A coordinated team of specialized agents that watches public healthcare data, decides what actually matters, and explains it in plain language, with every claim traceable back to its source.
Healthcare Intelligence DashboardWhat does healthcare leadership need to know this month?And can every claim be traced to its source?
32 public sources13 agentsevidence behind every finding
Healthcare Intelligence DashboardWhat does healthcare leadership need to know this month?And can every claim be traced to its source?
32 public sources · 13 agentsadamdustin.me/healthcare-intelligence
1:43 · starts on press · captions on
Transcript
0:00What does healthcare leadership need to know this month?
0:04And can every claim be traced to its source?
0:09The answers are spread across 32 public federal sources.
0:13Each one updates on its own schedule.
0:1713 AI agents to watch them.
0:23Each month, a pipeline pulls fresh data and flags what changed.
0:28Ten specialists. Code computes every number, trend and diff.
0:34A model picks what's noteworthy among the computed candidates.
0:38An executive analyst agent ranks the findings and explains why they matter.
0:44A grounding check drops any model-written number that doesn't trace to computed facts.
0:50What passes is published to the dashboard, with evidence behind every finding.
0:56One finding: 10 finalized CMS rules take effect within 260 days.
1:03The evidence shows the dataset, its data date, and a confidence rating.
1:08Confidence is high, because these are finalized rules taken straight from the Federal Register.
1:15486 automated tests, and CI on every push.
1:21A six-model benchmark picked each job's model on measured cost and quality.
1:27The same pattern works on your own data:
1:31CRM, deal desk, customer success, finance.
1:36Explore the dashboard, or talk to me about your operations.
What it watches
Public data from Medicare's administrator (CMS): hospitals, home health, physician payments, Medicare Advantage enrollment, ACA Marketplace pricing and federal rules, plus public FDA, NIH, ClinicalTrials.gov and SEC records. No private or proprietary data.
What the agents do
A team of specialized reviewers, each focused on one part of the picture. A calculator does the math; the reviewer explains what it means and why it's worth a leader's attention.
How a claim gets traced
Every finding has an "Inspect evidence" drawer: which dataset produced the number, when that data was published, and how much confidence it deserves, with a plain-language reason why.
Agent findings
What the agents flagged as worth your attention
Specialist agents run this cycle
12
Evidence-backed insights
83
Real data sources in use
32
Layers with real findings
7 / 7
Meet the team
The agents behind every finding
CI
Chief Intelligence Officer
Executive Orchestrator
Runs every specialist each cycle, reconciles what they report, and turns it into a single briefing for leadership.
MR
Market Research Specialist
Market Growth & Geography
Maps where healthcare capacity and demand are concentrated, and where they are growing, across states.
HC
Healthcare Cost Analyst
Claims, Utilization & Cost
Breaks Medicare spending growth into more care versus costlier care, down to specialties, service categories and individual codes.
RS
Reimbursement Strategy Manager
Reimbursement & Payment
Benchmarks what Medicare actually pays against what providers bill, by provider type.
PN
Provider Network Analyst
Provider & Network
Tracks the hospital supply side: ownership concentration, facility entries and exits, and whether star ratings match real quality outcomes.
MA
Medicare Advantage Market Analyst
Medicare Advantage & Part D
Follows Medicare Advantage and Part D enrollment, market share by parent organization, and the shift between plan types.
MP
Medicaid Program Analyst
Medicaid, CHIP & Duals
Tracks Medicaid and CHIP enrollment by state from the states' monthly reports, labeling each state's report month and status, and which health plan companies hold Medicaid managed care membership.
MP
Marketplace Pricing Analyst
ACA Marketplace
Watches ACA Marketplace premiums, deductibles and insurer entries and exits across HealthCare.gov states.
RA
Regulatory Affairs Specialist
Policy & Regulation
Separates CMS rules that are final and binding from ones still only proposed, and flags what takes effect soon.
SI
Strategic Insights Analyst
Emerging Trends
Looks across the other specialists' data for patterns no single dataset shows on its own.
CI
Competitive Intelligence Analyst
Market Catalysts
Logs dated market events: health-industry SEC filings and private funding notices, FDA new-drug approvals, NIH research awards and Phase 3 trial results.
DQ
Data Quality Engineer
Data Source Monitoring · infrastructure
Checks every source for new, late or changed data so the rest of the team never reports on stale numbers.
DA
Data Architect
Semantic Model · infrastructure
Keeps shared definitions of populations, geographies and entities consistent, so findings from different datasets stay comparable.
Executive briefing · reasoned autonomously on 2026-10-01
Executive Pulse
Swept 12 of 12 specialist agents this cycle; 83 insights returned. Open a category, then a finding, for its details, chart and evidence.
The benchmark silver premium for a 40-year-old rose +21.7% for plan year 2026 in the median HealthCare.gov state, rising in 28 of 30 states and breaking from a 2014-2025 median yearly change of +1.4%. Silver moving with bronze (+17.6%) points to market-wide repricing, and the spread is wide: AR stands apart at +69.1% and SD at -5.4%, against a 21.7% group median.
Medicaid and CHIP enrollment fell 5.9% year over year to 73.17M in June 2026 (preliminary), falling in 49 of 51 states, so the membership base for Medicaid managed care is still shrinking. IN (-18.5%) and LA (-14.0%) stand apart from the -3.8% state median, and adult enrollment fell faster in expansion states (-7.0%) than in non-expansion states (-2.4%).
Medicare Part B payment for Skin and Soft Tissue Substitutes went from $3.58B to $8.08B (+125.9%) in 2024, the largest dollar gain of any service category. Individual codes such as Q4205 (+324.9%, against a 0.7% median across service codes) stand far apart from peers, which marks this as a priority area for payment-integrity and utilization review.
1,956 of 2,945 hospitals (66.4%) take a net Medicare inpatient payment cut in FY2026 from the readmissions, hospital-acquired condition and value-based purchasing programs combined, and 325 are cut by all three. The average hospital adjustment is -0.47% and the deepest is -4.16%, which matters for health-system margins and for plan-provider quality contracting.
Biosimilars reached 51.2% of Adalimumab (Humira) Part D claims in Q1 2026 (preliminary), up from 21.1% in 2025 (preliminary). Tocilizumab (Actemra) went to 42.1%, so biosimilar substitution is moving quickly enough to reshape formulary and rebate strategy.
Medicaid and CHIP enrollment was already down 5.9% to 73.17M by June 2026. CMS has also finalized "Medicaid Program; Community Engagement Requirement for Certain Individuals" (effective 2026-07-31), which adds an eligibility condition for the same program after the latest enrollment data. The rule cannot explain the decline already measured, but Medicaid plans face a further eligibility change on top of an ongoing contraction. (links Q056, Q074)
The +6.4% rise in 2024 Medicare Part B professional payment came mostly from volume (services +4.6%, payment per service +1.7%). At 2024 volumes, 2026 fee schedule prices move Procedure payment -5.9% and E&M +1.0%, and none of these changes are in claims yet. For procedure-heavy practices, price will not support revenue in 2026; volume will have to. (links Q012, Q035)
Reasoned by
anthropic:claude-opus-5-5
Per-agent picks
openai:gpt-6-luna
Grounding check
numbers and names checked before publishing
Next run
Nov 1, 2026
Market & Growth24 findings
The benchmark silver premium for a 40-year-old changed +21.7% for plan year 2026 in the median HealthCare.gov state (rising in 28 of 30 states); the lowest-cost bronze premium changed +17.6%.
ACA Marketplaceanomalycurrent
The benchmark premium sets subsidy amounts for most Marketplace enrollees and is the standard yardstick for exchange pricing; its year-over-year change is the headline premium signal for plan pricing and competitive positioning.
30 HealthCare.gov states · period covered: 2014–2026Affects: Exchange insurers · Members
Median benchmark silver premium, age 40, 28 states on HealthCare.gov every year
Meets persistence and has a full baseline window, but lacks independent external corroboration. Every plan in every HealthCare.gov state for 13 plan years, not a sample.
Evidence
CMS Marketplace Rate and Plan Attributes PUFs, plan years 2014-2026, every HealthCare.gov state (30 in 2026), individual-market medical plans, summarized at pull time — ACA Marketplace Rate and Plan Attributes PUFs vintage 2026-09-25
Limitations
HealthCare.gov states only: states running their own exchange (including CA, NY, WA, PA, NJ, IL and GA in recent years) are not in CMS's federal files, and the set changes as states leave. Every comparison uses only states present in the years compared.
Filed list premiums before subsidies; most enrollees receive a premium tax credit that lowers what they pay.
No enrollment in these files, so every median is across plans or rating areas, not weighted by how many people bought them.
Benchmarks are computed per rating area; the official benchmark is set per county, and not every plan in a rating area serves every county in it.
State values are medians across rating areas, not population-weighted.
Next signal
CMS publishes plan year 2027 files in the fall; watch whether the benchmark keeps moving in the same direction.
Generated by commercial-marketplace-intelligence · sig-marketplace-2026-benchmark-trend
Medicaid and CHIP enrollment fell 5.9% year over year to 73.17M in June 2026 (preliminary), falling in 49 of 51 states; it peaked at 94.66M in April 2023.
Medicaidtrendcurrent
Medicaid is the largest coverage program by enrollment; a sustained national decline shrinks managed Medicaid membership for every state contractor and can push people toward the Marketplace or uninsurance.
All 50 states and DC · period covered: Jun 2017–Jun 2026Affects: Medicaid health plans · States · Members
Medicaid and CHIP enrollment, all states and DC (final reports)
Meets persistence and has a full baseline window, but lacks independent external corroboration. Every state and DC reported; the newest month is preliminary and is compared only with the preliminary report a year earlier.
Evidence
CMS State Medicaid and CHIP Applications, Eligibility Determinations, and Enrollment Data (data.medicaid.gov), every state and DC, June 2017 to June 2026 — State Medicaid and CHIP monthly enrollment reports vintage 2026-10-01
Limitations
State-reported counts from each state's eligibility system, not T-MSIS claims data; states differ in reporting cadence and definitions, so each state figure names its own report month and status.
Preliminary reports run lower than the final reports that replace them about a month later; every comparison here pairs preliminary with preliminary or final with final.
No disenrollment or renewal counts in this file, so these are net changes, not churn.
The file has a September 2013 baseline and then continuous months from June 2017, so monthly history starts in June 2017.
The file doesn't say why enrollment changed; policy and eligibility changes are context, not measured causes.
Next signal
Final reports for June 2026 arrive about a month after the preliminary ones; check whether they confirm the same year-over-year change.
Generated by medicaid-chip-dual-eligible-intelligence · sig-medicaid-2026-06-enrollment-trend
Certified hospital beds fell 0.5% nationally from 2015 to 2025, to 1,135,672; 19 of 51 states added beds, with the biggest gain in NV (+21.3%) and the biggest drop in RI (-13.3%).
Cross-populationtrendcurrent
Certified hospital beds are the inpatient capacity plans contract for; states adding beds are where hospital supply is growing, and states losing them are where access and negotiating leverage may shift.
51 states and DC with at least 1,000 certified hospital beds in 2015 · period covered: Oct 2015–Jun 2026Affects: Hospitals · Health plans
Certified hospital bed change, 2015-2025, 8 states selected as most noteworthy
Meets persistence and has a full baseline window, but lacks independent external corroboration. 10 years of fourth-quarter files; a single public source.
Certified beds are the beds in Medicare- or Medicaid-certified areas of a facility, not staffed or occupied beds.
Counts only facilities active in each file; a facility that changes owner keeps its certification number and isn't counted as a closure and an opening.
No population figures are used, so this shows where capacity changed, not whether it kept up with demand.
Hospital beds combine short-term acute, critical access, psychiatric, rehabilitation, long-term care and children's hospitals.
Next signal
CMS publishes the next quarter's file about a month after the quarter ends; watch whether the states at either end keep moving the same way.
Generated by market-growth-geographic-intelligence · sig-market-growth-2026-06-30-hospital-beds-by-state
Since 2019, hospices grew fastest (+37.6%, to 6,823) and federally qualified health centers next (+37.2%), while 3 of 11 facility types shrank, most of all intellectual disability (ICF/IID) facilities (-10.1%).
Cross-populationtrendcurrent
Shows which care settings are adding sites and which are consolidating: growth in community clinics and outpatient surgery, and shrinking nursing home and inpatient hospital counts, shift where members can get care and where plans need network coverage.
United States · period covered: Oct 2019–Jun 2026Affects: Health plans · Providers
Change in active certified facilities by type, 2019-2025
Every certified facility in each file, compared across 6 years; a single public source with no independent corroboration.
Evidence
CMS Provider of Services files (both CMS survey systems), fourth quarter 2019 to 2025, plus dated openings and closures in the 2026-Q2 and 2025-Q2 files — Provider of Services files (certified facilities and beds) vintage 2026-06-30
Limitations
Compared against 2019: short-term hospital counts rose in 2020 as temporary pandemic sites were certified and fell in 2023 when they closed, so a later starting year would overstate hospital decline.
Several facility types moved from CMS's older survey system to the newer one between late 2023 and 2026; each year counts a type from one system only, and the older system's home health and surgery center counts stopped updating in 2022-2023 and are skipped.
Closures keep being recorded for months after they happen, so the latest year is compared only with the prior year as it was recorded at the same point a year earlier.
Counts facilities, not capacity within them: a home health agency or surgery center can be any size.
Next signal
Watch whether 2026's openings and closures, as recorded a year from now, keep the same direction for nursing homes and hospices.
Generated by market-growth-geographic-intelligence · sig-market-growth-2026-06-30-facility-count-by-type
Certified nursing home beds fell 5.5% nationally from 2015 to 2025, to 1,571,105; 8 of 50 states added beds, with the biggest gain in NV (+15.6%) and the biggest drop in WI (-26.4%).
Cross-populationtrendcurrent
Nursing home beds are post-acute and long-term care capacity for Medicare, Medicaid and dual-eligible members; where beds are shrinking, discharge options narrow and plans may need other post-acute settings.
50 states and DC with at least 1,000 certified nursing home beds in 2015 · period covered: Oct 2015–Jun 2026Affects: Nursing homes · Medicaid health plans · Medicare Advantage plans
Certified nursing home bed change, 2015-2025, 8 states selected as most noteworthy
Meets persistence and has a full baseline window, but lacks independent external corroboration. 10 years of fourth-quarter files; a single public source.
Certified beds are the beds in Medicare- or Medicaid-certified areas of a facility, not staffed or occupied beds.
Counts only facilities active in each file; a facility that changes owner keeps its certification number and isn't counted as a closure and an opening.
No population figures are used, so this shows where capacity changed, not whether it kept up with demand.
Certified nursing home beds moved to CMS's newer survey system in 2025-Q3; beds changed -0.3% that quarter, against a typical quarterly move of 0.18% over the prior two years.
Next signal
CMS publishes the next quarter's file about a month after the quarter ends; watch whether the states at either end keep moving the same way.
Generated by market-growth-geographic-intelligence · sig-market-growth-2026-06-30-nursing-home-beds-by-state
Medicare Advantage enrollment reached 35.97M in 2026-09, +2.6% year over year, slower than +3.8% the year before; standalone Part D plans +8.2% to 25.42M.
Medicare Advantagetrendcurrent
The growth rate of the Medicare Advantage market itself - whether it's still expanding, and how fast, sets the size of the prize every MA plan and MA-contracting provider is competing for.
United States · period covered: Sep 2024–Sep 2026Affects: Medicare Advantage plans · Providers
Meets persistence and has a full baseline window, but lacks independent external corroboration. Based on 37 real monthly files; year-over-year compares the same calendar month to avoid the January enrollment jump.
National plan-level file - no state or county geography.
CMS suppresses plans with 10 or fewer enrollees each month; those are excluded, so totals run slightly under CMS's published contract-level counts.
Enrollment counts, not penetration - growth here includes growth in the overall Medicare population; the penetration finding measures share of all Medicare beneficiaries.
MA enrollment jumps each January when Annual Election Period choices take effect; month-over-month changes across January mostly reflect that.
Next signal
Watch the January report after the next Annual Election Period - it shows how much of the coming year's MA growth has already happened.
Generated by medicare-advantage-part-d-intelligence · sig-ma-partd-2026-09-ma-enrollment-trend
Medicare Advantage and other health plans covered 51.1% of Medicare beneficiaries in 2026-06, +0.18 pts from a year earlier, a smaller gain than the +0.78 pts the year before; their enrollment grew +2.2% while Medicare overall grew +1.8%.
Medicare Advantagetrendcurrent
Penetration separates Medicare Advantage winning members from Medicare simply getting bigger - the share of the whole market still open to conversion, and the base that traditional Medicare payment rules still govern.
United States · period covered: Jun 2024–Jun 2026Affects: Medicare Advantage plans · Providers
Medicare Advantage and other health plans, share of Medicare (June each year)
Meets persistence and has a full baseline window, but lacks independent external corroboration. Same month a year apart, from CMS's beneficiary-level count. The plan-level enrollment check confirms the Medicare Advantage count but is from the same agency and does not measure the share, so it is not counted as independent corroboration.
CMS counts Medicare Advantage together with cost plans, PACE and some demonstrations ("MA and other"), so these shares run slightly above Medicare Advantage alone.
This file runs about three months behind the plan-level enrollment file, so the latest enrollment count elsewhere on this page can be newer than this share.
Shares compare the same calendar month a year apart; January jumps after the Annual Election Period are not a trend on their own.
Next signal
Watch the January report after the Annual Election Period: it shows whether the share keeps rising more slowly.
Generated by medicare-advantage-part-d-intelligence · sig-ma-partd-2026-06-ma-penetration-trend
NJ shows the highest home-health episode volume per agency (4615) among states with above-median quality and favorable spending efficiency — a capacity-pressure signal, not a confirmed growth trend.
Medicare FFSbaselinecurrent
Episodes-per-agency is a real utilization/demand proxy (no external population data was available or used) — high volume per agency, combined with above-median quality and spending near the CMS benchmark, is a defensible starting point for where added home-health capacity or provider-network investment might be worth investigating further, not a confirmed growth opportunity.
NJ, MS, NY, MD, SC · period covered: Oct 2026Affects: Home health agencies · Health plans
Home-health episodes per agency, top candidate states
Single snapshot (cross-sectional, not a trend across periods) and no independent corroborating source for this specific combined read - meets none of the three confidence criteria (history, persistence, corroboration) yet.
Evidence
CMS Home Health Care Agencies, all 12460 agencies nationally, aggregated by state (states with fewer than 20 agencies excluded) — Home Health Care Agencies vintage 2026-10-01
Limitations
Episodes-per-agency is a demand proxy derived from this dataset itself (total episodes billed), not an independent population or Medicare-eligible-beneficiary denominator — it says a state's existing agencies are handling high volume, not that the state is under-served relative to true demand.
Not a growth trend — this is a single-snapshot ranking; it becomes a trend only once it persists across several pulls.
States with fewer than 20 agencies are excluded to avoid a noisy per-agency average from a handful of agencies.
"Full-service-line" means an agency reports offering all 6 tracked service types (nursing, PT, OT, speech, medical social, home health aide) - it does not measure service quality or capacity within each service line.
Next signal
Watch whether these states' episodes-per-agency and spending ratios hold or shift once a second real snapshot exists, and consider adding a real Medicare-eligible-population-by-state source to convert this into a true demand-adjusted read.
Generated by market-growth-geographic-intelligence · sig-market-growth-2026-10-01-home-health-capacity-signal
Medicare Prescription Drug Plan is the largest plan type across Medicare Advantage and standalone Part D combined (41% of 61,715,324 total enrollees), per CMS's 2026-09 report.
Medicare Advantagebaselinecurrent
How Medicare Advantage and Part D enrollment splits across plan structures (HMO, PPO, standalone drug plans) - the starting point for any plan-design or network-strategy read.
United States · period covered: Sep 2026Affects: Medicare Advantage plans · Part D plans
Medicare Advantage/Part D enrollment by plan type, 2026-09
A plan 'offering' Part D is a benefit-design flag, not proof every enrollee is actively using a drug benefit.
National aggregate only - no state-level or plan-type cross-tab included in this specific reading.
Single snapshot - a baseline reading, not yet a trend across periods.
Next signal
Watch the January report for a shift in how many members choose plans with a drug benefit.
Generated by medicare-advantage-part-d-intelligence · sig-ma-partd-2026-09-part-d-attachment
UnitedHealth Group, Inc. leads combined Medicare Advantage and standalone Part D enrollment with 13,067,487 enrollees (21.2% of 61,715,324 MA and Part D enrollees), per CMS's 2026-09 report.
Medicare Advantagebaselinecurrent
Who holds the most Medicare members across Medicare Advantage and standalone drug plans - the same kind of ranking industry directories publish from this public CMS file. For Medicare Advantage alone, see the share-shift finding.
United States · period covered: Sep 2026Affects: Medicare Advantage plans · Part D plans
Medicare Advantage/Part D enrollment by parent organization, 2026-09
enrollees
UnitedHealth Group, Inc.13,067,487
Humana Inc.11,066,732
Centene Corporation9,870,314
CVS Health Corporation8,034,415
Health Care Service Corporation3,673,965
Kaiser Foundation Health Plan, Inc.2,119,062
Elevance Health, Inc.1,974,677
BlueCross BlueShield Association, Federal Employee885,150
Blue Cross Blue Shield of Michigan Mutual Ins. Co.698,946
Plan-level enrollment aggregated nationally by parent organization - no state/county geography in this file.
A parent organization can own multiple consumer-facing brands; this reading rolls all of a parent's contracts together, not brand-by-brand.
1512 of 7608 real plan rows were suppressed by CMS itself (≤10 enrollees, HIPAA small-cell rule) and excluded, not imputed.
Single snapshot - a baseline reading, not yet a trend across periods.
A real, sourced market-share reading, not this project's own claim to any carrier's internal/proprietary data - see this file's header.
Next signal
Watch the January report: Annual Election Period switching is when most enrollment moves between organizations.
Generated by medicare-advantage-part-d-intelligence · sig-ma-partd-2026-09-parent-org-ranking
Humana Inc. gained the most Medicare Advantage share from 2025-09 to 2026-09 (+3.43 pts to 20.00% of MA enrollment); UnitedHealth Group, Inc. lost the most (-3.26 pts to 25.98%), while MA overall grew +2.6%.
Medicare Advantagestructural-changecurrent
Who is winning and losing Medicare Advantage members - the competitive picture behind network contracting, broker strategy and benefit design, measured on MA alone rather than blended with standalone drug plans.
United States · period covered: Sep 2025–Sep 2026Affects: Medicare Advantage plans · Providers
Change in Medicare Advantage share, 2025-09 to 2026-09
Does not yet meet the 2-consecutive-period persistence rule. A one-year share change between two monthly files; direction over further years would make it a trend.
National plan-level file - no state or county geography.
CMS suppresses plans with 10 or fewer enrollees each month; those are excluded, so totals run slightly under CMS's published contract-level counts.
Shares use CMS's Parent Organization field as filed each month; an acquisition or renamed parent moves members between names.
A real, sourced market-share reading from public CMS data, not any carrier's internal data.
Next signal
Watch the January report: Annual Election Period switching is when most share moves.
Generated by medicare-advantage-part-d-intelligence · sig-ma-partd-2026-09-ma-share-shift
HMO plans moved from 31.40% to 35.06% of Medicare Advantage enrollment between 2025-09 and 2026-09 (+3.66 pts), the largest shift among MA plan types.
Medicare Advantagestructural-changecurrent
Whether members are moving toward open-network PPOs or tighter HMO designs changes how much leverage a plan's network gives it - and what kind of contracts providers will be asked to sign.
United States · period covered: Sep 2025–Sep 2026Affects: Medicare Advantage plans · Providers
Change in share of Medicare Advantage enrollment by plan type, 2025-09 to 2026-09
National plan-level file - no state or county geography.
CMS suppresses plans with 10 or fewer enrollees each month; those are excluded, so totals run slightly under CMS's published contract-level counts.
Plan type is CMS's plan-level category; Special Needs Plans are not broken out in this file.
Next signal
Check whether the shift continues in the next January report, when new plan-year choices take effect.
Generated by medicare-advantage-part-d-intelligence · sig-ma-partd-2026-09-ma-plan-type-shift
Utah's Medicare Advantage and other health plan share rose most from 2025-06 to 2026-06 (+1.29 pts to 54.9%); Vermont's fell most (-18.28 pts to 12.7%), against +0.18 pts nationally; the share fell in 20 of the 50 states and DC.
Medicare Advantagestructural-changecurrent
Where Medicare Advantage is still gaining ground and where it has stalled or reversed - the state markets where plan exits, benefit cuts or star-rating changes are moving members back to traditional Medicare.
United States (by state) · period covered: Jun 2025–Jun 2026Affects: Medicare Advantage plans · Providers
Change in Medicare Advantage and other health plans' share of Medicare, 2025-06 to 2026-06
CMS counts Medicare Advantage together with cost plans, PACE and some demonstrations ("MA and other"), so these shares run slightly above Medicare Advantage alone.
Beneficiary counts by where the beneficiary lives, not where the plan is based.
Territories are left out; a state's change can come from a few large counties (see the county finding).
Next signal
Watch the January report after the Annual Election Period, when plan exits and benefit changes show up in state shares.
Generated by medicare-advantage-part-d-intelligence · sig-ma-partd-2026-06-ma-penetration-by-state
25 counties' Medicare Advantage and other health plan share moved at least 5 points more than their state's from 2025-06 to 2026-06 (19 fell further than their state, 6 fared better); furthest out: Carroll County, NH, 33.3% to 13.6% while New Hampshire moved -4.4 pts.
Medicare Advantageanomalycurrent
County swings this far from their state usually mean a large plan entered or left that county - the local markets where members are suddenly up for grabs, or where providers just lost or gained a Medicare Advantage payer.
United States (by county) · period covered: Jun 2025–Jun 2026Affects: Medicare Advantage plans · Providers
Change in Medicare Advantage and other health plans' share minus the state's change, 2025-06 to 2026-06
Does not yet meet the 2-consecutive-period persistence rule. 4 of 25 flagged counties also moved the same way the year before; the rest are one-year moves.
CMS counts Medicare Advantage together with cost plans, PACE and some demonstrations ("MA and other"), so these shares run slightly above Medicare Advantage alone.
Rule: a county is flagged when its share changed at least 5 percentage points more (up or down) than its state's between the same month a year apart. Only counties with at least 5,000 Medicare beneficiaries in both months are compared, and counties where CMS suppressed a small count are left out.
This file shows that a county's share moved, not why; plan entries and exits are the usual cause but are not in this data.
Beneficiary counts by where the beneficiary lives, not where the plan is based.
Next signal
Check the January report after the Annual Election Period: a plan exit shows up as a one-time drop that does not reverse.
Generated by medicare-advantage-part-d-intelligence · sig-ma-partd-2026-06-ma-unusual-counties
IN had the steepest Medicaid and CHIP enrollment change over the year to June 2026 (-18.5%) and MO the strongest (+1.1%); CA lost the most people, 1,650,639.
Medicaidbaselinecurrent
Where Medicaid membership is shrinking or growing fastest - the states where managed Medicaid contractors' volume, and the risk mix of who stays enrolled, are shifting most.
51 states and DC · period covered: Jun 2025–Jun 2026Affects: Medicaid health plans · States · Members
Medicaid and CHIP enrollment change by state, June 2025 to June 2026 (preliminary reports)
Does not yet meet the 2-consecutive-period persistence rule. Each state's own count; one year's state ranking is a baseline, not a trend.
Evidence
CMS State Medicaid and CHIP Applications, Eligibility Determinations, and Enrollment Data (data.medicaid.gov), every state and DC, June 2025 to June 2026 — State Medicaid and CHIP monthly enrollment reports vintage 2026-10-01
Limitations
State-reported counts from each state's eligibility system, not T-MSIS claims data; states differ in reporting cadence and definitions, so each state figure names its own report month and status.
Preliminary reports run lower than the final reports that replace them about a month later; every comparison here pairs preliminary with preliminary or final with final.
No disenrollment or renewal counts in this file, so these are net changes, not churn.
The file has a September 2013 baseline and then continuous months from June 2017, so monthly history starts in June 2017.
Next signal
Watch whether the steepest-declining states keep falling in the next monthly reports; two more months in the same direction meets the persistence rule.
Generated by medicaid-chip-dual-eligible-intelligence · sig-medicaid-2026-06-enrollment-by-state
Comprehensive Medicaid managed care enrollment fell 11.9% to 65.89M in 2024 across the 42 states that use it in both years; ND changed most (-33.6%).
Medicaidbaselinecurrent
Comprehensive managed care is how most Medicaid enrollees get coverage, so this is the size of the contracted Medicaid market each state offers health plans.
42 states with comprehensive managed care in 2023 and 2024 · period covered: 2023–2024Affects: Medicaid health plans · States · Members
Comprehensive Medicaid managed care enrollment change by state, 2023 to 2024
Annual and published with a lag of more than a year (2024 data came out in February 2026).
Counts only comprehensive managed care; dental, behavioral-health-only, transportation and other limited programs are left out because they count the same people again.
States report plans and parents in their own words, and a state's list of programs can change between years.
Next signal
CMS's 2025 release (expected early 2027, as 2024 data arrived in early 2026) will show whether the states that fell most kept falling.
Generated by medicaid-chip-dual-eligible-intelligence · sig-medicaid-managed-care-2024-by-state
Among parent organizations as named in states' 2024 reports, Centene had the most comprehensive Medicaid managed care enrollees (6.97M in 20 states, -19.5% from 2023); states name parents inconsistently, so company totals are understated.
Medicaidbaselinecurrent
Which health plan companies hold the most Medicaid managed care membership, and whose footprint grew or shrank; the same companies appear in Medicare Advantage, so this links the two programs.
States with comprehensive Medicaid managed care · period covered: 2023–2024Affects: Medicaid health plans · States · Members
Comprehensive Medicaid managed care enrollees by parent organization as named by states, 2024 (millions)
Annual and published with a lag of more than a year (2024 data came out in February 2026).
Counts only comprehensive managed care; dental, behavioral-health-only, transportation and other limited programs are left out because they count the same people again.
States report plans and parents in their own words, and a state's list of programs can change between years.
The same company can appear under several parent names (for example "UnitedHealthcare" and "UnitedHealth Group"), and some subsidiaries are listed as their own parent. Only spellings of one name are merged, never different names, so a company's true total can be larger than shown.
A change for one parent name can come from a state renaming the parent rather than from members moving.
Next signal
Compare with the Medicare Advantage parent-organization ranking to see which companies are growing in one program and shrinking in the other.
Generated by medicaid-chip-dual-eligible-intelligence · sig-medicaid-managed-care-2024-parents
AR had the largest benchmark premium change for plan year 2026 (+69.1%) and SD the smallest (-5.4%); WV has the highest benchmark at $1,094 a month for a 40-year-old.
ACA Marketplacebaselinecurrent
Where exchange premiums are rising fastest - the states where pricing pressure, and the subsidy cost that follows the benchmark, is concentrated.
30 HealthCare.gov states · period covered: 2025–2026Affects: Exchange insurers · Members
Benchmark silver premium change by state, plan year 2025 to 2026
Does not yet meet the 2-consecutive-period persistence rule. Every plan in each state; one year's state ranking is a baseline, not a trend.
Evidence
CMS Marketplace Rate and Plan Attributes PUFs, plan years 2014-2026, every HealthCare.gov state (30 in 2026), individual-market medical plans, summarized at pull time — ACA Marketplace Rate and Plan Attributes PUFs vintage 2026-09-25
Limitations
HealthCare.gov states only: states running their own exchange (including CA, NY, WA, PA, NJ, IL and GA in recent years) are not in CMS's federal files, and the set changes as states leave. Every comparison uses only states present in the years compared.
Filed list premiums before subsidies; most enrollees receive a premium tax credit that lowers what they pay.
No enrollment in these files, so every median is across plans or rating areas, not weighted by how many people bought them.
Small states have one or a few rating areas and few issuers, so one issuer's rate change can move the state's benchmark.
Next signal
Check whether the states with the largest 2026 increases also lost issuers (see the issuer participation insight).
Generated by commercial-marketplace-intelligence · sig-marketplace-2026-benchmark-by-state
The median silver-plan deductible in HealthCare.gov states is $5,925 for plan year 2026, +$900 from 2025 in the median state (bronze +$0); in the states on HealthCare.gov since 2014 it went from $2,500 to $5,925.
ACA Marketplacebaselinecurrent
Deductibles are the affordability signal the premium doesn't show: a lower premium bought with a higher deductible shifts cost to members at the point of care.
30 HealthCare.gov states · period covered: 2014–2026Affects: Members · Exchange insurers
Median silver deductible, 28 states on HealthCare.gov every year
Does not yet meet the 2-consecutive-period persistence rule. Every on-exchange plan's filed deductible, 13 plan years.
Evidence
CMS Marketplace Rate and Plan Attributes PUFs, plan years 2014-2026, every HealthCare.gov state (30 in 2026), individual-market medical plans, summarized at pull time — ACA Marketplace Rate and Plan Attributes PUFs vintage 2026-09-25
Limitations
HealthCare.gov states only: states running their own exchange (including CA, NY, WA, PA, NJ, IL and GA in recent years) are not in CMS's federal files, and the set changes as states leave. Every comparison uses only states present in the years compared.
Filed list premiums before subsidies; most enrollees receive a premium tax credit that lowers what they pay.
No enrollment in these files, so every median is across plans or rating areas, not weighted by how many people bought them.
Standard plan versions only: lower-income enrollees in cost-sharing-reduction silver variants have much lower deductibles.
Deductibles are medians across plans offered, not what enrollees chose.
Next signal
Watch whether plan year 2027 deductibles move with or against premiums - a premium cut paired with a deductible rise is a benefit buy-down, not cheaper coverage.
Generated by commercial-marketplace-intelligence · sig-marketplace-2026-deductible-trend
22 issuer exits and 10 entries across 30 HealthCare.gov states for plan year 2026; issuers per state range from 2 in AK to 18 in TX.
ACA Marketplacestructural-changecurrent
Issuer entry and exit is the clearest read on exchange competitive intensity: exits leave fewer choices and tend to precede premium increases, and entries mark markets issuers see as profitable.
30 HealthCare.gov states · period covered: 2025–2026Affects: Exchange insurers · Members
Marketplace issuers per state, plan year 2026 (fewest first)
issuers
AK2
HI2
WV2
WY2
DE3
MT3
ND3
SD3
0123
Issuer participations (state-issuer pairs), 28 states on HealthCare.gov every year
Does not yet meet the 2-consecutive-period persistence rule. Every filed issuer in each state; counts, not market shares.
Evidence
CMS Marketplace Rate and Plan Attributes PUFs, plan years 2014-2026, every HealthCare.gov state (30 in 2026), individual-market medical plans, summarized at pull time — ACA Marketplace Rate and Plan Attributes PUFs vintage 2026-09-25
Limitations
HealthCare.gov states only: states running their own exchange (including CA, NY, WA, PA, NJ, IL and GA in recent years) are not in CMS's federal files, and the set changes as states leave. Every comparison uses only states present in the years compared.
Filed list premiums before subsidies; most enrollees receive a premium tax credit that lowers what they pay.
No enrollment in these files, so every median is across plans or rating areas, not weighted by how many people bought them.
An issuer that files under a new HIOS id (after a merger or restructuring) counts as one exit and one entry.
Issuers are counted, never named, and without enrollment an exit by a small issuer counts the same as one by a large issuer.
Next signal
Plan year 2027 filings will show whether the 2026 exits continue.
Generated by commercial-marketplace-intelligence · sig-marketplace-2026-issuer-participation
Marketplace plan selections fell 4.9% to 23.13M in the 2026 open enrollment, from 24.32M, and fell in 41 of 51 states; since 2017 they went from 12.22M.
ACA Marketplacebaselinecurrent
Marketplace enrollment is the size of the individual-market opportunity for every exchange issuer; a national decline after years of growth changes membership forecasts and risk-pool assumptions.
All 50 states and DC · period covered: 2017–2026Affects: Exchange insurers · Members
Marketplace open enrollment plan selections, all states and DC
Plan selections during open enrollment, not effectuated enrollment: some people who pick a plan never pay the first premium, and people also enroll outside open enrollment.
CMS publishes no state-level open enrollment file before 2017, so history starts there. 2017-2019 come from CMS's report workbooks, which carry fewer fields than the 2020+ files.
CMS suppresses small cells and some state-run exchanges don't report every field; those are left out, never treated as zero.
Open enrollment counts don't say why people left; the premium-after-subsidy change is shown alongside, not proven as the cause.
Next signal
CMS's effectuated enrollment report for early 2026 will show how many of these selections became paid coverage.
Generated by commercial-marketplace-intelligence · sig-marketplace-2026-enrollment-trend
NC had the steepest 2026 Marketplace enrollment change (-21.9%) and NM the strongest (+18.1%); NC lost the most people, 213,653.
ACA Marketplacebaselinecurrent
Where exchange membership is growing or shrinking fastest - the states where issuers' individual-market volume, and the pool's average risk, are shifting most.
51 states and DC · period covered: 2025–2026Affects: Exchange insurers · Members
Marketplace plan selection change by state, 2025 to 2026
Does not yet meet the 2-consecutive-period persistence rule. CMS's count of every plan selection in each state; one year's state ranking is a baseline, not a trend.
Plan selections during open enrollment, not effectuated enrollment: some people who pick a plan never pay the first premium, and people also enroll outside open enrollment.
CMS publishes no state-level open enrollment file before 2017, so history starts there. 2017-2019 come from CMS's report workbooks, which carry fewer fields than the 2020+ files.
CMS suppresses small cells and some state-run exchanges don't report every field; those are left out, never treated as zero.
A state that moved to its own exchange is compared with its own prior total, but state-run exchanges can run longer open enrollment periods than HealthCare.gov, which can lift their counts.
Next signal
Check whether the states that lost the most enrollees in 2026 also lost issuers (see the issuer participation insight).
Generated by commercial-marketplace-intelligence · sig-marketplace-2026-enrollment-by-state
The average Marketplace premium after subsidy changed +58.2% in the median state for 2026; across 51 states, larger increases weakly went with larger enrollment losses (correlation -0.29).
ACA Marketplacebaselinecurrent
What enrollees actually pay after subsidy drives who stays covered; a strong link between net-premium increases and enrollment losses marks the states where price sensitivity, and the risk of healthier members leaving, is highest.
51 states and DC reporting net premiums · period covered: 2025–2026Affects: Exchange insurers · Members
Change in premium after subsidy vs. change in plan selections, by state, 2025 to 2026
Average premium after subsidy, % change vs Plan selections, % change
Does not yet meet the 2-consecutive-period persistence rule. A cross-state correlation for one year, not evidence that the price change caused the enrollment change.
Plan selections during open enrollment, not effectuated enrollment: some people who pick a plan never pay the first premium, and people also enroll outside open enrollment.
CMS publishes no state-level open enrollment file before 2017, so history starts there. 2017-2019 come from CMS's report workbooks, which carry fewer fields than the 2020+ files.
CMS suppresses small cells and some state-run exchanges don't report every field; those are left out, never treated as zero.
The average premium after subsidy also shifts when the mix of who enrolls changes (for example, if lower-income enrollees with the largest subsidies leave), not only when prices change.
Correlation across states, not a causal estimate.
Next signal
Watch whether the same states lose enrollment again in 2027; a second year in the same direction meets the persistence rule for a trend.
Generated by commercial-marketplace-intelligence · sig-marketplace-2026-net-premium-vs-enrollment
Claims & Cost12 findings
Treatment: Skin and Soft Tissue Substitutes added the most Medicare Part B payment in 2024: $3.58B to $8.08B (+125.9%); 14 of 49 service categories broke from their own history.
Medicare FFSanomalycurrent
Which kinds of care are driving Medicare's cost growth - procedures, imaging, tests, drugs or visits - and whether each is more services or costlier ones. The service-level view a utilization-management or medical-cost team works from.
United States · period covered: 2023–2024Affects: Health plans · Providers
Medicare Part B payment change by service category, 2023 to 2024
% change
Treatment: Skin and Soft Tissue Substitutes125.9
E&M: Care Management/Coordination55.7
Test: Molecular Testing-24.1
Imaging: Nuclear16
Treatment: Injections and Infusions (nononcologic)11.6
Does not yet meet the 2-consecutive-period persistence rule. National totals for every code across 12 years; a single year's break is an anomaly, not yet a trend.
Evidence
CMS Medicare Physician & Other Practitioners - by Geography and Service, national totals for every procedure code, data years 2013-2024; categories from CMS's Restructured BETOS Classification System (latest assignment) — Medicare Physician & Other Practitioners by Service vintage 2024-12-31
Limitations
Medicare fee-for-service Part B professional services only - not Medicare Advantage, hospital facility fees, or commercial insurance.
CMS publishes each data year with a lag of more than a year.
For drugs, a "service" is a billing unit of the drug (for example, per milligram), not a visit, so drug volume swings when unit definitions or products change.
Categories use each code's latest RBCS assignment for every year, so a code's history stays in one category even if CMS reclassified it.
Where services and payment per service swing more than 25% in opposite directions, the split is marked unreliable rather than reported.
Next signal
Watch whether the categories that broke from their history in 2024 do so again in 2025.
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2024-service-category-growth
Biosimilars were 51.2% of Adalimumab (Humira) Part D claims in Q1 2026 (preliminary), up from 21.1% in 2025 (preliminary); on a 5-year lookback the share went from 0.0% in 2020 to 21.1% in 2025 (preliminary). Part D biosimilar claims covered 14 reference biologics in Q1 2026 (preliminary).
Part Dtrendcurrent
Biosimilars cost less than the reference biologic, so a rising share is a direct cost-reduction signal, and where it lags is where formulary and contracting choices still hold savings.
United States · period covered: Jan 2020–Mar 2026Affects: Health plans · Providers
Biosimilar share of Part D claims by reference biologic, Q1 2026 (preliminary)
% of claims
Filgrastim92.6
Trastuzumab87.1
Bevacizumab83.8
PEGfilgrastim78.4
Rituximab75.3
Adalimumab51.2
Infliximab49.8
Epoetin Alfa45
Ustekinumab43.4
Tocilizumab42.1
Denosumab31.6
Insulin Glargine9.9
Insulin Aspart2.9
Aflibercept0.9
0255075100
Adalimumab (Humira) biosimilar share of Part D claims (points after the annual file are preliminary)
Meets persistence and has a full baseline window, but lacks independent external corroboration. Claim shares from every Part D drug in CMS's files; the adalimumab share rose in every year shown.
Evidence
CMS Medicare Part D Spending by Drug, data years 2020-2024 (every drug sold in 2024, gross drug cost) — Medicare Part D Spending by Drug vintage 2024-12-31
Spending is gross drug cost (what plans and beneficiaries paid at the pharmacy), before manufacturer rebates and other price concessions, which CMS is not allowed to publish; net cost is lower, by different amounts for different drugs.
Biosimilars are identified by the FDA naming rule (the reference name plus a four-letter suffix, e.g. adalimumab-adbm), counted only where the unsuffixed reference is also sold and the suffixed product comes from a different manufacturer. Unsuffixed follow-on products and unbranded versions sold by the reference maker count with the reference.
Shares use claims (prescriptions), since biosimilars' lower prices make spending shares understate adoption; biosimilars and references can differ in days' supply per claim.
State is where the prescriber practices, not where the beneficiary lives; CMS leaves out any state and drug with fewer than 11 claims.
2025 and 2026 figures come from CMS's preliminary quarterly file (pulled three months after each quarter); CMS advises caution comparing them with the annual file, and they may change in later releases.
Next signal
Watch the next preliminary quarter for adalimumab and the newer biosimilar molecules.
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2024-part-d-biosimilar-share
Medicare paid $120.1B for Part B physician and practitioner services in 2024, +6.4% from 2023: services +4.6%, payment per service +1.7%.
Medicare FFSanomalycurrent
The national growth rate of Medicare professional spending, and how much of it is more care versus costlier care - the benchmark any payer or provider group compares its own cost trend against.
United States (all providers) · period covered: 2013–2024Affects: Health plans · Providers
Meets persistence and has a full baseline window, but lacks independent external corroboration. Based on 12 years of full-population CMS data, not a sample.
Evidence
CMS Medicare Physician & Other Practitioners - by Provider, every provider in each data year 2013-2024 (1,296,739 providers in 2024), summarized at pull time — Medicare Physician & Other Practitioners by Provider vintage 2024-12-31
Limitations
Covers Medicare fee-for-service Part B professional services only - not Medicare Advantage, hospital facility payments, or commercial insurance.
CMS publishes each data year with a lag of more than a year, so the latest year here is not the current year.
Payment per service mixes fee-schedule changes with a shift toward costlier services, so the price component is payment per service, not a pure fee change.
2020's pandemic drop is part of the history; the median-based check resists it, but year-over-year rates around 2020-2021 are distorted.
Next signal
Watch whether CMS's 2025 data continues the +1.7% change in payment per service, which tracks fee-schedule updates and service mix.
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2024-physician-payment-trend
Part B drugs billed by practitioners were 24.5% of Medicare Part B professional payment in 2024 ($29.55B), up from 10.7% in 2013; drug payment grew +25.4% in 2024 alone.
Medicare FFSanomalycurrent
Physician-administered drugs (infusions, injections, biologics) are the fastest-growing slice of Part B spending and a major target of medical-benefit drug management and site-of-care programs.
United States · period covered: 2013–2024Affects: Health plans · Providers · Drug makers
Part B drugs as a share of Part B professional payment
Meets persistence and has a full baseline window, but lacks independent external corroboration. 12 years of national totals for every code.
Evidence
CMS Medicare Physician & Other Practitioners - by Geography and Service, national totals for every procedure code, data years 2013-2024; categories from CMS's Restructured BETOS Classification System (latest assignment) — Medicare Physician & Other Practitioners by Service vintage 2024-12-31
Limitations
Medicare fee-for-service Part B professional services only - not Medicare Advantage, hospital facility fees, or commercial insurance.
CMS publishes each data year with a lag of more than a year.
For drugs, a "service" is a billing unit of the drug (for example, per milligram), not a visit, so drug volume swings when unit definitions or products change.
Covers drugs billed by practitioners under Part B, not Part D pharmacy drugs or drugs billed by hospital outpatient departments.
Uses CMS's drug indicator, which includes some biologic products such as skin substitutes.
Next signal
Watch whether the drug share keeps rising in 2025 data, and CMS rules on Part B drug payment (inflation rebates, skin substitute payment reform).
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2024-part-b-drug-share
The average risk-adjusted Medicare spending ratio for home health episodes is 0.97 across 8603 agencies with reported data (1.0 = spending as risk-adjusted expected).
Medicare FFSbaselinecurrent
A national baseline for home-health episode spending efficiency, directly comparable to CMS's own published national benchmark of 1.0.
United States · period covered: Sep 23, 2026–Oct 1, 2026Affects: Home health agencies · Health plans
Spending-ratio distribution by state, 6 states selected as most noteworthy
ratio
Middle half (IQR)
Median
Outlier (beyond 1.5× IQR; a triangle is off this chart's visible scale)
Does not yet meet the 2-consecutive-period persistence rule. Based on 3 snapshot(s) spanning 8 day(s); full confidence needs a 2-year baseline.
Evidence
CMS Home Health Care Agencies, 3 real snapshot(s) from 2026-09-23 to 2026-10-01 — Home Health Care Agencies vintage 2026-10-01
Limitations
This is a risk-adjusted ratio published by CMS, not a raw dollar cost-per-episode figure.
3857 of 12460 agencies had suppressed/unavailable data for this measure and are excluded from the average.
Baseline window is 8 real day(s) so far, well short of the 24-month window this framework requires for full confidence - same-day re-pulls overwrite rather than adding history (see homeHealthCareAgencies.ts), so this grows only as real calendar days pass between pulls.
Next signal
Watch subsequent real pulls (on later calendar days) for this ratio's distribution to shift, and for state-level variation once geography-level aggregation is added.
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2026-10-01-home-health-spending-ratio
16 of 63 major provider types had 2024 Medicare payment growth that broke from their own history; Nurse Practitioner had the largest dollar change, +35.5% ($7489M to $10151M; services +31.1%, payment per service +3.4%).
Medicare FFSanomalycurrent
Separates specialties where Medicare is buying more care from those where each service costs more - different signals for network, contracting and utilization-management planning.
United States (all providers) · period covered: 2023–2024Affects: Health plans · Providers
Medicare payment growth by provider type, 2023 to 2024
Does not yet meet the 2-consecutive-period persistence rule. Full-population data with 12 years of history per provider type, but a single year's break is an anomaly, not yet a trend.
Evidence
CMS Medicare Physician & Other Practitioners - by Provider, every provider in each data year 2013-2024 (1,296,739 providers in 2024), summarized at pull time — Medicare Physician & Other Practitioners by Provider vintage 2024-12-31
Limitations
Covers Medicare fee-for-service Part B professional services only - not Medicare Advantage, hospital facility payments, or commercial insurance.
CMS publishes each data year with a lag of more than a year, so the latest year here is not the current year.
Payment per service mixes fee-schedule changes with a shift toward costlier services, so the price component is payment per service, not a pure fee change.
Provider types are CMS's self-reported specialty labels; a provider that changes its label moves its payments between types.
With dozens of provider types checked at 2 MADs, one or two can break from their history by chance.
Where services and payment per service swing more than 25% in opposite directions, the volume/price split is marked unreliable rather than reported - a change in how CMS counted services, not a real shift.
Next signal
Watch whether the types breaking from their history in 2024 do so again in 2025 - two years in the same direction meets this framework's persistence rule for a trend.
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2024-physician-type-growth
NV had the fastest growth in standardized Medicare Part B payment in 2024 (+16.3%) and IL the slowest (-4.0%), against +6.5% nationally.
Medicare FFSbaselinecurrent
Where Medicare professional spending is growing fastest once regional price differences are removed - a proxy for where utilization and intensity are rising.
51 states and DC · period covered: 2023–2024Affects: Health plans · Providers
Standardized Medicare Part B payment growth by state, 2023 to 2024
Does not yet meet the 2-consecutive-period persistence rule. Full-population state totals; a single year's state ranking is a baseline, not a trend.
Evidence
CMS Medicare Physician & Other Practitioners - by Provider, every provider in each data year 2013-2024 (1,296,739 providers in 2024), summarized at pull time — Medicare Physician & Other Practitioners by Provider vintage 2024-12-31
Limitations
Covers Medicare fee-for-service Part B professional services only - not Medicare Advantage, hospital facility payments, or commercial insurance.
CMS publishes each data year with a lag of more than a year, so the latest year here is not the current year.
Payment per service mixes fee-schedule changes with a shift toward costlier services, so the price component is payment per service, not a pure fee change.
States are where the billing provider is located, not where the patient lives; a provider serving several states counts in one.
Fee-for-service totals fall as more beneficiaries move to Medicare Advantage, so slow growth can reflect MA enrollment rather than lower use.
Next signal
Check whether the fastest-growing states in 2024 were also fast in prior years, using the 5-year rates above.
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2024-physician-state-growth
Q4205 Membrane graft or membrane wrap, per square centimeter added the most Medicare Part B payment of any established service in 2024: $270M to $1.15B (+324.9%); 24 near-new services each passed $50M, led by Q4271 at $931M.
Medicare FFSbaselinecurrent
The specific services behind cost growth - the level where payers set coverage policy, prior authorization and site-of-care rules.
United States · period covered: 2023–2024Affects: Health plans · Providers
Largest Medicare Part B payment growth by service, 2023 to 2024
Does not yet meet the 2-consecutive-period persistence rule. National totals for every code; a single year's growth is a baseline, not a trend.
Evidence
CMS Medicare Physician & Other Practitioners - by Geography and Service, national totals for every procedure code, data years 2013-2024; categories from CMS's Restructured BETOS Classification System (latest assignment) — Medicare Physician & Other Practitioners by Service vintage 2024-12-31
Limitations
Medicare fee-for-service Part B professional services only - not Medicare Advantage, hospital facility fees, or commercial insurance.
CMS publishes each data year with a lag of more than a year.
For drugs, a "service" is a billing unit of the drug (for example, per milligram), not a visit, so drug volume swings when unit definitions or products change.
CMS adds and retires codes every year; growth that moves from a retired code to a new one shows up as a new code, not as growth.
Next signal
Watch these codes in 2025 data, and any CMS rule changing their payment (the Federal Register agent tracks those).
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2024-service-code-growth
Part D gross drug cost rose 4.6% from 2023 to 2024 across prescribers in the 50 states and DC, to $284.89B; District of Columbia grew fastest (+14.7%) and Maine least (-2.0%).
Part Dbaselinecurrent
Where Part D cost is growing fastest, and whether that's more prescriptions or costlier ones - the split a regional pharmacy or bid team works from.
United States · period covered: 2023–2024Affects: Health plans · Providers
Part D gross drug cost change by prescriber state, 2023 to 2024
Spending is gross drug cost (what plans and beneficiaries paid at the pharmacy), before manufacturer rebates and other price concessions, which CMS is not allowed to publish; net cost is lower, by different amounts for different drugs.
State is where the prescriber practices, not where the beneficiary lives. CMS leaves out any state and drug with fewer than 11 claims, so state totals run slightly below the national drug file.
Totals are not adjusted for each state's Part D enrollment, which also changes year to year.
Next signal
Watch whether the same states lead in CMS's 2025 prescriber file.
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2024-part-d-state-growth
The costliest 1% of Medicare Part D prescribers accounted for 23.4% of 2024 gross drug cost and the costliest 10% for 71.1%; for the 8 largest high-cost drugs, each drug's costliest tenth of prescribers held 34.0% to 50.8% of its cost.
Part Dbaselinecurrent
Concentrated prescribing means a small group of prescribers drives most high-cost drug use - where prescriber outreach, utilization management and specialty-pharmacy programs get the most leverage.
United States · period covered: 2023–2024Affects: Health plans · Providers
Share of each high-cost drug's Part D cost from its costliest 10% of prescribers, 2024
Does not yet meet the 2-consecutive-period persistence rule. Every prescriber in CMS's file; concentration is a structural baseline, compared across 2 year(s).
CMS Medicare Part D Prescribers - by Provider and Drug, data year 2024, filtered to the 8 drugs with the most 2024 Part D spending among those above CMS's specialty-tier threshold ($950 per 30-day fill) — Medicare Part D Prescribers by Provider and Drug vintage 2024-12-31
Limitations
Spending is gross drug cost (what plans and beneficiaries paid at the pharmacy), before manufacturer rebates and other price concessions, which CMS is not allowed to publish; net cost is lower, by different amounts for different drugs.
CMS leaves out any prescriber and drug with fewer than 11 claims, so per-drug counts cover prescribers with 11 or more claims of that drug.
Specialty is CMS's label for each prescriber, from their claims or their national provider registration. Costlier prescribers often treat sicker patients; concentration is not a sign of inappropriate prescribing.
Prescriber names are as CMS publishes them. A prescriber's total reflects their patients' needs and practice size (often a specialty clinic or a physician treating many patients), not a judgment about their prescribing.
Next signal
Watch whether concentration and the specialty mix shift in CMS's 2025 prescriber files.
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2024-part-d-prescriber-concentration
Medicare Part D gross drug spending was $288.67B in 2024; Semaglutide (Ozempic, Rybelsus) added the most, $10.86B to $15.16B, and the 10 largest dollar gainers added $19.93B between them.
Part Dbaselineaging
The specific drugs behind Part D cost growth - the level where formulary tiering, prior authorization and rebate contracting happen.
United States · period covered: 2023–2024Affects: Health plans · Providers
Largest Part D gross spending change by drug, 2023 to 2024
Does not yet meet the 2-consecutive-period persistence rule. Every drug in CMS's annual file; one year's change is a baseline, not yet a trend.
Evidence
CMS Medicare Part D Spending by Drug, data years 2020-2024 (every drug sold in 2024, gross drug cost) — Medicare Part D Spending by Drug vintage 2024-12-31
Spending is gross drug cost (what plans and beneficiaries paid at the pharmacy), before manufacturer rebates and other price concessions, which CMS is not allowed to publish; net cost is lower, by different amounts for different drugs.
CMS's files carry no therapeutic category, so this reports individual drugs (by generic name, summed across brands) rather than drug classes.
The annual file lists only drugs still sold in 2024, so earlier years leave out drugs that have since left the market; this compares drugs, not an all-drug growth rate.
2025 and 2026 figures come from CMS's preliminary quarterly file (pulled three months after each quarter); CMS advises caution comparing them with the annual file, and they may change in later releases.
Next signal
Watch the next preliminary quarter for these drugs, and CMS's annual 2025 file (about 18 months after the year ends).
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2024-part-d-drug-spend-growth
Drugs above CMS's specialty-tier threshold ($950 per 30-day fill) were 46.1% of 2024 Part D gross spending ($133.13B) but 1.5% of claims; among established ones, Belzutifan (Welireg) grew fastest, $60M to $205M (+243.4%).
Part Dbaselineaging
High-cost (mostly specialty) drugs are a small share of prescriptions and a large share of cost, so their growth moves Part D budgets and plan bids more than anything else.
United States · period covered: 2023–2024Affects: Health plans · Providers
Fastest-growing high-cost Part D drugs, 2023 to 2024
Does not yet meet the 2-consecutive-period persistence rule. Every drug in CMS's annual file; one year's growth is a baseline, not yet a trend.
Evidence
CMS Medicare Part D Spending by Drug, data years 2020-2024 (every drug sold in 2024, gross drug cost) — Medicare Part D Spending by Drug vintage 2024-12-31
CMS Medicare Part D Prescribers - by Geography and Drug, national rows, data year 2024 (gross cost and 30-day fills per drug) — Medicare Part D Prescribers by Geography and Drug vintage 2024-12-31
Spending is gross drug cost (what plans and beneficiaries paid at the pharmacy), before manufacturer rebates and other price concessions, which CMS is not allowed to publish; net cost is lower, by different amounts for different drugs.
CMS's files carry no therapeutic category, so this reports individual drugs (by generic name, summed across brands) rather than drug classes.
"High-cost" uses CMS's Part D specialty-tier threshold ($950 per 30-day supply for 2024; $1,080 from 2027). CMS applies it to ingredient cost per 30-day supply on most of a plan's claims; this uses each drug's average gross cost per 30-day fill nationally, which also includes dispensing fees.
2025 and 2026 figures come from CMS's preliminary quarterly file (pulled three months after each quarter); CMS advises caution comparing them with the annual file, and they may change in later releases.
Next signal
Watch these drugs in the next preliminary quarter and CMS's annual 2025 file.
Generated by claims-utilization-cost-intelligence · sig-claims-cost-2024-part-d-high-cost-growth
Reimbursement & Provider Economics7 findings
1,956 of 2,945 hospitals (66.4%) take a net Medicare inpatient payment cut in FY2026 from the readmissions, hospital-acquired condition and value-based purchasing programs combined; 325 are cut by all three.
Medicare FFSbaselinecurrent
Shows how widely Medicare's quality penalties reach into hospital inpatient payment this year, and which kinds of hospitals are most often on the losing side: a starting point for contracting and quality-partnership conversations.
United States (hospitals paid under the inpatient prospective payment system) · period covered: Oct 2025–Sep 2026Affects: Hospitals · Health plans
Hospitals with a net Medicare inpatient payment cut, FY2026, by ownership
Adjustments apply to base inpatient operating payments, not a hospital's total revenue; without each hospital's Medicare inpatient volume here, dollar amounts are not estimated and every hospital counts equally.
Scores reflect performance periods one to three years before the payment year.
Critical access, children's, psychiatric and other hospitals not paid under the inpatient prospective payment system are outside these programs; Maryland hospitals are paid under the state's own model.
Next signal
Watch the FY2027 readmissions and value-based purchasing factors CMS posts with the inpatient final rule, to see which hospitals move in or out of a net cut.
Generated by reimbursement-payment-intelligence · sig-reimbursement-2026-hospital-penalty-exposure
Medicare's physician conversion factor is $33.4009 for 2026: +3.3% on 2025, and -1.8% since 2013 ($34.0230) before inflation.
Medicare FFStrendcurrent
The conversion factor turns every physician service's relative value into dollars, so it moves all Medicare physician rates at once, and many commercial and Medicaid fee schedules are set as a percent of Medicare.
United States · period covered: 2013–2026Affects: Physicians · Health plans
The conversion factor is one lever; relative values for individual services change every year as well, so a service's price can move against the factor.
Nominal dollars, not adjusted for inflation.
Uses each year's last quarterly release, the rate in effect at year end, which includes mid-year corrections.
Next signal
Watch the final Physician Fee Schedule rule for 2027, which sets next year's conversion factor.
Generated by reimbursement-payment-intelligence · sig-reimbursement-2026-conversion-factor-trend
Fee schedule price changes carried into Medicare claims: when a service's price moved 2% or more, its allowed amount per service moved the same way 88% of the time, 2013-2024 (r = 0.64).
Medicare FFSbaselinecurrent
Shows that a final fee schedule rule is a reliable early read on next year's Medicare physician payment by service, well before claims data arrives, so rate changes can be planned for rather than discovered.
United States · period covered: 2013–2024Affects: Physicians · Health plans
How closely claims followed fee schedule price changes, by year (correlation)
Meets persistence and has a full baseline window, but lacks independent external corroboration. Every priced service with enough volume, across 11 year pairs.
CMS Medicare Physician & Other Practitioners - by Geography and Service, national totals for every procedure code, data years 2013-2024 — Medicare Physician & Other Practitioners by Service vintage 2024-12-31
Limitations
Covers only services priced by the Physician Fee Schedule; Part B drugs, anesthesia, lab tests and ambulance are paid on other schedules and are left out.
Allowed amount per service also moves with where services are performed, geographic mix and billing modifiers, so the link is strong but not one to one.
Uses each year's last quarterly release; services billed earlier in a year may have been paid at an earlier release's rate.
Next signal
When 2025 claims data arrives, check whether services priced lower for 2025 show lower allowed amounts per service.
Generated by reimbursement-payment-intelligence · sig-reimbursement-2024-fee-schedule-leading-indicator
Across the 5 provider types selected as most noteworthy in 2024, Medicare paid 23% of submitted charges — Nurse Practitioner received the most Medicare payment.
Medicare FFSbaselinecurrent
Establishes a CMS-published payment benchmark by provider type - the public benchmark a plan or provider group would compare its own contracted rates against.
United States (all providers) · period covered: 2024Affects: Providers · Health plans
Medicare payment as % of submitted charges, by provider type (2024)
Does not yet meet the 2-consecutive-period persistence rule. Full national population for 2024; a single year's ratio is a baseline, not a trend.
Evidence
CMS Medicare Physician & Other Practitioners - by Provider, every provider in data year 2024 (1,296,739 providers), summarized at pull time — Medicare Physician & Other Practitioners by Provider vintage 2024-12-31
Limitations
Covers every Medicare fee-for-service Part B provider nationally, but not Medicare Advantage or commercial payments.
Payment-to-charge ratio reflects Medicare's fee-schedule-driven payment, not a negotiated commercial rate - submitted charge is a provider's list price, not a market price.
Provider types with fewer than 100 providers are left out, since a ratio over a few providers swings on a handful of large billers.
Next signal
Watch whether these ratios shift when CMS publishes 2025 data - a falling ratio means charges are growing faster than Medicare pays.
Generated by reimbursement-payment-intelligence · sig-reimbursement-2024-payment-vs-charge-by-provider-type
At 2024 volumes, 2026 fee schedule prices move Medicare physician payment -1.2% (-$818M) from 2024; 66984 takes the largest cut (-13.9%, -$225M) and 99214 the largest gain (+4.9%, $422M).
Medicare FFSpolicycurrent
Names the individual services whose Medicare price changed most in dollar terms since the latest claims year, the service lines where providers' revenue and payers' Medicare-indexed costs will shift first.
United States · period covered: 2024–2026Affects: Physicians · Health plans
Payment effect of 2024-2026 fee changes at 2024 volume, by service
Does not yet meet the 2-consecutive-period persistence rule. Prices are published; the dollar effect assumes 2024 volumes and site mix, which will have moved.
CMS Medicare Physician & Other Practitioners - by Geography and Service, national totals for every procedure code, data years 2013-2024 — Medicare Physician & Other Practitioners by Service vintage 2024-12-31
Limitations
Covers only services priced by the Physician Fee Schedule; Part B drugs, anesthesia, lab tests and ambulance are paid on other schedules and are left out.
National prices before geographic adjustment; a practice's actual rate differs by locality.
A projection at 2024 volume, not observed payment: services and site of care will have changed since.
Next signal
Watch 2025 and 2026 claims data for these services' allowed amount per service, and whether volume shifts as prices change.
Generated by reimbursement-payment-intelligence · sig-reimbursement-2026-procedure-fee-exposure
2026 fee schedule prices have not reached Medicare claims data yet: at 2024 volumes they move Procedure payment -5.9% and E&M +1.0% compared with 2024 prices.
Medicare FFSpolicycurrent
Sets what to expect in the next two years of physician claims: which kinds of care will cost Medicare less per service and which more, before any utilization response.
United States · period covered: 2024–2026Affects: Physicians · Health plans
Fee schedule price change 2024 to 2026 by kind of service, at 2024 volume
Does not yet meet the 2-consecutive-period persistence rule. Published prices applied to observed 2024 volume; the realized effect depends on volume and site of care after 2024.
CMS Medicare Physician & Other Practitioners - by Geography and Service, national totals for every procedure code, data years 2013-2024 — Medicare Physician & Other Practitioners by Service vintage 2024-12-31
Limitations
Covers only services priced by the Physician Fee Schedule; Part B drugs, anesthesia, lab tests and ambulance are paid on other schedules and are left out.
National prices before geographic adjustment; a practice's actual rate differs by locality.
Grouped by kind of service, not by specialty: the public claims data here does not split each service by the specialty that billed it.
Next signal
When CMS publishes 2025 claims data, check whether Procedure payment per service fell as these prices imply.
Generated by reimbursement-payment-intelligence · sig-reimbursement-2026-fee-change-by-category
WV's hospitals face the heaviest Medicare quality-program cuts in FY2026, averaging -1.19% of base inpatient payment with 79.2% of hospitals cut, against -0.47% nationally; MN fares best at +0.52%.
Medicare FFSbaselinecurrent
Points to the markets where hospital inpatient margins are under the most Medicare quality-penalty pressure this year, which shapes hospitals' leverage and priorities in payer negotiations.
WV · period covered: Oct 2025–Sep 2026Affects: Hospitals · Health plans
Average combined Medicare quality-program adjustment per hospital, FY2026, by state
Adjustments apply to base inpatient operating payments, not a hospital's total revenue; without each hospital's Medicare inpatient volume here, dollar amounts are not estimated and every hospital counts equally.
Scores reflect performance periods one to three years before the payment year.
Covers facility payment pressure from these three programs only, not physician fee changes or other payment policy.
States with fewer than 10 hospitals in the programs are left out; Maryland is not in them.
Next signal
Compare with the FY2027 factors when CMS posts them with the inpatient final rule, to see whether the same states stay at the top.
Generated by reimbursement-payment-intelligence · sig-reimbursement-2026-hospital-penalty-by-state
Provider & Network17 findings
Nursing homes reported 3.75 nurse hours per resident day in Q1 2026, up from 3.73 a year earlier (+0.4%); RN hours rose to 0.62, and staffing fell in 22 of 51 states.
Cross-populationtrendcurrent
Nurse staffing is the main capacity limit on nursing home admissions; where it falls, post-acute discharges get harder to place and quality risk rises.
United States · period covered: Jan 2025–Mar 2026Affects: Hospitals · Health plans
Change in nurse hours per resident day, Q1 2025 to Q1 2026
Meets persistence and has a full baseline window, but lacks independent external corroboration. Every reporting nursing home's daily payroll hours; same quarter compared year over year to avoid seasonal swings.
Nurse hours are RN (including the director of nursing and RN administrative time), LPN (including administrative time) and nurse aide hours (CNAs, aides in training, medication aides), divided by resident days from each day's MDS census. This is a straight ratio of totals, not CMS's case-mix-adjusted star-rating measure.
Hours per resident day doesn't account for how sick residents are; a state whose residents became more or less dependent can move without any change in care per need.
Facilities that didn't submit payroll data for a quarter are missing from that quarter.
Next signal
CMS's Q2 2026 payroll staffing file.
Generated by provider-network-intelligence · sig-provider-network-2026-03-31-nurse-staffing-trend
Contract (agency) staff worked 5.3% of nursing home nurse hours in Q1 2026, down from 6.2% a year earlier and 7.5% in Q1 2024; the share rose in 12 of 51 states.
Cross-populationtrendcurrent
Agency staff cost nursing homes more per hour and turn over faster; high or rising reliance points to operators under margin and continuity-of-care pressure.
United States · period covered: Jan 2024–Mar 2026Affects: Hospitals · Health plans
Contract share of nursing home nurse hours, Q1 2026
% of nurse hours
Vermont23.2
New Mexico14.2
New Hampshire13.1
Alaska12.8
North Dakota11.7
Montana11.6
New Jersey11.6
Pennsylvania11.3
0102030
Share of nursing home nurse hours worked by contract staff
Meets persistence and has a full baseline window, but lacks independent external corroboration. Every reporting nursing home's daily payroll and contract hours.
Nurse hours are RN (including the director of nursing and RN administrative time), LPN (including administrative time) and nurse aide hours (CNAs, aides in training, medication aides), divided by resident days from each day's MDS census. This is a straight ratio of totals, not CMS's case-mix-adjusted star-rating measure.
The file splits hours into employee and contract only; it has no wage or cost data, so this is reliance, not spending.
Next signal
Whether the contract share keeps falling in CMS's next quarterly payroll staffing file.
Generated by provider-network-intelligence · sig-provider-network-2026-03-31-contract-nurse-staffing
In PY2024, 360 of 476 Shared Savings Program ACOs (75.6%) earned $4.14B in shared savings and 16 owed $20.0M in losses; spending came in $6.61B (4.7%) under benchmarks, up from 69.1% of ACOs earning in PY2023.
Medicare FFStrendcurrent
Whether accountable-care contracts are paying off for provider groups, and which organizations are doing it at scale - partners, competitors or models for value-based contracting.
United States · period covered: 2023–2024Affects: Health plans · Providers · States
Spending below benchmark, largest savings rates, PY2024
% below benchmark
Sound Physicians Long Term Care Management, LLC20.9
LTC ACO, LLC20.2
ProHealth Physicians ACO, LLC17.9
Aledade Louisiana ACO, LLC17.3
EPC ACO, LLC16.7
Aledade Accountable Care 15, LLC16.6
0102030
Share of Shared Savings Program ACOs earning shared savings
Meets persistence and has a full baseline window, but lacks independent external corroboration. Every ACO's reported result; totals match CMS's own PY2024 results fact sheet.
Savings are measured against CMS's benchmark for each ACO, which CMS sets from past spending with regional and trend adjustments; a benchmark change moves savings without any change in care.
One-sided ACOs never owe losses, and in PY2020-PY2022 CMS's file shows no losses owed at all, so the owed count is not comparable across those years.
PY2019 is left out of year-over-year results: CMS split it into two six-month performance periods.
Next signal
CMS's PY2025 financial and quality results (the PY2024 file was published in 2026).
Generated by policy-regulation-cms-program-intelligence · sig-policy-2024-mssp-savings-losses
76.3% of Shared Savings Program ACOs are in two-sided risk tracks in 2026, up from 40.9% in 2021; by assigned beneficiary, the two-sided share rose from 43.8% in 2021 to 74.2% in 2024, and 44 of 46 states rose.
Medicare FFStrendcurrent
Two-sided ACOs repay Medicare when spending runs over benchmark, so this is where provider groups are taking real financial risk - the markets most ready for downside-risk contracts with other payers.
United States · period covered: Jan 2021–Jan 2026Affects: Health plans · Providers · States
Change in two-sided share of ACO-assigned beneficiaries, 2021 to 2024
percentage points
New Mexico77.8
Oregon68.9
Mississippi67.5
Colorado66
Oklahoma64.2
Kansas62.4
Maine-10.8
California31.6
-250255075100
Share of Shared Savings Program ACOs in two-sided risk
Meets persistence and has a full baseline window, but lacks independent external corroboration. National counts cover every ACO; state shares cover every unsuppressed county row.
Two-sided means BASIC levels C-E, ENHANCED, or the older Tracks 1+, 2 and 3; BASIC A-B and Track 1 are upside-only. From 2019 CMS moved ACOs onto a glide path toward risk, so part of the rise is required by program rules, not chosen.
State shares use CMS's county file, which hides counts of 1-10 (112,678 ACO-county rows in 2024; the rest cover 97.7% of that year's assigned beneficiaries), and each ACO's risk model from that year's results file. States with fewer than 10,000 assigned beneficiaries are left out.
The county file stops at 2024; the ACO track counts run through 2026.
Next signal
CMS's 2025 county assignment file, to see whether the state shift continued.
Generated by policy-regulation-cms-program-intelligence · sig-policy-2026-mssp-two-sided-risk
The top 4 hospital ownership types (Voluntary non-profit - Private, Proprietary, Government - Hospital District or Authority, Government - Local) account for 79.2% of all facilities nationally, as of the current CMS snapshot.
Medicare FFSbaselinecurrent
A CR4 concentration baseline for hospital ownership - no concentration change has been observed yet across the pulls collected so far.
United States · period covered: Sep 16–28, 2026Affects: Hospitals · Health plans
Hospital ownership type, share of facilities
facilities
Voluntary non-profit - Private2,32242.8%
Proprietary1,06319.6%
Government - Hospital District or Authority5129.4%
Does not yet meet the 2-consecutive-period persistence rule. Based on 4 snapshot(s) spanning 12 day(s); full confidence needs a 2-year baseline.
Evidence
CMS Hospital General Information, 4 real snapshot(s) from 2026-09-16 to 2026-09-28 — Hospital General Information vintage 2026-09-28
Limitations
Ownership-type concentration, not entity-level concentration - two facilities under the same ownership *type* are not necessarily the same owner.
Baseline window is 12 real day(s) so far, well short of the 24-month window this framework requires for full confidence.
Next signal
Watch subsequent real pulls for an ownership-type mix shift.
Generated by provider-network-intelligence · sig-provider-network-2026-09-28-ownership-concentration
Across 2961 real hospitals with both a CMS overall star rating and complete mortality/safety/readmission measure counts, star rating and net quality-outcome performance show a moderate-to-strong positive correlation (Pearson r = 0.56).
Medicare FFSbaselinecurrent
Tests whether CMS's headline star rating actually tracks the underlying real outcome measures it's meant to summarize - relevant to whether star rating alone is a sufficient quality proxy for network or quality-incentive decisions, or whether the underlying measure detail should be reviewed directly.
United States · period covered: Sep 16–28, 2026Affects: Hospitals · Health plans
Hospital overall star rating vs. net quality-outcome score
CMS overall star rating vs Net quality-outcome score
One record
Median at each CMS overall star rating
SOUTHEAST HEALTH MEDICAL CENTER (AL)CMS overall star rating 4 starsNet quality-outcome score 11.5 %
A cross-sectional correlation from a single real snapshot, not yet observed to persist across multiple pulls. Correlation alone is never sufficient grounds for a causal claim regardless of persistence - this insight's driver is explicitly marked "correlation", never "confirmed-causal".
Evidence
CMS Hospital General Information, real hospital_overall_rating and count_of_{mort,safety,readm}_measures_{better,worse,no_different} fields across 2961 hospitals, 2026-09-28 — Hospital General Information vintage 2026-09-28
Limitations
Correlation, not causation - the star rating methodology itself partly derives from these same measure groups, so some correlation is expected by construction, not solely reflecting an independently corroborating quality signal.
The star rating also incorporates real patient-experience and timely-and-effective-care measure groups (plus CMS's own weighting/imputation methodology) that this net quality-outcome score does not include, so a moderate rather than perfect correlation is the honestly expected result even if the two were measuring the same underlying construct.
Limited to the 2961 hospitals with both a numeric star rating and complete mortality/safety/readmission measure counts - hospitals with a suppressed/"Not Available" rating or measure count are excluded outright, never treated as zero.
Chart shows a deterministic, evenly-spaced sample of 371 of these 2961 real hospitals for rendering size - the correlation coefficient above is computed across the full 2961-hospital real sample, not the chart's sample.
Next signal
Watch this correlation across a second real pull once CMS's quarterly measure refresh next lands, to see whether the relationship's strength itself is stable.
Generated by provider-network-intelligence · sig-provider-network-2026-09-28-quality-correlation
Among the 8 states selected as most noteworthy, UT has the highest median (5.0★) and MS the lowest (2.0★).
Medicare FFSbaselinecurrent
A real, state-level distributional read on hospital quality star ratings - a starting point for network-quality-by-geography questions, not a claim about why any state's distribution sits where it does.
UT, CO, PA, SC, CA, TX, WV, MS · period covered: Sep 16–28, 2026Affects: Hospitals · Health plans
Hospital overall star rating distribution, 8 states selected as most noteworthy
A single real cross-sectional snapshot - not yet observed to persist across multiple pulls, and this dataset refreshes quarterly so genuine week-to-week movement isn't expected yet at this cadence.
Evidence
CMS Hospital General Information, real hospital_overall_rating field across 810 hospitals in the 8 selected states, 2026-09-28 — Hospital General Information vintage 2026-09-28
Limitations
Limited to the 8 states with at least 20 hospitals reporting a real, non-suppressed star rating - smaller states/territories are excluded to avoid a noisy few-hospital read, not because their data is less real.
Star rating is CMS's own composite methodology (incorporating mortality, safety, readmission, patient experience, and timely-and-effective-care measure groups plus its own weighting/imputation) - this box shows the rating's real distribution, not a re-derivation of it.
Next signal
Watch this state ranking across CMS's next quarterly refresh for the first genuine multi-period shift.
Generated by provider-network-intelligence · sig-provider-network-2026-09-28-star-rating-by-state
Among the 8 states selected as most noteworthy, UT has the best median net quality-outcome score (16.7%) and MS the worst (-7.7%).
Medicare FFSbaselinecurrent
A state-level read on hospital quality *outcomes* specifically (not the composite star rating) - useful alongside the star-rating-by-state finding to see whether a state's star-rating standing and its underlying outcome-measure performance tell the same story.
UT, CO, AZ, WI, CA, TX, WV, MS · period covered: Sep 16–28, 2026Affects: Hospitals · Health plans
Net quality-outcome score distribution, 8 states selected as most noteworthy
A single real cross-sectional snapshot - not yet observed to persist across multiple pulls, and this dataset refreshes quarterly so genuine week-to-week movement isn't expected yet at this cadence.
Evidence
CMS Hospital General Information, real count_of_{mort,safety,readm}_measures_{better,worse,no_different} fields across 766 hospitals in the 8 selected states, 2026-09-28 — Hospital General Information vintage 2026-09-28
Limitations
Limited to the 8 states with at least 20 hospitals with a complete, non-suppressed real measure-group count - smaller states/territories are excluded to avoid a noisy few-hospital read.
Net quality-outcome score only covers the mortality, safety, and readmission measure groups this dataset reports a better/worse/no-different count for - it does not include patient-experience or timely-and-effective-care measures.
Next signal
Watch this state ranking across CMS's next quarterly refresh for the first genuine multi-period shift.
Generated by provider-network-intelligence · sig-provider-network-2026-09-28-quality-outcome-by-state
No state has yet shown a persistent real improvement in net quality-outcome score across the 4 real pull(s) collected so far (this CMS dataset refreshes quarterly, and only 12 real day(s) of history exist); currently, UT has the best net quality-outcome score (16.7%) among states with enough hospitals to assess.
Medicare FFSbaselinecurrent
Directly answers whether any state shows real, sustained improvement in hospital quality outcomes over time, alongside a current-state ranking of where quality is best right now - both genuinely useful reads for network-investment and quality-incentive prioritization, reported honestly rather than forcing a 'trend' claim before one is real.
UT, CO, AZ, WI, OK · period covered: Sep 16–28, 2026Affects: Hospitals · Health plans
Current net quality-outcome score, 5 states selected as most noteworthy
Does not yet meet the 2-consecutive-period persistence rule. Based on 4 snapshot(s) spanning 12 day(s), short of the 2-year baseline full confidence needs; this CMS dataset itself refreshes quarterly, so genuine movement isn't expected at a sub-quarterly cadence regardless of pull frequency.
Evidence
CMS Hospital General Information, real per-state median net quality-outcome score across 4 real pull(s), 2026-09-16 to 2026-09-28 — Hospital General Information vintage 2026-09-28
Limitations
Limited to the 40 states with at least 20 hospitals with a real, non-suppressed net quality-outcome score in every real pull collected so far.
This dataset is CMS's own quarterly-refresh hospital quality data - real week-to-week pulls are not expected to show movement until the next quarterly refresh actually lands; a 'no persistent improvement yet' finding reflects that real refresh cadence, not a system limitation.
Tracks each state's median net quality-outcome score across all qualifying hospitals in that state, not any individual hospital's own trajectory.
Next signal
Watch every eligible state's median net quality-outcome score across CMS's next quarterly refresh for the first genuine, assessable direction of change.
Generated by provider-network-intelligence · sig-provider-network-2026-09-28-quality-by-geography-trend
Zero real facility entries observed across the 4 real Hospital General Information pulls collected so far (2026-09-16 to 2026-09-28).
Medicare FFSbaselinecurrent
A facility entry is a market-expansion signal worth an executive's attention - new capacity entering a market can affect network-adequacy and competitive positioning.
United States · period covered: Sep 16–28, 2026Affects: Hospitals · Health plans
Does not yet meet the 2-consecutive-period persistence rule. Based on 4 snapshot(s) spanning 12 day(s), short of the 2-year baseline full confidence needs; this CMS dataset itself refreshes quarterly, so an entry or exit is not expected at a sub-quarterly cadence regardless of pull frequency.
Evidence
CMS Hospital General Information, real facility_id diff across 4 real pulls, 2026-09-16 to 2026-09-28 — Hospital General Information vintage 2026-09-28
Limitations
Bounded to what this project has actually observed across 4 pull(s) spanning 12 day(s). CMS's live data API keeps no archive of past versions of this directory, so entries and exits before this project's first pull can't be recovered; the list grows only from here.
A facility disappearing from this dataset means it stopped appearing in CMS's own published file - it does not by itself confirm the facility physically closed (e.g. a real ownership/ID change could also cause this), and a facility appearing does not by itself confirm a genuinely new physical location rather than a reporting change.
Never a claim about why a facility entered or exited - only that its real facility_id appeared or disappeared between two real pulls.
Next signal
Watch CMS's next quarterly refresh of the hospital directory for the first facility added or removed.
Generated by provider-network-intelligence · sig-provider-network-2026-09-28-facility-entries
Zero real facility exits observed across the 4 real Hospital General Information pulls collected so far (2026-09-16 to 2026-09-28).
Medicare FFSbaselinecurrent
A facility exit is an access-risk signal worth an executive's attention - lost capacity can affect network adequacy and shift utilization to remaining facilities.
United States · period covered: Sep 16–28, 2026Affects: Hospitals · Health plans
Does not yet meet the 2-consecutive-period persistence rule. Based on 4 snapshot(s) spanning 12 day(s), short of the 2-year baseline full confidence needs; this CMS dataset itself refreshes quarterly, so an entry or exit is not expected at a sub-quarterly cadence regardless of pull frequency.
Evidence
CMS Hospital General Information, real facility_id diff across 4 real pulls, 2026-09-16 to 2026-09-28 — Hospital General Information vintage 2026-09-28
Limitations
Bounded to what this project has actually observed across 4 pull(s) spanning 12 day(s). CMS's live data API keeps no archive of past versions of this directory, so entries and exits before this project's first pull can't be recovered; the list grows only from here.
A facility disappearing from this dataset means it stopped appearing in CMS's own published file - it does not by itself confirm the facility physically closed (e.g. a real ownership/ID change could also cause this), and a facility appearing does not by itself confirm a genuinely new physical location rather than a reporting change.
Never a claim about why a facility entered or exited - only that its real facility_id appeared or disappeared between two real pulls.
Next signal
Watch CMS's next quarterly refresh of the hospital directory for the first facility added or removed.
Generated by provider-network-intelligence · sig-provider-network-2026-09-28-facility-exits
57 hospital ownership changes took effect in 2025 as of CMS's 2026-Q2 release, down 37.4% from 91 for 2024 at the same point; BOWLING GREEN-WARREN COUNTY COMMUNITY HOSPITAL CORPORATION was the most active buyer in 2023-2025 (6 hospitals).
Cross-populationbaselinecurrent
Hospital ownership changes reshape who a plan negotiates with: an acquired hospital usually moves onto the buyer's contracts and rates. Knowing which systems are buying, and where, is an early read on contracting leverage.
United States · period covered: Jan 2016–Jun 2026Affects: Hospitals · Health plans
Most active hospital buyers, 2023-2025
hospitals
BOWLING GREEN-WARREN COUNTY COMMUNITY HOSPITAL CORPORATION6
One year against one year at the same reporting lag; counts this small move a lot from year to year, so this is a baseline, not a trend.
Evidence
CMS Hospital Change of Ownership, every change effective 2016 onward (2026-04-01 to 2026-06-30 release), with each past quarterly release's counts for the same-lag comparison — Hospital and SNF Change of Ownership vintage 2026-06-30
CMS Hospital General Information, current overall star rating, joined by CMS certification number — Hospital General Information vintage 2026-09-28
Limitations
A change of ownership keeps appearing in CMS's files for 12-18 months after it takes effect, so the most recent year is always incomplete and is compared only with the prior year as it was reported at the same point.
Buyers are counted by the legal name on each enrollment; one health system or chain can buy under several names, so totals per organization are understated.
Covers changes CMS processed for Medicare-enrolled hospitals; deals that don't change the Medicare enrollment (for example a purchase of a parent company above the hospital) may not appear.
Star ratings are from the current Hospital General Information file, not from before the change.
Next signal
CMS's next quarterly release will add more 2025 changes; watch whether 2025 ends above or below 2024 once both are fully reported.
Generated by provider-network-intelligence · sig-provider-network-2026-06-30-hospital-ownership-changes
2,009 skilled nursing facilities changed owners in 2023-2025, 13.6 for every 100 nursing homes nationally; OR had the highest rate (41.4 per 100), and 2025 is running -6.0% against 2024 at the same reporting lag.
Cross-populationbaselinecurrent
Nursing home ownership turns over far faster than hospital ownership, and a new owner can change staffing, quality and which plans a facility contracts with. States with high turnover are where post-acute networks need the closest watch.
45 states with at least 50 nursing homes · period covered: Jan 2023–Jun 2026Affects: Nursing homes · Health plans
SNF ownership changes per 100 nursing homes, 2023-2025, 8 states selected as most noteworthy
A three-year count against today's nursing home count, from two public CMS files; not yet compared across periods by state.
Evidence
CMS Skilled Nursing Facility Change of Ownership and its owner information file (2026-04-01 to 2026-06-30 release), with each past release's counts for the same-lag comparison — Hospital and SNF Change of Ownership vintage 2026-06-30 (this record)
A change of ownership keeps appearing in CMS's files for 12-18 months after it takes effect, so the most recent year is always incomplete and is compared only with the prior year as it was reported at the same point.
Buyers are counted by the legal name on each enrollment; one health system or chain can buy under several names, so totals per organization are understated.
Rates divide changes over three years by today's nursing home count; a home that changed owners twice counts twice.
Owners are the 5%-or-greater organization owners on file for each buyer today; individuals, trusts and public hospital districts that own nursing homes appear under their own names.
Next signal
Watch whether 2025 ends above or below 2024 once both are fully reported, and whether the high-turnover states stay on top.
Generated by provider-network-intelligence · sig-provider-network-2026-06-30-snf-ownership-changes-by-state
41 hospitals and 97 skilled nursing facilities report a private equity owner to Medicare as of 2026-08; most of the rise since CMS added the field is older ownership being reported, not new deals, and 55 of the 2,009 nursing homes that changed owners in 2023-2025 now report one.
Cross-populationbaselinecurrent
Private equity ownership of hospitals and nursing homes draws regulatory attention and can change a facility's staffing, pricing and contracting posture. This is the first federal flag that names it directly, so it's the baseline to track.
United States · period covered: Nov 2024–Aug 2026Affects: Hospitals · Nursing homes · Health plans
Private equity owners by facilities on file, 2026-08
facilities
ONE EQUITY PARTNERS VII-A LP (hospitals)38
ONE EQUITY PARTNERS VII LP (hospitals)37
ONE EQUITY PARTNERS VII-B LP (hospitals)36
OMEGA HEALTHCARE INVESTORS INC (SNFs)9
EMSIR LLC (SNFs)8
LADS AVENUE ASSOCIATES LLC (SNFs)6
2020 GSR DYNASTY LLC (SNFs)5
AH DYNASTY LLC (SNFs)5
010203040
Skilled nursing facilities reporting a private equity owner
The private equity field is new and self-reported, and most of its growth so far is reporting catching up, so the count can't yet be read as a trend.
Evidence
CMS Hospital All Owners and Skilled Nursing Facility All Owners, every monthly release since each added a private equity owner field (hospitals 2025-04, SNFs 2024-11) through 2026-08 — Hospital and SNF All Owners vintage 2026-08-31
Facilities report their own owner types; a private equity firm that owns through a holding company or operating company may not be flagged, so these counts are a floor, not the full extent of private equity ownership.
Months before the field existed can't be compared, so the history starts when CMS added it.
Private equity firms often hold facilities through several fund entities; each fund name is counted separately as CMS records it.
Owners are as of the latest month, not at the time of each ownership change.
Next signal
Watch the count of private equity owner links with a recent start date: once late reporting catches up, that is the measure of new private equity acquisitions.
Generated by provider-network-intelligence · sig-provider-network-2026-08-31-private-equity-owners
The 269 nursing homes that changed owners between Q1 2025 and Q1 2026 moved a median -0.06 nurse hours per resident day over the year, against +0.02 for the 14,036 that didn't; staffing fell at 55.4% of the homes with new owners vs. 46.4% of the rest.
Cross-populationbaselinecurrent
Whether new owners add or cut nursing staff in their first months - an early read on how an acquisition is likely to change care at facilities in a network.
United States · period covered: Jan 2025–Mar 2026Affects: Hospitals · Health plans
Median change in nurse hours per resident day, Q1 2025 to Q1 2026
Does not yet meet the 2-consecutive-period persistence rule. One year-over-year window; a correlation between two CMS files, not a measured effect of the ownership change.
CMS Skilled Nursing Facility Change of Ownership (version through 2026-06-30), changes effective after 2025-03-31 through 2026-01-01 — Hospital and SNF Change of Ownership vintage 2026-06-30
Limitations
Correlation only: facilities that change hands may differ from others before the sale (size, finances, prior staffing), and the change may land at any point in the window.
Facility-quarters under 1.5 or over 12 nurse hours per resident day are treated as reporting errors and left out. CMS's change-of-ownership file keeps filling in for a year or more, so recent changes are undercounted.
No facility or owner is named; private-equity ownership is not compared because CMS's owner file identifies facilities by enrollment ID, not CCN.
Next signal
Repeat with the next quarter's payroll file to see whether the gap holds.
Generated by provider-network-intelligence · sig-provider-network-2026-03-31-staffing-after-ownership-change
511 ACOs are in the Medicare Shared Savings Program in 2026, up 35 from 2025 and below the 2018 peak of 561; in PY2024 the program covered 10.3 million assigned beneficiaries.
Medicare FFStrendcurrent
How many provider groups are taking Medicare accountable-care contracts - the base that value-based arrangements with commercial and MA payers often build on.
United States · period covered: Jan 2014–Jan 2026Affects: Health plans · Providers · States
ACOs in the Shared Savings Program (as of January)
ACO counts are as of each January (2019's file is the July list, after that year's mid-year start); ACOs that leave mid-year are still counted.
Covers the Shared Savings Program only, not CMS Innovation Center models such as ACO REACH.
Next signal
CMS's 2027 ACO list (January 2027).
Generated by policy-regulation-cms-program-intelligence · sig-policy-2026-mssp-participation
From PY2021 to PY2024, 9 of 11 comparable Shared Savings Program quality measures improved on average across ACOs and 2 declined; the biggest gain was Health Promotion and Education (61.6 to 65.5) and the biggest drop Access to Specialists (78.8 to 75.8).
Medicare FFStrendcurrent
Whether accountable-care savings come with better or worse patient experience and readmissions - the quality side a value-based contract is also judged on.
United States · period covered: 2021–2024Affects: Health plans · Providers · States
Change in average ACO patient-experience score, PY2021 to PY2024
points (out of 100)
Health Promotion and Education3.9
Health Status/Functional Status2.4
Stewardship of Patient Resources2.3
Shared Decision Making1.4
Courteous and Helpful Office Staff1
How Well Your Providers Communicate0.4
Patients' Rating of Provider0.2
Care Coordination0.2
Getting Timely Care, Appointments, and Information-1
Only measures collected the same way every year are compared: the CAHPS patient-experience survey and the claims-based readmission measure. The clinical measures (diabetes control, depression screening, blood pressure control) moved from the CMS Web Interface to other collection types in PY2024, so their values aren't comparable.
Averages are across ACOs, not weighted by beneficiaries, and the set of ACOs changes each year.
The quality performance standard itself changed in PY2024 (a higher percentile threshold), so the share meeting it is not compared across years.
Next signal
CMS's PY2025 results.
Generated by policy-regulation-cms-program-intelligence · sig-policy-2024-mssp-quality-measures
Policy & Program Watch3 findings
CMS rules finalized per month (last 3 months, through Oct 1)
rules
Days since comment period closed (negative = still open), most recent proposed CMS rules
days
Request for Information; Medicare Part D Reasonable and Rel…-53
Amending the Indirect Hold Harmless Threshold of Health Car…10
CY 2027 Payment Policies Under the Physician Fee Schedule a…17
Request for Information; Clinical Laboratory Improvement Am…17
Hospital Outpatient Prospective Payment and Ambulatory Surg…31
Calendar Year 2027 Home Health Prospective Payment System (…31
-75-50-2502550
CMS finalized 59 rules in the last 730 days — most recently "Medicare Program; Inpatient Rehabilitation Facility Prospective Payment System for Federal Fiscal Year 2027 and Updates to the IRF Quality Reporting Program; Correction" on 2026-10-01.
N/Apolicycurrent
Distinguishes what CMS has actually finalized — and is therefore binding — from what's merely proposed (see the proposed-rules finding). Final rules are what budgets and operations must plan around; proposed ones are scenarios.
United States · period covered: Oct 2024–Oct 2026Affects: Health plans · Providers · States
Recently finalized CMS rules selected as most noteworthy
This interim final rule with comment period (IFC) interprets and implements the community engagement requirement in Medicaid under section 1902(xx) of the Social Security Act.
This final rule revises the Medicare Advantage (Part C), Medicare Prescription Drug Benefit (Part D), and Medicare cost plan regulations to implement changes related to Star Ratings, marketing and communications, drug coverage, enrollment processes, special needs plans, and other programmatic areas.
This final rule will revise the Medicare hospital inpatient prospective payment systems (IPPS) for operating and capital-related costs of acute care hospitals; make changes relating to Medicare graduate medical education (GME) for teaching hospitals; update the payment policies and the annual payment rates for the Medicare prospective payment system (PPS) for inpatient hospital services provided…
This interim final rule with comment period repeals provisions of the final rule titled "Medicare and Medicaid Programs; Minimum Staffing Standards for Long-Term Care Facilities and Medicaid Institutional Payment Transparency Reporting." This action is taken in view of changes made by by public law, which precludes HHS from implementing, administering, or enforcing certain provisions of the final…
This major final rule addresses: changes to the physician fee schedule (PFS); other changes to Medicare Part B payment policies to ensure that payment systems are updated to reflect changes in medical practice, relative value of services, and changes in the statute; codification of establishment of new policies for: the Medicare Prescription Drug Inflation Rebate Program under the Inflation…
Bounded to CMS-attributed documents published in the trailing 730 days, not full regulatory history.
Does not yet determine which domain(s) (payment, providers, beneficiaries, utilization) each rule actually affects — that sorting is not done yet.
A rule being 'finalized' here means it appeared in the Federal Register with type=Rule — it does not yet confirm downstream claims-level or provider-behavior impact.
Next signal
Watch for whether these rules' effective dates arrive as scheduled, and for the next batch of finalizations in a future pull.
Generated by policy-regulation-cms-program-intelligence · sig-policy-2026-10-01-finalized-rules
48 CMS rules are currently proposed but not yet final — 1 still has an open comment period.
N/Apolicycurrent
These are explicitly proposals, not decided policy — CMS can and does change substance between a proposed and final rule, per this system's standing rule to never present a proposal as settled. Tracking them separately from finalized rules prevents premature operational planning against a rule that could still change.
United States · period covered: Oct 2024–Oct 2026Affects: Health plans · Providers · States
This proposed rule addresses: changes to the physician fee schedule (PFS); other changes to Medicare Part B payment policies to ensure that payment systems are updated to reflect changes in medical practice, relative value of services, and changes in the statute; codification of establishment of new policies for: the Medicare Prescription Drug Inflation Rebate Program under the Inflation…
Published 2026-07-16 · comment period closed 2026-09-14
This proposed rule would codify the Medicare Drug Price Negotiation Program ("Negotiation Program") and would establish certain new policies for the Negotiation Program and the Medicare Prescription Drug Benefit Program as required by the Inflation Reduction Act of 2022.
Published 2026-06-16 · comment period closed 2026-08-17
These proposals are intended to improve the electronic exchange of health care data and streamline processes related to prior authorization by increasing the interoperability of systems used across the health care industry.
Published 2026-04-14 · comment period closed 2026-06-15
This proposed rule would revise the Medicare hospital inpatient prospective payment systems (IPPS) for operating and capital- related costs of acute care hospitals; make changes relating to Medicare graduate medical education (GME) for teaching hospitals; update the payment policies and the annual payment rates for the Medicare prospective payment system (PPS) for inpatient hospital services…
Published 2026-04-14 · comment period closed 2026-04-10
This proposed rule contains provisions to improve implementation of the Patient Protection and Affordable Care Act, including payment parameters and provisions related to the HHS-operated risk adjustment and risk adjustment data validation (HHS-RADV) programs, as well as 2027 user fee rates for issuers offering qualified health plans (QHPs) through Federally-facilitated Exchanges (FFEs) and…
Published 2026-02-11 · comment period closed 2026-03-13
A closed comment period does not mean a final rule is imminent — CMS sets no fixed deadline between comment close and finalization, and this system does not estimate one.
Proposed rules with no published comments_close_on (e.g. requests for information) are included in the count but excluded from the 'still open' figure.
Must never be read as a forecast of what the final rule will say — only that a proposal exists and is not yet decided.
Next signal
Watch whether each proposed rule here is later matched by a real finalized rule with the same or a related title in a future pull — that pairing, once observed, is the actual proposed-to-final signal this system can then report on.
Generated by policy-regulation-cms-program-intelligence · sig-policy-2026-10-01-proposed-rules
15 finalized CMS rules become effective in the next 258 days — the nearest is "Medicare Program; FY 2027 Inpatient Psychiatric Facilities Prospective Payment System-Rate Update; Correction" on 2026-10-01.
N/Apolicycurrent
A forward operational-readiness calendar — rules on this list are already finalized (not proposals) and will become binding on their stated date, giving a real lead time to prepare rather than reacting after the fact.
United States · period covered: Oct 1, 2026–Jun 16, 2027Affects: Health plans · Providers · States
Upcoming finalized CMS rules selected as most noteworthy
This final rule will revise the Medicare hospital inpatient prospective payment systems (IPPS) for operating and capital-related costs of acute care hospitals; make changes relating to Medicare graduate medical education (GME) for teaching hospitals; update the payment policies and the annual payment rates for the Medicare prospective payment system (PPS) for inpatient hospital services provided…
This final rule finalizes changes and updates to the policies and payment rates used under the Skilled Nursing Facility (SNF) Prospective Payment System (PPS) for fiscal year (FY) 2027.
This final rule updates the prospective payment rates, the outlier threshold, and the wage index for Medicare inpatient hospital services provided by Inpatient Psychiatric Facilities (IPFs), which include psychiatric hospitals and excluded psychiatric units of an acute care hospital or critical access hospital.
In accordance with the Privacy Act of 1974, as amended, the Department of Health and Human Services (HHS) is modifying an existing system of records maintained by the Centers for Medicare & Medicaid Services (CMS), titled "Hospice Item Set (HIS) System," System No. 09-70-0548.
Only rules already finalized are included — proposed rules with no effective date are covered separately (see the proposed-rules signal) and are explicitly excluded here.
Limited to the 730-day publication window this pull covers — a rule published earlier with a still-future effective date would not appear here if it fell outside that window.
Next signal
Watch each of these effective dates arrive on schedule in a future pull — a rule that fails to take effect as stated would itself be a notable signal.
Generated by policy-regulation-cms-program-intelligence · sig-policy-2026-10-01-upcoming-effective
Emerging Signals1 finding
Across 51 states with data in both sources, hospital ownership concentration (CR4) and standardized Medicare physician payment per service show a weak positive correlation (r=0.15).
Medicare FFSbaselinecurrent
Unlike a facility-count comparison (mechanically expected to track physician staffing), hospital ownership concentration has no definitional link to physician payment levels - this is a market-power question (does a more consolidated hospital market coincide with higher or lower physician reimbursement?) worth an executive's attention whichever direction the result points. It takes two agents' data together: neither dataset can answer it alone.
51 states and DC with data in both sources · period covered: Jan 1, 2024–Sep 28, 2026Affects: Hospitals · Physicians
Hospital ownership concentration vs. standardized physician payment per service, by state
Hospital ownership concentration (CR4) vs Standardized Medicare payment per service
One record
ALHospital ownership concentration (CR4) 85.9 %Standardized Medicare payment per service 29.1 USD
51 states are comparable across both sources, but this is cross-sectional (one point in time per state), not a trend across periods, and the two sides are from different years.
Evidence
CMS Hospital General Information, real ownership-concentration (CR4) for 51 states and DC — Hospital General Information vintage 2026-09-28
CMS Medicare Physician & Other Practitioners - by Provider, standardized payment per service by state, every provider in data year 2024 — Medicare Physician & Other Practitioners by Provider vintage 2024-12-31
Limitations
Correlation, not causation - either variable could track a real underlying driver (e.g. state cost-of-living or regulatory environment) rather than one influencing the other directly.
The hospital side is the current CMS directory and the physician side is CMS's 2024 data year - the latest each publishes, not the same year.
States are where the billing provider is located, not where the patient lives.
Hospital ownership concentration (institutional) and individual-physician payment (professional) are different entity types - this is a market-level comparison, not a claim that the same providers appear in both datasets.
Next signal
Watch whether this correlation direction holds once more states or more periods are added to either source.
Generated by emerging-trends-signal-detection · sig-emerging-2026-09-28-ownership-concentration-vs-payment-correlation
Market Catalysts19 findings
openFDA recorded 104 new molecular entity (Type 1) approvals between 2024-10-01 and 2026-10-01; most recently ONSWIK (INSULIN EFSITORA ALFA-GOBE), sponsor ELI LILLY AND CO, approved 2026-09-23 under STANDARD review.
N/Apolicycurrent
A real, regulatory-approval-based catalyst signal - a new molecular entity approval marks the start of a real commercial product's market entry, directly relevant to a payer's formulary/utilization-management planning horizon.
United States · period covered: Oct 2024–Oct 2026Affects: Drug makers · Health plans
Official record, pulled directly from the issuing government source.
Evidence
openFDA drugsfda, 104 real Type 1 (new molecular entity) approvals with submission_status AP, 2024-10-01 to 2026-10-01 — openFDA novel drug approvals vintage 2026-10-01 (this record)
Limitations
Never uses the term "breakthrough" — openFDA's drugsfda dataset has no such field. Only the real `submission_class_code` (e.g. "TYPE 1") and `review_priority` (e.g. "PRIORITY" vs. "STANDARD") fields are used verbatim.
openFDA's search matches at the application level, not the submission level — this adapter re-filters to only the specific real submission(s) whose own status date falls in this window and whose class code/status actually match, per this agent's data adapter header.
Bounded to a single 1000-application API request per pull (a real, disclosed cap, raised 2026-09-24 alongside the window widening to 730 days - verified live to comfortably cover the real matching population) — if a real window ever exceeds 1000 matching applications, this insight would undercount rather than silently paginate.
An approval record only, never a claim about the drug's eventual clinical or commercial success.
Next signal
Watch this approval count across a second real pull for the first genuine period-over-period shift in NME approval activity.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-fda-nme-count
57% of the 104 new molecular entity approvals between 2024-10-01 and 2026-10-01 (59 of 104) received PRIORITY review; the remainder were STANDARD.
N/Apolicycurrent
A PRIORITY-review designation is a real, FDA-stated signal that a drug may offer a significant improvement over available therapies - directly relevant to how soon a payer should expect formulary/utilization-management decisions to matter for a given approval.
United States · period covered: Oct 2024–Oct 2026Affects: Drug makers · Health plans
Review priority split, new molecular entity approvals, 2024-10-01 to 2026-10-01
Official record, pulled directly from the issuing government source.
Evidence
openFDA drugsfda, real `review_priority` field across 104 Type 1 approvals, 2024-10-01 to 2026-10-01 — openFDA novel drug approvals vintage 2026-10-01
Limitations
PRIORITY review reflects FDA's own review-timeline designation, not a clinical-efficacy or safety judgment about the drug.
Small real sample (104 approvals this window) — a single approval or two can swing this percentage meaningfully.
Next signal
Watch this priority/standard split across a second real pull for a genuine period-over-period shift.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-fda-priority-split
New molecular entity approvals were spread across 24 month(s) between 2024-10-01 and 2026-10-01, ranging from 1 to 8 approvals per month.
N/Apolicycurrent
A real month-over-month cadence read on new molecular entity approval activity — the same shape as this dashboard's CMS-rules-finalized-per-month chart, applied to the FDA approval pipeline instead.
United States · period covered: Oct 2024–Oct 2026Affects: Drug makers · Health plans
New molecular entity approvals per month (trailing window)
Official record, pulled directly from the issuing government source.
Evidence
openFDA drugsfda, 104 real Type 1 approvals grouped by real submission_status_date month, 2024-10-01 to 2026-10-01 — openFDA novel drug approvals vintage 2026-10-01
Limitations
Small real monthly counts (104 approvals total across 24 months) — month-to-month variation here is not yet statistically meaningful.
Next signal
Watch this monthly cadence across a second real pull for the first genuine multi-period pattern.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-fda-by-month
NIH issued 100 award notices (sampled by dollar amount from 141,371 total) between 2024-10-01 and 2026-10-01; the largest single award is $56,578,999 to NATIONAL OPINION RESEARCH CENTER for "Partnering for Precision Medicine: A Collaborative Services Center for the All of Us Program", project 3OT2OD036464-01S5, notice dated 2026-01-15.
N/Astructural-changecurrent
Real, dollar-ranked NIH award activity is a direct catalyst signal for the biopharma/research ecosystem a payer executive tracks - large awards can precede new competitive research programs, novel therapy pipelines, or shifts in where NIH-funded research capacity is concentrating.
United States · period covered: Oct 2024–Oct 2026Affects: Research institutions · Life sciences
Top NIH award notices by dollar amount (trailing window)
Does not yet meet the 2-consecutive-period persistence rule. Each monthly pull recounts the trailing two-year window; a change counts as a trend only once it holds across consecutive pulls.
Evidence
NIH RePORTER projects/search, 100 real sampled award notices (of 141,371 total), 2024-10-01 to 2026-10-01 — NIH RePORTER award notices vintage 2026-10-01 (this record)
Limitations
Bounded to the top 100 awards by dollar amount out of 141,371 real total award notices in this window - this list names the largest awards; the NIH total and by-institute findings cover every award.
An award notice is a real, real-dollar commitment, not a claim about the research's eventual clinical or commercial outcome.
Single snapshot — a baseline reading, not yet a trend across periods.
Next signal
Watch this award count and its largest-award figure across a second real pull for the first genuine period-over-period shift in NIH funding activity.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-nih-award-count
NIH issued 141,370 award notices totaling $79.93B between 2024-10-01 and 2026-10-01; 2026-09 alone had 8,948 notices worth $5.70B.
N/Astructural-changecurrent
The overall scale and monthly rhythm of NIH research funding - the upstream pipeline for the therapies and devices a payer will eventually cover. Sudden gaps show up here first.
United States · period covered: Oct 2024–Oct 2026Affects: Research institutions · Life sciences
Does not yet meet the 2-consecutive-period persistence rule. Full-population counts, but NIH funding follows the federal fiscal year, so month-to-month swings are mostly seasonal.
Evidence
NIH RePORTER projects/search, every award notice 2024-10-01 to 2026-10-01 (141,370 of 141,370 reported), summarized at pull time — NIH RePORTER award notices vintage 2026-10-01
Limitations
3,123 notices carry no award amount and count as $0.
Award notices are obligations, not money spent; a multi-year project can get several notices.
The first and last months of the window are partial.
NIH awards cluster before the September 30 fiscal year-end and slow sharply when federal funding lapses, so compare months year over year, not in sequence.
Next signal
Watch the year-over-year comparison of the same months for a real change in NIH funding pace.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-nih-total-dollars
NCI administers $11.38B (14%) of the $79.93B NIH awarded between 2024-10-01 and 2026-10-01, the most of any of 39 institutes and centers.
N/Astructural-changecurrent
Which disease areas NIH is funding most heavily - cancer, infectious disease, heart and lung, aging - a leading indicator of where new therapies and the trials behind them will come from.
United States · period covered: Oct 2024–Oct 2026Affects: Research institutions · Life sciences
NIH award dollars by administering institute, 2024-10-01 to 2026-10-01
Does not yet meet the 2-consecutive-period persistence rule. Full-population totals across the window; a single two-year split is a baseline, not a trend.
Evidence
NIH RePORTER projects/search, agency_ic_admin field, every award notice 2024-10-01 to 2026-10-01 (141,370 notices) — NIH RePORTER award notices vintage 2026-10-01
Limitations
Grouped by the administering institute. Co-funded awards count fully toward the administering one.
3,123 notices carry no award amount and count as $0.
Institute budgets are set by Congress, so shares move slowly; a large shift usually reflects appropriations, not research demand.
Next signal
Watch each institute's share across future pulls for a shift in where NIH funding concentrates.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-nih-by-agency
Among the 100 sampled NIH awards, "Research Specimen" is the most common non-generic research term, appearing in 20 of 100 awards (20%).
N/Abaselinecurrent
Where NIH-funded research attention is concentrating within this sample - distinct from the dollar, organization and institute rankings, this shows *what* the funded research is actually about, which is the signal a payer executive tracking pipeline/research trends would want.
United States · period covered: Oct 2024–Oct 2026Affects: Research institutions · Life sciences
Most common non-generic research terms, sampled NIH awards (2024-10-01 to 2026-10-01)
awards
Research Specimen20
Specimen20
Assay20
Bioassay20
Biological Assay20
Age20
MR Imaging20
MR Tomography20
MRI20
MRIs20
Magnetic Resonance Imaging20
Medical Imaging, Magnetic Resonance / Nuclear Magnetic Resonance20
05101520
Inspect evidence
Confidence
low
Why low confidence:
Does not yet meet the 2-consecutive-period persistence rule. Each monthly pull recounts the trailing two-year window; a theme counts as a trend only once it holds across consecutive pulls.
Evidence
NIH RePORTER projects/search, real `terms` field across 100 sampled awards, 2024-10-01 to 2026-10-01 — NIH RePORTER award notices vintage 2026-10-01
Limitations
NIH RePORTER's real `terms` field is a keyword-tagging system that generates many near-synonymous phrasings for the same underlying real concept (e.g. "AD risk", "Alzheimer risk", and "Alzheimer's disease risk" all appear as separate entries below when present) - this insight does NOT merge synonyms, so a single real research theme may appear as several adjacent, similar-looking entries rather than one combined count. Read the list as raw real keyword tags, not a finalized topic taxonomy.
Covers only the 100 sampled awards (the largest by dollar amount), not every award for this window - a theme common among smaller awards outside this sample would not appear here.
The 20%-of-sample generic-term cutoff is a disclosed, documented statistical rule (excludes administrative boilerplate like "Research" or "Data"), not a hand-picked stoplist of specific words.
A term's presence describes what the award's own real record was tagged with, not this project's independent judgment about the research's actual scientific content.
Next signal
Watch whether this same theme leads again across a second real pull, which would be the first genuine persistence signal for a funding-theme concentration.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-nih-research-themes
FRED HUTCHINSON CANCER CENTER leads sampled NIH award recipients with $225,892,177 in total real award dollars between 2024-10-01 and 2026-10-01.
N/Astructural-changecurrent
A real, sourced ranking of which universities/health systems are attracting the most NIH funding right now - the same kind of finding a real research-funding tracker publishes; legitimate to name real organizations here under this project's revised naming rule since it's computed directly from real data, not implied-proprietary.
United States · period covered: Oct 2024–Oct 2026Affects: Research institutions · Life sciences
Top recipient organizations by sampled NIH award dollars, 2024-10-01 to 2026-10-01
Does not yet meet the 2-consecutive-period persistence rule. Each monthly pull recounts the trailing two-year window; a change counts as a trend only once it holds across consecutive pulls.
Evidence
NIH RePORTER projects/search, real `organization.org_name` field, award dollars summed across 100 sampled award notices, 2024-10-01 to 2026-10-01 — NIH RePORTER award notices vintage 2026-10-01 (this record)
Limitations
Covers only the 100 sampled awards (top by dollar amount) out of 141,371 total real award notices in this window - an organization with many small awards outside this sample would be undercounted here.
A real, sourced ranking computed from public NIH RePORTER data, not this project's own proprietary read on any institution's finances.
Single snapshot — a baseline reading, not yet a trend across periods.
Next signal
Watch this organization ranking across a second real pull for the first genuine period-over-period shift in funding concentration.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-nih-top-orgs
Of the 6 tracked health insurers, UnitedHealth Group, Inc., CVS Health Corporation, Humana Inc., Centene Corporation, The Cigna Group, Elevance Health, Inc. filed at least one Item 5.02 (departure or election of directors/principal officers) 8-K in the trailing window (38 filings total).
N/Astructural-changecurrent
Item 5.02 filing activity is a real, disclosed leadership-change signal (departure OR appointment of an officer/director) at a tracked competitor — directly relevant to competitive/talent-market awareness, never itself a claim about why the change happened.
United States · period covered: Oct 2024–Oct 2026Affects: Health insurers · Investors
Item 5.02 8-K filings by tracked health insurer (trailing window)
Does not yet meet the 2-consecutive-period persistence rule. Each monthly pull recounts the trailing two-year window; a change counts as a trend only once it holds across consecutive pulls.
Bounded to a fixed 6-company watchlist (UnitedHealth Group, Inc., CVS Health Corporation, Humana Inc., Centene Corporation, The Cigna Group, Elevance Health, Inc.) — not the full health-insurance sector or every publicly traded healthcare company.
Item 5.02 covers BOTH departure AND appointment of officers/directors — this insight never characterizes a filing as a firing or a resignation, only as "departure or election of directors/principal officers," the real SEC item description.
A filing count is a real disclosure-activity signal, not a materiality or business-impact judgment about what was filed.
Next signal
Watch this per-company Item 5.02 filing count across a second real pull for a genuine period-over-period shift.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-sec-leadership-change-by-company
Across the 6-company watchlist, 300 total 8-K item codes were filed between 2024-10-01 and 2026-10-01: Item 5.02 accounted for 13%, Item 1.01 for 3%, Item 2.02 for 17%, and all other item codes for 67%.
N/Astructural-changecurrent
An industry-wide 'what kind of activity is happening' read across the whole tracked watchlist at once - distinct from the per-company findings, which show who is filing, not what kind of activity dominates market-wide.
United States · period covered: Oct 2024–Oct 2026Affects: Health insurers · Investors
8-K item-code mix across the tracked watchlist, 2024-10-01 to 2026-10-01
item occurrences
Item 2.02 (results of operations and financial condition)5117.0%
Item 5.02 (departure/election of directors or principal officers)3812.7%
Does not yet meet the 2-consecutive-period persistence rule. Each monthly pull recounts the trailing two-year window; a change counts as a trend only once it holds across consecutive pulls.
Bounded to a fixed 6-company watchlist — not the full health-insurance sector.
A single 8-K filing can carry more than one item code, so these are item-code occurrences, not filing counts — the two totals are not the same and must not be conflated.
Item 5.02 covers both departure and appointment of officers/directors; Item 1.01 covers far more than partnerships (credit facilities, leases, and other material contracts) — never characterized beyond the real item description.
Next signal
Watch this item-code mix across a second real pull for a genuine shift in the kind of disclosure activity dominating the watchlist.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-sec-filing-mix
Of the 6 tracked health insurers, CVS Health Corporation, Humana Inc., Centene Corporation, The Cigna Group filed at least one Item 1.01 (entry into a material definitive agreement) 8-K in the trailing window (9 filings total); UnitedHealth Group, Inc., Elevance Health, Inc. filed none.
N/Astructural-changecurrent
Item 1.01 filing activity is a real, disclosed material-agreement signal (which covers far more than partnerships — credit facilities, leases, and other material contracts) at a tracked competitor — worth monitoring even though the agreement's specific content isn't captured by this structured field alone.
United States · period covered: Oct 2024–Oct 2026Affects: Health insurers · Investors
Item 1.01 8-K filings by tracked health insurer (trailing window)
Does not yet meet the 2-consecutive-period persistence rule. Each monthly pull recounts the trailing two-year window; a change counts as a trend only once it holds across consecutive pulls.
Bounded to a fixed 6-company watchlist (UnitedHealth Group, Inc., CVS Health Corporation, Humana Inc., Centene Corporation, The Cigna Group, Elevance Health, Inc.) — not the full health-insurance sector or every publicly traded healthcare company.
Item 1.01 is "entry into a material definitive agreement" and covers far more than partnerships (credit facilities, leases, and other material contracts) — this insight never characterizes a filing as "a partnership" without confirming the agreement's actual content.
A filing count is a real disclosure-activity signal, not a materiality or business-impact judgment about what was filed.
Next signal
Watch this per-company Item 1.01 filing count across a second real pull for a genuine period-over-period shift.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-sec-material-agreement-by-company
Health-industry companies filed 472 8-Ks reporting a completed acquisition or disposition of assets (Item 2.01) between 2024-10-01 and 2026-10-01, from 354 companies; pharma & biotech filed the most (338).
N/Astructural-changecurrent
Completed deals across drug makers, device makers, distributors, insurers and providers: consolidation and divestitures that can change who a plan contracts with, what a drug or device costs, and which competitors are growing by acquisition.
United States · period covered: Oct 2024–Oct 2026Affects: Life sciences · Providers · Investors
Most recent Item 2.01 8-Ks (completed acquisition or disposition of assets), health-industry companies
Does not yet meet the 2-consecutive-period persistence rule. Each monthly pull recounts the trailing two-year window; a change counts as a trend only once it holds across consecutive pulls.
Item 2.01 is "completion of acquisition or disposition of assets": it reports disposals as well as acquisitions, so a filing is a completed deal of either kind, not necessarily a purchase.
Covers companies whose SEC industry code (SIC) is one of 24 health codes. Health businesses filed under other codes (conglomerates, research labs, holding companies) are not included, and the code is SEC's classification, not a judgment about the business.
Counts original 8-Ks only; amendments (8-K/A) restate a report already counted. A filing with several filers counts once, under the first filer with a health code.
A filing count is disclosure activity. It is not evidence of a company's enrollment, revenue or performance, and is never read as one.
The filing says a deal closed; its size and counterparty are in the filing text, which is not read here.
Next signal
Watch the monthly count and which sectors lead it on the next pulls; read the linked filings for deal size and counterparty.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-sec-health-asset-deals
Health-industry companies filed 4,373 8-Ks reporting entry into a material definitive agreement (Item 1.01) between 2024-10-01 and 2026-10-01, from 1,053 companies; IMAC Holdings, Inc. filed the most (43).
N/Astructural-changecurrent
Material contracts signed across the health industry: licensing and supply deals, credit facilities and merger agreements that often precede a product launch, a financing or an acquisition a plan or provider will feel.
United States · period covered: Oct 2024–Oct 2026Affects: Life sciences · Providers · Investors
Health-industry companies with the most Item 1.01 8-Ks, 2024-10-01 to 2026-10-01
filings
IMAC Holdings, Inc.43
Jaguar Health, Inc.32
Scilex Holding Co29
PMGC Holdings Inc.28
Venus Concept Inc.28
MANGOCEUTICALS, INC.25
Scorpius Holdings, Inc.25
Wellgistics Health, Inc.24
Vireo Growth Inc.24
NKGen Biotech, Inc.23
0204060
Item 1.01 8-Ks by month, health-industry companies
Does not yet meet the 2-consecutive-period persistence rule. Each monthly pull recounts the trailing two-year window; a change counts as a trend only once it holds across consecutive pulls.
Item 1.01 is "entry into a material definitive agreement" and covers far more than partnerships (credit facilities, leases, licensing and supply contracts, merger agreements); a filing is never described as a partnership without reading it.
Covers companies whose SEC industry code (SIC) is one of 24 health codes. Health businesses filed under other codes (conglomerates, research labs, holding companies) are not included, and the code is SEC's classification, not a judgment about the business.
Counts original 8-Ks only; amendments (8-K/A) restate a report already counted. A filing with several filers counts once, under the first filer with a health code.
A filing count is disclosure activity. It is not evidence of a company's enrollment, revenue or performance, and is never read as one.
Companies that fund themselves through frequent share sales, loans or board turnover file many of these, and larger companies file more 8-Ks of every kind, so a high count reflects financing and deal volume, not the company's standing.
Next signal
Watch which companies and sectors lead Item 1.01 filings on the next pulls; read the filings behind any cluster.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-sec-health-material-agreements
Health-industry companies filed 4,374 8-Ks reporting departure or election of directors or principal officers (Item 5.02) between 2024-10-01 and 2026-10-01, from 1,087 companies; Lifeward Ltd. filed the most (18).
N/Astructural-changecurrent
Board and executive changes across the health industry. A cluster at one competitor or supplier is worth reading, because leadership changes often come before shifts in strategy.
United States · period covered: Oct 2024–Oct 2026Affects: Life sciences · Providers · Investors
Health-industry companies with the most Item 5.02 8-Ks, 2024-10-01 to 2026-10-01
filings
Lifeward Ltd.18
Canton Strategic Holdings, Inc.15
Vor Biopharma Inc.15
Matinas BioPharma Holdings, Inc.15
Shuttle Pharmaceuticals Holdings, Inc.14
IMMUNIC, INC.14
ACCURAY INC14
ImageneBio, Inc.14
180 Life Sciences Corp.13
Longeveron Inc.13
05101520
Item 5.02 8-Ks by month, health-industry companies
Does not yet meet the 2-consecutive-period persistence rule. Each monthly pull recounts the trailing two-year window; a change counts as a trend only once it holds across consecutive pulls.
Item 5.02 covers appointments as well as departures of directors and principal officers, and routine board changes; a filing is never described as a firing or a resignation without reading it.
Covers companies whose SEC industry code (SIC) is one of 24 health codes. Health businesses filed under other codes (conglomerates, research labs, holding companies) are not included, and the code is SEC's classification, not a judgment about the business.
Counts original 8-Ks only; amendments (8-K/A) restate a report already counted. A filing with several filers counts once, under the first filer with a health code.
A filing count is disclosure activity. It is not evidence of a company's enrollment, revenue or performance, and is never read as one.
Companies that fund themselves through frequent share sales, loans or board turnover file many of these, and larger companies file more 8-Ks of every kind, so a high count reflects financing and deal volume, not the company's standing.
Next signal
Watch which companies and sectors lead Item 5.02 filings on the next pulls; read the filings behind any cluster.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-sec-health-leadership-changes
Health-care companies reported 3,221 new private offerings on Form D from Q3 2024 to Q2 2026, with $39.99B sold; biotechnology issuers accounted for $19.60B (49%); Q2 2026 had 465 offerings and $3.37B sold, against 398 and $5.97B in Q2 2025.
N/Astructural-changecurrent
Private capital flowing into health-care companies before they are public or between public offerings: the funding behind the next wave of drugs, devices and care-delivery competitors a plan or provider will meet.
United States · period covered: Jul 2024–Jun 2026Affects: Life sciences · Investors
Health-care Form D amount sold by industry group, Q3 2024 to Q2 2026
Does not yet meet the 2-consecutive-period persistence rule. Eight quarters of full-population notices; a quarter-to-quarter swing counts as a trend only once it holds across later quarters.
Form D covers exempt offerings of every kind: venture and growth rounds, but also private placements by listed companies, debt and other securities. It is a notice, not a record of a closed round.
Health care is the five health industry groups on the form (biotechnology, pharmaceuticals, health insurance, hospitals and physicians, other health care), as each issuer chose them. Health-focused investment funds file as pooled investment funds and are not included.
Amounts sold are as the issuer reported them. An offering counts once, in the quarter of its first notice, at the amount sold in its latest amendment; 315 amendments to offerings first reported before the window are left out.
Offerings with nothing sold yet count as $0, and issuers can report late, so the latest quarter may rise when amendments arrive.
Next signal
Watch the next quarterly file, and compare each quarter with the same quarter a year earlier rather than the one before.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-formd-health-raises
The largest health-care private offering reported on Form D from Q3 2024 to Q2 2026 was by Fovea Investment Holdings, LLC (Other Health Care, Pennsylvania): $800.0M sold, first notice filed 2025-01-17.
N/Astructural-changecurrent
The biggest private bets on health-care companies, named from their own SEC notices: well-funded entrants and expanding competitors worth knowing before they show up in a network, formulary or bid.
United States · period covered: Jul 2024–Jun 2026Affects: Life sciences · Investors
Largest health-care private offerings by amount sold (Form D), Q3 2024 to Q2 2026
Does not yet meet the 2-consecutive-period persistence rule. Named from the issuers' own notices; amounts are as reported and not independently checked.
Form D covers exempt offerings of every kind: venture and growth rounds, but also private placements by listed companies, debt and other securities. It is a notice, not a record of a closed round.
Health care is the five health industry groups on the form (biotechnology, pharmaceuticals, health insurance, hospitals and physicians, other health care), as each issuer chose them. Health-focused investment funds file as pooled investment funds and are not included.
Amounts sold are as the issuer reported them. An offering counts once, in the quarter of its first notice, at the amount sold in its latest amendment; 315 amendments to offerings first reported before the window are left out.
Offerings with nothing sold yet count as $0, and issuers can report late, so the latest quarter may rise when amendments arrive.
Naming an issuer reports its own filing; it is not an assessment of the company.
Next signal
Watch for amendments that raise these amounts and for new large notices in the next quarterly file.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-formd-health-largest-raises
Issuers based in California reported the most health-care private offering dollars on Form D from Q3 2024 to Q2 2026: $10.54B across 505 offerings, 26% of $39.99B.
N/Astructural-changecurrent
Where private health-care capital is concentrating: the clusters most likely to produce new therapies, devices and care-delivery companies, and where new competitors and partners are forming.
United States · period covered: Jul 2024–Jun 2026Affects: Life sciences · Investors
Health-care Form D amount sold by issuer state, Q3 2024 to Q2 2026
Does not yet meet the 2-consecutive-period persistence rule. Full-population totals over eight quarters; a single large offering can move a smaller state's rank.
Form D covers exempt offerings of every kind: venture and growth rounds, but also private placements by listed companies, debt and other securities. It is a notice, not a record of a closed round.
Health care is the five health industry groups on the form (biotechnology, pharmaceuticals, health insurance, hospitals and physicians, other health care), as each issuer chose them. Health-focused investment funds file as pooled investment funds and are not included.
Amounts sold are as the issuer reported them. An offering counts once, in the quarter of its first notice, at the amount sold in its latest amendment; 315 amendments to offerings first reported before the window are left out.
Offerings with nothing sold yet count as $0, and issuers can report late, so the latest quarter may rise when amendments arrive.
Located by the primary issuer's principal place of business as filed, not where the money is spent.
Next signal
Watch whether the leading states hold their share in the next quarterly files.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-formd-health-by-state
1277 industry-sponsored Phase 3 trials had results newly posted to ClinicalTrials.gov between 2024-10-01 and 2026-10-01; most recently Abcuro, Inc.'s "A Study to Evaluate the Efficacy and Safety of ABC008 for Inclusion Body Myositis" (272 enrolled), posted 2026-10-01.
N/Abaselinecurrent
A real posted-results event is a concrete, dated catalyst for a payer's clinical/formulary-planning watchlist - it marks that a real Phase 3 program has reached a reportable milestone, independent of what the results actually show.
United States · period covered: Oct 2024–Oct 2026Affects: Drug makers · Health plans
Does not yet meet the 2-consecutive-period persistence rule. Each monthly pull recounts the trailing two-year window; a change counts as a trend only once it holds across consecutive pulls.
A results posting encodes NO success/failure judgment — this insight never says "positive result," only that results were posted, by whom, with what real enrollment number.
Bounded to industry-sponsored trials only (leadSponsorClass = INDUSTRY) — academic and other-government-sponsored Phase 3 results in the same window are excluded by design, see this agent's data adapter.
Bounded to a real, disclosed 10-page pagination cap (up to 10,000 raw studies) per pull, raised from a single 100-record page on 2026-09-24 when the real window widened to 2 years — see this agent's data adapter.
Next signal
Watch this results-posting count across a second real pull for a genuine period-over-period shift in industry Phase 3 reporting activity.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-ct-results-count
Enrollment across the 1277 industry-sponsored Phase 3 trials with results posted between 2024-10-01 and 2026-10-01 has a median of 304 participants (IQR 119–609).
N/Abaselinecurrent
Trial scale (enrollment) is a proxy for a program's evidentiary weight and eventual real-world population relevance - a useful read alongside the results-posting count for prioritizing which newly reported trials merit closer clinical/formulary review.
United States · period covered: Oct 2024–Oct 2026Affects: Drug makers · Health plans
Enrollment distribution, industry-sponsored Phase 3 trials with newly posted results, 2024-10-01 to 2026-10-01
participants
Middle half (IQR)
Median
Outlier (beyond 1.5× IQR; a triangle is off this chart's visible scale)
Does not yet meet the 2-consecutive-period persistence rule. Each monthly pull recounts the trailing two-year window; a change counts as a trend only once it holds across consecutive pulls.
Evidence
ClinicalTrials.gov v2 studies API, real `enrollmentInfo.count` field across 1277 industry-sponsored Phase 3 trials, 2024-10-01 to 2026-10-01 — ClinicalTrials.gov Phase 3 results postings vintage 2026-10-01
Limitations
Enrollment size is a real proxy for trial scale, not a measure of clinical significance or result direction.
Bounded to industry-sponsored trials only, and to a real, disclosed 10-page pagination cap per pull — see this agent's data adapter.
Only built when at least 20 real trials are available in the sample (this cycle: 1277) — a real size-gating rule, same pattern as this dashboard's other boxplot insights.
Next signal
Watch this enrollment distribution across a second real pull for a genuine shift in typical Phase 3 trial scale.
Generated by market-catalyst-intelligence · sig-market-catalyst-2026-10-01-ct-enrollment-distribution
Calendar
What to watch next
Coming events that will update the findings above. Exact dates come from the data itself (rule effective dates, comment deadlines) or a fixed calendar; a month marked "~" is an estimate from the source's own release history.
Oct 1, 20261 day ago10 finalized CMS rules take effect — FY 2027 Inpatient Psychiatric Facilities Prospective Paymen…; Inpatient Rehabilitation Facility Prospective Payment Syste…; FY 2027 Hospice Wage Index and Payment Rate Update and Hosp…; Prospective Payment System and Consolidated Billing for Ski…; Hospital Inpatient Prospective Payment Systems for Acute Ca…; Hospital Inpatient Prospective Payment Systems for Acute Ca…; FY 2027 Hospice Wage Index and Payment Rate Update and Hosp…; Inpatient Rehabilitation Facility Prospective Payment Syste…; FY 2027 Inpatient Psychiatric Facilities Prospective Paymen…; Prospective Payment System and Consolidated Billing for Ski… See the finding.scheduled
~Oct 2026July 2026 Medicaid and CHIP enrollment — States' next monthly reports, plus final figures replacing the preliminary month shown now. See finding 1, finding 2.expected
~Oct 2026July 2026 Medicare enrollment by state and county — The next month of Medicare Advantage share of all Medicare beneficiaries, by state and county. See finding 1, finding 2, finding 3.expected
~Oct 2026Plan year 2027 Marketplace rate and plan files — Filed premiums, deductibles and issuers for next year, published before open enrollment. See finding 1, finding 2, finding 3.expected
~Oct 2026Q2 2026 nursing home payroll staffing — The next quarter of nurse hours per resident day and contract-staff share, by state. See finding 1, finding 2, finding 3.expected
~Oct 2026Q2 2026 preliminary Part D spending by drug — The next quarter of CMS's preliminary drug spending: biosimilar share and the early read on this year's fastest-growing drugs. See finding 1, finding 2, finding 3.expected
Oct 13, 2026in 11 daysA finalized CMS rule takes effect — Prohibition on Federal Medicaid and Children's Health Insurance Program Funding for Sex-R… See the finding.scheduled
Oct 28, 2026in 26 daysA finalized CMS rule takes effect — Updates to the Master List of Items Potentially Subject to Face-to-Face Encounter and Wri… See the finding.scheduled
Nov 1, 2026in 30 daysMarketplace open enrollment for 2027 begins — CMS's state-level results for it typically follow in spring 2027. See finding 1, finding 2.scheduled
Nov 1, 2026in 30 daysMonthly data refresh — Every source is pulled again; if any data changed, the agents and the executive analyst rerun and this page updates.scheduled
~Nov 2026Final Hospital Outpatient (OPPS) and ASC payment rule — Proposed Jul 2026. Last year's final came out in Nov 2025, 4 months after its proposal. It sets next year's payment rates. See the finding.expected
~Nov 2026Final Physician Fee Schedule rule — Proposed Jul 2026. Last year's final came out in Nov 2025, 4 months after its proposal. It sets next year's payment rates. See finding 1, finding 2.expected
Data Explorer
The underlying data, unfiltered
Everything above is what the agents judged worth an executive's attention. This section is the broader picture behind it — real numbers from the same public datasets, not just the ones that became a finding.
Hospitals analyzed
5,419
States/territories covered
56
Top-4 ownership concentration (CR4)
79%
Home health agencies analyzed
12,460
Medicare Part B providers (2024)
1,296,739
CMS Federal Register documents (2-year window)
487
Rules finalized (2-year window)
59
Rules currently proposed
48
MA/Part D enrollment (2026-09)
61,715,324
Marketplace plan selections, all states and DC (2026 open enrollment)
23,130,860
HealthCare.gov states with Marketplace plans (2026)
30
Hospital type
facilities
Acute Care Hospitals3,10757.3%
Critical Access Hospitals1,38225.5%
Psychiatric62411.5%
Acute Care - Veterans Administration1322.4%
Childrens941.7%
Other801.5%
Hospital ownership type
facilities
Voluntary non-profit - Private2,32242.8%
Proprietary1,06319.6%
Government - Hospital District or Authority5129.4%
Government - Local3937.3%
Voluntary non-profit - Other3536.5%
Other77614.3%
Hospital overall star rating
facilities
1★198
2★661
3★985
4★946
5★384
02505007501,000
Home health quality star rating
facilities
1★161
1.5★614
2★778
2.5★977
3★1,378
3.5★1,298
4★1,320
4.5★1,047
5★477
05001,0001,500
Home health agencies offering each service, nationally
agencies
Nursing care10,276
Physical therapy9,846
Occupational therapy9,429
Home health aide9,122
Speech pathology8,716
Medical social services8,062
05,00010,00015,000
Medicare payment as % of submitted charge, by provider type
% of charge
Nurse Practitioner33
Ophthalmology30
Internal Medicine31
Clinical Laboratory28
Hematology-Oncology27
Ambulatory Surgical Center16
Family Practice34
Ambulance Service Provider21
010203040
Home health spending-ratio distribution by state
ratio
Middle half (IQR)
Median
Outlier (beyond 1.5× IQR; a triangle is off this chart's visible scale)
Benchmark silver premium across each state's rating areas, age 40, plan year 2026 (states with 5+ rating areas) — full quartile datausd/month
Category
Median
Q1
Q3
Whisker low
Whisker high
Outliers
n
WV
1093.95
1050.77
1093.95
1036.38
1157.29
944.25, 944.25, 1174.56
11
AR
773.96
773.96
773.96
773.96
773.96
—
7
UT
737.07
593.64
876.19
573.56
905.19
—
6
TN
707.50
689.99
759.57
654.32
793.80
—
8
KS
703.15
693.33
732.81
688.66
753.57
579.41
7
FL
699.04
664.39
769.66
571.85
898.81
949.75, 991.19, 1263.70, 1277.62
67
LA
684.53
612.12
729.50
561.86
734.52
—
8
AZ
668.52
492.66
772.93
448.94
921.26
—
7
TX
667.85
639.13
725.25
563.51
779.39
898.13, 924.74
27
MO
665.96
638.95
706.68
580.77
790.49
531.51
10
NC
644.93
594.52
689.97
545.64
758.07
836.64, 844.92
16
MS
641.50
635.35
654.26
634.33
657.48
602.32, 684.48
6
AL
639.41
628.53
655.70
597.95
686.28
—
13
WI
627.10
566.71
659.13
542.15
746.29
—
16
OK
569.27
567.94
663.92
564.27
683.63
—
5
SC
556.31
530.13
585.52
460.78
664.65
—
46
OR
536.00
527.00
559.50
510.49
605.00
—
7
OH
525.17
503.81
538.05
483.07
552.17
592.25
17
MI
481.63
448.77
529.73
388.03
632.60
705.99, 910.03
16
IA
477.40
459.95
493.53
452.73
518.28
—
7
IN
475.58
463.58
499.19
446.14
515.75
—
17
Marketplace issuers per HealthCare.gov state, plan year 2026 (most first)
issuers
Benchmark silver premium change since 2014, age 40: the 5 fastest-rising HealthCare.gov states
% since 2014
Scroll the chart sideways for earlier years →
WV
TN
UT
KS
NE
2026WV: +317%TN: +259%UT: +229%KS: +224%NE: +205%
Medicare Advantage enrollment
37 real pulls
2026-09-0136M enrollees
Total Medicare Part B professional payment
12 real pulls
2024-12-31120.1B USD
The two line charts above are real multi-period histories: CMS's monthly Medicare Advantage enrollment reports and each published year of Medicare Part B professional payment. Not shown: a geographic map (this project doesn't have verified U.S. state boundary data to plot against, and won't guess at one) and a patient-flow diagram (no public CMS source here provides patient-level clinical flow data).
Case study
Why this project exists, and how it's built
01 · Stakeholder question
What does a health plan's executive team need to know this month, and can every claim be traced to its source?
Markets, cost and utilization, reimbursement, provider networks, Medicare Advantage and Part D, Medicaid, the ACA Marketplace, policy, and market events. The questions were catalogued up front: 135 executive questions.
02 · Problem
Most AI projects fail on the operations side.
Healthcare leadership teams have more data available than they have attention. The bottleneck isn't access to data — it's knowing which changes actually matter, and being able to trust the answer enough to act on it.
03 · Architecture
Code computes, models judge, a person reviews.
A coordinated system of specialized agents turns public data into evidence-backed executive intelligence — deterministic code handles the math (rates, trends, thresholds), and language-model reasoning is reserved for synthesis and interpretation, never for calculating a number that code can calculate more reliably.
source32 public sourcesCMS · Medicaid.gov · Federal Register · SEC · FDA · NIH · ClinicalTrials.gov
codeMonthly pipelineGitHub Actions on the 1st · a Data Quality Engineer agent flags new, late or changed data
code10 domain specialistsCode computes every number, trend and diff; a Data Architect agent keeps definitions consistent
modelPicks per agentopenai:gpt-6-luna chooses what's noteworthy among computed candidates
modelExecutive briefingChief Intelligence Officer agent · anthropic:claude-opus-5-5 ranks and explains
codeGrounding checkDrops any number or name not traced to computed facts
personDashboardRanked findings, evidence drawer on every one · Adam reviews
Dashboard → Executive Orchestrator → 12 agents: 10 domain specialists (Market Growth & Geography; Claims, Utilization & Cost; Reimbursement & Payment; Provider & Network; Medicare Advantage & Part D; Medicaid, CHIP & Duals; ACA Marketplace; Policy & Regulation; Emerging Trends; and Market Catalysts) and 2 infrastructure agents (Data Source Monitoring and Semantic Model) → data adapters → public sources. A model-provider interface sits behind every agent's reasoning step, so the same system can run on different language models without changing any business logic. 12 agents ran this cycle, producing 83 evidence-backed insights from real public data — every one traceable through each finding's "Inspect evidence" drawer above.
Currently live: 32 public sources, counted from what the agents actually cited this cycle. From CMS: Provider of Services files (certified facilities and beds); Home Health Care Agencies; Medicare Physician & Other Practitioners by Provider; Medicare Physician & Other Practitioners by Service; Medicare Part D Spending by Drug; Medicare Quarterly Part D Spending by Drug (preliminary); Medicare Part D Prescribers by Geography and Drug; Medicare Part D Prescribers by Provider; Medicare Part D Prescribers by Provider and Drug; Physician Fee Schedule national relative value files; Hospital Readmissions, HAC Reduction and Value-Based Purchasing programs; Hospital General Information; Hospital and SNF Change of Ownership; Hospital and SNF All Owners; Payroll-Based Journal Daily Nurse Staffing; Medicare Advantage/Part D Monthly Enrollment by Plan; Medicare Monthly Enrollment (national, state and county); State Medicaid and CHIP monthly enrollment reports; Medicaid Managed Care Enrollment by Program and Plan; ACA Marketplace Rate and Plan Attributes PUFs; Marketplace Open Enrollment state-level PUFs; Shared Savings Program Accountable Care Organizations; Shared Savings Program ACO Participants; Shared Savings Program Performance Year Financial and Quality Results; Number of ACO Assigned Beneficiaries by County; and Federal Register documents published by CMS. Beyond CMS, behind the Market/Catalyst agent: NIH RePORTER award notices, openFDA novel drug approvals, SEC EDGAR 8-K filings for a 6-company health-insurer watchlist, SEC EDGAR 8-K filings by health-industry companies, SEC Form D data sets (health-care private offerings), and ClinicalTrials.gov Phase 3 results postings. The physician data counts every Part B provider nationally (1,296,739 in 2024), not a sample. The Federal Register feed and the Market/Catalyst sources each use a rolling 730-day window, not full history. Files CMS publishes one year or month at a time are backfilled, for example: physician payments 2013–2024, 37 monthly Medicare Advantage reports since 2023-08, and Marketplace plan years 2014–2026. Hospital and home health data come from CMS's Provider Data Catalog API, which exposes only the current dataset vintage (no historical-vintage parameter exists), so those two can only accumulate real history forward from each pull. The Marketplace file exposes only category fields and an opaque issuer/plan identifier used solely as a count, never a carrier name; the MA/Part D file does name a real parent organization on every row, and this project surfaces that name only for a genuine, sourced finding (an enrollment ranking), never a fabricated or implied-proprietary claim (see "Known limitations" below).
A single data point is never called a trend. A finding only earns that label once it holds across multiple periods, or a specific, cited event explains it — otherwise it's labeled a baseline, plainly. Confidence levels are computed from whether a finding has enough history, persists, and is corroborated by an independent source, and each finding's evidence drawer shows the level with its reasoning. Laws and policies (CMS rules, FDA approvals) come straight from the government's own record, so they read high confidence and list each rule or approval with a link to its source. Every number is still computed by deterministic code, never an LLM — but which of several real, computed candidates is worth an executive's attention is a judgment call, so agents route that specific decision through a reasoning layer that can only choose among and briefly explain real candidates it's given, never invent one. The current briefing was reasoned on 2026-10-01.
Every data adapter in this system is built behind the same interface a real enterprise data source would implement — membership, claims, provider, reimbursement, pharmacy, and network. Swapping a public CMS adapter for an internal one wouldn't require changing the agents, the evidence model, or the dashboard — only the adapter itself.
04 · Guardrails
What stops a wrong number from reaching a decision-maker.
Code computes every number
Deterministic code handles every number, trend and diff. Models never do arithmetic.
Models only pick and explain
The per-agent model chooses among real computed candidates; the analyst explains why they matter.
Grounding check before publishing
Auto-published text passes a code-level check that drops any model-written number or name (insurer, prescriber, ACO, facility, owner) not traced to computed facts.
Stale data never leads
The Data Quality Engineer flags new, late or changed sources, and findings rank by recency: current, aging, stale.
Models chosen by measurement
Six models from two providers ran the same 12 benchmark tasks; routing follows measured score and cost.
Tests are the definition of done
486 unit and integration tests, a Playwright e2e suite on desktop and mobile with an axe accessibility scan, and CI on every push.
top mean benchmark score of 6 models (Claude Opus 5.5, 2026-09-25 evaluation)
486
automated tests passing (2026-09-28)
Known limitations
Stated plainly rather than left for a visitor to discover — a demo that hides its own gaps is less trustworthy than one that names them:
Medicaid coverage uses the states' public monthly enrollment reports and CMS's annual managed care enrollment by plan. The claim-level T-MSIS research files need a CMS data use agreement and aren't used, so Medicaid churn, provider networks and service use aren't covered.
The model evaluation's own scorer was wrong on its first live runs. It counted "not the full national file" as an overclaim and missed several correctly worded refusals, so every model looked like it had fabricated. That was caught by reading the raw answers, then fixed and re-scored from the saved outputs. Automated evaluation needs the same auditing as the models it grades.
The Federal Register feed and the 6 corporate/regulatory/research sources behind the Market/Catalyst agent are each a rolling 730-day window, not full history. Hospital and home health data have only about 2 weeks of history so far and can only build it forward from each pull, since CMS's Provider Data Catalog API exposes no historical-vintage parameter to backfill from.
61 of 83 findings read low confidence — genuinely, not a bug. Outside laws and policies, confidence needs a change that holds across consecutive periods. Most findings rest on years of history (physician payments 2013–2024, 37 monthly Medicare Advantage reports since 2023-08, and Marketplace plan years 2014–2026) but show one period of change so far, and hospital and home health data have only about 2 weeks of snapshots; the 16 that reach medium or higher have a change that persists. The 6 law and policy findings read high because they come straight from the official government record.
No proprietary insurer data is used anywhere — every number on this page is public data. A real carrier name (e.g. in the Medicare Advantage enrollment ranking above) only ever appears when it's a genuine, sourced finding computed from that public data, the same kind of reading a real industry directory publishes — never a fabricated claim or an implied look at any carrier's real internal systems.
Every insight carries a "contradictoryEvidence" field in its schema, but no agent has populated it yet — named directly here rather than left as a silently unused field.
06 · Skills demonstrated
What this project shows I can deliver.
AI solution architecture
A business question decomposed into a 13-agent team with defined responsibilities, inputs and escalation rules.
Multi-agent system design
Specialists report to an orchestrator; an executive-analyst step ranks findings and looks for cross-domain patterns.
LLM evaluation and model selection
Six models, two providers, one benchmark; routing chosen per job from measured score and cost.
Grounding and guardrail design
Numbers from code, models choose among real candidates, a grounding check gates publishing.
Data pipeline design
32 public sources on a monthly schedule, with a monitoring agent for new, late or changed data.
Executive dashboard design
Findings ranked by recency and importance, a jump menu, and an evidence drawer behind every finding.
Public healthcare data
CMS (Medicare, Medicare Advantage, Part D, ACOs, hospitals, nursing homes), Medicaid.gov, federal regulatory and market sources.
Test and CI discipline
486 tests, e2e with accessibility checks, and CI on every push.
Cost control
Each job routed to a model by measured score and cost, not prestige.