Fabrício Kury, MD · New York, NY

Health data, made usable.

I'm a physician and health-data scientist. I help teams get access to claims and EHR data, map it and analyze it, with 10+ years inside the CMS VRDC and the OMOP Common Data Model.

Portrait of Fabrício Kury
10+ yrshands-on with CMS VRDC Medicare claims
87M+patients' claims under my custody at NIH
MDphysician trained, informatics postdocs at NIH and Columbia
OMOPETL, data quality and OHDSI tooling since 2015

Flagship service

CMS VRDC Data Use Agreements

I write your CMS Data Use Agreement from start to finish, including the study design, the IRB protocol and the DUA request itself, and see it through to approval by the IRB and the CMS Privacy Board.

  • Secured the CMS Privacy Board approval that gave Sparx its first VRDC access (May 2026)
  • Rebuilt Wellvana's DUA to cover files the team was eligible for but had never requested
How it works
  • Study design and minimum-necessary data scope
  • IRB protocol and submission (Sterling IRB)
  • DUA request package and supporting documents
  • Answering CMS questions through to approval

Also available

Data analytics & engineering

Data-model mapping & OMOP ETL

Move data from where it lives (FHIR, CCLF, RIF or a homegrown schema) into the model you need, OMOP CDM or your own, with the vocabularies mapped correctly.

Claims & EHR reporting

Recurring or one-off reports and pipelines on Medicare and Medicaid claims (RIF, CCLF) and EHR data, built by someone who knows what the codes mean clinically.

VRDC setup, DUA audits & training

Get a new VRDC team productive fast. I set up the workspace, check your DUA's scope against what CMS offers today, and train your team hands-on on RIF data.

Recent work

A startup's first VRDC access

At Sparx I wrote the study, got it through the IRB and drafted the DUA request. The CMS Privacy Board approved it in May 2026.

Annual Wellness Visits, nationwide

A 100% RIF study for Wellvana of 26–27M beneficiaries per year-pair. AWVs were associated with $923–$1,271 lower total cost of care the following year.

FHIR → OMOP, continuously

Led a small team that built a cloud ETL from FHIR endpoints into OMOP CDM v5.4, with 10,000+ patients and ongoing ingestion.

More work and open-source projects →

Have a project in mind?

Email me a few lines about your data, your question and your timeline. I reply personally.