USC, Highland Hospital) as the ultimate backstop.
Incentive structure: Counties and MCOs are financially incentivized to minimize procedural disenrollment — counties to avoid downstream uncompensated-care cost, MCOs because capitation revenue depends on retained enrollment.

Montana — Early Adopter

Exposure: Enforcement begins July 1, 2026 — six months ahead of the federal deadline. A 3-month non-enforcement grace period (Jul-Oct 2026) precedes actual disenrollment risk.
Action plan: DPHHS launched a public online resource hub (March 2026) and is "moving forward... despite questions" (per its own framing in state press coverage) about system readiness.
Strengths vs. weaknesses: Early move gives MT (and CMS) real operational data before most states reach their 2027 deadline (strength) — but the Montana Budget & Policy Center and Montana Hospital Association have both publicly flagged eligibility-system readiness gaps (weakness).
Tipping point: First disenrollment wave, December 31, 2026. DPHHS's own estimate is ~13,000 adults (~18% of the expansion population) losing coverage by the 2027 enforcement completion (Montana Free Press, citing DPHHS). This page uses DPHHS's own number rather than the wider 10,000-30,000 span that circulates in advocacy estimates.
Cost-shift cascade: Rural Montana hospitals, already thin-margin, absorb uncompensated care first; the state general fund backstops eligibility-system costs.
Incentive structure: Moving early lets Montana shape its own implementation terms ahead of a one-size-fits-all federal framework, per its budget-policy-center reporting.

Georgia — Waiver-Only / Non-Expansion

Exposure: Pathways to Coverage has run since July 2023 — two-plus years before OBBBA's national requirement — and was extended through Dec 31, 2026 despite its own cost findings.
Action plan: State and federal government (the Trump administration extended the program in Sept 2025) have doubled down rather than restructured, despite the GAO's admin-cost findings.
Strengths vs. weaknesses: GA is the only state with real multi-year operational data on a Medicaid work-requirement program — genuinely useful evidence as CMS designs the national rollout (strength) — but GAO documented a 2:1 admin-to-care spending ratio (GAO-25-108160: $54.2M vs. $26.2M, FY2021 through Q2 FY2025 (ending March 2025), ~88% federally funded) and enrollment has run far below target (7,400-9,175 vs. a Pathways-eligible population commonly cited at ~240,000-250,000, missing even its own ~25,000 first-year goal per GAO) (weakness). Deloitte's $27M Pathways marketing contract (grown from an initial $10.7M award) produced enrollment still under 3% of the a Pathways-eligible population commonly cited at ~240,000-250,000 even after the campaign ran.
Tipping point: The Dec 31, 2026 extension deadline — if Pathways is folded into the 2027 national framework rather than renewed standalone, GA's cost structure becomes either the template CMS exports or the cautionary tale other states cite to avoid it.
Cost-shift cascade: The federal government has covered ~88% of Pathways' administrative spend to date, so the "vendor spending > care" pattern has so far been a federal cost, not a Georgia one — that changes if OBBBA's national framework shifts more administrative burden onto states. The 2:1 admin-to-care ratio is also front-loaded, not a steady state: the administrative share fell from 96.5% of spend in FY2023 to 58.8% in FY2024 as the initial system buildout cost was absorbed — a caution against extrapolating GA's early ratio flatly onto the national rollout's steady-state cost.
Incentive structure: State political incentive to preserve Pathways as proof-of-concept for work requirements nationally, even as its own numbers argue against the model's cost-effectiveness.

Any-State Drill-Down

The three cards above are the deep-dive states. Select any other state or DC below for its category assignment, expansion status, work-requirement timeline, litigation posture, and whichever coverage-loss figure is actually verifiable for it. Most states have no verified state-specific coverage-loss estimate yet — where that is the case this says so plainly rather than extrapolating one, and points at the tracker to check.

Select a state or DC:

The Bay Area as Bellwether

Within California, the Bay Area's dense MCO/FQHC safety-net infrastructure diverges sharply from Central Valley and rural CA, where thinner provider networks and longer travel distances likely translate the same paperwork-failure rate into a higher effective coverage-loss rate. A standalone page maps potential navigation-network options to explore and ranked at-risk and resilient organization tables for healthcare-and-housing navigation networks in the region, limited to organizations with direct fiscal exposure to OBBBA-related cuts:

→ Open: Bay Area Navigation Networks — Potential Actionable Insights (standalone page)

Notable people & entities shaping this conversation: Michelle Baass (Director, CA Dept. of Health Care Services); Rob Bonta (CA Attorney General, co-plaintiff counsel in Commonwealth of Massachusetts et al. v. Oz — Massachusetts is the named lead plaintiff); California Health Care Foundation (CHCF, implementation-guidance publisher); San Francisco Health Plan, Alameda Alliance for Health, Santa Clara Family Health Plan, Contra Costa Health Plan, Health Plan of San Mateo (regional Medi-Cal MCOs — not Partnership HealthPlan of CA, which serves Northern California counties outside this footprint). Statewide advocacy organizations (California Pan-Ethnic Health Network, Western Center on Law & Poverty, California Primary Care Association) are active on this issue nationally but their specific Bay Area positions were not independently verified in this research pass — treat as directionally relevant, confirm before citing specifics.

Works Cited by Topic: A Reading Guide

Organized by topic rather than as one flat list, so this doubles as a reading guide into the underlying sources — separate from the main repo's Data Freshness sources list (README) to keep this deep-dive self-contained.

OBBBA Text & Overview

H.R. 1 / P.L. 119-21 full textcongress.gov/bill/119th-congress/house-bill/1

Signing & provision summaryPwC — President Trump signs H.R. 1

Medicaid Work Requirements — National Tracking

1115-waiver implementation trackerKFF

State readiness trackerGeorgetown CCF — H.R. 1 Readiness Tracker

National implementation frameworkCMS Newsroom

Implementation guidance for statesCBPP

State Early Adopters (NE, MT, IA, AR)

Montana rollout & system gapsMontana Free Press · Montana Budget & Policy Center

Nebraska first-in-nation rolloutNebraska Public Media · CBPP

Iowa Dec 2026 rollout & coverage-loss estimateIowa Legislative Services Agency (LSA) projection, via Becker's Payer Issues

Arkansas early-adopter statusHolland & Knight — CMS IFR summary

North Dakota readiness strainGeorgetown CCF Readiness Tracker · ND HHS

Texas backlogHouston Public Media

Georgia Pathways to Coverage

GAO admin-cost findingsThe Current GA · KFF Health News

Two-year program trackingGeorgia Budget & Policy Institute

Immigrant Eligibility & Premium Tax Credits

§71301/§71302 summaryGeorgetown CHIR

Enhanced-subsidy expiration (separate from OBBBA)CBPP · KFF calculator

§71305 clawback cap repealDisability Belongs

Provider Tax / Rural Health Transformation Fund

Provider-tax phase-down mechanicsBDO

RHTF distribution & blueprint processNational Rural Health Association

California / Bay Area

CA implementation policy briefCalifornia Health Care Foundation

DHCS federal-impact tracking hubCA DHCS

Litigation filingsCA DOJ

Massachusetts v. Oz docket & PI rulingGeorgetown Litigation Tracker · AJMC

Cost Trend & Dual-Eligible Context

2027 employer cost trend surveysAon · PwC

Dual-eligible disability prevalenceCMS

Static/Dynamic Refresh Protocol

This entire tab is static HTML with no backing JSON (same pattern as the Data Taxonomy reference sections) — edited in place. It is not automatically refreshed by the standard "update all tabs" workflow; see CLAUDE.md Step 10 for the explicit check-and-update triggers added for this tab.

Element	Source to check each refresh cycle
CBO coverage-loss projections	CBO.gov new scores on OBBBA-related legislation
KFF / Georgetown CCF tracking data	KFF work-requirements tracker; Georgetown CCF H.R. 1 Readiness Tracker
State case-study numbers (NE, MT, GA, TX)	State Medicaid agency releases; GAO/state auditor reports
Litigation status	CA DOJ federal-accountability healthcare page
Individual mandate penalty amounts	State revenue department guidance (annual, calendar-year change)

Limitations & Caveats

State case-study numbers are mostly projections, not realized outcomes — evidence grade varies by row (see Categories of Impact table): confirmed for NE, MT, GA (GAO), and ND (all independently re-verified against state-agency/GAO/tracker sources across multiple research passes); modeled for IA (state Legislative Services Agency projection); insufficient for the Expansion+Waiver category (HI/MA/NY/OR/UT) pending a verified state-specific figure. Texas's confirmed application-backlog figure is real but was removed from the Categories of Impact table itself, since Texas is a non-expansion state with no work-requirement population — see the note below the Categories of Impact table.
Threading state individual-mandate penalty interactions with the federal enhanced-subsidy expiration (Individual Mandate Variation section) is original synthesis for this repo, not a claim mirrored from any single cited source — treat it as an analytical framing, not an established consensus finding.
The Georgia Pathways-to-national-rollout comparison (Case Studies) is illustrative of work-requirement administrative-cost mechanics, but GA's waiver program and OBBBA's national requirement are legally distinct instruments — GA's numbers are a preview, not a direct prediction, of national-rollout costs.
This research pass could not independently verify: the spec's original Hawaii 193-minute call-center wait-time figure, or Bay Area advocacy organizations' specific published positions on OBBBA. Both are flagged inline rather than asserted as fact. (Iowa's and North Dakota's figures, previously flagged as unverified here, were confirmed in a later independent verification pass — see Categories of Impact table. North Dakota was reinstated to that table after initially being dropped in error.)
The specific ESRD/disability dual-eligible disparity ratio (this project has at different points used both "2x" and "4.7x") could not be confirmed against a single CMS source pairing both a dual-eligible and a Medicare-only figure — treat any such multiplier as unverified until re-sourced.
The CBO ~10M coverage-loss figure is a combined, cross-provision estimate that varies by score vintage and methodology across CBO, KFF, and Georgetown CCF — used here as an order-of-magnitude anchor, not a precise number.
Litigation status is a snapshot as of this pass (early September 2026): a preliminary injunction was denied in Massachusetts v. Oz on July 29, 2026, and a motion to reconsider filed Aug 26, 2026 was still pending as of this update — check current docket status before citing "depends on injunction outcome" as the live state of play.

Healthcare Policy Ontology Graph

Reimbursement & Value-Based Care Roadmap