How Much Does Custom EHR Software Cost in 2026?

The bands a custom electronic health record build falls into

There are two shapes to this project and the one you choose is decided in the first fortnight, not by the software. The first is a focused release that leaves your billing rails alone: specialty charting built on your actual clinical objects, one scheduling engine for the whole group, and a patient portal, reading from the incumbent through its application programming interface during transition. That runs $60,000 to $130,000 over 12 to 16 weeks. The second is a full platform that eventually owns the record: everything above plus e-prescribing, lab and imaging feeds, billing integration and an analytics layer. That runs $150,000 to $400,000 phased over 6 to 12 months.

Here is what the individual components cost when we scope them separately, drawn from Digital Heroes delivery work with multi provider groups.

• Specialty charting module, $40,000 to $90,000. Your specialty's real objects rather than a generic encounter form: a lesion record pinned to a body map with photo timelines for dermatology, an operative episode linking injury, imaging, procedure, implants and post operative milestones for orthopaedics.
• Group wide scheduling engine, $32,000 to $70,000. One rules engine holding provider skills, visit durations by type, room and equipment constraints and payer rules, offering the next best slot across every location.
• Patient portal and digital intake, $18,000 to $45,000. Self scheduling that exposes real capacity, forms that land as structured data rather than a scanned page.
• Lab and imaging interfaces, $12,000 to $28,000 each. HL7 version 2 results feeds from Quest, LabCorp or a hospital system, each with its own message profile and its own testing window.
• E-prescribing through a Surescripts certified partner module, $20,000 to $45,000. Integration and workflow, not the certification itself.
• Billing integration, $30,000 to $70,000. X12 837 claim and 835 remittance flows through a clearinghouse such as Availity, with reconciliation you can inspect.
• Analytics warehouse and dashboards, $24,000 to $55,000. Every appointment, charge, message and status change landing in your own store so a new question is a query.
• Data migration, $20,000 to $80,000. Discrete problems, medications, allergies and results mapped forward, legacy charts archived as searchable documents.

Nobody buys every line at once. A group that keeps its billing rails, runs one lab interface and defers analytics lands near the bottom of the full platform band. That is not a compromise, it is the sequence we recommend.

What drives an electronic health record build up

• Interface count. This is the dominant driver in this category. Every additional feed is $12,000 to $28,000 plus a testing cycle whose pace is set by the counterparty. Five interfaces is not five times the work of one, it is five separate schedules you have to hold open at the same time.
• Controlled substance prescribing. Electronic prescribing of controlled substances brings identity proofing, two factor authentication at the moment of signing and specific audit requirements. Budget $20,000 to $45,000 and add real weeks, because the identity proofing step involves your providers individually and cannot be compressed.
• Certification scope. If your providers report under a programme that requires a certified electronic health record, such as the Promoting Interoperability category of the Merit based Incentive Payment System, the certified components have to be in scope or kept alongside as a certified module. Decide this before design, not after.
• Migration depth. Years of discrete data is a different project from a document archive. In our delivery experience migration lands at 10 to 20 percent of total project cost for a multi location group, and the variable is how much history the physicians insist on seeing as structured fields.
• Location count. Each location is its own parallel run, its own training, its own rollback plan. Cutting over four locations is not four times harder than one, but it is not the same job either.
• Specialty count. A multi specialty group needs more than one charting model, and the second model is rarely a copy of the first.
• Multi state telehealth. Consent language and licensing rules vary by state, and each variation is a rule with an effective date.

What keeps the number down

• Keep certified billing rails through phase one. Never put clinical go live and billing cutover in the same phase. Leaving billing alone removes $30,000 to $70,000 and most of the risk from the first release.
• One specialty first. Build the charting model for the specialty that generates the most volume, prove it, then extend. Groups that try to serve every department in release one produce something shallow everywhere.
• Read from the incumbent instead of migrating up front. During transition the new system can pull demographics and history through the existing application programming interface, which lets you delay the migration decision until the clinical workflow is trusted.
• Archive rather than migrate. Map discrete problems, medications, allergies and results forward. Everything else becomes a searchable document. That distinction alone can halve the migration line.
• Defer controlled substance prescribing. If your controlled substance volume is low, keep prescribing in the incumbent for phase one.
• Feed an existing analytics platform. If finance already runs a warehouse, push data to it rather than building dashboards inside the clinical system.
• Phase interfaces one per quarter. Same total cost, far less concurrent coordination, and each one goes live with attention on it.

A worked example that adds up

A 22 provider orthopaedic group across four locations in one state, currently on athenahealth, migrating four years of discrete data, two lab feeds, no controlled substance prescribing in phase one.

• Discovery and clinical workflow mapping across four locations: $12,000
• Specialty charting with operative episodes and implant tracking: $68,000
• Group wide scheduling engine with waitlist and recall: $54,000
• Patient portal and digital intake: $31,000
• Two HL7 version 2 lab interfaces: $34,000
• E-prescribing through a Surescripts certified partner module: $28,000
• Billing integration, X12 837 and 835 through the clearinghouse: $46,000
• Analytics warehouse and administrator dashboards: $37,000
• Migration of four years of discrete data plus document archive: $44,000

That totals $354,000. Add a 12 percent contingency, because at least one lab counterparty will take twice as long to schedule testing as they promise, and the committed number is $396,000 across roughly 10 months. That sits near the top of the full platform band, which is where a four location group with real interface requirements belongs.

How the spend phases across the year

• Weeks 1 to 4, about $12,000. Discovery with physicians, the practice administrator and the biller in the room together, working from real charts rather than a process diagram.
• Weeks 3 to 18, about $122,000. Charting and scheduling, the two components clinicians feel immediately. At the end of this phase one location is closing charts in the room.
• Weeks 8 to 42, about $44,000. Migration runs long and in the background, because discrete data mapping surfaces questions nobody can answer in a single sitting.
• Weeks 12 to 22, about $31,000. Patient portal and intake, once scheduling is stable enough to expose to patients.
• Weeks 16 to 28, about $62,000. Lab interfaces and e-prescribing, scheduled around the counterparties rather than around you.
• Weeks 22 to 34, about $46,000. Billing integration, deliberately after clinical trust exists.
• Weeks 26 to 40, about $37,000. Analytics, built last because it needs a full quarter of real data to be worth anything.

The ongoing costs nobody quotes

• Support and maintenance, 15 to 20 percent of build. On a $396,000 platform that is roughly $59,000 to $79,000 a year, and it is not optional in a clinical system.
• Interface maintenance, $4,000 to $9,000 per interface per year. Message profiles change, hospital systems upgrade, and a silently broken results feed is a patient safety issue rather than an inconvenience.
• Clearinghouse, e-prescribing and lab pass through fees. These are whatever your existing vendors charge and they do not disappear when you build. Read them off your current invoices and carry them forward unchanged.
• Annual security risk analysis, $8,000 to $25,000. A documented risk analysis is a standing obligation under the HIPAA Security Rule, and building your own system moves that work in house.
• Hosting, backup and disaster recovery, $12,000 to $35,000 a year. Higher than a typical business application because recovery time expectations in a clinic are measured in minutes.
• Payer rule and code set updates, $10,000 to $30,000 a year. Code sets change annually and payer edits change more often than that.
• Provider onboarding and training, $5,000 to $15,000 a year. Every new physician needs the system taught properly, and this is the line most groups forget until the third hire complains.

Comparing the build against your current renewal

Do the arithmetic before the debate. Percentage of collections pricing means your software bill grows with revenue rather than with usage, so take last year's effective spend, apply your own growth forecast for three years, and total it. Groups reaching us at $25 million to $40 million in annual collections are typically carrying a seven figure three year total on that basis. Against that, put the build plus its running cost: on the worked example above, $396,000 once plus roughly $70,000 a year, so about $606,000 over three years and falling as a share of revenue every year after.

The crossover for groups above roughly $20 million in collections usually lands between 18 and 36 months in our experience. Two other lines belong in the comparison and rarely appear in a vendor renewal discussion. First, staff whose actual job is working around the software: scribes hired because templates cannot keep pace, an analyst re keying exports, a biller reworking claims the system coded wrong. Second, adding provider number 31 costs an onboarding session on a platform you own and a permanent licensing tail on one you rent.

When buying beats building

Stay with athenahealth, or move to a strong niche system for your specialty, if you are under roughly 10 providers. At that scale the percentage of collections bill is irritating but survivable, and a build would consume management attention you cannot spare. Stay if you have no operations lead who can own a software project through a full year, because a custom clinical system without an internal owner fails regardless of who builds it. Stay if you need certified quality reporting next quarter, because certification scope is not something to rush.

Epic through a Community Connect arrangement is genuinely right for some independent groups, particularly where your referral relationships run through one hospital system and clinical data sharing with that system is your main constraint. Be clear eyed about what it costs you in exchange: per provider fees, template changes that queue behind the host organisation's own work, and upgrade windows you do not control.

Build when your annual electronic health record spend has crossed $400,000 and climbs with revenue, when you employ two or more people whose real job is working around the software, and when a workflow that wins you referrals depends on capability the vendor will never prioritise for a group your size. Any two of those and the case closes on arithmetic. Even then, do not replace the whole system at once. Build the layer where you are differentiated, keep certified billing rails, and replace the core last.

If you would rather scope this before committing budget, Digital Heroes contracts through India LLP, US LLC and UK LTD entities, so the agreement and the intellectual property assignment sit under law your own advisers already read. The document is yours whichever way you go.