How to Switch EHR Systems Without Losing a Year of Patient Records (2026 Guide)

You don’t hate your EHR because you read a bad review. You hate it because you’re charting at midnight, your billing keeps getting denied, and every real complaint comes back as a “feature request” that ships in 2027.

So why are you still on it? Probably the same reason most practices are: the migration scares you more than the system does. You’ve heard the horror stories — practices that switched and lost two years of charts, billing chaos that took months to untangle, staff revolt at go-live. The cost of staying stuck is real (1.4 unpaid hours per day on after-hours charting, per Stanford Medicine). The cost of switching badly is also real. Both are true.

This guide is how to switch EHR systems without losing the records, the revenue, or your staff. The replacement market alone is now $9 billion (Black Book Research, January 2025) — a lot of practices figured this out, and most of the ones who got burned made the same handful of mistakes. The fixes are mechanical. Read how it works if you want the 60-second version. The longer version is below.

Your data is yours — and the law is finally on your side

Before you talk to any vendor — old or new — understand that you are not as locked in as your current vendor wants you to think.

The ONC’s information-blocking rule (effective April 2021, enforcement penalties strengthened through 2024 under the 21st Century Cures Act) prohibits health IT vendors from “blocking” your access to your own electronic health information. That includes patient records you generated and own. Vendors who interfere with reasonable, timely access can face civil monetary penalties of up to $1 million per violation. HIPAA, layered on top, gives you additional obligations and rights around patient records in your care.

In practice: if your current vendor charges $5,000–$10,000 to export your patient data in usable format — common, and the “held hostage” feeling practices keep describing in reviews is real — that fee is worth questioning. Document the fee in writing. Document the format they’ll provide versus the format you need. Then weigh that documentation against the framework above.

(Informational only — this is not legal advice. If you believe your vendor is engaging in information blocking, consult your healthcare attorney before filing any formal complaint with HHS or ONC.)

The point isn’t to wage a fight. The point is to walk into the conversation knowing the leverage isn’t all on their side.

Step 1 — Audit what you actually have

Before you talk to any new vendor, before you take a single demo, sit down for an hour and document what you have. Every migration quote you’ll receive from this point forward depends on these numbers. If you don’t have them, the new vendor will guess high and the old vendor will guess high, and you’ll pay for it twice.

The minimum audit checklist:

• Total patient count in the system (lifetime).
• Active patient count (seen in the last 18 months — this is what most new vendors actually migrate as “active”).
• Total chart count and average chart size, if your system reports it.
• Document and attachment count — this is the line item that gets forgotten. Lab PDFs, scanned consents, faxed referrals, imaging reports. They live in a separate database table than structured data and they migrate separately.
• Custom note templates — count them, list them, and identify the 10 that staff actually use.
• Active integrations — labs (which ones), e-Rx, billing/clearinghouse, telehealth, patient portal, referral network, payer connections.
• Staff users — total seats, by role.
• Custom report builds — anything your billing or operations team built inside the EHR that isn’t out-of-the-box.

This is the document you bring to every vendor conversation. Without it, you are not negotiating — you are getting sold to.

Step 2 — Get the data export quote in writing BEFORE signing the new vendor’s contract

This is the single most-skipped step, and it costs practices the most money.

The pattern: practice falls in love with a new vendor, signs the contract, then turns to the old vendor and says “we’re leaving — please export our data.” The old vendor, who knows the leverage just flipped, comes back with a $7,500 quote for a “complete” export plus another $2,000 to convert it into the format the new vendor needs. By then, the practice has already signed, paid the implementation deposit, and committed to a go-live date. They pay. There is no other choice.

Don’t do it in that order. Before you sign anywhere new:

1. Tell your current vendor — in writing, by email — that you are evaluating options and need a complete export quote, including format, timeline, and total cost.
2. Get the response in writing. Email is fine. A phone call is not.
3. Confirm the export format (commonly C-CDA, HL7, FHIR, or a structured database dump for documents/attachments).
4. Ask for a delivery commitment in business days from request.
5. Confirm what is and is not included — structured data, unstructured documents, attachments, audit logs, custom templates.

You may discover the export is “free.” You may discover it’s $10,000. Either way, you now know the actual cost of leaving before you make any other decision. That is leverage. (Want to know what realistic switching costs look like end-to-end? The medical software pricing guide walks through the typical hidden line items.)

> Want quotes from vendors that handle migration in their implementation fee? Get Free Quotes →

Step 3 — Pick a new vendor that includes migration support

Vendors fall into three tiers on migration:

• Top-tier: Migration is included in the implementation fee. They have an internal migration team that has done your old vendor before and has a documented playbook. Ask for it.
• Mid-tier: Migration is an add-on, but with a fixed cap and a documented scope. You pay extra, but you know the number going in.
• Bottom-tier: They hand you a CSV import template and wish you luck. Run.

When you talk to a new vendor, get answers to these five questions before you take a serious demo:

1. Will you accept data directly from [your current vendor], or do you require us to broker the conversation? (Top vendors have direct relationships and a tested import path. Bottom vendors make you the middleman.)
2. What format do you need the export in? (You’ll send this exact spec back to your current vendor when you request the export quote.)
3. Is migration in the implementation fee, or is it billed separately? If separately, get the cap in writing.
4. Who is responsible for verifying data integrity post-migration — you or us? (Hint: it should be them, with you signing off. If they put the verification burden on you, you will not catch the corrupted records until weeks after go-live.)
5. What’s the rollback plan if migration fails on cutover day? (If they don’t have one, that is the answer.)

If you get vague, sales-flavored responses to any of these — “yes, easy,” “we handle everything” — escalate the question to their migration team, not their sales rep. The sales team will scope migration as easy because their job is to close. The migration team will scope it honestly because their job is to deliver.

Step 4 — Run parallel for 30 days

Do not rip out the old system on go-live day. Run both side-by-side for at least 30 days. This is the step practices skip to “save costs,” and it’s the most expensive saving they ever make.

Three reasons parallel-run is non-negotiable:

1. Verify migrated data renders correctly under real workflows. Data that imports cleanly is not the same as data that displays correctly inside the new system’s templates. You’ll only find the gaps when staff opens real charts during real visits.
2. Catch missing custom templates and report builds. Your billing manager pulls the same denied-claims report every Tuesday. If it didn’t migrate, you find out on Tuesday, not on go-live Monday.
3. Keep billing operational while staff trains. Claims already in flight on the old system finish on the old system. Claims after cutover go on the new one. No revenue gap.

Yes, paying for two subscriptions for a month is annoying. Yes, it’s worth every cent. Software Advice data on successful switchers is consistent: the practices that get this right shortlist 3 vendors, decide within 12 weeks, and budget a parallel-run buffer. Practices that skip parallel-run land in the 81% of switchers (AAFP) who rate the time investment as a moderate-to-major problem.

Step 5 — Cut over with a documented checklist

Cutover day is not a vibe. It is a checklist.

Before the day arrives, write out every step. Assign every step to a named human. Estimate every step to a time block. Then on cutover day, work the checklist. A 12–15 item version looks like this:

• [ ] Final data sync from old → new confirmed and timestamped.
• [ ] Migration team sign-off on data integrity (record, document, and attachment counts all match).
• [ ] Lab integration cutover — old disconnected, new connected, test order placed.
• [ ] e-Rx integration cutover — DEA validation re-confirmed.
• [ ] Clearinghouse / billing cutover — test claim submitted.
• [ ] Payer enrollments confirmed live on the new system.
• [ ] Telehealth integration cutover — test session run.
• [ ] Patient portal cutover — login flow tested with a real (staff) account.
• [ ] Staff schedule confirmed — cutover happens on the lowest-volume day of your week.
• [ ] Staff phone tree posted — who to call when something breaks.
• [ ] Old system preserved in read-only mode for 90 days minimum.
• [ ] Rollback trigger defined in writing — if X breaks by Y time, abort and stay on the old system.
• [ ] Front desk script for the first three days.
• [ ] Patient communication sent (only if patient-facing portal/scheduling is involved).
• [ ] End-of-day reconciliation — chart, claim, and encounter counts match expectations.

Don’t cut over on a Monday. Don’t cut over the week of a holiday. Don’t cut over without a rollback trigger.

> Want quotes from vendors that handle migration in their implementation fee? Get Free Quotes →

Common mistakes that break migrations

Patterns from the practices that got it wrong:

• Cutting over on a Monday — when vendor support ticket volume is highest and your front desk is already buried.
• Letting the new vendor’s sales team scope migration — they’ll say “yes, easy.” Always escalate scoping to their migration team.
• Not migrating attachments — lab PDFs, scanned consents, faxed referrals. Practices migrate structured data, declare victory, and discover three weeks later they can’t find a single signed consent.
• Forgetting payer enrollments don’t auto-transfer — re-credentialing on the new clearinghouse runs 30–60 days per payer. Start 60 days before cutover or you have a revenue gap.
• Skipping the parallel-run period to save costs — covered above. Don’t.
• Migrating only structured data — losing custom note templates and turning the new system into a generic experience for staff who built workflow around the old ones.
• Not documenting custom workflows before training new staff — you trained your existing team on tribal knowledge over five years. New hires get 90 minutes and a tutorial video. Write the tribal knowledge down.

The 90-day post-migration checklist

Cutover day is not the finish line. It’s day zero of a 90-day stabilization window. Build calendar reminders for these checkpoints:

• Day 7 — Data parity. Pull a random sample of 25 charts. Compare old (read-only) against new. Flag anything missing or rendered incorrectly. Most data integrity bugs surface in week one.
• Day 30 — Billing reconciliation. Compare claim volume, denial rate, and days-in-A/R against your three-month pre-cutover average. If denials are up more than 10%, audit the new system’s claim scrubber.
• Day 60 — Staff confidence and portal adoption. Survey staff on remaining pain points. Confirm patients are logging into the new portal at a similar rate to the old one.
• Day 90 — Reporting parity. Make sure every report your billing manager, operations lead, and physician owners ran on the old system runs on the new one with matching numbers. Custom reports are the last thing to migrate cleanly and the first thing people miss.

If all four checkpoints land green, the switch worked. If any land red, the issue is identifiable and fixable — far easier at day 30 with the migration team still on the hook than at day 180 after they’ve moved on. (More on what other practices ask after a switch in the FAQ.)

---

Considering a switch? Get free, side-by-side quotes from 2–3 vendors that handle migration well.

The hardest part of switching isn’t the technical migration. It’s the upfront work of figuring out which vendors are actually worth the conversation, before you commit your time to a sales gauntlet.

See Your Options →

60 seconds. No commitment. We earn a referral fee from the vendor you choose, never from you.