fhir r4 · us core · conformance

Charles Trinh — FHIR R4 Interoperability

Charles Trinh — FHIR R4 Interoperability
Charles Trinh — FHIR R4 Interoperability

The specification is not the problem

R4 is stable and well documented. Projects run long because it is a framework that leaves every consequential decision to the implementer, and two conformant servers can disagree about nearly everything a client cares about.

The capability statement is the only document a counterparty builds against without asking, so publish narrow and accurate rather than broad and hopeful.

Conformance that holds

Validate against US Core profiles and pinned terminology on every build rather than at the end of a quarter. A quarterly run produces a defect list measured in hundreds; a per-commit run produces ones.

Declare meta.profile on every instance, because a validator that cannot tell which profile you intended cannot validate you.

Performance and search

Every advertised search parameter needs a query plan someone has read. A parameter that is instant on ten thousand rows is a different query on ten million.

Treat Bulk Data as a job system with checkpointing, resumption and file expiry, because a partial export that cannot resume re-runs from zero.

The external bar

I have taken a platform to 317 of 317 subtests on the ONC (g)(10) suite. What that run teaches is that the last fifteen percent of failures are never algorithmic.

They are a required element left null, a profile not declared, or a search parameter returning the right resource with the wrong include.

Why conformance is a build gate

A conformance suite run at the end of a project produces a defect list measured in hundreds, and every item on it is expensive because the design has already set. Run on every commit, the same suite produces ones and the cost of each is a fix rather than a redesign. The difference is not the tool; it is where in the process the tool sits, and moving it earlier is the single highest-return change most FHIR programmes can make.

The failures the suite finds are rarely algorithmic. A required element left null because the source system never captured it. A profile not declared, so the validator cannot tell which one was intended. A terminology binding that passed against a value set loaded at the wrong version. A search parameter implemented but unindexed. Each is a small defect with a large consequence, and each is cheapest where it is introduced.

Search is the performance cliff

Returning a Patient by identifier is table stakes. Answering a date-ranged, chained search with includes against a large population is a different engineering problem, and it is where most implementations fall over in production having passed their own tests on small data. Every advertised search parameter needs a query plan that somebody has read, and the plan should be re-read when the table crosses a million rows, because the planner changes its mind.

Advertise narrow and accurately. A capability statement that lists every parameter the library supports produces a client that queries one you never planned for, and the resulting latency is indistinguishable from an outage. Publish what you can serve at the latency the client expects, and add parameters when the plan exists rather than when the specification allows.

Bulk Data is a job system

A Bulk Data export is not a large query; it is a job with a lifecycle, and treating it as a request produces an implementation that times out and cannot resume. The job needs a status endpoint, file expiry, chunking and a resume point, because an export interrupted at eighty percent that restarts from zero is an export that never completes for the largest patients.

Design the storage side too. Export files are copies of PHI under weaker controls than the primary store unless somebody decides otherwise, and the retention of those files is a compliance question that arrives with the first export rather than with the design. Set the expiry, log the download, and keep the file in a store with the same access model as the records it contains.

The test account is the product

Every one of those concerns is easier to resolve when a counterparty can test against you without a meeting. Keep a test account, with representative patients, kept current, and documented in the capability statement. Organisations that invest an afternoon in a test account remove weeks of integration support, because a client that can try the sequence themselves does not need to ask what it should be.

The same account makes the conformance run reproducible, which matters when a result is quoted. A conformance claim that cannot be re-run against a named commit is a claim about a Tuesday, and the reviewer across the table knows the difference.

What the 317 of 317 run taught

The last fifteen percent of failures in a full (g)(10) run are almost never the interesting architectural ones. They are declarations, required elements and includes: the difference between a server that returns each resource correctly and a server that performs the sequence a real client performs against a real patient. That distinction is worth carrying into any FHIR programme, because it changes what you test and who you ask to test it.

The run is also a milestone rather than an exam. Teams that schedule it in week two ship in month two, because the findings still change the design. Teams that schedule it as a final gate discover the same things with a customer attached, and the same findings cost far more.

Version seams and the next standard

Build for R4 and keep one seam: an adapter between your internal model and the FHIR representation. The eventual move to a later version is then a translation exercise rather than a rewrite, and the same seam absorbs profile variance when two counterparties require different extensions. Teams that map database columns directly onto resource elements discover the seam does not exist and cannot be added cheaply.

The seam is also where you can enforce the things the specification leaves to the implementer: which profiles are declared, which terminology is pinned and which search parameters are actually indexed. Putting those decisions in one place makes them reviewable, which is what turns a conformance claim into something a reviewer can check.

Working with counterparties who are behind

Most of the estate a new FHIR surface talks to is not itself conformant, and a client that sends US Core profiles to a server that has never heard of them will fail in ways that look like your defect. Build for the counterparties you have while keeping the seam for the ones you want, and record which profile each partner actually validated against.

Where a counterparty cannot meet the profile, the honest options are a documented relaxation for that partner or a translation layer at the edge. Both are defensible and only one leaves a record, and the record is what stops the relaxation from becoming the standard.

Keep a record of which profile version each partner validated against and when. The record is what lets a later change be assessed for its actual blast radius rather than for the size of the specification revision.

Proof

  • ONC (g)(10) self-test PASS, 317 of 317 subtests, April 2026 — public suite URL
  • 67 public hostnames, 51 serving, one AWS account
  • 104 clearinghouse enrolments across 53 jurisdictions
  • USCDI v7 proposal — Patient Digital Identity Credential — listed at isp.healthit.gov
  • SAM.gov UEI MVKQW9DKKGB8 · CAGE 20KM5 · NPI 1194645564
  • AWS Business Associate Addendum active since 18 August 2026

Where this work is delivered

Engagements run from Las Vegas and reach across the valley: clients in Las Vegas, NV; Las Vegas, Nevada; Henderson; Clark County, Nevada; and North Las Vegas, Nevada — plus remote delivery nationwide.

How an engagement runs

1. Trace one real case

Follow a single patient or claim end to end with a timestamp at every hop, before any document is written.

2. Written findings

A findings list naming the defect, the system, the owner and the date.

3. Monthly review

Review the list monthly and check whether the dates held.

Frequently asked

What is this service?

This page describes charles trinh — fhir r4 interoperability — the scope, the failure modes it prevents and how the work is measured. It sits in the Interoperability category at trinh.io.

Where is the work delivered?

Engagements run from Las Vegas, Nevada across the valley — Las Vegas, Henderson, North Las Vegas, Spring Valley, Paradise, Winchester, Sunrise Manor, Enterprise, Summerlin, Clark County — and remotely anywhere in the United States. On-site work in the Las Vegas valley, remote delivery elsewhere.

What does it cost?

Standard rate $195 per hour, committed $165 per hour at twenty hours or more per week for three months or more; $1,600 standard day, $1,350 committed day. Scoped in writing before work starts, fixed fee where the deliverable is defined.

Who is Charles Trinh?

A technology architect in Las Vegas, Nevada, working in healthcare identity, consent and interoperability. Not a physician, not a medical practice, and not affiliated with any other Charles Trinh.

How does an engagement start?

With one trace of a real case and a written findings list naming the defect, the system, the owner and the date. The first call is twenty minutes and free: 702-559-3489.

Engagement

Scoped in writing before work starts; fixed fee where the deliverable is defined. First call is 20 minutes and free. 702-559-3489 · charles@trinh.io

Talk to Charles

Record

Profile

Healthcare technology architect working where federal regulation becomes production infrastructure. Built and operates a FHIR R4 clinical platform — SMART App Launch authorization server, patient and provider portals, SQLite — on a hardened single-host AWS estate of 67 public hostnames. The platform returned a self-test PASS on the ONC (g)(10) Standardized API test suite in April 2026, 317 of 317 subtests, publicly verifiable. Designed a consent architecture in which a permissioned ledger records patient grant, refusal and revocation as immutable events while all clinical data stays off-chain and encrypted per patient — no PHI on the ledger by design. Filed a proposed USCDI v7 data element with ONC on verifiable patient identity. Translates the Federal Register, HHS/CMS guidance and HITRUST requirements into architecture, controls and audit evidence.

Selected engineering scale

  • Self-managed estate of 67 public hostnames, 51 serving, on one AWS account.
  • FHIR R4 API with a SMART App Launch authorization server; patient and provider portals; an AI service layer; Hyperledger Fabric peer, orderer and CA; IPFS; a 337-table SQLite clinical store.
  • Two permissioned distributed ledgers in production, one operating a consent registry that records grant, refusal and revocation with no PHI on chain.
  • X12 005010 pipelines across eight transaction sets.
  • Designed, built, deployed and operated end to end by one person.

Credentials and standards work

  • Inferno (g)(10) self-test PASS, 317 of 317 subtests, April 2026 — the ONC Standardized API test suite. Standards validated: US Core 6.1.0, USCDI v3, SMART App Launch 2.0.0, Bulk Data 2.0.0. Public suite URL.
  • USCDI v7 public comment submission, April 2026 — proposed Patient Digital Identity Credential under the Provenance data class, listed publicly on the ONC Interoperability Standards Platform.
  • SAM.gov UEI MVKQW9DKKGB8 · CAGE 20KM5 · NPI 1194645564. AWS Business Associate Addendum active 18 August 2026.

Technical environment

AWS us-east-1 (compute, storage, messaging, SMS, email, voice) · Linux and systemd hardening · nginx · Fabric and EVM ledgers · FHIR R4 · X12 005010 · HL7 · SQLite · self-hosted mail, drive, search and monitoring · Git-based change control and CI release gates.

TRINH — Charles Trinh, technology architect
TRINH

Charles Trinh

Fractional CTO/CIO and Technology Architect — Healthcare Data, Identity & Regulated Infrastructure · Las Vegas, NV

Las Vegas, Nevada · US citizen · Fractional and project engagements

(702) 559-3489 Email Charles Trinh