The Data Science team at the Southern California Clinical and Translational Science Institute (SC CTSI) recently collaborated with multiple universities, as well as industry and healthcare organizations, to publish a manuscript in the journal Science. The authors argue that real-world health data, like electronic health records, wearable data, and patient-generated data, should be managed as a public utility, similar to infrastructure like water or electricity.

Patient-related health data and information is scattered across many separate systems that don’t easily connect or share data, such as various Electronic Health Record systems. This siloing leads to delays in research and innovation, and has limited benefit for patients. The authors argue against real-world health data being managed as fragmented, privately controlled assets.

"When health data is disconnected and siloed, clinicians and researchers don't see the whole story which leads to missed diagnoses, duplicate tests, and medical errors," said Dario Kuzmanović, MHSc, Program Director of Data Science at the SC CTSI. "If we addressed this gap, we would truly have patient-centered care.”

They proposed a solution in the manuscript to tackle persistent issues like data silos, patient privacy, and similar challenges. They used historical examples of public utility development, such as electrification, federal drinking water standards, and the rise of the internet, and suggest that adopting a public utility model could improve how health data is governed and ensure it flows where it’s most needed–from individuals to innovative solutions, and back again.

The SC CTSI Data Science team first got involved with this project when their program director, Dario Kuzmanović, MHSc, served on a working group with the Principal Investigator Melissa Haendel, PhD, FACMI. This working group is called the Real World Data Workforce Development Across the Translational Spectrum and is hosted by the National Center for Advancing Translational Sciences, the entity that funds the Clinical and Translational Science Awards. As part of this working group, they were charged with creating a Real World Data (RWD) needs assessment, resource inventory, and designing a generalizable competency framework. RWD is information about patient health status or healthcare delivery that is routinely collected from various sources outside of conventional randomized controlled trials. The working group explored topics like data collection and provenance, clinical data models, evaluation of bias in the data, and best practices and analytic approaches that support ethical research analytics using RWD, with a focus on Artificial Intelligence and machine learning modeling. The Working Group also established a training community to assist with dissemination.

In the future, the working group plans to continue their ongoing collaboration through the RWD Workforce Development Across the Translational Spectrum at NCATS.