Enabling the Use of Standardized Clinical Data to Advance Discovery
Enabling the Use of Standardized Clinical Data to Advance Discovery
On April 22, 2024, HSI, in collaboration with the Martin Trust Center, presented the next seminar in their “Sparking the Data Revolution in Healthcare” series. The speaker, Teresa Zayas-Caban, PhD, Assistant Director for Policy Development at the National Library of Medicine (NLM), focused on the role of FHIR (Fast Healthcare Interoperability Resources) in facilitating the sharing and utilization of healthcare data. Advancing standardized clinical data and data for discovery has been a central focus of Dr. Zayas-Caban's work at both the Office of the National Coordinator for Health IT (ONC) and the National Library of Medicine (NLM) at the National Institutes of Health (NIH).
Advancing Standardized Clinical Data
Dr. Zayas-Caban's work at ONC emphasized the office's role in certifying health information technology systems. While certification is voluntary for developers, healthcare providers are required by agencies such as the Centers for Medicare and Medicaid Services to purchase certified systems to participate in specific programs.
Notably, this regulatory lever mandated the availability of patient data through APIs using modern FHIR standards. This requirement applies to health providers and health insurance companies. A significant achievement driven by ONC's work was making doctors' notes available via APIs and portals, a development heavily supported by patient advocacy groups. Doctor's notes contain crucial details for the continuity of care and patient/care partner management, and their availability, even if unstructured, enables applications such as AI analysis.
The NLM, where Dr. Zayas-Caban currently works, also plays a vital role in standards development. NLM manages and maintains the standard technology RXNORM (for drugs) and supports the standards SNOMED (for diagnoses) and LOINC (for laboratory tests and results). These controlled vocabularies are essential for interoperability, ensuring data are consistently coded and interpreted. NLM integrates standards into its products, such as providing a FHIR API for ClinicalTrials.gov.
Historically, standardizing communication between payers and providers has been challenging, dating back to the implementation of HIPAA in the 1990s. Current efforts, influenced by Dr. Zayas-Caban's work at ONC, include regulations anticipated for 2026 or 2027 that aim to facilitate data sharing between payers and providers to automate prior authorization. Population-level data standards, such as those developed in collaboration with Boston Children's Hospital, are crucial to automating this process.
Advancing Data for Discovery
At ONC, Dr. Zayas-Caban and her team spearheaded ONC's involvement in several programs, including the Precision Medicine Initiative, specifically the Sync for Science pilot and the Sync for Genes project. Her team also managed a $20 million portfolio of Patient-Centered Outcomes Research projects funded by the Affordable Care Act.
Moving to the NLM, Dr. Zayas-Caban works within an institution that is a leader in computational health, data science, research, and informatics. NLM's goal is to facilitate the interoperable movement of data and information across systems for research, discovery, and the improvement of health.
NLM influences the advancement of data for discovery by engaging in standards development with community input and encouraging the broader use of standards in research. They have published guide notices encouraging NIH-funded researchers to leverage FHIR to exchange research data and obtain clinical data. NLM's strategic vision emphasizes advancing data science, open science, and biomedical informatics while accelerating data-driven discovery toward a sustainable digital ecosystem.
The speaker and the subsequent discussion also touched on the need for policy to evolve in response to the rapid growth in AI, real-world evidence, and patient-generated health data. The conversation also addressed transparency in AI model development and data, privacy concerns, and how to conceptualize results while still promoting innovation.