TEFCA’s second draft gave every connected provider a duty to answer

ONC’s April 2019 draft of the Common Agreement said every network, participant and participant member must respond to requests for permitted purposes. For a health system, that turned interoperability from sending records into running a service that answers, matches, honors choices and proves who asked.

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

ONC’s April 2019 draft of the Common Agreement said every network, participant and participant member must respond to requests for permitted purposes. For a health system, that turned interoperability from sending records into running a service that answers, matches, honors choices and proves who asked.

  1. The week On April 19, 2019, the Department of Health and Human Services, through the Office of the National Coordinator for Health Information Technology, released three documents for public comment on nationwide exchange of health information between networks.
  2. Answering: a service that receives questions The draft defined three ways information would move.
  3. Matching: deciding who the request is about Every query starts with a person.
  4. Honoring choice: making an opt-out travel The draft gave individuals a right to opt out.
  5. Proving: who asked, what was answered, and what went wrong The last task is evidence.
  6. Objections: voluntary, draft and handled by the vendor Nothing in the 2019 draft forced a health system to join.
  7. How the four tasks fit in one request Seen from inside a health system, the four tasks happen in a fraction of a second for each request.
  8. Sending: the other half of the bargain A health system that joins also gains the right to ask.
  9. Preparing to connect The draft gave QHINs 18 months to update their agreements and technical requirements.
  10. Where the cost of answering falls A duty to answer has a cost, and the draft said something about who could charge whom.
  11. With hindsight The framework moved from draft to operation over the following years.

The week

On April 19, 2019, the Department of Health and Human Services, through the Office of the National Coordinator for Health Information Technology, released three documents for public comment on nationwide exchange of health information between networks. ONC’s announcement that day listed them as second drafts of the Trusted Exchange Framework and of the Minimum Required Terms and Conditions, which had previously been one document, and a first draft of the Qualified Health Information Network Technical Framework. The work carried out a provision of the 21st Century Cures Act calling for a trusted exchange framework and common agreement. ONC said the documents would together form the basis of a single Common Agreement creating baseline technical and legal requirements for sharing electronic health information nationally.

ONC also announced a Notice of Funding Opportunity for a Recognized Coordinating Entity to develop, update, implement and maintain the Common Agreement and the technical framework, under a four-year cooperative agreement. It said it had received more than 200 comments on the first draft of January 2018, and that its advisory committee had made 26 recommendations. In response, it moved most standards requirements out of the agreement and into the technical framework, to be incorporated by reference, and broadened the definition of a Qualified Health Information Network, or QHIN, so that more types of organizations could apply.

The draft package set out the structure of exchange. QHINs connect to each other. Participants connect to a QHIN, and Participant Members connect through a Participant. The draft defined three exchange modalities, QHIN Targeted Query, QHIN Broadcast Query and QHIN Message Delivery, and a minimum set of Exchange Purposes: treatment, public health and benefits determination, plus a subset of payment and health care operations activities, and Individual Access Services. It said QHINs, Participants and Participant Members have a duty to respond to all requests for electronic health information they receive for any of the Exchange Purposes with the information they have available, with a narrower duty for those that only provide Individual Access Services.

The draft extended privacy and security terms to organizations outside HIPAA that chose to take part, required Breach notification in line with the HIPAA Breach Notification Rule regardless of HIPAA status, and required identity proofing at a minimum of IAL2 before credentials were issued. It gave individuals a right to exercise Meaningful Choice that their information not be used or disclosed through the Common Agreement, except as required by law. It gave QHINs 18 months, rather than the first draft’s 12, to update agreements and technical requirements. Comments were due on June 17, 2019.

The draft also described how a network would become a QHIN. An applicant had to be operating a network that could locate and transmit electronic health information among multiple parties in a live clinical environment, and give the coordinating entity a written plan for meeting all the requirements. Once approved, it would become a Provisional QHIN, assigned to a cohort, and be designated only when the coordinating entity confirmed it had met the requirements. Federal agencies not subject to HIPAA could take part as Participants or Participant Members, complying with the privacy and security requirements of applicable federal law.

Much of the discussion of the April 2019 drafts concerned the network at the top: who could become a QHIN, what the Recognized Coordinating Entity would do, which standards belonged in the agreement and which in the technical framework. Those were real questions, and they shaped the program. This article is about one sentence lower down, in the description of Exchange Purposes, and what it meant for a health system that would sit at the edge of the network rather than at its center.

The sentence said that QHINs, Participants and Participant Members have a duty to respond to all requests for electronic health information they receive for any of the Exchange Purposes, with the information they have available. Without a common agreement, a health system could choose which networks to join and answer on each network’s own terms. A duty to respond changed the direction of the work. Interoperability stopped being mainly about sending records out and became about running a service that receives questions from parties the health system had never dealt with, and answers them correctly.

Our argument is that a duty to answer was an architecture decision for every health system that would join. Answering correctly breaks into four tasks, which we use as the sections of this article, followed by a fifth on objections:

  • Answering: receiving and responding to inbound requests at the volume and speed a national network brings.
  • Matching: deciding which patient a request is about, from demographic data supplied by someone else.
  • Honoring choice: knowing whether the patient has opted out, and making that choice travel to every place a request might come from.
  • Proving: knowing who is asking, keeping a record of what was answered, and reporting when something goes wrong.
  • Objections: participation was voluntary, the text was a draft, and the EHR vendor would handle it.

The four tasks happen in sequence for each request and do not overlap: receive it, find the patient, check the patient’s choice, and record the transaction. Together they cover what a responder does. We describe the draft as it stood in 2019; the hindsight section covers what was finalized.

Answering: a service that receives questions

The draft defined three ways information would move. QHIN Targeted Query sends a request to a known location. QHIN Broadcast Query requests information from all other QHINs. QHIN Message Delivery pushes information to a recipient. The draft gave an example of why both kinds of query matter: a provider following up after a hospitalization may want records from that hospital only, while a provider seeing a new patient may want records from every provider the patient has seen.

Broadcast changes the load

Broadcast Query is the modality that changes a responder’s architecture. A health system answering targeted queries hears from parties who already know it holds the patient’s records. A health system reachable by broadcast may hear from any requester in the network looking for any patient, including requests for patients it has never seen. Each still has to be matched and answered, even if the answer is that no records exist. We think a health system planning to join needed to treat inbound query volume as a capacity question with a number attached, not as a feature its EHR would absorb.

Which purposes, and who decides

The duty covered the Exchange Purposes. The draft initially required treatment, public health and benefits determination; a subset of payment, namely utilization review; a subset of health care operations, namely quality assessment and improvement and business planning and development; and Individual Access Services. It said that commenters had found the full payment and operations purposes too burdensome to implement immediately, and that new modalities and purposes would be phased in over time.

For the responder, each purpose is a policy rule. A request for treatment from a provider is a familiar disclosure under HIPAA. A request for benefits determination from a government agency, or for individual access from an app the patient chose, may raise different questions about what can be released and what must be logged. We think the responder needed a clear rule for each purpose, built into the system that answers, so that the answer did not depend on a person reading each request.

Answering with what you have

The duty was to respond with the information the responder has available. That phrase is easy to read past. For a health system, the information available may be spread across an EHR, a separate laboratory system, a radiology archive and records from acquired practices on older platforms. What the exchange interface can reach is what the health system will be able to answer with. A health system that wanted its answers to be complete needed to know which of its records were reachable through the exchange interface and which were not, and to decide whether that gap was acceptable.

Availability and speed

A requester treating a patient in an emergency department needs an answer while the patient is still there. A responder that answers slowly, or is down during its own maintenance windows, gives that requester nothing. The draft’s technical framework focused on exchange among QHINs, and said it did not address the methods QHINs use to query and deliver messages within their own networks. We think a health system should have set its own targets for response time and availability for inbound requests, monitored them, and planned for its own outages, including telling its QHIN when it could not answer.

Requests from public health and benefits agencies

Two of the required purposes bring requesters a health system may not have dealt with through exchange before. Public health requests come from health departments. Benefits determination requests come from government agencies deciding eligibility for programs. Each has its own legal basis and its own expectations about what is returned. We think a health system needed to confirm with counsel what it could release for each purpose, configure its answers accordingly, and make sure its logs recorded the purpose stated in each request, so that any later accounting or review could show why each disclosure was made.

Records with stricter rules

Some records carry protections stricter than the HIPAA baseline, under other federal or state laws. A responder answering automatically needs to know which of its records fall into those categories and whether its interface can hold them back or segment them when a request’s purpose or the patient’s consent does not permit release. Where it cannot, the health system faces a choice between excluding those records from what the interface can reach and building the controls to release them correctly. Either choice should be made deliberately and written down.

Matching: deciding who the request is about

Every query starts with a person. The requester sends demographic information, and the responder has to decide which of its patients, if any, that information describes. The draft’s technical framework described this step: in the patient discovery workflow, the initiating QHIN shares patient demographic information, and each responding QHIN uses it to resolve the patient’s identity, which the framework calls patient matching, and returns a resolved identity with a local patient identifier.

Two ways to be wrong

Matching can fail in two directions. A missed match means the responder holds the patient’s records but does not return them, so the requester treats the patient without information that existed. A false match means the responder returns another person’s records. The first is a care quality problem. The second is a privacy breach and a safety problem, because a clinician may act on someone else’s allergies or medications. The two failures pull against each other: loosening the match to find more true records finds more false ones too.

When there is more than one candidate

Some queries will match more than one patient, or match one patient weakly. The system then has three options: return nothing, return the best candidate, or send the request to a person to resolve. Returning nothing protects privacy and fails the requester. Returning the best guess risks a false match. Sending to a person costs staff time and may be too slow for the requester. We think the health system should set the rule deliberately, by purpose: a treatment request in an emergency may justify a person checking quickly, while a bulk or administrative request may justify returning nothing until the match is clear.

Data quality at the front desk

The draft placed weight on demographic data. Its principles said that sophisticated algorithms using demographic data are the primary method for connecting data to an individual, that networks should agree on and consistently share a core set of demographic data with each request, and that participants should ensure that core set is consistently captured for all individuals so it can be exchanged in a standard format. It pointed to ONC’s Patient Demographic Data Quality Framework as a tool for evaluating demographic data quality.

For a health system, that means matching quality starts at registration. Names spelled inconsistently, missing middle names, outdated addresses and placeholder dates of birth all reduce the chance that an inbound query finds the right patient. We think the registration process is part of the exchange architecture, and the people who run it should know that their accuracy determines what the organization can answer.

Measuring match quality

A health system can measure how its matching performs on inbound queries: how many requests found a match, how many found more than one candidate, and how many were resolved by a person. It can sample matches for accuracy. False matches are hard to detect from inside the system, so a route for requesters and patients to report a wrong record, and a process to investigate each report, is part of the measurement. Those measures tell the organization whether a duty to answer is being met correctly as well as promptly.

Honoring choice: making an opt-out travel

The draft gave individuals a right to opt out. It required QHINs, Participants and Participant Members to give individuals the opportunity to exercise Meaningful Choice to request that their information not be used or disclosed through the Common Agreement, except as required by applicable law. Participants and Participant Members were responsible for communicating that choice up to their QHIN, which must then communicate it to all other QHINs. The choice had to be respected on a prospective basis.

The path of a choice

Read as architecture, that is a message that has to travel. A patient tells a registration clerk, a portal or a call center that they want to opt out. The choice is recorded in the health system’s systems. It goes to the Participant, if the health system connects through one, and on to the QHIN, which tells every other QHIN. From then on, any request about that patient, from anywhere in the network, should be handled according to the choice.

Each step is a place where the choice can be lost. A clerk may record it in a free-text note. An interface may not carry it. A Participant may batch updates. We think a health system should have traced that path end to end before joining, and tested it by recording a test opt-out and confirming how it was handled.

Prospective, not retroactive

The draft said the choice must be respected on a prospective basis. Information already disclosed before the choice is not recalled. That limits what a choice can do, and patients may not expect the limit. We think the health system’s explanation to patients should say plainly what opting out does and does not do, and the explanation should match how the systems behave.

Privacy summaries for individual access

The draft also required those providing Individual Access Services to publish a written notice describing their privacy practices regarding the access, exchange, use and disclosure of electronic health information, mirroring ONC’s Model Privacy Notice. For a health system that offered its own patient access app through the network, that notice was its own task. For one answering requests from apps patients chose, it was a reason to understand what those apps had promised.

Proving: who asked, what was answered, and what went wrong

The last task is evidence. A responder needs to know who is asking, keep a record of what it released, and report when something goes wrong. The draft addressed all three.

Identity of the requester

The draft’s terms required identity proofing before credentials were issued. Each QHIN had to identity proof its own staff and users who could initiate a query or message delivery at a minimum of IAL2. Each QHIN had to require that Participants be identity proofed at a minimum of IAL2, and require each Participant to identity proof its Participant Members at the same level. Individual users with a direct relationship to a QHIN also had to be proofed at a minimum of IAL2, with policies that could be commensurate with risk: the draft said procedures for someone getting credentials to see their own records could be less rigorous than for someone getting credentials to see records on multiple patients.

That is a chain. The responder trusts that a request came from a proofed and authenticated person because each link in the chain agreed to proof the next. A health system answering requests relies on other organizations’ proofing. A health system sending requests carries duties of its own toward its staff. We think a health system needed to know both sides and to confirm its own credentialing met the level the agreement required.

Audit of every transaction

The draft’s technical framework said each QHIN involved in a query maintains audit logs of all activities and transactions it performed in resolving the query, according to the IHE Audit Trail and Node Authentication profile. A health system’s own systems keep their own logs. If a patient later asks who saw their record, or a false match is reported, the answer depends on joining those logs: the request that arrived, the match the system made, the choice it checked and the records it returned.

We think a health system should be able to reconstruct any single inbound request from its logs within a reasonable time, and should test that it can. A log that exists but cannot be searched is little help when a patient complaint or a breach investigation needs an answer.

Breach notification regardless of status

The draft required QHINs, Participants and Participant Members to comply with breach notification requirements based on the HIPAA Breach Notification Rule regardless of whether they were covered entities or business associates. Each QHIN was to notify the Recognized Coordinating Entity and others who may have been affected without unreasonable delay. The draft said these requirements did not replace any HIPAA or FTC breach reporting obligations. For a health system already subject to HIPAA, the new element was the network path: a breach involving exchange might need to reach the QHIN and through it other parties, alongside the reports the health system already made.

Answering the patient who asks who looked

A patient who learns their records travel through a national network may ask who has requested them. The draft’s audit requirements make that question answerable in principle. In practice, the answer is spread across the health system’s logs, its Participant’s logs and the QHINs’ logs. We think a health system should decide in advance how it will respond to such a request: which logs it can search itself, which it must ask others to search, and how long that takes. A patient who asks and waits a long time for a partial answer may conclude the network is less trustworthy than it is.

Objections: voluntary, draft and handled by the vendor

Participation was voluntary

Nothing in the 2019 draft forced a health system to join. ONC described a Common Agreement to which QHINs would voluntarily agree to adhere. A health system could stay with the networks it already used. Our answer is that the choice to join was voluntary but the consequences of joining were not. A health system that joined took on the duty to respond, and the four tasks came with it. The time to design for them was before signing, while the health system could still decide how and through whom to connect.

The text was a draft

The documents were released for comment, and the final terms could change. A health system might reasonably have waited. We think the details would change more than the shape. Any national exchange that lets parties query each other needs responders who answer, match, honor choices and keep records. A health system that began measuring match quality, tracing its opt-out path and testing its logs in 2019 was doing work that any final version would need.

The EHR vendor would handle it

The EHR vendor will supply much of the machinery: the interface that receives queries, the matching algorithm, the place to record a patient’s choice. The vendor does not control the health system’s registration accuracy, its staff credentialing, its decision about which records are reachable or its response to a false match. The draft’s definitions also made clear that a Participant Member is a party to its own agreement, with an order of precedence among applicable law, the agreement and other terms. We think the useful step was to read the vendor contract alongside the network terms and write down which duties the vendor performs and which stay with the health system.

Our existing networks already do this

A health system connected to a regional or national exchange before 2019 may already have answered queries, and some of the four tasks were familiar. ONC’s own description of the problem was that national networks and local or regional exchanges “are often not connected and offer varied services and use cases.” A common agreement aimed to connect them, which meant the volume and variety of requests reaching any one responder could rise. Work that was adequate for a regional exchange might need to scale.

The volume will be small at first

In 2019, no QHIN had been designated, and a health system might expect early volume through any new framework to be modest. A health system might plan to scale its service when the traffic arrived. We think that sequencing is reasonable for capacity and risky for the other three tasks. Matching quality, opt-out propagation and log reconstruction are hard to fix under load, and errors made at low volume are easier to find and correct. Building those correctly while traffic was light was cheaper than retrofitting them later.

Answering exposes more patient data

A privacy officer might argue that a duty to answer every permitted request increases the exposure of patient data, and that a health system should join cautiously or not at all. The concern is legitimate, and the draft acknowledged it by extending privacy and security terms to parties outside HIPAA and by giving individuals a right to opt out. Our view is that the exposure is managed by doing the four tasks well rather than by answering less. A responder that matches accurately, honors choices, proofs requesters and keeps searchable logs is exposing data on terms it can defend. A responder that answers reluctantly and inconsistently risks both missed care and poor records of what it released.

How the four tasks fit in one request

Seen from inside a health system, the four tasks happen in a fraction of a second for each request. A query arrives. The system checks the requester’s credentials and the Exchange Purpose. It runs the match. It checks the patient’s choice. It assembles the available records, returns them and logs every step. If any step fails, the request either goes unanswered, which fails the duty to respond, or is answered wrongly, which may be a breach.

That is why we read the duty to respond as an architecture decision. It turned a set of separate concerns, owned by registration, health information management, IT, privacy and compliance, into one service with one path. A health system that left each concern with its usual owner, without someone owning the whole path, could have each part working and the service still failing.

Failing safely

Each step needs a defined behavior when it cannot complete. If the credential check cannot reach the directory, the request should be refused and logged, not answered. If the match is uncertain, the rule set for that purpose applies. If the system cannot read the patient’s choice, it should treat the patient as opted out until it can. If logging fails, answering should stop, because an answer without a record cannot be defended later. We think those rules should be written down, tested and reviewed by privacy and compliance as well as IT, because they decide what happens on the bad day rather than the ordinary one.

An owner for the service

We think the inbound exchange service should have a single owner with authority across the path: someone who sees the volume, the response times, the match rates, the choice exceptions and the audit results together, and can fix a problem wherever it sits. The owner could sit in health information management or in IT. The title matters less than the scope.

Measures for the board

The service produces measures a board or a privacy committee can follow without technical detail: inbound requests received and answered, the share answered within the target time, match outcomes and confirmed false matches, opt-outs recorded and propagated, and the results of the latest log reconstruction test. Those measures show whether the health system is meeting its duty and protecting its patients at the same time. They also give internal audit a basis for testing the service, by sampling requests and following each one from arrival to log, which is the same reconstruction the health system would need to perform for a patient complaint or a breach review.

Sending: the other half of the bargain

A health system that joins also gains the right to ask. Its clinicians can query for records from providers they have never worked with, and its patients can reach their records through apps they choose. That benefit carries duties of its own, and they mirror the four tasks from the requester’s side.

Asking for the right reason

Each request carries an Exchange Purpose. The health system sending it is asserting that the purpose is true. A clinician querying for a patient they are treating is making a treatment request. A query for any other reason under a treatment label would be a misuse. We think a health system should limit who can initiate queries, require the purpose to be chosen rather than defaulted, and review a sample of outbound queries against the clinical record to confirm the stated purpose.

Sending good demographics

The draft’s call for consistent capture of a core set of demographic data applies to requests as well as responses. A query with incomplete or wrong demographics produces missed matches at every responder. The registration accuracy that makes a health system a good responder also makes it a good requester.

Credentials for its own people

The identity proofing chain runs through the health system. Its staff who can initiate queries need credentials issued after proofing at the level the agreement required, and those credentials need to be removed when people leave or change roles. A responder elsewhere is trusting that the health system did this. We think joiner, mover and leaver processes for exchange access deserve the same attention as those for the EHR itself.

Using what comes back

Records retrieved from other organizations arrive in the health system’s own environment. The health system then has to decide how they are stored, how they are shown to clinicians, and how a clinician can tell an outside record from a local one. Retrieved records may also contain errors from the other organization’s matching. A clinician who suspects a record belongs to someone else needs a way to say so, and the report needs to reach the organization that sent it.

Preparing to connect

The draft gave QHINs 18 months to update their agreements and technical requirements. A health system’s own preparation could run alongside. We think the preparation falls into a short sequence that does not depend on the final text.

Choose the route

The draft’s structure offered more than one route in. A health system could connect through a Participant, such as a regional exchange or a technology vendor, or seek to connect more directly. The route determines which agreement the health system signs, which party carries which duties, and how a patient’s choice or a breach notice travels. We think the route should be chosen with those questions in mind, as well as price and convenience.

Inventory what can answer

Before joining, the health system can list the record sources its exchange interface will reach, the sources it will not, and the categories of records it will hold back. That inventory sets expectations for requesters and for the health system’s own clinicians, who will be asked why a record was or was not returned.

Test the path

The four tasks can be tested before live traffic arrives: a test query with a known patient, a test query with a near-miss demographic, a test opt-out traced to the network, and a reconstruction of a test transaction from the logs. Each test has a pass or fail result and an owner for the fix. Running them before go-live finds the gaps when they are cheap to close.

Tell patients

Patients will be affected by the health system’s participation and have a right to opt out. A plain explanation of what joining means, what opting out does and does not do, and how to ask who has seen their records belongs in the preparation. The explanation should be written after the testing, so that it describes how the systems behave rather than how they were meant to.

Where the cost of answering falls

A duty to answer has a cost, and the draft said something about who could charge whom. Its comparison of the two drafts said QHINs must use reasonable and non-discriminatory criteria if they charge fees to another QHIN, that a QHIN may not charge another QHIN any amount to exchange information for Individual Access Services, and that a QHIN may not impose any other fee on the use or further disclosure of information once another QHIN has accessed it. It defined fees broadly, to include membership, testing, usage, transaction and data analytics fees, and said any additional requirements on fees would be specified later by the coordinating entity with ONC’s approval.

Costs the fee rules do not touch

Those rules govern charges between networks. The responder’s own costs sit elsewhere: the staff who resolve uncertain matches, the registration work that improves demographic data, the interface capacity to handle broadcast queries, the monitoring and the log searches. None of those are paid by the requester under the draft’s terms, and some of them grow with volume. We think a health system’s budget for joining should name each of them, so that the decision to join is made with the running cost in view as well as the connection fee.

Costs that fall on patients if the work is skipped

The costs of not doing the four tasks fall elsewhere. A missed match costs a patient a repeated test or a decision made without history. A false match costs a patient their privacy and may cost another patient their safety. A lost opt-out costs a patient the choice the draft promised them. Those costs do not appear in the health system’s budget, which is a reason for the board to see the measures described above alongside the financial ones.

With hindsight

The framework moved from draft to operation over the following years. ONC’s TEFCA page, as checked on November 15, 2026, says the three foundational documents, the Common Agreement, the Trusted Exchange Framework and the QHIN Technical Framework, were published in January 2022 and updated in November 2023. It says the Sequoia Project serves as the Recognized Coordinating Entity under a five-year contract with ONC awarded in August 2023, and that in December 2023 the first QHINs were designated and health data began flowing among them within days. The page lists the coordinating entity’s tasks as including managing QHIN applications and designation, monitoring designated QHINs and updating the technical framework.

The page describes the Common Agreement as a legal contract signed by all QHINs, and the technical framework as covering patient identity resolution, authentication and performance measurement, with requirements that flow down to participants and subparticipants. The page speaks of participants and subparticipants where the 2019 draft spoke of Participants and Participant Members. The page lists six Exchange Purposes: treatment, payment, health care operations, public health, government benefits determination and individual access services. It says each request must indicate its purpose and that ONC and the coordinating entity will expand the permissible purposes over time.

Federal rulemaking followed. On December 16, 2024, HHS published a final rule, Health Data, Technology, and Interoperability: Trusted Exchange Framework and Common Agreement, at 89 FR 101772, effective January 15, 2025. Its abstract says it amends the information blocking regulations by including definitions related to the TEFCA Manner Exception and implements provisions related to TEFCA that support its reliability, privacy, security and trust.

Use grew. ONC’s fact sheet dated February 11, 2026 says approximately 10 million documents were exchanged across TEFCA before 2025 and that by the end of 2025, 464 million documents had been exchanged. It says over 71,000 sites or organizations are participating through 11 QHINs, that the Indian Health Service was the first federal agency to select a QHIN, and that the Social Security Administration was involved in developing its own sub-purpose within government benefits determination.

The same fact sheet describes how organizations connect. An entity can connect directly to a QHIN of its choice as a participant, or through a health information network or technology provider already connected to a QHIN as a subparticipant, and the fact sheet’s steps for joining include getting listed in the coordinating entity’s directory. Individual access services providers chosen by patients must agree to the terms all participants sign and to additional transparency requirements on how they collect, share and use a person’s health information.

We have not compared the duty to respond in the current Common Agreement clause by clause with the 2019 draft, and we do not describe its current wording here. What held up, as we read it, is the architecture point. A network that went from approximately 10 million documents exchanged before 2025 to 464 million by the end of 2025 is one in which responders carry real load, and in which matching, choice and audit run at that scale. A health system that had treated inbound exchange as a service with an owner would have had measures in place when that volume arrived.

Spotted an error? Tell us and we will check it and correct the article where needed.

  • TEFCA
  • interoperability
  • QHIN
  • patient matching
  • identity proofing
  • ONC