INTEGRATION

Charles Trinh — EHR Data Migration | Las Vegas, NV · Las Vegas, Nevada · Henderson · Clark County, Nevada · North Las Vegas, Nevada

Diagram of a migration pipeline with source extraction, staging, transformation, reconciliation and cutover
Charles Trinh — EHR Data Migration

Data migration is the part of an EHR programme that cannot be retried. I run migrations as a sequence of rehearsals, each a dry run against a fully populated copy, with the cutover measured against a reconciliation report that the clinical team signs.

Why migration is harder than integration

An integration runs in parallel with the source system; a migration replaces it. The source side stops being authoritative at a defined moment, . The consequences of getting it wrong are not a failed interface that retries; they are a lost chart, a misaligned medication list or a history that a clinician cannot open. The severity is why I refuse to run a migration as a single cutover and insist on a sequence of rehearsals, each one a full dry run, each one measured against a reconciliation report the clinical team has signed.

The rehearsals are expensive in time and modest in direct cost. Each rehearsal runs against a fully populated copy of the source system, produces the target system, and is timed and measured. The rehearsals also run the cutover runbook end to end, so the cutover itself is the rehearsal that the team has already executed four or five times. This is the difference between a cutover that completes in the planned window and one that runs into the morning rounds. The customer pays for the rehearsals and the cutover is almost free.

The rehearsal also tests what the cutover will actually stress: the downstream systems, the printed reports, the clinician workflows and the patient-facing channels. A rehearsal that only validates the database is a rehearsal for half of the problem. I include at least one clinician review session per rehearsal, where a named clinical lead walks a sample of patients through the target system and reports back against a structured questionnaire. The questionnaire is reused between rehearsals, so the drift is visible.

Extraction: reading the source without breaking it

The source system is a production EHR, and the extraction has to read it without affecting its performance. I never extract from the primary database directly. The first step is a read replica, or a nightly logical dump, or a vendor-supported extract with a defined scope. The choice depends on what the customer's vendor allows. Every vendor I have worked with has a position on extraction, and the position is in the vendor agreement, not in the developer documentation. I read the agreement before designing.

The extraction scope is the harder decision. The clinical records are the obvious part; the patient identity, the scheduling history, accounts, documents and audit trails are also part of the record to keep. I list every table or table group the target needs and every one the customer's compliance team has a retention requirement on. The two lists are rarely the same. The reconciliation report covers the union of the two; tables that exist only in the compliance list still get migrated, and tables that exist only in the target list are migrated only if they can be mapped without loss.

The extraction has to be repeatable and incremental. A full extraction for each rehearsal takes hours on a mid-sized source; incremental extraction between rehearsals keeps the delta small. I build the extraction as a delta every time after the first. The delta extraction is the source system's own changed-records log if the vendor provides one, or a timestamp-based scan against a table-level watermark. The watermark is documented, because the first problem I see when a vendor changes their schema is a watermark that points at a column the vendor renamed.

Transformation and the mappings that decide what survives

The transformation is where the customer's clinical and operational knowledge lives. A source field that looks unambiguous ("primary diagnosis") becomes a multi-valued target field with a hierarchy and a timestamp, and the decision about which value becomes primary is the customer's, not the engineer's. I collect every such decision in a mapping document, one row per field, with the source representation, the target representation, the transformation rule and the data steward who approved it. The data steward is named, not generic. Decisions have owners.

The transformation has to handle nulls, ambiguity and truncation explicitly. A source field that is allowed to be empty in the source, and required in the target, needs a decision: drop the record, populate a default, or hold it for review. The default decision I have seen most is populate a default, and that default decision gets caught in the reconciliation as a clinical surprise. I write the null decisions into the mapping document, and the reconciliation report counts every defaulted value by field, so the customer can see the scale before the cutover.

The third case is truncation. A source field that allows more characters than the target, or a value that the target's code set does not accept, gets truncated or rejected. I prefer rejected. A rejected record that sits in a review queue is visible and recoverable. A truncated value is a silent clinical error. The review queue is part of the rehearsal's output, and the customer's data steward clears it during the rehearsal.

The reconciliation report the clinical team signs

A migration is not done when the target system loads. It is done when the clinical team has read a reconciliation report that proves the load is correct. The reconciliation report is a set of counts that compare the source and the target at the same point in time: patients, encounters, diagnoses, orders, results, documents, medication records. The counts are granular enough that a difference is diagnosable, not just noticeable. A breakdown by month tells the steward which records need investigation.

The report also includes the sampled clinical content. A random sample of patients, weighted to include the longest records, the most comorbidities and the most recent activity, is compared field by field in a printable format the clinical lead can annotate. Each field with a difference has a comment line. The clinical lead walks the sample and writes the comment: "acceptable, the medication list has been re-sequenced", "not acceptable, the allergy is missing from the target". The comments become the action list for the next rehearsal.

The sign-off is not a single signature by a director. It is a signature by the clinical lead for each specialty area the migration touches. The medicines, the diagnostics and the emergency department have different tolerances. A migration that five directors approved and the ED clinical lead has not seen is not ready. I insist on the specialty sign-off and I have never had a customer regret it; two customers have caught a specialty-specific problem in the sign-off that the rehearsal's summary would have missed.

Cutover, downtime and the runbook that decides the night

The cutover is one night, and the runbook is what stands between it and a morning of clinician frustration. The runbook is written before the first rehearsal and refined after each one. It has a timeline with minute-level tasks, a named owner for each task, a rollback point for every irreversible action and a communications plan for every stakeholder group. It has the contact list for every vendor the customer depends on, and the escalation path if a vendor does not answer. I have seen cutovers fail because the source system's vendor was out of office and the customer's engineer needed an account enabled.

The downtime plan is part of the runbook. The customer's clinical staff have to continue working during the cutover, and the workaround procedure (paper, a read-only view, a documented manual process) is agreed with the clinical leads before the rehearsals begin. The workaround procedure is tested in at least one rehearsal. A workaround that nobody has tested is a paper procedure written by a project manager. A workaround that the ED has tested is a working procedure that the ED will follow.

The rollback point is defined per phase. The cutover has phases: extraction frozen, source changes frozen, target loaded, reconciliation re-run, systems flipped. Each phase has a commitment that the team is not switching back without a specific trigger. The trigger is defined in advance. I have seen a cutover where the team panicked after a failed reconciliation that turned out to be an extraction artefact; they rolled back unnecessarily and lost a day. This is preventable with a clear criterion.

What the migration leaves behind and what you pay for

The migration leaves a set of artefacts: the extraction specification, the mapping document, the reconciliation reports, the cutover runbook and the post-migration defect list. They also want the source system's read-only view maintained for a settlement period, typically 90 days, during which the data steward can chase any missed record. The read-only view has a cost; the reconciliation report has made it shorter than it would otherwise be, but a settlement period is prudent. See the master patient index work for how the migrated identity records are kept consistent afterwards and the EHR vendor selection work if you are choosing the target system rather than replacing one you already run.

The cost of a migration is dominated by the rehearsal count, the data steward's time and the clinical leads' time. A realistic estimate for a mid-sized community hospital is four to five rehearsals over a quarter. The engineering cost is one substantial engineer per month. I quote the engineering and the rehearsal separately; the customer sees both and decides the rehearsal schedule against the clinical leads' availability. The migration is a programme deliverable, not a database project, and the budget should reflect that.

The last thing I leave behind is the customer's own migration competence. At least two of the customer's engineers are on every rehearsal, and one runs the final cutover with me watching. This is deliberate. The customer should never need to call me back for a second migration because they could not run one. The handover is the migration's final test.

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 — ehr data migration — the scope, the failure modes it prevents and how the work is measured. It sits in the Security and Compliance 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