EHR interoperability is the ability of different electronic health record systems to exchange patient data and use that data meaningfully. When it works, your medical history, lab results, medications, and allergies can follow you from one doctor’s office to another, regardless of which software each provider uses. When it doesn’t, clinicians end up making decisions with incomplete information, and patients end up repeating the same tests or manually relaying their own medical history.
The Four Levels of Interoperability
The Healthcare Information and Management Systems Society (HIMSS) breaks interoperability into four progressively complex levels, each building on the one before it.
Foundational (Level 1) is the simplest: one system can send data and another system can receive it. Think of it as establishing a phone line between two buildings. The data arrives, but the receiving system doesn’t necessarily know how to organize or interpret it.
Structural (Level 2) adds a shared format. The data is packaged in a standardized way so the receiving system can recognize its structure, preserving the purpose and framework of the information. The phone line now carries a language both sides can parse.
Semantic (Level 3) is where meaning is preserved. A “blood glucose level” recorded in one system means exactly the same thing when it arrives in another. Without semantic interoperability, two systems might use different codes or terminology for the same condition, creating confusion or errors.
Organizational (Level 4) goes beyond technology entirely. It encompasses governance, legal agreements, consent policies, and shared workflows that allow institutions to trust and act on each other’s data. This is the level where a hospital system and an independent clinic not only can share records but have agreed on the rules for doing so.
How Data Actually Moves Between Systems
The technical backbone of modern health data exchange is a standard called FHIR (Fast Healthcare Interoperability Resources), developed by the standards organization Health Level 7. FHIR uses web-based tools called APIs, the same technology that lets apps on your phone pull data from remote servers, to move health information between systems quickly and securely.
At the core of FHIR are modular building blocks called “Resources.” Each Resource defines a specific type of clinical or administrative data, such as a medication list, an allergy record, or a lab result, along with its structure and relationships. Rather than requiring systems to exchange an entire patient file at once, FHIR lets them request and send discrete pieces of information as needed. This modular approach makes it far easier for developers to build apps that work across different EHR platforms.
What Data Gets Shared
In the United States, a standardized set called the United States Core Data for Interoperability (USCDI) defines which categories of health information must be exchangeable nationwide. The core data classes include allergies and intolerances, medications, lab results, immunizations, vital signs, clinical notes, procedures, problems (diagnoses), care team members, encounter information, patient demographics, and provenance (a record of where the data came from and who entered it).
The USCDI is updated regularly, with newer versions adding data elements as they become widely adopted. More recent additions include categories like adverse events, appointment information, nutrition assessments, and advance care plans. The goal is a continuously expanding floor: a minimum set of data that every certified EHR system must be able to send and receive.
The Law Behind the Push for Interoperability
The 21st Century Cures Act, signed into law in 2016 and implemented through a final rule by the Office of the National Coordinator for Health Information Technology (ONC), created two major mandates. First, it requires the healthcare industry to adopt standardized APIs so patients can securely access all of their electronic health information, structured and unstructured, at no cost, including through smartphone apps. Second, it established information blocking rules that make it illegal for healthcare providers, health IT developers, and health information networks to unreasonably prevent the access, exchange, or use of electronic health information.
The law does recognize nine exceptions where withholding data is permitted. These include situations where sharing could cause harm to a patient or another person, where privacy protections require it, where security would be compromised, or where fulfilling a request is genuinely infeasible. Other exceptions cover the manner in which data is provided, reasonable fees, licensing terms, participation in the national Trusted Exchange Framework (TEFCA), and temporary system downtime needed for maintenance or performance.
Where Adoption Stands Now
Electronic data exchange among U.S. hospitals has reached near-universal levels for certain types of reporting. In 2024, 97% of non-federal acute care hospitals reported immunization data electronically, 94% reported lab data, and 94% participated in syndromic surveillance (tracking patterns of illness in communities). Electronic case reporting, which was required in Medicare’s incentive program starting in 2022, jumped from 53% in 2021 to 84% in 2024.
These numbers reflect public health reporting specifically. Broader clinical data exchange between hospitals and outpatient providers remains less uniform. The ability to send data is widespread, but the ability to receive outside data and meaningfully integrate it into clinical workflows still varies significantly depending on the EHR vendor, the institution’s size, and its technical infrastructure.
Why It Matters for Patient Safety
A systematic review of 12 studies on interoperability’s impact in high-income countries found that EHR interoperability positively influenced medication safety, reduced patient safety events, and lowered costs. The effects on time savings and clinical workflow were more mixed.
Medication errors are a particularly clear example of what poor interoperability costs. One study tracked what happened when patients arrived at a hospital and their medication lists had to be assembled from multiple disconnected systems. Roughly 64% of patients had discrepancies in their hospital medication lists. Of 442 total discrepancies found among 111 participants, about 45% were medications listed that shouldn’t have been, 44% were current medications missing from the list entirely, and 11.5% were duplicates. Each of those errors represents a point where a clinician could prescribe the wrong drug, miss a dangerous interaction, or unknowingly discontinue a medication the patient needs.
An analysis of over 1.7 million patient safety event reports from the Pennsylvania Patient Safety Reporting System found that 8% of health IT-related safety reports were specifically tied to interoperability failures between the EHR and other health IT systems.
Barriers That Slow Progress
Cost is the most frequently cited obstacle. System upgrades, ongoing maintenance, and the technical work of mapping data between incompatible formats all require sustained investment. Smaller practices and rural hospitals face this most acutely, often lacking the IT staff or budget to implement and maintain interoperable systems. Limited internet infrastructure, insufficient hardware, and restrictive software licensing add to the burden.
Beyond money, organizational challenges are significant. Research based on key informant interviews with healthcare stakeholders found recurring themes: poor user interfaces that discourage clinicians from using data exchange tools, lack of leadership support for interoperability initiatives, and mismatches in technical capability between organizations that are supposed to be exchanging data. A large academic medical center and a small community practice may both use certified EHR systems, but their ability to actually connect those systems can differ enormously.
There’s also a strategic tension. Some health systems have historically viewed data as a competitive asset. Sharing patient records freely with rival hospitals can feel counterintuitive, even when federal law now requires it. Aligning interoperability efforts with value-based payment models, where providers are rewarded for outcomes rather than volume, has helped shift this calculus, but the cultural resistance hasn’t fully disappeared.
How AI Is Filling the Gaps
Artificial intelligence is increasingly being used to address the messiest parts of the interoperability problem: unstructured data. A large portion of what’s in an EHR isn’t neatly coded. Physician notes, discharge summaries, and patient narratives are written in free text, which traditional data exchange standards can’t easily move or interpret. Natural language processing tools can extract structured, usable information from these notes, turning a paragraph about a patient’s symptoms into coded data that another system can work with.
AI also plays a role in data cleaning and normalization. When records arrive from multiple sources, they often contain missing values, duplicates, and inconsistencies. Machine learning models can flag and correct these errors, standardize measurements to a common scale, and reconcile records that describe the same patient or condition in different ways. The net effect is data that’s more accurate and more comparable across systems, which is the practical definition of semantic interoperability.