Fifty billion dollars, no scoreboard
Federal money is buying software for small rural hospitals. The program already demands four numbers for every initiative. One of them should be a patient, and it should be published.
Listen to this special topic
read by an AI voiceFifty billion dollars is going into rural health care, and the Centers for Medicare and Medicaid Services, which is handing it out, has put in writing that the results will not be published. The supporting statement says the information “is intended to be publicly available upon request.” It also says “the results of this collection will not be published.” A report you have to know exists before you can ask for it is not a public record of anything.
The Rural Health Transformation Program carries $50 billion, $10 billion a year through 2030, and every state has an award. In August the agency announced $93.3 million for Georgia, including money for telehealth, cybersecurity and surgical robotics, and for preparing 87 rural hospitals for value-based care. First-year awards averaged about $200 million a state, so the Ohio announcement that follows spends part of a sum already granted rather than adding to it.
On Oct. 16 the agency, known as CMS, announced $92 million for rural Ohio. The money splits three ways: $31.4 million for coordinated rural care networks, and $30.3 million each for maternal care and for equipment and facility upgrades at 22 rural hospitals. The agency's administrator, Dr. Mehmet Oz, said the focus now is on making sure “this early progress translates into lasting results for rural patients.” The release names no measure and no target.
That is the whole problem, and it is not for want of paperwork. The program's funding notice already tells each state to identify at least four quantifiable metrics for every initiative it proposes. It describes the technology category as information technology advances “designed to improve efficiency, enhance cybersecurity capability development, and improve patient health outcomes.” So the numbers exist. States propose them, the agency approves them, and nothing in the program's paperwork makes anyone publish how they turn out.
Some of it is going to artificial intelligence. Daniel O'Neil of Civic Operator, which tracks the program's grants, told Healthcare IT News in May that he had “counted approximately 23 states that specifically reference it in their plans,” a count of plans and not of dollars. Vermont offered its rural practices money for AI scribes. Its own list of activities sells the grant with a claim it does not source, that scribes “automate an average of over two hours of daily administrative work.” Its notice, KFF Health News reported, asks grantees to report how many clinicians and patients the software serves. Not how much time it saves. The sales claim is about hours. The reporting is about headcount.
The rules go further than the announcements do. One of the four metrics is meant to show how the effect is spread within the state. States are asked for baselines and targets “where possible” and milestones “if possible.” The post-award questions and answers repeat the four and require the agency's approval to add or drop one. States file three quarterly reports a year and an annual report. The first annual report was due on Aug. 31. The first quarterly report is due on Nov. 30.
And the scores matter. The Government Accountability Office reported on Oct. 6 that CMS requires those reports “that CMS will use to measure states' performance against states' milestones,” and plans to “convert states' progress implementing initiatives and policy actions into numerical scores” when it allots up to $40 billion more over four years. Half of that is split equally by law and a quarter follows rural factors fixed in 2025, so up to ten billion dollars will move on numbers that nobody is required to publish. CMS's state snapshots name each state's initiatives without their metrics. States must send their reports in but are not required to post them proactively. A CMS spokesman said the agency will publish an annual report on state progress. So far it has posted a tally of what states have obligated, which it says “is not a full view of state progress.”
State officials have a fair answer to all this. The money came with deadlines, and a small hospital closing a service cannot wait for a trial it has no staff to run. That is true. It is not an argument for measuring nothing. The reporting runs to 2031 and four metrics are required anyway, so the only live question is which four. Pick one a clinician would recognize. The share of eligible patients screened. The days from an abnormal result to the visit that followed. Either costs nothing extra at signing and neither can be reconstructed afterwards. That long-term outcomes take years is a reason to publish the near-term ones, not to skip them.
So here is the ask, and it is small. Make one of the four metrics for every technology initiative a clinical measure at the hospitals getting the tool, record a baseline before it arrives, and post it when the contract is signed. New Hampshire already posts its recipients' contracts, so that is the floor, not the ceiling. Few critical access hospitals have anyone whose job is to judge a surgical robot or a scribe, so pool the judging. One state agency or university center measuring two or three outcomes across twenty hospitals can answer in a year what none of them can answer alone. CMS holds the score that divides a quarter of the money still to come, and can withhold any of it from a state that strays from its plan. It should publish the reports behind that score, not only a summary of them, and set one short list of measures for technology so a result in Ohio means something in Georgia.
The 22 Ohio hospitals will have their new equipment for years. Whether rural patients were better off for it is a question someone will ask in 2031, when the last report comes due. On the paperwork as it stands, nothing guarantees anyone will be able to answer. Four numbers for each initiative, chosen now and published as they arrive, would be enough. That is the entire reform, and the four numbers are already in the rules.
Correction (Oct. 26): This article has been corrected. An earlier version said up to $40 billion could move on states' progress scores and that CMS's score releases the rest of the money. By law half of the remaining money is split equally among the states, and a quarter follows rural factors fixed in 2025. The scores set up to $10 billion, and CMS can withhold funds from a state that strays from its plan.
- CMS award announcement for OhioCMS press release, Oct. 16, 2026, headed "Trump Administration Announces $92 Million to Coordinate Rural Hospital Networks" and continuing on improving maternal care and expanding medical services across Ohio. The release describes $31.4 million to establish or expand coordinated rural care networks, including Clinically Integrated Networks and Rural Health Innovation Hubs; $30.3 million for hospital networks and health centers to improve maternal care; and $30.3 million for one-time equipment and facility upgrades at 22 rural hospitals, including a new behavioral health unit, on-site MRI services and expanded diagnostic capacity. CMS Administrator Dr. Mehmet Oz is quoted: "Our focus now is on making sure this early progress translates into lasting results for rural patients." The release states no performance measure and does not mention reporting or artificial intelligence. Ohio's first-year Rural Health Transformation Program award was $202,030,262.Sources: CMS newsroom, CMS, first-year awards
- CMS award announcement for GeorgiaCMS press release, Aug. 27, 2026, announcing $93.3 million to expand telehealth services, advance surgical robotics and transform rural healthcare access across Georgia. The release says the award completes Georgia's first-year allotment of $218 million and funds pre-implementation work for 87 rural hospitals preparing for the AHEAD Model, workforce training for nurses, EMTs and paramedics, telehealth access points, maternal and behavioral health, newborn screening, transportation, dementia care capacity, hospital cybersecurity and surgical robotics. It gives no dollar breakdown by project and no performance measures. Georgia's first-year award was $218,862,170. CMS's obligated funds report, with data states submitted as of July 31, 2026, lists Georgia's Budget Period 1 award as $218,862,169.63 and total funds put to work as the same amount, and says the data "is not a full view of state progress."Sources: CMS newsroom, CMS, obligated funds report
- The program and its first-year awardsThe Rural Health Transformation Program was authorized by Section 71401 of Public Law 119-21 and provides $50 billion, with $10 billion available each fiscal year from 2026 through 2030. Only the 50 states are eligible. Half of each year's funds is distributed equally among approved states and half on factors including the state's rural population, the proportion of rural health facilities in the state, the situation of certain hospitals in the state, and other factors CMS specified in the funding notice. States must use funds for at least three of ten approved uses. CMS announced first-year awards on Dec. 29, 2025, averaging about $200 million and ranging from $147,250,806 for New Jersey to $281,319,361 for Texas; Ohio received $202,030,262 and Georgia $218,862,170.Sources: Medicaid.gov program page, CMS newsroom
- What the funding notice requires states to measureCMS's Notice of Funding Opportunity instructs applicants: "Identify at least four quantifiable metrics for each initiative described in your proposed initiatives and use of funds section." "At least one of the four should allow you to demonstrate how the impact is distributed in different parts of the State." Baselines and targets are sought "where possible" and milestones "if possible," along with data sources and the state's ability to collect and analyze them; applicants acknowledge they "will be required to report on performance metric progress during annual reporting." Approved use category F covers "significant information technology advances designed to improve efficiency, enhance cybersecurity capability development, and improve patient health outcomes"; category D covers training and technical assistance "including remote monitoring, robotics, artificial intelligence, and other advanced technologies."
- Reporting schedule, metrics and publicationSupporting Statement A for CMS-10949, Rural Health Transformation Program Reporting, describes five annual reports plus a final report, with Annual Report #1 due Aug. 31, 2026, the first quarterly report due Nov. 30, 2026 (the table prints Nov. 29, a Sunday, and moves weekend deadlines to the next business day) and the final report due Feb. 27, 2031, and three quarterly reports a year because the annual report replaces the fourth. Section 16 states: "The results of this collection will not be published." Section 10 states: "The information being requested for this data collection is intended to be publicly available upon request." CMS does not post the document itself; the copy read here is the one posted by KFF Health News. A Federal Register notice of March 11, 2026 confirms the collection, its CMS number and its quarterly and yearly frequency. CMS's post-award FAQ of September 2026 says states "are required to track at least four quantifiable metrics for each initiative" and requires CMS approval for any addition or deletion of a metric.
- GAO on monitoring and the remaining $40 billionGovernment Accountability Office report GAO-27-108853, "Rural Health Transformation Program: Early Implementation Actions and Monitoring Plans," published Oct. 6, 2026. GAO reviewed CMS documentation, interviewed CMS officials and took written responses from the agency, and interviewed three rural health or state government stakeholder groups, whose views it says are not generalizable. It found the planned monitoring activities "generally consistent with selected requirements for monitoring federal awards." In states' notices of award CMS "required states to submit quarterly and annual progress reports that CMS will use to measure states' performance against states' milestones." CMS plans to "allot up to $40 billion of the remaining RHT program funds to states over the next 4 years" based on its assessment of satisfactory progress, and to "convert states' progress implementing initiatives and policy actions into numerical scores" to calculate allotments. Half of each year's money is split equally by law; of the rest, CMS scores rural factors once and recalculates technical scores each year. Budget period 2 funding determinations were to begin in August 2026. The authorizing law limits states' ability to appeal funding changes.Sources: GAO product page, GAO report (PDF)
- What is public, state by stateKFF Health News, Aug. 27, 2026, reported that states must send progress reports to the federal government but are not required to post them proactively, and that a CMS document says the reports are "intended to be" shared on request while CMS will not publish them proactively. CMS spokesman Timothy Foster said the agency "will publish an annual report on state progress." The article reports that Alaska, Kansas and Oklahoma list recipients, award amounts and spending descriptions; New Hampshire posts recipient contracts; Florida and Nebraska do not share what awardees plan to do with funds; Mississippi's governor vetoed a transparency bill; and South Dakota released a nearly fully redacted budget. CMS's own state snapshots, drawn "from State RHT Program applications," give initiative names and descriptions rather than performance metrics.Sources: KFF Health News, CMS state snapshots
- Artificial intelligence in the state plansHealthcare IT News, May 21, 2026, quoted Daniel O'Neil, a principal at Civic Operator, which operates the Rural Health Transformation Grant Tracker: "On the AI front, I've counted approximately 23 states that specifically reference it in their plans." The article reports that Vermont's notice of funding opportunity sought to provide AI scribe technology to rural practices, with applications due June 5. Vermont's application activity details, posted by the state legislature, describe an ambient documentation grant and state that "Scribes automate an average of over two hours of daily administrative work"; the document gives no source for that figure and lists no metrics or targets. KFF Health News reported on Aug. 11, 2026 that Vermont's AI scribe funding opportunity "requires organizations to report only how many clinicians and patients are served by the tech, not how much time they save."