JMIR Publications

ABSTRACT

Objective:

The complexity of the Fast Healthcare Interoperability Resources (FHIR) standard limits its direct usability for clinicians despite its transformative potential in healthcare data management. To bridge this gap, we aimed to describe the development of an interactive dashboard enabling non-technical users to intuitively build and analyze oncologic patient cohorts. By leveraging FHIR, we aimed to enhance data accessibility and interoperability in clinical practice.

Methods:

DermaDashboard builds on a Structured Query Language (SQL) database using a relational FHIR model, which ensures data compliance with the FHIR schema. A materialized view was assembled and optimized performance by providing only relevant data. The user interface was built with Grafana and supports intuitive data exploration. We applied DermaDashboard to the use case of melanoma, demonstrating its utility in real-world oncologic cohort analyses.

Results:

DermaDashboard was successfully built and integrated into the clinical environment, identifying 3,949 melanoma patients and corresponding to 82,783 electronic health records. The primary FHIR resources used were Patient, DiagnosticReport, and QuestionnaireResponse, and captured 54 data attributes, including demographics, histological classifications, genetic mutations, clinical and pathological staging, treatments, and procedures. Clinicians can filter the data using 29 variables to create specific subcohorts. The dashboard also enables operational insights by tracking annual trends in procedures and drug administrations.

Conclusions:

DermaDashboard enhances data accessibility for non-technical clinical users while showcasing the power of FHIR standardization in healthcare applications. By enabling oncological insights and identifying cohort discrepancies, it enhances both clinical decision-making and data quality.