REMOTE: A Framework to Create Fast Healthcare Interoperability Resources (FHIR) from Unstructured Clinical Data

Area of Science:

• Health Informatics
• Clinical Data Management
• Biomedical Data Standards

Background:

• Unstructured clinical notes in electronic health records (EHRs) contain valuable patient information.
• The unstructured nature of clinical notes hinders their use in population health, genomic research, precision medicine, quality improvement, and clinical decision support.

Purpose of the Study:

• To introduce REMOTE, a systematic framework for generating Fast Healthcare Interoperability Resources (FHIR) from unstructured clinical notes.
• To demonstrate the feasibility of transforming clinical notes into interoperable and actionable data.

Main Methods:

• Developed REMOTE, a stepwise pipeline: Reformat, Extract, Map, Transform, Exchange.
• Processed 1,618 unstructured clinical notes.
• Extracted chief complaints, signs/symptoms, diagnoses, and medications.
• Mapped concepts to SNOMED CT codes.
• Converted data into FHIR Observation, Condition, and MedicationRequest resources.
• Uploaded resources into FHIR repositories.

Main Results:

• Successfully demonstrated the feasibility of the REMOTE framework.
• Generated FHIR resources from unstructured clinical notes.
• Enabled data exchange across health repositories.

Conclusions:

• Information in free-text clinical notes can be transformed into discrete FHIR resources.
• This transformation makes clinical data interoperable, exchangeable, and actionable.
• REMOTE facilitates the utilization of underutilized patient information within EHRs.