CMS would switch off a nursing home data feed but keep the switch
CMS’s CARE proposal, published November 10, would end nursing homes’ routine respiratory reporting, keep emergency reporting, ask hospitals about automation and drop three state Medicaid reports. Each change moves work from submitting data to staying able to submit it.
Looking Back revisits past weeks one at a time, up to the most recent, and asks what it meant for the banks, health systems and public agencies we serve. The series was written in 2026, with the benefit of hindsight, and each article says which week it looks back on.
Two-minute summaryThe dek and the first sentence of each section, quoted from the article
CMS’s CARE proposal, published November 10, would end nursing homes’ routine respiratory reporting, keep emergency reporting, ask hospitals about automation and drop three state Medicaid reports. Each change moves work from submitting data to staying able to submit it.
- The week On Tuesday, November 10, 2026, the Centers for Medicare & Medicaid Services published a proposed rule in the Federal Register titled “Medicare and Medicaid Programs; Cutting Administrative Requirements for Excellence in Patient Care” (CMS-3484-P, 91 FR 72634).
- The feed CMS would switch off Under 42 CFR 483.80(g)(1), long-term care facilities report respiratory illness data to NHSN on an ongoing basis.
- The switch CMS would keep The current rule gives the Secretary authority, during a declared federal, state or local public health emergency, to require long-term care facilities to report additional or modified data on an acute infectious illness.
- The feed CMS wants automated Hospitals and critical access hospitals are treated differently.
- The report CMS would derive The fourth data flow belongs to state Medicaid agencies.
- Where the argument is weakest, and what to file by January 11 The strongest objection to our reading is that CMS has measured the burden and found it unjustified.
- With hindsight Very little time has passed.
The week
On Tuesday, November 10, 2026, the Centers for Medicare & Medicaid Services published a proposed rule in the Federal Register titled “Medicare and Medicaid Programs; Cutting Administrative Requirements for Excellence in Patient Care” (CMS-3484-P, 91 FR 72634). CMS calls it the CARE proposed rule. Its fact sheet says CMS issued the proposal on November 5. Comments must be received by January 11, 2027.
The proposal comes out of what CMS describes as a comprehensive review of the conditions of participation for all 21 provider and supplier types. Conditions of participation are the health and safety rules a hospital, nursing home or other facility must meet to take part in Medicare and Medicaid. CMS says it looked for obsolete and burdensome requirements, and that the proposal also relates to a 2025 deregulatory request for information. The fact sheet puts the expected savings at $320 million and more than 3.26 million hours of staff time.
Most of the rule is about clinical and facility requirements. For hospitals and critical access hospitals, CMS would remove recently finalized staff training requirements for transfer protocols, emergency services readiness and obstetric care, and obstetric quality requirements. The fact sheet notes that the obstetric provisions are otherwise effective January 1, 2027. CMS would also adopt newer editions of the National Fire Protection Association’s Life Safety Code and Health Care Facilities Code, and the 2022 edition of its standard for emergency and standby power systems.
Four provisions concern data that facilities and states send to the federal government. For long-term care facilities, CMS would remove the ongoing requirement to report respiratory illness data to the Centers for Disease Control and Prevention’s National Healthcare Safety Network, known as NHSN. It would keep the requirement to report specified data during a declared public health emergency. For rural emergency hospitals, it would remove expired COVID-19 and influenza reporting requirements and the requirement to report during an emergency.
For hospitals, including critical access hospitals, CMS proposes no such removal. It asks instead, in a request for information, how to cut the burden of respiratory and infectious disease reporting through automation, including the HL7 FHIR data standard. And for state Medicaid agencies, it would rescind three of the annual facility-level reports on how much of Medicaid payment goes to worker compensation, set by a 2024 rule and not yet in effect.
The long-term care reporting requirement being removed is recent. CMS finalized it in the calendar year 2025 home health payment rule. It requires facilities to report a facility census and data on acute respiratory illnesses, including influenza, COVID-19 and RSV, through NHSN in a format and frequency the Secretary specifies.
CMS’s case for removing it rests on burden and duplication. The proposal says commenters described the reporting as overlapping with state and local requirements and with the Minimum Data Set, and as creating technical problems with reporting software. It says facilities have reported continuing to struggle to the extent that they are unable to comply. CMS estimates that keeping the requirement would cost about $61 million a year on average and over 600,000 staff hours a year.
It is natural to read a deregulatory rule for what goes away. The CARE proposal invites that reading, with its $320 million estimate and its list of removed training, vaccination and building requirements.
For the people who run clinical data platforms and reporting pipelines, the proposal does something more specific. It sorts federal health data flows into different treatments. One routine feed would be switched off. The authority to switch it back on during a declared emergency, without new rulemaking, would be kept. A second feed, from hospitals, would stay and become the subject of questions about automation. And three state reports that have not started yet would be rescinded, with CMS saying it may be able to approximate one of the measures from existing data.
Our argument is that each of these moves the work from submitting data to staying able to submit it. A nursing home that stops weekly reporting still needs a reporting capability that works on the day an emergency is declared. A hospital asked about FHIR is being asked how much of its reporting can come straight from its electronic health record. A state that no longer files a compensation report may still feed the data CMS would use instead. The burden changes shape. It does not disappear, and organizations that treat it as gone will pay for that later.
The article takes the proposal in five parts. The first four are the four data flows the proposal treats differently: the feed CMS would switch off, the switch it would keep, the feed it wants automated and the report it would derive. They do not overlap, and together they cover the CARE provisions about data reported to the federal government. The fifth tests the argument against the strongest objections and sets out what to file by January 11. Organizations should confirm their reading of the proposal with counsel and their compliance teams.
The feed CMS would switch off
What long-term care facilities report today
Under 42 CFR 483.80(g)(1), long-term care facilities report respiratory illness data to NHSN on an ongoing basis. CMS describes the content as a facility census and information on acute respiratory illnesses, including influenza, COVID-19 and RSV. The reported data covers resident vaccination status, confirmed resident cases and hospitalized residents.
The requirement replaced a heavier pandemic regime. CMS says the 2025 rule extended a modified form of the earlier COVID-19 and influenza reporting. It added RSV and reduced how often facilities report compared with prior rules.
Why CMS says the feed should end
The proposal gives three reasons. The first is duplication. Commenters told CMS that facilities already report clusters of respiratory symptoms and confirmed cases to local public health agencies under infection control rules, and send data through the Minimum Data Set. The second is technical difficulty. Commenters described problems operating reporting software, and CMS says facilities have continued to struggle to the point of being unable to comply. The third is cost. CMS puts the cost of keeping the requirement at about $61 million a year and over 600,000 staff hours, and says it has not identified matching improvements in resident health or safety.
CMS is candid that it heard similar complaints when it finalized the requirement. At that time, the proposal says, it believed facilities could overcome the problems. It now concludes they have not.
How CMS counts the cost
The $61 million figure is a labor estimate. CMS says the burden of the reporting includes labor costs for developing policies and procedures, and for gathering, reviewing and organizing the data used to compile each report. CMS counts it as staff time, not software cost.
That matters for what a facility actually saves. Where reporting was largely manual, ending it frees staff time directly. Where a facility had automated most of it, the saving is smaller, and the cost that remains is keeping the automation alive. The proposal’s estimate does not separate the two cases.
What the change would not remove
The proposal keeps the rest of the long-term care infection control framework. Facilities must still run an infection prevention and control program with an antibiotic stewardship program and at least one infection preventionist. That program must include a system for identifying, reporting, investigating and controlling infections. The facility assessment and the quality assessment and performance improvement program remain. State and local reporting rules are untouched.
CMS also says it encourages long-term care facilities to keep submitting data to NHSN voluntarily, to support wider public health work. The proposal notes that CDC continues to promote electronic collection directly from health records, such as FHIR-enabled reporting from hospitals.
Rural emergency hospitals
Rural emergency hospitals are a Medicare provider type created by the Consolidated Appropriations Act, 2021, for former critical access hospitals and rural hospitals with not more than 50 beds that provide no acute inpatient care. CMS counts 51 operating as of April 2026. Their conditions of participation, set in a November 2022 rule, included data reporting requirements that CMS describes as similar to those for hospitals and critical access hospitals at the time. The proposal would remove expired COVID-19 and influenza reporting provisions and the requirement to report during a public health emergency.
The vendor contract
Where a facility reports through a tool supplied by its health record or analytics vendor, the contract terms for that tool matter when a federal requirement ends. A contract may price the reporting module separately or bundle it. A facility should check whether it can keep the module in a dormant but working state, and what it would cost to reactivate it if canceled.
Vendors face their own choice. We think a reporting module with no routine users is hard to justify maintaining, and a vendor may retire it. A facility that relies on one should ask the vendor directly how it will support emergency reporting if the requirement changes as proposed.
Differences by type of operator
A single, independent nursing home is the case CMS seems to have most in mind. It may have had one or two staff who compiled the weekly report by hand. For it, the change returns time to care, and the readiness question is mostly about keeping one person trained and one account active.
A multi-facility operator is different. We would expect it to be more likely to have built a central extract that feeds every facility’s submission. Its saving is smaller per facility, and its readiness can be centralized: one pipeline, one test, one owner. The risk for an operator is that a central team is cut once the routine work ends, taking the knowledge with it.
Operators in states with their own facility reporting rules sit in a third position. Their state reports continue. Commenters told CMS the federal report overlapped with state and local requirements, which may give these operators a natural way to keep the federal capability exercised.
What switching off a feed means for a facility’s systems
Ending a federal reporting requirement looks like pure relief to a facility. For its technology staff, it raises practical questions that are easy to skip.
The first is what to do with the pipeline. A nursing home chain that built an extract from its electronic health record to NHSN, or bought one from a vendor, has to decide whether to keep it running. Keeping it costs money. Turning it off saves money but means the extract, credentials and mappings will decay.
The second is access. Submitting to a federal system depends on accounts and credentials held by named staff. Where those staff leave and are not replaced, the facility can lose the ability to submit at all. A chain should decide who keeps each facility’s NHSN access current, even if routine submissions stop.
The third is the state. Several of the duplicated reports CMS cites go to state and local health departments. Those requirements stay, and some states may rely on federal data that will no longer arrive. A facility should check whether its state reporting depends on anything it currently sends only to NHSN.
The switch CMS would keep
Reporting during a declared emergency
The current rule gives the Secretary authority, during a declared federal, state or local public health emergency, to require long-term care facilities to report additional or modified data on an acute infectious illness. Reporting can be required up to daily, without new notice-and-comment rulemaking. The Secretary may also require reports of supply shortages, staffing shortages, use of medical countermeasures and therapeutics, and confirmed staff infections.
The CARE proposal keeps that authority and moves some data elements into it. Because the census and case counts sit in the ongoing section being removed, CMS proposes to restate them in the emergency section. The proposed text lists seven elements for a declared emergency, “including but not limited to” these:
- facility census;
- confirmed resident cases;
- hospitalized residents with confirmed cases;
- confirmed infections among staff;
- supply inventory shortages;
- staffing shortages;
- medical countermeasures and therapeutic inventories.
Resident vaccination status would no longer be among the listed elements.
The element that would drop out
Resident vaccination status is the one element of the current ongoing report that CMS would not carry into emergency reporting. The fact sheet says CMS would also streamline the remaining COVID-19 vaccine requirements for residents, while removing the staff vaccine education and offer requirements.
For data teams, the effect is that vaccination status stays in the facility’s own records but would not be part of the federal emergency picture. An emergency involving a vaccine-preventable illness could prompt HHS to ask for it again, since the proposed list of elements is not exhaustive. A facility that keeps its vaccination records structured, rather than in scanned documents, keeps that option cheap.
How the switch would be thrown
The proposal says HHS would notify facilities to activate emergency reporting, and set the frequency and data elements for the circumstances. CMS expects to use an existing channel such as a Quality, Safety and Oversight memorandum. The Secretary would provide standardized definitions for each data element and instructions on the effective reporting date.
CMS gives its reasoning plainly. Commenters on the 2025 rule said the value of reporting during an emergency outweighed the burden. Retaining the emergency requirements, the proposal says, “enables facilities to maintain a functional reporting capacity.”
The gap between a dormant duty and a working capability
That phrase is where the argument of this article starts. A requirement that applies only in an emergency does not, by itself, maintain a capacity. Capacity comes from use. Under the ongoing requirement, facilities submitted regularly, so problems with users, mappings and software showed up in normal weeks. Under the proposal, a facility may go years without submitting anything and then be told to report at whatever frequency the Secretary specifies.
The pandemic showed what that start-up looks like. CMS’s own history in the proposal describes an “evolving series of requirements” issued by interim final rules during COVID-19. Its account of the period describes long-term care facilities facing severe staffing shortages and limited access to protective equipment and critical supplies.
A facility that wants the switch to work when it is thrown has three practical choices:
- keep submitting voluntarily, as CMS encourages, at a frequency that keeps the pipeline and the users alive;
- keep the pipeline but stop submissions, and run a scheduled test, for example twice a year, that produces a submission-ready file for the seven emergency elements;
- switch everything off and accept a manual start-up in an emergency, with the staffing cost that implies.
We judge the first the most reliable. The third is the cheapest until it is needed. Each is a legitimate choice, and the right one depends on the facility’s size, its vendor and its state. What is not legitimate is making the third choice by default.
Owners and evidence
We expect a surveyor assessing emergency reporting after the change to look at whether the facility could comply once notified. The evidence a facility would want to hold is simple to list: a named owner of NHSN access, a current list of registered users, a documented procedure for each of the seven elements, and the date and result of the last test. For a multi-facility operator, a workable split, as good practice rather than a requirement, is for the corporate data team to own the extract and each facility to own its users and its clinical counts.
The feed CMS wants automated
What hospitals are being asked
Hospitals and critical access hospitals are treated differently. CMS does not propose to end their respiratory illness reporting. Instead, it asks for comment on how to reduce the burden of that reporting and of reporting during an emergency.
The request describes where things stand. CMS says that, with CDC, it has worked to reduce burden through data standards, interoperability work and a single reporting system, NHSN. NHSN’s hospital respiratory data reporting offers flexible options, including CSV import, application programming interfaces and FHIR-based approaches. CMS says these options exist because hospitals vary in how far they have modernized their data systems.
The questions are specific. CMS asks how it and CDC could support wider adoption of HL7 and FHIR standards to reduce reporting burden, particularly during an emergency. It asks what training, technical guidance, pilots or assistance would help. It asks what rural hospitals need to adopt automated reporting. It asks which variables collected by CDC are hard to collect, and which are easy to pull from clinical workflows and electronic health records. And it asks how CMS could revise its rules to fit clinical workflows better while keeping federal monitoring.
Three ways to submit, three levels of effort
The options CMS lists, CSV import, application programming interfaces and FHIR-based approaches, are not equal. Each puts the work in a different place.
A CSV import is the simplest to start. Someone produces a file, perhaps from a report in the health record, and uploads it. The effort sits with staff every time a report is due, and errors come from manual steps.
An application programming interface moves the submission from a person to a system. The hospital’s integration engine sends the data on a schedule. The effort shifts to building and maintaining the interface, and errors come from mappings that drift as the record system changes.
A FHIR-based approach goes one step further. The data is expressed in standard resources that other systems can read the same way. In principle, that lets the same build serve several reporting programs and makes it easier to change what is collected. In return, it asks more of the hospital’s record system and of the receiving system.
A hospital’s answer to CMS will be more useful if it says which of these it uses today, what it costs to run and what would make the next step worth taking.
Why the hospital feed is treated differently
The proposal does not explain the difference between hospitals and nursing homes at length. Read together, the two sections suggest a reason. CMS sees hospitals as able to move toward automated reporting from their health records, and nursing homes as struggling to report at all. One feed is being invested in. The other is being retired.
For hospital technology leaders, that makes the request for information more important than its form suggests. It is a chance to shape what automated reporting will mean. The answers CMS receives could inform a later proposal that changes the hospital requirement itself.
What automation actually involves
Automated reporting from an electronic health record is sometimes described as a switch. It is a project with several parts.
The data has to be defined. A census or a count of confirmed cases needs a rule about which patients, which locations and which tests count. Those rules have to be written into queries against the record, or into the FHIR resources the hospital exposes.
The data has to be validated. A count produced automatically is still wrong if a laboratory result is coded inconsistently or a unit is mapped to the wrong location. Someone has to compare automated output with a manual check, at least at first.
The connection has to be secured and maintained. An interface to a federal system carries credentials, certificates and monitoring. It needs an owner in the integration team, change control when the record system is upgraded, and alerting when a submission fails.
And the people have to stay involved. Infection prevention staff know what the numbers should look like. An automated feed that no clinician reviews can send wrong data on time.
Aggregate counts or patient-level records
The request does not say what an automated feed should carry. There are two broad designs. In one, the hospital computes the counts and sends aggregates. In the other, it exposes patient-level FHIR resources and the receiving system computes the counts.
The second design is more flexible. When definitions change in an emergency, the receiver can recompute without each hospital rewriting queries. It also moves more identifiable data out of the hospital, which raises questions about minimum use, retention and security that the aggregate design avoids.
Hospitals answering the request should say which design they are describing and what safeguards they would expect. A privacy officer’s view belongs in the response alongside the informatics team’s.
Changing definitions in an emergency
CMS’s questions return several times to emergencies. It notes that reporting requirements may need to be revised during one, and asks what would help hospitals adjust. That is a design requirement as much as a policy one.
A reporting pipeline that hard-codes each definition will be slow to change when a new pathogen or a new test appears. One that keeps definitions as versioned configuration, with a tested way to add a data element, can change in days. Hospitals that have built the second kind should describe it in their comments, because it shows what federal guidance could ask of everyone.
Differences by type of hospital
A large system on a single enterprise health record has the easiest path. It can build once and deploy to every hospital, and it may already use its vendor’s reporting tools. A system with several record platforms, for example after acquisitions, faces a mapping exercise per platform.
Critical access hospitals and small rural hospitals are a separate case. CMS asks specifically about their needs. Where a small hospital runs a smaller record system with few interface staff, the useful answers to the request may be about shared services, vendor-hosted reporting and federal technical assistance, more than about standards.
The report CMS would derive
What the 2024 rule required of states
The fourth data flow belongs to state Medicaid agencies. The 2024 minimum staffing final rule required states to report each year, by facility, on Medicaid payments for nursing facility and ICF/IID services. The reports would show the share spent on compensation for direct care workers and for support staff. ICF/IID stands for intermediate care facilities for individuals with intellectual disabilities. The requirement aimed at transparency on pay for the workforce that delivers institutional care.
The proposal gives the scale. In 2023, 1.29 million Medicaid beneficiaries received nursing facility services and about 86,000 received ICF/IID services. Those services accounted for over $79.4 billion in Medicaid spending.
What would be rescinded, and what would stay
CMS proposes to rescind three of the reports: the nursing facility share for direct care workers, the nursing facility share for support staff and the ICF/IID share for support staff. The proposed regulatory text for 42 CFR 442.43 keeps the state’s duty to report, by facility, the share of Medicaid payments for ICF/IID services spent on direct care worker compensation.
The requirements have an applicability date of June 21, 2028. CMS says states’ costs to date are likely minimal, and that beneficiaries and workers have not yet relied on the reports.
The public website that stays
The proposed text keeps the duty that comes with the remaining report. The state must operate a website, consistent with 42 CFR 435.905(b), that makes the reported information available. A state should confirm with counsel exactly what that cross-reference requires. For the ICF/IID direct care report, then, a state still needs a data model, a calculation, a review step and a public page.
The remaining report is a small build compared with the original four. It is still a public, facility-level number with the state’s name on it. The controls a state would apply to any published figure apply here: documented method, reconciliation with source payments and a correction process.
From reported to derived
The reason CMS gives is the most interesting part for data leaders. It says it now believes it may be able to approximate the nursing facility direct care share using existing data. It lists payroll-based journal data, the Minimum Data Set, Bureau of Labor Statistics data, Medicare cost reports and Medicaid claims and encounter data.
That is a shift from a reported measure to a derived one. A reported measure puts the definitional work on each state: which payments count, which costs are excluded, how managed care payments are treated. A derived measure puts it on CMS, which must combine sources built for other purposes and accept the approximation that results.
For states, the change removes a reporting build on which, by CMS’s estimate, they have so far spent little. It does not remove their role. In our understanding, Medicaid claims and encounter data reach CMS from state systems, so their quality would affect a federal estimate of how Medicaid money reaches the workforce. A state that had planned a reporting pipeline for 2028 could, we suggest, redirect some of that effort to the completeness of the data it already sends.
For nursing facilities, the derived approach rests on data they already submit, such as payroll-based staffing records and cost reports. Errors in those submissions would flow into a public estimate without the facility seeing a separate report first.
What derived data can and cannot show
The sources CMS lists were each built for a different purpose. Staffing records describe hours worked. Cost reports describe a facility’s finances over a fiscal year. Claims and encounter data describe what Medicaid paid for services. Wage statistics describe labor markets, not individual facilities.
Combining them can produce a useful estimate. It also brings problems a reported measure would handle through definitions. The sources cover different periods and arrive on different schedules. Managed care payments may appear differently from fee-for-service payments. The proposal acknowledges the result would be an approximation.
For a state Medicaid agency, the practical question is how its encounter data will be used. Encounter data that lacks payment amounts, or links poorly to facility identifiers, will weaken any federal estimate. A state’s data quality program is now part of how this measure will be judged.
Where the remaining ICF/IID report goes
States still face the ICF/IID direct care report. CMS says it intends to issue subregulatory guidance and technical specifications, and will put a reporting method and template out for comment under the Paperwork Reduction Act. The information collection carries the CMS identifier CMS-10851. States should watch for those notices, because the template will set the data model they have to build.
Where the argument is weakest, and what to file by January 11
The case that the burden really is gone
The strongest objection to our reading is that CMS has measured the burden and found it unjustified. If long-term care facilities could not comply, the feed was producing weak data at real cost. Ending it releases staff for resident care, which is the rule’s stated aim.
That objection is strong for the routine feed. It is weaker for the capability. CMS itself keeps emergency reporting because the data matters in an emergency. The facilities that, by CMS’s account, struggled with routine reporting are, we think, the ones least likely to report smoothly after years without practice. The burden of readiness is smaller than the burden of weekly reporting, but it is not zero, and it falls on the same organizations.
The case that the hospital request is only a request
A second objection is that a request for information commits CMS to nothing. Hospitals may reasonably decide to wait for a proposal.
Waiting has a cost, though. The questions CMS asks, about which variables are hard and how FHIR could be used, are ones hospital informatics teams can answer from experience. If they do not, the agencies will design the next step with less evidence from the organizations that have to run it.
The case that derived data is better data
A third objection favors CMS’s approach to the Medicaid report. Derived measures avoid separate state definitions, and they rely on data with existing validation, such as cost reports.
That may prove right. The risk is transparency about method. A reported figure can be checked against the state’s report. A derived figure is only as clear as CMS’s published method. Commenters who care about the measure should ask CMS to say how it would combine the sources, how it would treat managed care payments and how often it would publish.
The case that this belongs to clinicians, not technology
A fourth objection is that public health reporting is an infection prevention matter, and that framing it as a data platform question puts the wrong people in charge. Infection preventionists know what the counts mean and whether they are plausible.
We agree that they should own the content. The argument here is about the plumbing beneath it: accounts, extracts, interfaces, definitions and tests. Those are technology responsibilities, and when they lapse the clinicians cannot report however well they understand the data. Readiness needs both owners, with a clear line between them.
What readiness costs
Readiness has a price, and it should be stated honestly. Keeping a pipeline alive means license or hosting fees where a vendor tool is used, some integration staff time and periodic clinical review of test output. It is far less than weekly reporting. It is also the kind of spend that disappears in the first budget round after a requirement ends, because nothing visible breaks.
The trade-off is between a small recurring cost and a large, unpredictable one in an emergency, when staff are scarce and deadlines are short. An organization can choose either. It should make the choice in writing, with the chief information officer and the chief nursing or medical officer both signing, so the next emergency does not reveal a decision nobody remembers making.
Before the comment period closes
Organizations that want to comment have limited time, and the holiday period falls inside it. A workable sequence is to assign each data flow to an owner first, then gather evidence, then draft.
Evidence matters more than opinion in a comment on burden. A long-term care operator that can say how many staff hours its reporting took, how often submissions failed and why, will be heard. So will a hospital that can name the variables that never map cleanly from its record. A state that can describe the gaps in its encounter data, and what it would take to close them, helps CMS design a better derived measure.
What to file, and who should file it
Each data flow suggests a different comment from a different author:
- Long-term care operators should say what a workable activation would look like: notice period, test submissions, and whether CMS should allow a periodic test in place of routine reporting.
- Hospital informatics and infection prevention leaders should answer the FHIR questions with specifics: which variables are hard, which come cleanly from the record and what help small hospitals need.
- Rural emergency hospitals and critical access hospitals should describe their actual reporting tools and staffing.
- State Medicaid agencies should comment on the derived measure’s method and on the timetable for the ICF/IID template.
Comments must be received by January 11, 2027. The docket also covers the obstetric and emergency readiness training changes, which clinical leaders will want to address separately.
With hindsight
Very little time has passed. As of November 14, 2026, CARE is a proposed rule with its comment period open until January 11, 2027. No part of it is in effect. The ongoing respiratory reporting requirement for long-term care facilities in 42 CFR 483.80(g)(1) remains in the regulations unless a final rule removes it.
Some dates are already fixed and are worth setting beside the comment deadline. The fact sheet says the obstetric training and obstetric quality requirements CMS would remove are otherwise effective January 1, 2027. That date comes before the comment period closes. Hospitals and critical access hospitals offering obstetric services should check with counsel how to plan for those requirements while the proposal is pending. The Medicaid compensation reporting requirements have an applicability date of June 21, 2028, so states have more room.
Several things are unknown. The proposal does not say when CMS expects to finalize the rule. The number and content of comments are not yet public. We do not know whether CMS will change the list of emergency data elements, add a test or readiness requirement for long-term care facilities, or follow the hospital request for information with a proposal. The proposal does not set out a methodology for approximating the nursing facility direct care share.
The same week brought another sign of how CMS is approaching data and systems. On November 13, CMS announced an optional “hold harmless” period through June 30, 2027 for states implementing the Medicaid work requirement that begins January 1, 2027. States that use it will assess whether people meet or are excluded from the requirement but will not deny applications or disenroll otherwise eligible people solely for failing to meet it. CMS paired it with advice to maximize automated verification using data states already hold, and to test end-to-end how systems handle missing, outdated or conflicting information. That is a different program and a different rule. In our reading, the common thread with CARE is a preference for data already in hand over new submissions.
In the meantime, the rules in force are unchanged for every organization the proposal touches. The proposal says long-term care reporting currently occurs weekly, and that requirement stays until a final rule says otherwise. Hospitals report under their current rules. Our reading is that states should keep the 2028 compensation reports in their plans until CMS finalizes or withdraws the rescission.
The comment docket will be the next public evidence. The proposal says CMS posts comments it receives on regulations.gov as soon as possible. Comments from long-term care operators may show how many facilities had automated their NHSN submissions and how many reported by hand. Comments from hospitals may show how far FHIR-based reporting has spread. Neither figure is in the proposal.
What would show whether our reading is right lies ahead. If CMS finalizes the long-term care change as proposed, the test will come with the next declared emergency that triggers reporting. In our view, facilities that kept a working pipeline will be able to report quickly, and those that let it lapse will rebuild it under pressure. The hospital request will matter if it produces a proposal that ties reporting to data drawn directly from health records. We will revisit the rule when the final version appears.
Spotted an error? Tell us and we will check it and correct the article where needed.