The Future of Precision Health: Bridging Clinical Care and Everyday Life

Just attended an eye-opening session on CHIL 2025 - Association for Health Learning and Inference UCSF × UC Berkeley Computational Precision Health with Dr. Ida Sim. Three key insights emerged that are reshaping how we think about healthcare delivery in the digital age.

About Dr. Ida Sim

Dr. Ida Sim is UCSF's inaugural Chief Research Informatics Officer, Professor of Medicine and Computational Precision Health, and Co-Director of the UCSF UC Berkeley Joint Program in Computational Precision Health. She is a global leader in the technology and policy of large-scale health data sharing and co-founder of the Open mHealth/IEEE 1752 family of global mobile health data standards. She is also co-founder of Vivli, the world's largest data sharing platform for participant-level clinical trial data.

1️⃣ Healthcare Beyond Hospital Walls: Precision Health in Everyday Settings

While medical innovation thrives in hospitals, precision health's full potential requires systems supporting care across everyday settings—where much of health actually happens. As Dr. Sim notes, "Ninety-nine percent of the time patients are by themselves. They're their own chronic care physician."

This insight is particularly critical given that nearly 40% of the U.S. population has multiple chronic conditions. Dr. Sim emphasizes, "With sensors and with apps, we can be out there. We can get beyond the walls of the clinic or the hospital." The paradigm shift recognizes that effective healthcare must extend beyond episodic clinical encounters to continuous monitoring and support in patients' daily lives.

2️⃣ The Rise and Challenges of Patient-Generated Data (PGD)

Patient-Generated Data encompasses health-related information created, recorded, or gathered by or from patients through sensors and mobile applications. This data stream is transforming healthcare by providing continuous insights into patients' health status between clinical visits.

However, significant challenges remain. PGD remains personally identifiable and fragmented compared to clinical EHRs, with current integrations tending to be one-off efforts entailing high costs to build and maintain custom connections with each device and their proprietary data formats.

The integration of PGD into clinical systems requires a comprehensive pipeline that includes acquisition (the person-facing component with sensors and apps), aggregation (pooling, standardizing, and structuring data), and consumption (use by different solutions in different care settings for clinicians and patients).

The Critical Role of Metadata and Context

Clinically relevant contextual information is necessary for making clinical decisions. For example, a blood glucose reading of "138" is clinically meaningless unless the units, relationship to meals or sleep, and effective time are made clear. Beyond the data itself, metadata—such as the device name, model, unique ID, and the app that acquired the data—is essential for quality assurance, auditing, and regulatory oversight.

3️⃣ Emerging Solutions: Standards and Open Platforms

IEEE P1752 Standards Family

The IEEE P1752 family of standards defines specifications for standardized representations for mobile health data and metadata. IEEE 1752.1-2021 was officially published in September 2021, covering metadata, sleep, and physical activity measures. Work continues with IEEE P1752.2, which addresses cardiovascular, respiratory, and metabolic measures.

The purpose is to provide standard semantics to enable meaningful description, exchange, sharing, and use of mobile health data, making data aggregation across multiple mobile health sources easier and more accurate while reducing the costs of using this data for biomedical discovery, improving health, and managing disease.

Open mHealth schemas are open source, free to all, and are the output of a global community of stakeholders consisting of developers, data scientists, informaticians, researchers, and clinicians.

JupyterHealth: Open-Source Middleware for Composable AI

JupyterHealth is an open-source, modular platform that eliminates data silos by integrating real-time health data from wearables, IoT devices, mobile apps, and Electronic Health Records into a unified, secure, and AI-powered ecosystem.

The platform is modular, consisting of open-source components that span the entire lifecycle of digital health research, development, and deployment—from data ingestion involving wearable and clinical data to advanced data analysis and presentation via JupyterHub. Each component is designed to operate independently or in conjunction, supporting rigorous health data standards like HL7 FHIR and Open mHealth.

JupyterHealth is currently in early development with use cases drawn from real-world primary and subspecialty care from UCSF and Weill Cornell, with a participatory design process that includes frontline clinicians and lay users of wearable sensors.

Real-World Implementation: Agile Metabolic Health Project

The Open Platforms for Health initiative, led by Dr. Sim and Dr. Fernando Pérez of UC Berkeley, is using JupyterHealth to transform data collection and interpretation in the treatment of metabolic disorders. The Agile Metabolic Health Project compiles and analyzes continuous glucose monitoring data and other health metrics, like sleep and exercise, from patients with diabetes, using AI to learn better ways to manage diabetes in real time.

The Path Forward: Integration and Impact

Dr. Sim envisions JupyterHealth accelerating companies that draw on multiple sensor streams and electronic health record data to generate individualized metabolic profiles and personalized decision support for patients. By lowering the technical barrier to entry and backend operating costs while creating an open community for sharing and co-creating new computational approaches, more robust digital health markets can emerge that will impact patient outcomes.

However, significant challenges remain. Clinical workflows are currently anchored in EHR systems, which often limit the ability of third-party computational tools to access care context or write back into the official medical record. While FHIR and SMART-on-FHIR standards are important mechanisms for enabling greater workflow integration, they are not by themselves sufficient to ensure large-scale effective integration into everyday care.

The integration of everyday health data with clinical systems represents the next frontier in precision medicine, bridging the gap between healthcare facilities and our daily lives 🩺💫

Sources

• UCSF Medicine Department. "Ida Sim, PhD, MD." https://medicine.ucsf.edu/people/ida-sim
• UCSF Precision Medicine. "Ida Sim, PhD, MD." https://precisionmedicine.ucsf.edu/people/ida-sim-phd-md
• JupyterHealth. "About." https://jupyterhealth.org/about/
• Zeng et al. "Standardized Integration of Person-Generated Data Into Routine Clinical Care." JMIR Mhealth Uhealth. 2022. https://mhealth.jmir.org/2022/2/e31048
• IEEE P1752 Working Group. https://sagroups.ieee.org/1752/
• Open mHealth. "IEEE 1752.1 Standard Published." https://www.openmhealth.org/ieee-1752-1-standard-published/
• UC Berkeley CDSS. "Ida Sim: A healthcare pioneer transforming how data works for patients." https://cdss.berkeley.edu/news/ida-sim-healthcare-pioneer-transforming-how-data-works-patients
• PMWC. "Interview Questions for Ida Sim." https://pmwcintl.com/interview-questions-for-ida-sim/

---

BERSUI Inc. supports corporate productivity improvement and innovative business creation through standardization of AI-driven business operations.