If all the ways to improve healthcare were written down, nobody knows how long the list would be. But somewhere on it, probably high up, you’d find EHR interoperability. If clinicians could reliably access complete, accurate patient records across systems, we’d see fewer delays, fewer repeat tests, and a lot less fragmented care. But like many well-known wishes in healthcare, this one still hasn’t come true. In this article, we’ll walk you through five major challenges of EHR interoperability that we’ve seen in more than nine years of healthcare software development.
Why EHR interoperability is an illusion in healthcare
The 21st Century Cures Act, passed in the U.S. in 2016, defined interoperability as the ability to access, exchange, and use “all electronically accessible health information… without special effort.” The law charged the Office of the National Coordinator for Health Information Technology (ONC) with advancing these goals, including enforcement against information blocking.
What is the main advantage of interoperability within an electronic health records (EHR) system? It’s supposed to ensure that clinicians can access complete, timely medical records from across healthcare settings, thereby improving patient care, reducing duplications, and enabling better coordination.
Nearly all certified EHR vendors now support FHIR to some degree. Most health information systems can technically send, receive, and integrate health data. But in practice, interoperability in healthcare remains elusive because many EHR vendors still fail to deliver usable information when and where it’s needed. External patient records often arrive too late, buried in unreadable formats, or stripped of essential context.
Clinicians know this. That’s why they still re-enter medications by hand, call external clinics, or rely on PDF printouts. Health information exchange is possible, but it’s rarely seamless.
Meanwhile, more than 500 million Continuity of Care Documents (C-CDAs) are exchanged annually, and 7 billion lab tests are run in the U.S. each year. Without clear mappings, normalization, or clinician training, these volumes only worsen signal-to-noise problems.
Clearly, so far, we've treated EHR interoperability as a box to check, not an operational function to maintain. Until health information technology is aligned with how care is actually delivered, the benefits of EHR interoperability will remain mostly theoretical, promised by healthcare information policy but undermined by reality.
5 EHR interoperability challenges
Challenge 1: Standards ≠ standardization
In theory, FHIR (Fast Healthcare Interoperability Resources) was designed to create a shared language for healthcare data exchange. But in our experience working across multiple health systems, we’ve found that FHIR isn’t FHIR everywhere.
The standard exists, but how it’s implemented varies widely between EHR vendors, and even between organizations using the same vendor. That’s one of the biggest challenges in healthcare interoperability: everyone’s speaking “FHIR,” but not in the same dialect. Here's why:
a. Vendors support different versions of EHR standards
Vendors support different FHIR versions (such as DSTU2, STU3, R4, and now R5) at the same time. For example, Epic supports up to R4, but implementations differ by resource. Its AllergyIntolerance resource, for instance, supports only limited interactions in older versions. Cerner supports DSTU2 and R4, but is phasing out the former. This version fragmentation creates compatibility gaps when trying to exchange health data across platforms.
Even when versions align, vendors expose different FHIR resources with different capabilities. Some support only read access. Others limit which fields can be queried.
b. Differences in system performance
In situations where systems support FHIR, they don’t all perform the same. Some respond faster while some deliver more complete data. Others can’t handle large requests without slowing down or dropping fields.
A 2024 evaluation of the SMART/HL7 Bulk FHIR Access API found just how uneven that can be: Epic sites exported data at 1,555 to 2,500 resources per minute, while a custom-built HIE hit 12,000. If data can’t move fast enough to keep up with care, it doesn’t matter how “interoperable” the system claims to be.
Interoperable EHR systems aren’t just expected to connect. They need to move data at scale, respond quickly under load, and deliver complete records when they’re needed most.
c. The need for customization
Many of the FHIR APIs require customization to be usable. One study found that only 51% of data elements needed across clinical trials were fully supported by FHIR. The rest required custom development. In practice, this means every project begins with mapping mismatched data fields, sometimes down to individual value sets like how “gender” is coded.
To deal with these gaps, healthcare organizations often use middleware like Rhapsody, Mirth Connect, or PilotFish. These tools help normalize FHIR outputs across vendors. However, their very necessity highlights the lack of true interoperability among EHR systems.
d. Differences in data storage methods
In some cases, different sites using the same EHR software don’t even store data the same way. A study of clinical decision support experts found that this kind of implementation variation (even within identical EHR platforms) was a major constraint. Data might technically conform to FHIR, but fields are stored, labeled, or interpreted differently, making it harder for (clinical support decision) CDS tools to operate reliably.
Solution: These structural interoperability issues affect everything from patient care to public health reporting. And they won’t be solved by checking the "FHIR-supported" box. Interoperability requires more than standards. It requires consistency, transparency, and shared assumptions across the healthcare ecosystem. That’s what’s missing.
Challenge 2: Semantic mismatches
In our experience, one of the most overlooked ehr interoperability challenges isn’t whether data arrives. It’s whether it still means what it’s supposed to. Semantic interoperability is the ability of different systems to exchange data and interpret it consistently. Without it, systems may technically communicate, but the message gets garbled. That confusion introduces risk into clinical decision-making, analytics, and patient care.
Here are some examples of how mismatched data across EHR systems prevents true interoperability.
a. Inconsistent diagnosis coding across systems
Diagnoses are a prime example of how meaning gets lost early. In one study, only 54% of ICD-10 codes entered by clinicians were appropriate to the clinical scenario, and 26% were completely missing. If that’s the variation at the point of entry, imagine what happens when data moves between systems with different structures and rules.
b. Medication errors from mismatched naming and units
Medication data isn’t immune either. After implementing a new electronic medical record, one hospital saw wrong-dose errors triple and wrong-drug errors nearly triple as well. In addition to being input problems, these are downstream consequences of inconsistent drug naming, dosage units, and coding across platforms.
c. Lab data fails without proper LOINC mapping
Even structured lab data can be unreliable. Health Information Exchanges report that nearly half of incoming lab results rely on local, proprietary codes, data that must be mapped to the national LOINC (Logical Observation Identifiers Names and Codes) standard before it’s usable. But this process is far from seamless. Between 6% and 19% of lab tests can’t be mapped due to missing or inaccurate details, and manual mapping carries error rates of up to 20%.
d. FHIR fields misaligned across implementations
We've seen firsthand how FHIR data mappings fail when different systems use the same resource, say, MedicationRequest, but define critical fields differently. In one case, gender values were coded as M/F in one system and full SNOMED concepts in another. The mismatch wasn’t flagged. It just broke downstream logic. This variability affects everything from clinical decision support to public health reporting.
Solution: The point is this: interoperability isn’t just about pipelines. It's about trust in what flows through them. Without aligned semantics across different EHR systems, we risk building faster connections to flawed interpretations. And no amount of speed improves patient safety and health if the data itself can’t be trusted.
Challenge 3: Vendor lock-in and gatekeeping
One of the less visible (but most frustrating) barriers to EHR interoperability is how vendors control access to their systems. In our work with healthcare organizations, we’ve seen how these roadblocks slow down integration, drive up costs, and limit the potential of even the best-designed solutions in the following ways:
a. High pricing
Some EHR vendors charge steep fees just to access their APIs, i.e., the tools that allow outside apps to connect and share data. Cerner has charged up to $15,000 per year for apps, while Epic has billed thousands just to get started. These pricing models turn interoperability into a pay-to-play game, discouraging smaller vendors from building tools that could improve patient care.
b. API limitations
Other vendors limit what the APIs can actually do. Many only allow read-only access, so third-party apps can pull EHR data but can’t push anything back, like notes, orders, or updates. That kind of one-way access is often useless in real clinical workflows.
On top of that, vendors may throttle data volume, meaning you can only make a limited number of requests per minute. That might work for demos, but not for real-world systems that need to handle thousands of records at once, especially in large healthcare facilities.
c. Vendors still obstruct integration despite being certified
Even certified systems can fall short. Despite meeting interoperability standards, some vendors still provide incomplete or poorly documented APIs. In fact, 40% of digital health companies say they can’t get the data they actually need. This creates a gap between certification and real functionality.
We’ve seen these constraints play out directly. One client needed to unify access across multiple EHR systems for patient scheduling. But each vendor had different license terms, different access rules, and different integration delays. What should’ve been a single connection turned into a patchwork of exceptions, emails, and workarounds.
These interoperability challenges often stem from business choices, not technical ones. In the healthcare industry, vendor lock-in and restrictive data policies limit the electronic exchange of patient data. This slows progress on foundational interoperability and forces healthcare professionals to work around gaps using manual processes. These patterns persist because many vendors see interoperability as a threat to their business model, not a baseline requirement for modern healthcare.
Solution: To overcome these challenges, the healthcare ecosystem needs stronger healthcare interoperability standards, more transparent licensing, and greater accountability. Interoperability helps, but only if it’s usable at a large scale, without friction, and without surprise fees. Until that happens, the benefits of EHR interoperability will remain out of reach for many.
Challenge 4: Operational and workflow misalignment
Technology alone doesn’t deliver interoperability. In our experience, even with the right tools in place, the day-to-day clinical workflows often don’t support the seamless use of external data. That gap between capability and usability is where many EHR interoperability solutions fall short. The examples below show how breakdowns in workflow undermine interoperability at the point of care.
a. Clinicians complain of insufficient integration
Clinicians don’t just need access. They need usable data at the point of care. But only 44% say their systems actually integrate outside information as expected. Even when external data is available, only 42% of clinicians routinely use it, often because it’s buried, delayed, or not trusted.
b. Poor data usability leads to ignored alerts
Inconsistent integration creates real-world consequences. Between 54% and 91% of clinical decision support alerts are ignored, frequently due to alert fatigue caused by irrelevant or duplicative information. Poor data interoperability leads to mistrust, which leads to inaction.
c. Duplicate notes clog the medical record
Redundant documentation is another symptom. In a review of over 100 million clinical notes, half the content was duplicated. With clinicians relying heavily on copy-paste, patient records become bloated and harder to work through, especially when pulling in data from multiple systems.
d. When systems fail, fax machines step in
When data doesn’t flow where it needs to, the fallback is manual workarounds. As of 2021, 70% of hospitals still used fax machines to exchange records. In rural and smaller hospitals, nearly half rely primarily on fax or mail to respond to electronic requests for patient information. Despite widespread electronic health record adoption, this remains a major bottleneck to efficient healthcare.
e. Poor real-time data integration leaves clinicians guessing
Timeliness is also a barrier. While 62% of hospitals say external data is available at the point of care, that means 38% still can’t count on it. Without real-time data integration, healthcare professionals are left making decisions without a full picture.
f. Medication lists still rely on manual patchwork
In many systems, even reconciling a patient’s medication list still requires a manual review of fragmented sources. That kind of task, which is common across healthcare facilities, drains time, increases risk, and shows how fragile the connection is between systems.
Solution: These challenges of healthcare interoperability aren’t new. But they remain persistent because they cut across technology, policy, and clinical practice. To improve EHR interoperability, healthcare organizations need more than tools. They need alignment. Alignment between systems and workflows. Between data and trust. And between what the technology can do and what the people using it need it to do.
Challenge 5: Fragmented consent and privacy models
Even with technical connections in place, sharing patient data across healthcare systems hits a wall when consent models don’t align. From our experience, this isn’t just a policy problem, but it’s one of the toughest operational barriers to overcome in achieving an interoperable EHR.
The following examples show how fragmented consent and privacy rules create uncertainty, stall data sharing, and expose patients and providers to real risk.
a. Different states have different consent laws
Consent laws vary widely across jurisdictions. As of the latest nationwide scan, only 7 U.S. states follow an “opt-in” model where patient permission is required before sharing health records. Fifteen states default to “opt-out.” The remaining 29 either have hybrid or unclear policies. For organizations working across state lines, or even across systems within a state, this creates constant uncertainty about what data can be shared, when, and with whom.
b. Sensitive data adds layers of restriction
Privacy rules around specific types of personal health data raise even more complexity. For example, regulations like 42 CFR Part 2 restrict sharing of substance use treatment records unless patients give explicit consent. Also, reproductive and mental health data are now subject to increasingly strict protections in states like California and Washington.
As a result, what one EHR system considers routine information might be off-limits in another due to health insurance portability and accountability rules and state-specific laws.
c. EHR systems can't cater to dynamic consent preferences
Most systems can’t keep up with dynamic consent preferences. While patients increasingly expect control over how their data is shared, by time, data type, or recipient, many EHR systems simply can’t handle that level of granularity. A major study found that even when patients wanted to restrict specific data, the technical infrastructure didn’t exist. That gap between patient expectation and system capability puts both privacy and patient experience at risk.
d. Without consent logic, over-blocking or oversharing takes over
When consent logic isn’t built into the system, organizations face blunt choices. We’ve seen sites default to over-blocking (withholding critical data) out of caution, or oversharing because they lack finer-grained controls. In either case, patient trust suffers.
In one documented breach, a mental health clinic in Finland accidentally exposed tens of thousands of therapy notes due to flawed access safeguards. Closer to home, hospitals have suspended digital tools after realizing sensitive data was left accessible online.
Consent fragmentation is still one of the most persistent barriers to interoperability and patient access across the healthcare ecosystem. As long as healthcare technology systems can’t support the complexity of dynamic consent and varying state laws, every healthcare organization will keep facing difficult trade-offs, between protecting privacy and enabling care, between compliance and usability.
Solution: Solving this requires more than better interfaces. It demands clear federal guidance on interoperability standards in healthcare and coordination among healthcare organizations to apply them consistently.
Why are these barriers to interoperability still there?
Despite advances in health information technology systems, the challenges of healthcare interoperability are rarely just technical. Based on our work across different healthcare settings, three persistent issues continue to undermine progress.
a. Providers aren't reaping the financial benefits of EHR interoperability
Interoperability helps healthcare providers deliver more coordinated care, but financial returns remain uncertain. Nearly two-thirds of initiatives fail to meet ROI expectations. Often, the organization investing in a solution (typically a provider) doesn’t reap the full benefit.
Payers may gain from reduced duplication or hospital readmissions, but providers absorb most of the cost. This disconnect discourages deeper investment, especially when leaders prioritize projects like revenue cycle optimization instead.
b. Regulation sets the floor, not the bar
Certification through the Office of the National Coordinator for Health Information Technology (ONC) focuses on baseline functionality. But over 60% of certified vendors skipped or underperformed on important usability tests. While there are penalties for information blocking under the Department of Health and Human Services, enforcement has been inconsistent, and the rules don’t demand meaningful real-world outcomes.
That gap between compliance and effectiveness leaves too many health information technology systems underperforming across the healthcare ecosystem.
c. Switching vendors is cost-prohibitive
Replacing an EHR is often a multi-year, multi-million-dollar project. One health system recently budgeted $160 million for a new EHR transition. With costs that steep, many providers stay locked into disparate systems that make integration harder. The result is a patchwork of platforms that struggle to communicate, despite health level interface standards, and leave patients without easy access to their health data.
Addressing these interoperability challenges in healthcare will require coordinated action: better incentive alignment, tougher enforcement, and clearer strategies for modernizing legacy infrastructure. Every healthcare stakeholder has a role to play because the cost of doing nothing is an increasingly fragmented patient experience.
