A Billion Points of Data… and a Trillion Reasons for Excitement for UCSF Researchers!
[email protected]@atulbutte
Atul Butte, MD, PhDDirector, Bakar Computational
Health Sciences InstituteUniversity of California, San Francisco
We Heard You Last YearWe’re Still Listening. Slido today.
• New Services:‒ Lots of new workshops, training courses (Library,
Bakar Institute)‒ Consultation for APeX-enabled research (CTSI)‒ NLP symposium, training, nlp.ucsf.edu, Slack online
community• New Data: Geocoded address, CA Death Registry, ZSFG data,
other data from Dept. of Public Health• (Very) distinguished computational speakers: Ed Boyden,
Sandrine Dudoit, Andrew Moore, Peter Norvig• New Tools and Platforms
It’s an Exciting Time to be a Clinical Researcher!
3
Clinical Devices go Consumer GradeSystems getting more manageable
Industry partnerships are booming
Need more & better ones
AI opening up new avenues
Clinical Notes for Information Commons
4
RadiologyNotes
PathologyReports
NurseNotes
EncounterNotes
OrderNarratives
OrderImpressions
PatientEmails
DoctorApprovals
SecureHIPAA-
compliantServer
Info Extraction
Information Commons
NLP
71 million notes at UCSF
200 department specialties
150 provider specialties
“…there is great potential [in clinical notes] to inform treatment choices in ways that improve patient care and health outcomes”
- S. Schneeweiss, N Engl J Med 2014 nlp.ucsf.edu
Imaging Commons – On Our Way
Images
EMR
NLP
ML
Information Commons
6 million exams1.5 billion images
12000 new exams/week
Total: One Petabyte Notes
EMR
PatientExploreR
>
PatientExploreR: Deidentified Patient Data Brought Alive
Tableau: Analytics at Everyone’s FingertipsWide Availability in Fall
Scaling up to the Cloud for Real
8
AWSAmazon
responsible for security OF the
cloud UCSF Research Cloud on AWSResponsible for security IN the
cloud
Ensures platform
compliance
Researchers (You)Responsible for
application/services IN the cloud
Requires IT Security Review
You Can Use Our Managed AWS Cluster
9
AWSAmazon
responsible for security OF the
cloud UCSF Research Cloud on AWSResponsible for security IN the
cloud
Ensures platform
compliance
Researchers (You)FAST
ultrasound study
Image management using BisQue
Clinical Notes DeID for
Commons
Pregnancy Ultrasound
Study
Health Data Warehouse
The near future… The UC Health Data Warehouse:Combining healthcare data from across the six
University of California medical schools and health systems
11
Center for Data-Driven Insights and Innovation
Compliance / LegalPartnerships &
ProjectsResearch
Population Health Management
System-wide Initiatives Strategic Planning Health SciencesClinical Quality
Clinical / EHR Financial (limited) OSHPD Other SW Tools Health Plan Claims
Health Data
• Combined data from UCSF, UCLA, UC Irvine, UC Davis, UC San Diego, and UC Riverside
• Central database built using OMOP (not Epic) as a data backend– Structured data from 2012 to the present day: 4.7 million patients with
“modern” data, total 15 million with an MRN– 101M encounters, 307M procedures, 283M med orders, 684M vital signs,
453M lab test results, 422M diagnosis codes– Claims data from our self-funded plans now included– Continually harmonizing elements
• Quality and performance dashboards12
UC will have an unprecedented view of the medical system
The basic researcher of the future will use EHR data…
The clinician of the future will use EHR data…
The patient of the future will need their EHR data…
“With great power comes great responsibility”
“With great power comes great responsibility”
— Uncle Ben, Spider-Man