Skip to main content

FHIR Preprocessing

Overview

Code on Github

In the Tuva Project, we have built a FHIR preprocessing mart designed to streamline the transformation of healthcare data into a FHIR-ready format. This mart transforms data from the Tuva Core Data Model into standardized tables aligned with FHIR resources, ready to be exported as CSVs, making it easier to generate valid FHIR patient bundles.

To support this pipeline, we will be releasing a companion open-source Python library that helps convert the exported CSVs into fully structured FHIR patient bundles.

Together, these tools are built for flexibility, enabling seamless integration into ETL workflows for anyone adopting FHIR for healthcare data exchange and interoperability.

Stay tuned for more updates on the open-source initiative and an announcement regarding our soon-to-be-released turnkey solution for NCQA-certified HEDIS measures.

How to run the mart

This data mart is disabled by default. To run the data mart, simply add the variable fhir_preprocessing_enabled to your dbt_project.yml file or use the --vars dbt command.

dbt_project.yml:

vars:
fhir_preprocessing_enabled: true

dbt command:

dbt build --select tag:fhir_preprocessing --vars '{fhir_preprocessing_enabled: true}'

Data Dictionary

The FHIR Preprocessing mart outputs a table per FHIR resource.

condition

Clinical data sources only.

Column NameData TypePrimary KeyDescription

coverage

Claims data sources only.

Column NameData TypePrimary KeyDescription

explanation_of_benefit

Claims data sources only.

Column NameData TypePrimary KeyDescription

medication_dispense

Clinical data sources only.

Column NameData TypePrimary KeyDescription

observation

Clinical data sources only.

Column NameData TypePrimary KeyDescription

patient

Column NameData TypePrimary KeyDescription

procedure

Clinical data sources only.

Column NameData TypePrimary KeyDescription