An EMR is a digital chart that lives inside one practice. An EHR is a patient record built to move between practices, labs, hospitals, and specialists. The difference is interoperability, and it is not only a matter of vocabulary: federal programs require certified EHR technology, so the distinction decides whether your practice can report for Medicare's Promoting Interoperability category.
Key takeaways
| EMR | EHR | |
|---|---|---|
| Stands for | Electronic Medical Record | Electronic Health Record |
| Scope | One practice or organization | The patient's full care journey |
| Built to share data? | No. Designed as an internal record. | Yes. Designed around exchange. |
| Who can see it | Clinicians and staff at that practice | Authorized providers across settings |
| Getting records out | Print, fax, or PDF export | Standards-based electronic exchange |
| Patient access | Usually limited or none | Patient portal, and an API |
| Typically certified? | Not necessarily | Often ONC-certified (CEHRT) |
| Qualifies for MIPS Promoting Interoperability | Generally no | Yes, if certified |
| Think of it as | A digital filing cabinet | A connected record |
An electronic medical record is the digital version of the paper chart that used to sit in a filing cabinet behind the front desk. ONC's own description is that an EMR is a digital version of the paper charts in a doctor's office: notes, orders, and results collected by the clinicians in one office, for use by the clinicians in that same office.
The gains over paper are immediate. Nothing is illegible, nothing gets physically lost, and for the first time the practice can ask a question of its own records — find every diabetic patient overdue for an A1c — and get an answer in seconds instead of an afternoon.
What an EMR cannot do is let the data leave. When a patient sees a specialist, moves out of state, or turns up in an emergency department, the EMR's answer is a records request. Somebody prints the chart, faxes it, or exports a PDF, and somebody at the other end re-keys whatever survived the trip. ONC has made the point bluntly: for care coordination purposes, an EMR is not much better than paper. The chart is legible now. It is still stuck in one building.
An EHR is built on the opposite assumption: the record should follow the patient, rather than the patient chasing the record.
It holds the same clinical documentation an EMR does, and adds the connective tissue that lets it exchange information with laboratories, pharmacies, imaging centers, hospitals, health information exchanges, and other practices. A patient walks in, and the EHR has already pulled the cardiology consult, the hospital discharge summary, and the medication history from the urgent care they visited on vacation.
That exchange capability is the entire distinction. Everything else in this article follows from it.
Data portability. An EMR exports; an EHR exchanges. Exporting produces a document a human has to read and re-enter. Exchanging produces structured data that populates the receiving chart on its own. That difference is one person's job.
Medication safety. This is the one that keeps clinicians up at night, and it deserves more than a line. An isolated EMR only knows what your practice prescribed. If a patient's psychiatrist started them on something six weeks ago and your system has no connection to an e-prescribing network, your interaction checking is running on incomplete data and giving you a clean result anyway. That is worse than no check, because it reads as reassurance. EHRs connected to networks like Surescripts pull external medication history into the chart, so the interaction alert fires on what the patient is actually taking rather than on what you happen to have written down.
Care coordination. With an EMR you know what happened in your office. With an EHR you can see the specialist consult and the hospital admission the patient forgot to mention, which is most of them.
Patient access. EHRs include a patient portal, and under current federal rules they are expected to make records available to patients through an API without special effort or special charge. A lot of EMRs have no patient-facing layer at all.
Certification. EHRs are frequently certified under the ONC Health IT Certification Program. EMRs typically are not. This carries more weight than it sounds like, and the next section is about why.
Reporting. Because an EHR aggregates data from across the care continuum, its reports describe what happened to the patient. EMR reports describe what happened in your building. For quality measures tied to outcomes, those are not the same number.
Federal incentive and reporting programs do not require "an electronic system." They require certified electronic health record technology, usually shortened to CEHRT. To participate in the Merit-based Incentive Payment System, and specifically in its Promoting Interoperability performance category, a practice needs software certified under the ONC Health IT Certification Program by an authorized testing body.
An uncertified EMR generally does not qualify, and that is not a paperwork detail. A practice running one can be shut out of a performance category that affects its Medicare payment adjustment, no matter how well the software handles a Tuesday morning clinic.
The policy history explains why the gap exists at all. The HITECH Act of 2009 created incentive payments, and from 2011 those payments were tied specifically to certified technology. Vendors who built for certification are the ones who are certified now. Vendors who did not, largely are not, and fifteen years of divergence is difficult to close with a software update.
There is also the information-blocking side. Rules established under the 21st Century Cures Act set expectations around making electronic health information available when it is requested. A system with no real mechanism for releasing data electronically puts a practice in an uncomfortable position the first time a patient or another provider asks.
So "is my system an EMR or an EHR?" is really three questions. Can it exchange data? Is it certified? Does that certification cover what my reporting obligations actually require?
Vendor marketing will not settle this. Plenty of products are sold as EHRs and behave like isolated EMRs, and a handful of genuinely capable systems still carry the older name for historical reasons. Four checks will tell you what you are really running.
1. Look it up on the CHPL. The Certified Health IT Product List is a public federal database of every certified health IT product, searchable by developer and product name. If your system is not there, it is not certified, whatever the sales deck implied. Check the certification edition while you are in there. The current one is the 2015 Edition Cures Update.
2. Ask how a hospital discharge summary reaches your chart. If the answer involves a fax machine, a scanner, or a person typing, you have an EMR with a document management feature. If it arrives as structured data through an exchange, you have an EHR.
3. Ask to see the API documentation. Certified systems are expected to expose patient data through a standards-based API, which in practice means FHIR. Ask whether it exists and whether it is published publicly. A vendor who has to check is telling you something.
4. Ask which networks you are connected to. Health information exchanges, e-prescribing networks, Carequality, TEFCA-designated networks. A system connected to none of these is not exchanging much of anything, regardless of what the product is called.
A third acronym shows up in the same conversations. A personal health record is maintained by the patient rather than the practice. It may pull data from an EHR, and it often holds things no clinical system does: over-the-counter medications, home blood pressure readings, wearable data, notes on how someone actually felt between visits.
Put simply, an EMR belongs to the practice, an EHR is shared across the practice's care partners, and a PHR belongs to the patient. A patient portal sits on the boundary. It is the EHR's window into the patient's hands, which is a different thing from a record the patient owns and controls.
Epic is an EHR. So are Oracle Health (formerly Cerner), athenahealth, eClinicalWorks, NextGen, and the other major certified platforms. All are certified health IT and all support standards-based exchange.
The confusion comes from inside hospitals, where staff say "the EMR" as shorthand for whatever they chart in. That is ordinary usage and there is no reason to correct it over coffee. It matters when you are evaluating software, signing a contract, or checking whether you can report for a federal program.
More than it does for a large one.
A twenty-provider multi-specialty group refers internally, so a good share of a patient's care already happens inside one system. A three-provider independent practice sends patients out constantly, and every handoff is a place where an isolated record costs staff hours and creates clinical risk.
There is also a market reality worth knowing. ONC's 2024 data brief found that nine in ten office-based physicians were using a certified EHR, up from four in ten using any EHR in 2008. But the same analysis found that vendor market share varies sharply by practice size: large practices cluster on a handful of market-leading systems, while small practices are spread across a much wider field of vendors. The long tail of that field is where uncertified and thinly-certified products survive. If you are a small independent practice, you are statistically more likely than a large group to be running something that will not satisfy a Promoting Interoperability requirement, and less likely to have someone on staff whose job is to notice.
We build an EHR. Rather than ask you to take the word at face value, here is what it means in our case and how to check it.
PCIS GOLD is certified under the ONC Health IT Certification Program to the 2015 Edition Cures Update, tested by Drummond Group as an ONC-Authorized Certification Body. We are a Surescripts-certified e-prescribing partner, and we publish a FHIR API so patient data is reachable through a documented standard rather than an export routine somebody runs on Fridays.
None of that needs to be taken on trust. Search "PCIS GOLD" on the CHPL and read the record. Then do exactly the same for every other vendor on your shortlist, including the ones with larger marketing budgets than ours. A vendor who cannot produce a CHPL entry has answered the EHR vs EMR question for you, and saved you a demo.
If you want to see the exchange working in a live chart rather than on a slide, book a demo. We will open a real referral and follow it through.
No. An EMR is a digital chart confined to one practice. An EHR is built to share patient information across practices, labs, hospitals, and specialists. In conversation people use the terms interchangeably, and that is fine. For contracts, certification, and federal reporting the distinction is real.
Electronic Health Record. The word "health" is deliberate. It signals a record covering a patient's overall health across providers, rather than the medical encounters at one office.
Electronic Medical Record. The term came first, back when these systems were built purely for diagnosis and treatment inside a single practice.
For almost any practice operating today, an EHR, because patients move between providers and the record needs to move with them. The exception is a fully self-contained practice that never refers out and does not participate in Medicare reporting, which describes very few practices.
Only if it is certified health IT. Search your product on the ONC Certified Health IT Product List. If it is not listed, it is not CEHRT, and it generally will not satisfy the Promoting Interoperability requirements.
Sometimes, if the vendor has built exchange capability and gone through certification. Often not, because interoperability is an architectural decision rather than a feature that bolts on afterwards. Ask your vendor for their CHPL listing and their certification roadmap, and treat a vague answer as an answer.
An EHR handles clinical documentation. A practice management system handles the business side: scheduling, registration, eligibility, claims, and collections. Many vendors, PCIS GOLD included, offer both on one platform so clinical and financial data share a record rather than a nightly file transfer.
Longer than most practices assume, and the answer comes from three places at once. HIPAA requires covered entities to retain required documentation, including policies and certain compliance records, for six years from creation or last effective date, though HIPAA does not set a retention period for the medical record itself. CMS requires Medicare providers to keep patient records for a minimum of five years, and managed care program providers for ten. State medical board rules sit on top of both and commonly run five to ten years for adults, with records for minors typically held until some period past the age of majority. Because the longest applicable requirement governs, most practices default to ten years for adults. None of this changes based on whether your system is called an EMR or an EHR. What does change is how painful it is to produce those records when somebody asks.
Certification details reflect the ONC Health IT Certification Program, 2015 Edition Cures Update. Retention requirements summarised here are general; confirm your state board's rules and payer agreements.