Do you have to change your EHR to change how you bill?

Why are the clinical record and the claim separable?

Because they are different artifacts with different audiences. The record is what the clinician writes and what a payer, an auditor, or a court would read. The claim is a structured message — the industry’s standard claim transaction, the 837 — that says who was seen, by whom, on what date, under what diagnosis, for which procedure code, for how many units. It travels to the payer through a clearinghouse, and the answers come back the same way: an acknowledgment saying the claim was accepted or rejected before anyone adjudicated it, a status response when you ask where it stands, and an electronic remittance advice — the 835 — saying what was paid, what was adjusted, and why.

Those formats are standardized across the industry, which is the whole point of them. They do not care which EHR produced the encounter. So the question “can this billing company work with my system” is not really about your system at all — it is about whether the data it holds can get out cleanly and on a schedule. When a vendor tells you that billing only works properly inside its own platform, that is a packaging decision, and a reasonable one for them. It is not a technical constraint you have to accept.

The honest counter-case, since we are the ones arguing the other way: a system that does both can catch a coverage, authorization, or coding problem while the clinician is still in the note, and anyone billing from a scheduled export is at least one cycle behind that. That is a real trade, not a myth. Whether it matters to you depends on how often your export runs and how much gets caught before submission anyway — but it is the strongest argument against the position this page takes, and you should weigh it.

What does a billing company actually need from your EHR?

The fields a claim is built from, delivered on a predictable cadence. In plain terms:

• Who the patient is — demographics, and the insurance on file: payer, member ID, group, plan effective dates, and whether there is a secondary.
• Who rendered the service — the clinician, their NPI, their license level, and the billing entity and location the claim goes out under.
• What happened — the date of service, the procedure code and units, the diagnoses, any modifiers, the place of service, and whether it was delivered in person or by telehealth.
• What permission existed — the authorization number, the visits it covers, and how many of them are already used.
• Whether the note is finished — signed, and by whom. A claim built from an unsigned note is usually a recoupment with a delay in front of it: the payer rarely sees the note at adjudication, pays, and takes it back on audit.

Most practice systems can produce that, in some form. The honest variable is the form: a scheduled export, a report you can automate, a direct connection, or — at the low end — a spreadsheet somebody pulls every Monday. If your system cannot export cleanly, that gap does not disappear; someone has to build a connection to it. The question to ask a prospective biller is not whether they “integrate,” which everyone says, but who builds that connection, who maintains it when your vendor changes something, and who pays for both.

If the EHR does not change, where do the savings come from?

From the transactions themselves — specifically, from not doing them by hand. This is the part of the cost structure that most buyers never see itemized, and it is surveyed annually. The 2024 CAQH Index splits administrative transactions into three modes — fully manual (phone, mail, fax, email), partially electronic (payer web portals and phone trees), and fully electronic (the standard X12 transactions) — and prices them where the data allows. On the provider side of the medical industry, per transaction:

• Eligibility and benefit verification — $8.57 manual, $4.46 through a portal, $2.00 fully electronic.
• Claim status inquiry — $13.80 manual, $5.24 through a portal, $3.64 fully electronic. The widest manual-to-electronic gap here — and, from a portal, the narrowest of the three. Which number applies to you depends entirely on how you chase status today.
• Claim submission — $6.33 manual, $3.05 electronic.
• Remittance advice — the file you post payments from — $5.67 manual, $5.31 through a portal, $2.95 fully electronic. Note the portal column: the gain there is $2.36, the smallest on this list.
• Prior authorization — $12.88 manual, $8.93 through a portal, $5.38 fully electronic.

Two things are worth noticing about that list. The first is the middle column, which is the one that decides whether any of this matters to you: the saving depends on what you do now. If your staff is on the phone, the gap is large. If they are already in a payer portal — which is where a great deal of this work sits — it is much smaller, and on remittance advice it nearly closes. Quoting only the widest spread is how this argument usually gets made, ours included, and it is not the honest version.

The second is that CAQH measures time the same way, and it needs the same two baselines. Going fully electronic saves a medical provider about 12 minutes per eligibility check and 14 per prior authorization against phone or fax — but only about 4 minutes and 6 against a portal. Across every transaction it measures, CAQH puts the total provider time saving at 70 minutes from fully manual and 21 minutes from partially electronic. If someone sells you an automation case, ask which of those two numbers it was built on.

Treat all of it as a good survey rather than a measurement: CAQH collects these figures from plans and providers who report them, says the results “may be subject to response bias,” and counts only the labor time of the transaction itself — not the gathering and follow-up around it, and not system costs. None of the numbers is behavioral-health-specific either, and none of them is large on its own. That is the point. A billing operation’s price is mostly the labor of repeating small transactions across every visit — and an outpatient mental health practice runs a high count of small claims, which is the worst possible shape for a cost structure built on staff time and the best possible shape for one built on standard transactions.

What actually has to change when you switch billers?

Less than you fear in the clinical system, and more than most billers mention in the enrollment plumbing. Nothing about your charting changes: same EHR, same templates, same logins, same people writing the same notes. What moves is behind the scenes.

The data export has to start flowing to whoever is billing — that is the list above, on a schedule. Electronic remittance enrollment has to be moved payer by payer, and this is the item that deserves your attention: a provider can generally have only one clearinghouse receiving electronic remittance advice from a given payer at a time, so enrolling with a new one typically stops the old one from receiving those files. It is routine, it is reversible, and it is also exactly where a sloppy transition loses a fortnight of posted payments. It varies by payer, so confirm it payer by payer rather than assuming it — and ask for the sequence in writing, with dates.

Somebody has to own the claims already in flight — the ones submitted before the switch, the denials sitting in a work queue, the balances aging past 90 days. Those do not migrate on their own, and they are not covered by a going-forward arrangement unless someone says so. Get an explicit answer about who works claims with dates of service before the cutover, under what terms, and at what price. An old A/R balance that nobody has agreed to work is the most common thing lost in a biller change.

What still gets checked before a claim goes out?

The things that make a claim fail for reasons that have nothing to do with whether the care was necessary. In KFF’s analysis of 2024 HealthCare.gov marketplace plan data, only about 5% of in-network claim denials were for medical necessity, against 25% recorded as administrative, 13% for an excluded service, and 9% for a missing prior authorization or referral. Far more denials were procedural than clinical — though we should be careful how far that goes, because the largest single bucket was “other” at 36%, and KFF notes the data do not link denial reasons to the services denied. This covers HealthCare.gov non-group plans — not state exchanges, not employer coverage, and not behavioral health specifically — and among returning insurers, denial reasons were reported for only about three-quarters of claims.

The cost of finding out after the fact is the argument for checking first. An RCM software vendor’s summary of the AHIMA Journal puts the cost of reworking a denied claim at an average of $25 for a practice (its $181 figure is the hospital number), and reports the industry average that as many as 60% of returned claims are never resubmitted at all — averages the journal repeats from 2017–2019 trade sources, not measurements, and worth reading as a direction rather than a statistic. On a practice’s typical claim value, rework can cost a meaningful share of what the claim is worth, which is why catching the problem before submission beats appealing it afterward almost every time. That check is a function of the billing arrangement, not of the EHR — which is another way of saying you can improve it without changing where your clinicians work.

What about authorizations — do those move too?

The tracking moves; the clinical decision does not. Authorization work is administrative and deadline-shaped: knowing which payers require one for which service, how many visits are approved, how many are left, and when the window closes. It is also genuinely expensive in staff time. In the AMA’s 2025 prior authorization survey — 1,000 practicing physicians, 40% primary care and 60% specialists, surveyed in December 2025 — respondents reported that physicians and their staff spend about 13 hours a week completing prior authorizations. Read the denominator before you use that number: the survey screened for physicians who complete prior authorizations in a typical week, so it describes doctors who already do this work — not physicians generally, and not therapists or behavioral health. An outpatient mental health practice’s burden depends heavily on its payer mix, and a practice whose caseload is routine outpatient psychotherapy will sit well below it.

It also matters that payers get these wrong. Reviewing a stratified sample of 250 prior authorization denials issued by 15 of the largest Medicare Advantage organizations during a single week in June 2019, the HHS Office of Inspector General estimated that 13% met Medicare coverage rules — meaning, in OIG’s words, that those services likely would have been approved under original Medicare. That is a narrow sample of one program, not a general denial rate — but it is the reason the documentation and authorization records you keep are worth keeping properly. They are what an appeal is made of.

What should you ask before you sign?

Six questions, all of which have specific answers if the arrangement is real:

• Can you take our export as it comes out today? If not, what exactly is missing?
• Who builds the connection, who maintains it, and what does each cost? Including when our EHR vendor changes something.
• What is the payer-by-payer sequence for moving remittance enrollment, with dates?
• Who works claims with dates of service before the cutover, and on what terms?
• What will we be able to see without asking you — and how often?
• If we leave, what do we get back, in what format, and how quickly? An arrangement that is easy to leave is usually one built to be checked.

Where the percentage itself is concerned, the rate a billing company charges and what it should include is a separate question from this one, and worth reading before you compare quotes.

Key takeaways

• Changing billers does not require changing your EHR. Claims travel as standardized transactions that do not depend on which system produced the encounter.
• What a biller needs from your system is a clean, scheduled export of claim data — patient, insurance, clinician, encounter, codes, modifiers, units, authorization, and a signed note.
• The savings come from not doing routine transactions by hand. Per CAQH’s 2024 provider figures, eligibility runs $8.57 manual, $4.46 by portal, and $2.00 fully electronic; claim status runs $13.80, $5.24, and $3.64.
• The one thing that genuinely has to move is electronic remittance enrollment, generally one clearinghouse per payer at a time — get the sequence and dates in writing.
• Decide explicitly who works claims dated before the cutover. Old A/R with no owner is what usually gets lost in a switch.
• Checking a claim before submission is a function of the billing arrangement, not the EHR — so it can improve without your clinicians changing anything.