What an EHR for a cash-pay or direct primary care practice has to do differently

An EHR built for insurance billing is organized around producing a defensible claim. A cash-pay, direct primary care, or concierge practice never files one. That single fact propagates further through a clinical system than most vendors admit — into scheduling, documentation, messaging, and reporting — and it is the reason so many direct practices end up running their EHR alongside a second system that handles the part that actually pays them.

This is worth being precise about, because "we support DPC" has become a checkbox on a lot of vendor sites, and it usually means the billing module can be switched off.

The claim is the thing being optimized

Almost every design decision in a conventional ambulatory EHR traces back to reimbursement, and traces back correctly.

Discrete structured fields exist in the numbers they do because payers need discrete data. Templates exist in the shape they do because a note has to justify the level of service billed. The encounter is the atomic unit of the system because the encounter is the billable unit. Even the review-of-systems checkbox that everyone complains about is not a design failure — it is a rational response to an environment where an unjustified level of service is a financial and legal risk.

The result is a chart that is, in part, an evidence file assembled for a third party who was not in the room.

None of that is a criticism of those systems. It is the correct answer to the problem they were built to solve. The difficulty is only that it is not the problem you have.

What falls away

Strip out the claim and a surprising amount of a conventional EHR goes quiet:

• Coding for reimbursement. E/M level selection, the documentation required to support it, and the anxiety attached to both.
• Eligibility and benefits checks. No coverage to verify before a visit.
• Claim scrubbing and clearinghouse submission. No claim.
• Denials, appeals, and accounts receivable. No denials to work, and no thirty-to-ninety-day gap between doing the work and being paid for it.
• Prior authorization. For services you provide directly. Not for everything you refer out or prescribe, which is a real and separate caveat.
• RVU-based productivity reporting. The measure and the machinery that produces it.
• Payer quality reporting. Programs tied to participation you do not have.

One honest qualification. Very few practices are purely one thing. Plenty of direct practices still bill something — a Medicare-enrolled physician who has not opted out, an insurance line kept running during a transition, workers' compensation, a procedure billed to a plan. If you are hybrid, you need both halves working, and that is a materially harder software problem than either pure case. Say so to a vendor in the first conversation rather than the fifth, because "we support DPC" and "we support a practice running both models at once" are not the same claim.

What takes its place

The replacements are not smaller than what they replace. They are just different, and they tend to be the parts of a conventional system that were built last.

1. Recurring membership billing. The central financial event is a subscription, not a claim. That means monthly and annual terms, per-family and per-employer tiers, proration when someone joins mid-month, failed card retries, dunning, pauses, and cancellations that have to be honored quickly — many state DPC statutes require that a patient be able to terminate without penalty. This is a genuine billing system, not a lighter one, and most EHRs have no concept of it at all.
2. The panel rather than the schedule. In fee-for-service, an empty slot is lost revenue, so the system optimizes throughput. In direct care, revenue is fixed by panel size and a light schedule may mean the model is working. What you need to see instead is panel health: who has not been seen in a year, who is due for something, who is quietly disengaging before they cancel. Retention is the metric that matters, and almost no EHR reports on it.
3. Asynchronous care treated as real care. Text, email, and phone are not adjuncts in a direct practice; for many panels they are the majority of clinical contact. In a fee-for-service system these are largely unbillable and therefore close to invisible — they happen in a messaging sidebar and often never reach the chart. When they are the care, they have to land in the record like anything else.
4. Visit length. Thirty to sixty minutes changes the artifact. A template designed to capture a twelve-minute visit efficiently produces the wrong shape of note for an hour-long one, and clinicians end up fighting the form.
5. Transparent pricing for labs and medications. Wholesale lab panels and in-office dispensing at cost are core to the value proposition of most direct practices. That needs price transparency at the point of ordering, which is close to the opposite of how ordering works when a payer sits in between.
6. A record the patient can genuinely take with them. Patients who pay you directly tend to feel a stronger sense of ownership over their chart, and they are right to. Export that produces something usable, rather than a stack of PDFs, becomes a patient-facing feature rather than an IT afterthought.

The note changes shape

This is the part worth dwelling on, because it is the change clinicians feel first.

A note in a billing practice serves three masters at once: clinical communication, justification of the level of service billed, and defense in the event of an audit. Remove the second and third and what remains is the first — the part that was always actually about the patient.

In practice that means shorter notes, and better ones. No padding a history to reach a threshold. No copy-forward of a problem list that stopped being true two years ago because deleting it felt risky. No ten-point review of systems on a patient who came in about a knee. What is left is what a colleague seeing this patient next actually needs to know, which is the thing a note was for in the first place.

Two things do not go away, and it is worth being blunt about them. You still need an accurate, contemporaneous record for continuity of care and for your own protection, and your state's medical-record retention requirements apply to you regardless of who pays you. Removing the billing rationale for documentation does not remove the clinical or legal one.

This is also where the choice of system matters most, and where it is easiest to buy the wrong thing. An AI feature tuned to produce a coded, audit-ready E/M note is solving a problem you do not have, extremely well. It will keep suggesting codes, keep padding toward a level, and keep optimizing an artifact you have no use for. That is a specific enough trap that it is worth its own discussion — see free AI in your EHR versus an AI that owns the record.

What to actually ask a vendor

Six questions that separate a system that fits from one that has been told to say yes:

1. How does money actually flow through this system? Not "do you support DPC" — walk me through a patient joining mid-month, a card failing in month four, and a family upgrading a tier. If the answer is "you'd use Stripe alongside us," that is a legitimate answer, but you should hear it now rather than in month two.
2. Where does an asynchronous encounter live? If a patient texts a photo of a rash and I answer, is that in the chart or in a messaging tool bolted to the side of it?
3. Can I see my panel rather than my schedule? What does the system tell me about who I have not seen?
4. What does the documentation look like without a billing rationale? Show me a note this system produces when nothing is being coded.
5. How do I get my data back out, in what format, and at what cost? This is the term that determines the price of every future decision, and it belongs in the first conversation rather than the last. It is worth reading what to check in an EHR contract before signing before you get to a redline.
6. What does the first year genuinely cost, including my own time? The subscription is rarely the largest number. If you are moving from an existing system, what switching EHRs actually costs a small practice breaks down where the rest of it hides. If you are starting from nothing, what has to be running before the first patient is the more useful list.

Where Zenthea fits

Zenthea is built as an AI-native EHR, designed around a clinical AI assistant, Thea, from the workflow foundation rather than as a feature layered onto a legacy record. The principle is that the AI prepares and a clinician decides: Thea drafts documentation, prepares orders, and surfaces context, while a human clinician reviews and signs every clinical action.

Being straight about the boundary, since this post has spent a thousand words arguing that the financial layer is where these systems diverge: Zenthea's billing is built around claims and coding. Recurring membership billing is not part of it. For a practice whose central financial event is a subscription, that is a real gap, and you should weigh it accordingly rather than discover it later.

What we would say is that the argument above holds whoever you buy from. The question that actually separates systems for a direct practice is not which one has the longest feature list — it is which one is organized around the way your practice earns its living, and which one is quietly still organized around a claim you are never going to file.