ADT · ORU · ORM · MLLP · ACK

Charles Trinh — HL7 v2 Interfaces | Las Vegas, NV · Las Vegas, Nevada · Henderson · Clark County, Nevada · North Las Vegas, Nevada

A service card reading HL7 v2 — ADT, ORU and ORM over MLLP with real acknowledgement handling
Charles Trinh — HL7 v2 Interfaces

HL7 v2 still runs the hospital

FHIR is the future and v2 is the present. ADT feeds, order messages, results, scheduling and documents still move through interface engines built on v2, and any realistic integration strategy treats v2 as a permanent fixture rather than a legacy to be retired on a roadmap slide.

The practical work is therefore not choosing between them. It is building a translation layer that does not lose clinical meaning in either direction, and operating it as a product rather than a script.

The message set that carries the load

ADT for admissions, discharges and transfers — the event stream every downstream system depends on, and the one whose ordering assumptions cause the most damage. ORM and ORU for orders and results. SIU for scheduling. MDM for documents. RDE and RDS at the margins for pharmacy. The rest exist and matter occasionally; these four families carry the overwhelming majority of traffic and almost all of the failures.

Z-segments are where the real information lives

Local extensions carrying site-specific meaning. No standard parser helps you with them, no public documentation describes them, and they frequently contain the clinical detail that makes a message useful — a triage note, a scheduling constraint, a payer-specific flag.

Enumerate them by interface and by sending application. Document the field meanings against a real sample. Then map them deliberately into FHIR extensions, or drop them consciously with a written reason. A translator that silently discards Z-segments is a translator that silently discards the interesting half of the message.

Identifier types are the matching trap

The same patient carries an MRN, an account number, a visit number, and potentially a payer member identifier, and v2 does not consistently tell you which is which across sending systems. Field CX-5 carries the type, but the codes are applied inconsistently and some senders populate the same identifier under different types on different message classes.

Getting this wrong produces exactly the failure the identity layer exists to prevent: automatic assignment of an encounter to the wrong record, with the system reporting success. Build an identifier normalisation step that maps each sending system's conventions to your own authority, and log the unmapped cases rather than dropping them.

Trigger events without ordering guarantees

An A08 update arriving out of order is normal, not exceptional. Networks reorder, queues retry, and a sending system that was down replays its backlog. A translation layer that assumes ordering produces a record reflecting the wrong version of the truth, and it does so quietly.

The fix is a monotonic marker — message control ID plus the sending system's event timestamp — and a rule that rejects or defers a message older than what has already been applied, with a dead-letter path for the ones that cannot be ordered.

ACK and NACK, at both levels

Transport-level acknowledgement tells you the socket accepted bytes. Application-level acknowledgement tells you the receiving application processed the message. An interface that logs transport success and fails application processing while recording no error has an error rate of zero and a data quality problem nobody can see.

Distinguish accept, reject and error, log the acknowledgement code and text per message, and expose a per-connection count of each. That single dashboard replaces most of the arguments about whether an interface is working.

The mapping into FHIR, deliberately

Do not translate field by field. Map to a model you own, then render FHIR from that model. Field-by-field translation produces resources that satisfy the letter of a profile and carry no clinical semantics, and it makes a second target — an analytics store, a second FHIR version — a second full project.

Operate it as a product

Versioned mappings, replay, dead-letter queues, per-connection monitoring, and a runbook that answers the only question anyone asks at 3 a.m.: what was the last successful message on this interface, and when.

Interfaces that live in a cron job and a Python file are the ones that fail on a holiday with nobody able to say how long they have been down or what has been lost in the interval.

What I do

I have built and operated interface layers feeding a 337-table clinical store, with X12 and FHIR surfaces on the other side of the same normalised model. I do not sell an interface engine. I fix the translation model and then make the engine observable.

Throughput, back-pressure and the queue you did not build

Interface traffic is bursty. An ADT flood during a downtime catch-up, a lab analyser replaying a backlog, a scheduling system resending after a restart — each produces a spike that a synchronous converter handles badly and a queue handles well.

Put a durable queue between receipt and processing. Acknowledge at the transport layer as soon as the message is persisted, then process asynchronously with retry and a dead-letter path. That ordering is what lets you answer the only question anyone asks during an incident: what has been received but not yet applied, and where is it.

Back-pressure matters too. An interface that accepts unlimited messages during a downstream outage converts a database problem into a memory problem. Bound the queue, reject beyond the bound with a NACK that the sender can retry, and log the rejection rather than dropping it.

MLLP is unencrypted by default

Lower Layer Protocol wraps a message in a frame and sends it over TCP. There is no authentication and no encryption in the base protocol. Most modern deployments wrap it in TLS, and the ones that do not are frequently the ones nobody has looked at since the interface was built.

Enumerate every MLLP listener, confirm which are TLS-wrapped, and confirm which have any peer authentication at all. An interface accepting messages from any host that can reach the port is an unauthenticated write path into a clinical record, and it is far more common than it should be.

The migration question, answered honestly

FHIR does not replace v2 by decree. What actually happens is that new integrations are built on FHIR, existing v2 interfaces are maintained until the vendor retires them, and the two coexist through a normalised model you own.

The mistake is to attempt a translation layer between v2 and FHIR that goes field to field. Map both into a model you control, then render either format from that model. Field-to-field translation produces FHIR resources that satisfy a profile and carry no clinical semantics, and it makes a third target — an analytics store, a second format version — a third full project.

Ownership and the runbook

Every interface needs a named owner, a documented message flow, a sample of a real message, and a stated maximum tolerable gap. When that interface goes quiet, the runbook should say what to check first and who to call. Absent that, the first hour of every interface incident is spent discovering who built it.

What I do

I have built and operated interface layers feeding a 337-table clinical store, with X12 and FHIR surfaces on the other side of the same normalised model. I do not sell an interface engine. I fix the translation model and then make the engine observable, because an interface that works and cannot be monitored is an interface that will fail silently.

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.

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

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 — hl7 v2 interfaces — 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.

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