Why Your EHR Systems Still Can’t Talk to Each Other (And What That Costs Your Practice Every Day)

Your patient had a cardiac event six months ago. They were admitted to a hospital across town, put on a new medication, and had three imaging studies done. Today they are sitting in your exam room. You pull up their chart. None of that information is there. You ask them what happened. They hand you a folded sheet of paper they pulled from their wallet.

This is not a rare scenario. This is Tuesday for most US providers.

The frustrating part is not that the technology to fix this does not exist. It does. 

First, Let Us Be Honest About Where Things Stand

Based on feedback from over 500,000 clinicians, the KLAS Arch Collaborative EHR Interoperability 2024 report found that among 11 metrics used to calculate the Net EHR Experience Score, clinicians are least satisfied with external integration. Only 44% of respondents agree their EHR provides expected integration with outside organizations. 

The problem is not just clinical. It is regulatory and financial, and organizations that ignore it are already paying for it through denials, duplicate testing, and staff burnout.

Most definitions of interoperability make it sound simple: systems share data. But in healthcare, sharing data can mean four very different things, and the gap between the lowest and highest level is enormous.

HIMSS, the Healthcare Information and Management Systems Society, defines these four levels. Understanding them tells you exactly where your organization is stuck.

This is the level most US practices think they have reached, and most have not. Lab values transfer correctly as numbers. 

But does the receiving system know what unit of measure was used, what reference range applied, or what clinical context generated that lab order? 

Often, no.

Organizational interoperability is the top level. It includes governance, policy, social, legal, and organizational considerations to facilitate the secure, smooth, and timely communication and use of data both within and between organizations, entities, and individuals, enabling shared consent, trust, and integrated end-user processes and workflows. Most organizations have not scratched the surface of this level, which is precisely why technically connected systems so often fail to produce coordinated care.

In 2026, healthcare organizations are moving beyond standard data exchange to handle the more complex challenge of semantic interoperability, ensuring that clinical information is not just shared but meaningfully understood across systems. This shift is not optional. It is what separates a data pipe from a tool that actually improves care.

The Problem Is Not Just Technical. It Is a Trust and Incentive Problem.

Here is something most interoperability blogs skip over: a significant number of data exchange failures are not accidents. 

They are choices.

For years, some EHR vendors and health systems treated data lock-in as a competitive strategy. If a hospital’s patients were difficult to refer elsewhere because their records could not travel, those patients stayed. The 21st Century Cures Act, passed in 2016 and enforced through ONC’s information blocking rules, was Congress’s direct response to this behavior.

For healthcare providers, enforcement takes a different form: financial disincentives tied to Medicare payment programs rather than direct fines. Providers found to have committed information blocking can lose a significant portion of annual Medicare payment increases, receive a zero score in the MIPS Promoting Interoperability performance category, or face exclusion from the Medicare Shared Savings Program. The full breakdown of each consequence is in the FAQ below. 

The scale of the problem is visible in enforcement data. Since information blocking rules took effect in April 2021 through June 2026, HHS received more than 2,000 complaints, OIG is actively hiring attorneys and conducting investigations, and some organizations have already acknowledged noncompliance. 

Struggling to comply and choosing not to comply are different problems with different solutions. Knowing which category your organization falls into is the first step.

What Happened After the Change Healthcare Breach

Interoperability became an even bigger challenge following the Change Healthcare breach in 2024, the largest healthcare data breach ever recorded, affecting approximately 190 million individuals. A March 2024 survey by the American Hospital Association found that 74% of nearly 1,000 hospitals reported direct patient care impacts, including delays in authorizations for medically necessary care, and 94% reported financial disruption.

Many organizations responded by severely restricting data connections and tightening access controls, an understandable reaction to security concerns, but one that compounded the clinical problem. Staff ended up with even less information than they needed when it mattered most.

This is the interoperability paradox: the same connectivity that enables better care is the connectivity that expands your attack surface. The answer is not to choose between security and access. It is to design systems that deliver both simultaneously, using modern identity management, API-level authentication, and audit logging rather than broad restrictions.

FHIR Is Not the Finish Line. It Is the Starting Line.

If you have heard one thing about healthcare interoperability in the last few years, it is probably that FHIR is the solution. FHIR, Fast Healthcare Interoperability Resources, is an HL7 standard that uses modern web APIs to exchange health data. It is far easier to implement than its predecessors, and it carries significant regulatory backing.

In the U.S., ONC’s HTI-1 Final Rule requires support for the U.S. Core Data for Interoperability version 3 via FHIR APIs, with FHIR R4 as the required version for all ONC-certified EHR systems. Every major EHR vendor, Epic, Oracle Health, athenahealth, and others, now exposes FHIR R4 endpoints. So the pipes exist.

But having a FHIR endpoint does not mean your data is actually interoperable.

A 2024 scoping review noted persistent challenges in FHIR adoption, including terminology mapping issues, inconsistent implementations, and limited scalability. Real-world platforms often require significant manual mapping and data transformation, hindering usability by domain experts despite promising tooling for EHR integration. 

For practice administrators, the practical question is not “do we have a FHIR endpoint?”

It is “what data flows through it, in what standard terminology, and does the receiving system actually do something with it?”

The answer to that last question is where most implementations fall short.

The Vocabulary Problem You Must be Aware of 

If you have ever seen a patient’s medication list imported from an outside system and noticed that drug names look different from what your system uses, you have seen a vocabulary problem in action.

Healthcare data relies on several standardized coding systems to carry shared meaning:

• SNOMED CT for clinical terms, diagnoses, and findings
• LOINC for laboratory tests and observations
• RxNorm for medications and drug interactions
• ICD-10-CM for diagnosis codes used in billing and documentation
• CPT for procedure codes

When a referring physician’s system sends a patient’s diabetes diagnosis using one SNOMED CT concept, and your system maps it to a slightly different one, the clinical record quietly becomes inaccurate. No error message. No alert. Just a wrong record.

This vocabulary alignment problem is invisible until it causes harm. As one nurse described in the KLAS Arch Collaborative report: “The EHR allows us to pull in information for some patients, but there are times when the information is not accurate. EHRs can interpret information for things like immunizations differently, so if we are not careful, we can incorrectly document that a patient has received a vaccine.” 

That is a patient safety issue, not just an IT inconvenience. And it is one a FHIR endpoint alone does not solve.

TEFCA: The National Framework Your Practice Needs to Know About

TEFCA, the Trusted Exchange Framework and Common Agreement, is the federal government’s most significant structural intervention in healthcare interoperability. It became operational in December 2023 and has grown at a pace few anticipated.

Before TEFCA, if your hospital was connected to a regional health information exchange, you could share data with members of that exchange. But if a patient’s records lived in a different network in a different state, you were often out of luck. Any organization, or patient, connected to one QHIN can now exchange data with TEFCA participants connected to any other QHIN, with QHINs routing queries and responses among all participating healthcare constituents. 

The growth has been substantial. According to ONC’s official TEFCA fact sheet, over 71,000 sites and organizations are now participating in TEFCA through 11 designated QHINs. The Sequoia Project’s TEFCA homepage reports that more than 1.2 billion documents have been shared through the network since go-live in December 2023. HealthITONC TEFCA RCE

For a small practice, joining a QHIN is not always a direct process. More commonly, your EHR vendor or regional health information exchange connects to a QHIN on your behalf, and you benefit through that connection. Whether that connection is active and what data flows through it are worth confirming, the practical audit steps are in the section below.

TEFCA enables query-based data exchange, meaning health information is shared only when a specific request is made. A healthcare provider can request, or query, a patient’s records from a connected organization, which retrieves them on demand. 

For healthcare providers, participation in TEFCA is currently voluntary. That status is shifting fast: a 2025 Black Book Research survey found 44% of US health IT buyers actively onboarding with QHINs, and the network’s value compounds with every organization that joins.

The CMS Prior Authorization Rule: Where Things Stand Right Now

The CMS Interoperability and Prior Authorization Final Rule, CMS-0057-F, is the most concrete interoperability deadline actively hitting provider workflows right now.

The operational requirements that took effect January 1, 2026, require impacted payers, including Medicare Advantage, Medicaid managed care, CHIP, and ACA marketplace plans, to:

• Return prior authorization decisions within 72 hours for urgent requests
• Return decisions within seven calendar days for standard requests
• Provide a specific reason for every denial
• Publish prior authorization metrics publicly each year, with first-cycle data for calendar year 2025 due March 31, 2026

For providers, the operational changes are in force now. Ask your EHR vendor specifically whether they support the Da Vinci Coverage Requirements Discovery (CRD) and Documentation Templates and Rules (DTR) implementation guides. These FHIR-based workflows enable real-time prior authorization checking at the point of care, before the patient leaves the room.

The Four Problems Driving Day-to-Day Interoperability Failure

Most interoperability problems at the practice level reduce to four root causes. Each has a different fix.

#1. Patient identity matching

Before any records can be exchanged, both systems must agree on which patient you are asking about. The US has no universal patient identifier. Systems rely on matching algorithms using name, date of birth, address, and other demographic data. When a patient has a common name, has moved, or has a data entry error in one system, the match fails.

Research by Audacious Inquiry, a contractor that has conducted research for ONC, found match rates as low as 50% even between organizations that share the same EHR vendor, because of the variability in technology and processes. Pew Charitable Trusts On the patient safety side, ECRI’s Patient Safety Organization reviewed more than 7,600 wrong-patient events from 181 healthcare organizations and found that about 9% led to temporary or permanent harm or even death.

The bipartisan MATCH IT Act of 2025, reintroduced in March 2025 by Reps. Mike Kelly and Bill Foster, would direct HHS to establish a standard definition and data set for patient matching and build accuracy standards into the ONC health IT certification program. The bill has not yet been enacted, but its bipartisan support reflects how seriously Congress now treats this problem.

#2. Training gaps that make working technology invisible

The KLAS Arch Collaborative found that 72% of clinicians who feel they received sufficient training on accessing outside data report their EHR has expected external integration. Among clinicians who feel they are not well trained in accessing external data, only 26% agreed that their EHR has expected integration. TechTarget

#3. Inconsistent implementation of standards

Even successfully delivered data is not necessarily usable data. A practice may receive an outside record that clears every technical hurdle only to find that the receiving EHR cannot trigger clinical decision support on it, reconcile it to the local medication list, or surface it at the right point in the workflow,  because the sending vendor and receiving vendor made different choices about optional FHIR elements, code system bindings, and resource structure.

#4. Unstructured documentation that poisons downstream exchange

When a physician dictates a note in free text without structured data fields, the information exists in the chart but cannot be read, queried, or exchanged by any downstream system. It is a PDF in a FHIR world. The consequences compound quickly: clinical decision support cannot fire on data it cannot read, population health tools cannot flag care gaps from unstructured notes, and AI summarization tools produce unreliable outputs when the underlying records are uncodified narrative.

Where AI Fits Into Interoperability (And Where It Does Not)

Artificial intelligence is showing up in interoperability conversations with increasing frequency. Some applications are useful. The enthusiasm often runs ahead of the evidence, so it is worth being precise.

The data quality problem makes all of this harder than it should be. According to the 2025 Healthcare Data Quality Report by Clinical Architecture, 82% of healthcare organizations are concerned about the quality of data from external sources, and 68% rated the quality of their own patient data as mixed or poor. AI tools operating on poor-quality source data produce poor-quality outputs, regardless of how sophisticated the model is. Yahoo Finance

Machine learning models can help identify and resolve data inconsistencies, automatically mapping different coding systems and clinical terminologies, offering intelligent insights that improve clinical decision-making and patient care coordination. But those gains are additive on top of, not a substitute for, foundational technical interoperability.

Practical Steps for US Healthcare Organizations Right Now

• Audit your current data exchange capabilities. Ask your EHR vendor to show you exactly what data flows through their FHIR API, which QHIN or health information exchange you are connected to, and which exchange purposes are enabled. Many organizations discover their FHIR endpoint is active but data sharing is limited to minimum compliance thresholds.
• Verify information blocking compliance. Make sure your organization does not have policies or contracts that restrict patient data access beyond the regulatory exceptions defined by ONC. Common violations include requiring patients to come in person to request records, charging fees beyond what is permitted, and placing contractual barriers on data export.
• Conduct a duplicate patient record audit. A crowdsourced poll of enterprise master patient index users found that, before implementing an EMPI tool, an average 18% of an organization’s patient records are duplicates. Survey respondents estimated that 33% of all denied claims result from inaccurate patient identification or information. Your duplicate record rate is a direct proxy for your external match rate. Organizations that clean their internal patient data first achieve dramatically better results when querying external networks.
• Invest in structured documentation upstream. The quality of data your practice sends through FHIR APIs is only as good as the structure of the data going in. Templated workflows, structured order entry, and AI-assisted documentation tools all improve the machine-readability of your records, which directly determines how useful your data is when it reaches another system.
• Connect interoperability to your value-based care contracts. According to the NAACOS and Innovaccer “State and Science of Value-Based Care 2025” report, based on a nationwide survey of 168 healthcare leaders across 142 organizations, 75% of respondents cite lack of interoperability as a barrier to value-based care adoption, making it the third most commonly cited obstacle after financial risk and provider readiness. If your organization is in any value-based contract, the ability to track patient outcomes, share care gap data with payers, and coordinate across care settings depends entirely on your data exchange infrastructure. 
• Work with a vendor that treats interoperability as a core function, not an add-on. FHIR implementation depth, QHIN connectivity, patient identity matching capability, and data quality controls are now baseline evaluation criteria for any health IT platform.

What OmniMD Brings to This Problem

Navigating this landscape, from FHIR implementation to TEFCA participation to prior authorization workflows, requires a platform built for it from the ground up.

If your practice is losing time, revenue, or clinical accuracy because your systems do not talk to each other, OmniMD is designed to close those gaps. Contact us to see how the platform connects your clinical and billing workflows with the broader health data ecosystem your patients are already part of.

Frequently Asked Questions

What is the difference between HL7 and FHIR?

HL7 is the standards organization that produces healthcare data exchange specifications. FHIR is one of those specifications, the most recent and most widely adopted for API-based exchange. Earlier HL7 standards like HL7 v2 and C-CDA are still in use in many hospital systems for lab results, ADT notifications, and clinical document exchange. FHIR uses modern REST-based APIs with JSON or XML formats, which makes it far easier to build applications on top of. The two coexist in most health systems today.

Does every US EHR now have a FHIR API?

Every ONC-certified EHR is required to support FHIR R4 APIs under the 21st Century Cures Act and the HTI-1 Final Rule. However, supporting an API is not the same as actively exchanging meaningful data through it. The quality and completeness of FHIR implementations vary significantly across vendors, and the data elements exposed may be limited to minimum compliance requirements rather than the full clinical picture.

What is TEFCA and do small practices need to care about it?

TEFCA is the national framework for health data exchange, built on a network of Qualified Health Information Networks. Small practices typically benefit through their EHR vendor’s or regional HIE’s QHIN connection rather than by joining directly. Ask your EHR vendor whether they are connected to a QHIN and what data you can access through that connection.

What counts as information blocking?

Under ONC rules, information blocking includes any practice that interferes with access, exchange, or use of electronic health information that is not covered by a defined regulatory exception. Common examples include charging unreasonable fees for record access, requiring in-person requests for records that could be handled electronically, and placing contractual barriers on data export.

What are the actual consequences for a provider that commits information blocking?

Providers do not face a per-violation fine the way health IT developers do. Eligible hospitals found to have committed information blocking lose meaningful EHR user status, which means they lose 75% of their annual market basket payment increase. Clinicians in MIPS receive a score of zero in the Promoting Interoperability performance category. Providers in ACOs may be ineligible to participate in the Medicare Shared Savings Program for at least one year.

What is the realistic timeline for full US healthcare interoperability?

There is no honest single-date answer. The regulatory infrastructure, FHIR mandates, TEFCA, information blocking enforcement, and the prior authorization API deadlines create a framework pushing the industry toward meaningful exchange over the next two to five years. The organizations that benefit earliest are the ones investing now in technical, training, and governance work rather than waiting for the industry to arrive at their door.

What happens to practices that cannot submit prior authorizations electronically?

Practices still relying on fax and phone for prior authorization will not face immediate regulatory penalties under CMS-0057-F, which targets payers rather than providers. But with the payer-side operational requirements already in force since January 2026 and the FHIR API deadline hitting January 2027, practices without electronic submission capability will experience growing turnaround time gaps, more administrative burden, and less visibility into authorization status compared with peers who have made the transition.