What switching EHRs actually costs a small practice in the first year

Ask a vendor what switching costs and you will get a subscription price. That number is real, and it is almost never the largest one.

The costs that actually determine whether a migration was worth it are the ones nobody invoices for: the weeks you see fewer patients, the evenings spent re-entering problem lists, the afternoon someone spends on hold re-enrolling your lab interface. This post is a way to estimate your own number rather than a claim about what the number is, because a credible single figure does not exist and anyone offering one without asking about your chart count and your data format is guessing.

The subscription is the smallest number on the list

Start by writing down the annual difference between the new subscription and the old one. Then set it aside, because for most small practices it is somewhere between a tenth and a third of the true first-year cost.

The rest divides into two piles: things that arrive as an invoice, and things that come out of your own hours. The second pile is bigger, and it is the one practices consistently fail to budget.

The costs that arrive as an invoice

• Overlapping subscriptions. You will pay for both systems for some period. Two months is optimistic if you are running parallel; three to six is common if the old contract has a term left on it. Multiply the overlap months by the old monthly fee and add it in.
• Data migration services. Whether the new vendor charges to import, and whether the quote covers the whole record or only demographics and a problem list. Get the scope in writing, because "migration included" frequently means the structured fields and not the notes.
• Export or extraction fees from the outgoing vendor. Sometimes zero, sometimes four figures, and occasionally priced per chart — which is trivial at 200 patients and serious at 4,000. If your contract says "the vendor's then-current rates" rather than a number, treat that as an unbudgeted line, and read what to check in an EHR contract before signing before you sign the next one.
• Implementation, configuration, and training packages. Often optional, often worth it, rarely free.
• Reconnecting third parties. Lab interfaces, imaging, e-prescribing, payment processing. Some charge a setup fee; all of them charge you time.
• Temporary help. Some practices hire a scribe, a temp, or extra front-desk hours through the transition. Cheaper than the alternative, but a real line.

The costs that come out of your hours

These do not appear anywhere in a budget, which is precisely why they are worth writing down at an hourly rate you would actually accept.

1. Chart abstraction. Migrated data almost never arrives in a state you would sign your name to. Somebody reconciles medication lists, allergies, problem lists, and preventive care history — usually a clinician, because it requires judgment. Estimate minutes per active chart and multiply by your active panel. This is the single largest hidden number for most practices.
2. Data validation. Separate from abstraction. Before go-live somebody checks that the import actually worked — that the allergy field went into the allergy field, that dates did not shift, that nothing silently truncated.
3. Template and workflow rebuilding. Whatever you had customized in the old system has to be rebuilt, and the rebuild is where most of the "this is worse than what we had" feeling comes from.
4. Training. Not the vendor's four-hour session. The three weeks afterwards when everything takes twice as long.
5. Administrative overhead. Re-enrolling for electronic prescribing, and for controlled substances in particular, is its own project — the DEA framework requires identity proofing and two-factor authentication at signing, and that is not something you complete on the morning of go-live. Add PDMP access, lab account changes, and payer or clearinghouse updates if you have any.
6. The productivity dip. Large enough to deserve its own section.

What the evidence actually says about the productivity dip

Honestly: the published evidence is thinner than you would hope, and what exists is not reassuring.

The most direct study we found is a 2016 paper in BMC Health Services Research examining an academic outpatient ophthalmology department before and after EHR adoption. It reported that patient volume per provider fell an average of 16.9% across the four years studied, with a statistically significant reduction in every year after implementation, and that "during the final year studied, no provider had returned to the pre-EHR number of patients per clinic session."

That finding needs heavy qualification before anyone applies it to their own practice. The sample was eight physicians at a single institution, dropping to six or seven as staff left. The analysis used four-month windows rather than full years. Ophthalmology is image-heavy and may be unusually sensitive to a system change. And critically, this measured adoption of an EHR by a department moving from paper-era workflows, which is not the same event as a practice moving from one modern EHR to another. The authors themselves list most of these limitations.

So it is not a number to plan against. What it is good for is calibrating a direction: the dip is real, it is not confined to the first fortnight, and the honest planning assumption is that recovery takes longer than an implementation timeline suggests.

The practical response is to make the dip deliberate rather than let it happen to you. Cut your own schedule for the first weeks on purpose — a reduced clinic you planned is far cheaper than a full clinic that collapses. For a direct practice this is genuinely easier than for a fee-for-service one, because membership revenue does not fall when the schedule thins. That is a real structural advantage of the model, and it is worth using.

The line people leave off

Every list like this makes switching look expensive, which biases the decision toward staying. So the honest counterweight belongs in the same spreadsheet: what does not switching cost?

Put a number on the hours per week the current system costs you that a better one would not. Multiply by fifty. Add whatever you are paying for tools that exist only to work around it. Add the renewal increase you will accept because leaving felt too hard — which is exactly the leverage an unclear export clause hands your vendor. And if the current system is a live factor in whether you or a colleague want to keep practising, that belongs in the calculation too, even though you cannot price it cleanly.

Sometimes that side wins decisively. Sometimes it doesn't, and the right answer is to stay another year and renegotiate. Both are legitimate outcomes of doing the arithmetic. The failure mode is not doing it.

Building your own number

A rough first pass, in the order the money appears:

1. Overlap months × old monthly fee.
2. Migration and export quotes, in writing, with scope stated.
3. Active charts × minutes of abstraction per chart × a real hourly rate.
4. Reduced-schedule weeks × the revenue or visits you are giving up. For a membership practice, count this in clinician hours rather than lost revenue.
5. Staff hours for training, validation, rebuilding, and re-enrolment.
6. Third-party setup fees and lead times.
7. Subtract the annual cost of staying.

The result will be uncomfortably large and considerably closer to true than the subscription difference. It is also the only version of this number that is about your practice rather than someone's average.

Where Zenthea fits

Zenthea is built as an AI-native EHR, and we have an obvious interest in practices deciding to switch, which is a reason to be careful about how we say this rather than a reason not to say it.

The honest version: a migration is expensive, the expense is mostly your own time, and no vendor's implementation package makes that go away. The questions above are worth putting to us and to everyone else on your shortlist, and the answers should be specific — chart counts, formats, lead times, and what happens to your data if it goes wrong.

If the wider question is what a cash-pay or direct practice should be looking for in the first place, that is covered in what an EHR for a cash-pay or direct primary care practice has to do differently. If you are not switching but starting, the relevant list is what has to be running before the first patient.