Electronic health records (EHRs). Insurance claims. Patient registries. Wearable tech. Never before has health data been so abundant.
However, the United States continues to struggle with siloed health data systems that make it difficult for critical data to be streamlined and used for the advancement of science and medicine-particularly public health, safety monitoring, and precision medicine.
A multidisciplinary team of researchers, clinicians, technologists, and policy experts, led by the UNC School of Medicine, have published a paper in Science calling for real-world health data (RWD) to be governed as a public utility, an essential public infrastructure akin to electricity, water, and the internet.
Melissa Haendel, PhD, FACMI
"Health data are now as fundamental to society as power grids and clean water," said Melissa Haendel, PhD, the Sarah Graham Kenan Distinguished Professor of genetics and first author on the paper. "Yet unlike other essential utilities, health data infrastructure is largely voluntary, fragmented, and economically misaligned. We need a model that treats data as shared infrastructure, governed for the public good, while protecting patient autonomy and trust."
To address data silos, patient privacy, and other related challenges, researchers looked to lessons learned from the creation of public utilities throughout history, such as the electrification movement, federal drinking water standards, and the advent of the internet.
They propose that a public utility model could help better govern health data and get data to where it matters: from the people and to new and innovative solutions and back.
"We take for granted having ready access to high-quality clinical data, but that access really depends on the existence of consistently funded, reliable, and secure data infrastructure," said co-author Emily Pfaff, PhD, who is an associate professor of medicine at the UNC School of Medicine.
"On the research side, that consistent funding is not guaranteed, leading to eventual degradation or disappearance of important data repositories. To prevent this, we can learn from the public utility model, which comes with a guarantee of that long-term infrastructure investment."
Emily R. Pfaff, PhD, assistant professor in the Division of Endocrinology and Metabolism at the UNC School of Medicine
The public utility approach would ensure that there is governmental regulatory oversight, ensuring fair access, transparency, and accountability for health data.
Governance would be distributed to utility boards, while the government would establish and enforce regulations and standards, enabling individual research entities and stakeholders the freedom to address their own research questions and challenges. Like other public utility boards, this information could be accessed by patients and the community, ultimately empowering with a formal role in oversight and data sharing.
The researchers also argue that current U.S. privacy laws, including HIPAA, are insufficient for today's digital health ecosystem. They call for extending HIPAA-like protections to include consumer health data (including wearables, genomics, and mobile apps), clarifying public health and research use cases, and more.
"HIPAA was written before mobile health apps, genomic sequencing, and AI-driven analytics," said Haendel. "We must modernize privacy protections while ensuring that patients, not just institutions, benefit from the evidence their data generate."
Haendel and the researchers also recognized that transforming health data governance will require both legal reform and cultural change. They underscore that healthcare organizations and policy-makers must recognize health data as shared infrastructure with collective value, not just a secondary byproduct of clinical care.
"To fully realize precision medicine, drug safety, and population health for everyone, we must treat real-world data as an essential resource," Haendel said. "A government-regulated, community-owned, federated model can balance innovation with trust, without resorting to centralized data ownership."
