Optum Labs Overview - National Academy of Medicine · 2019. 5. 21. · Optum and Mayo Clinic are...
Transcript of Optum Labs Overview - National Academy of Medicine · 2019. 5. 21. · Optum and Mayo Clinic are...
Optum Labs Overview Paul Wallace MD
Chief Medical Officer
Optum and Mayo Clinic are
partnering to launch Optum
Labs, a collaborative
research and development
center for the health care
industry with a singular goal:
improving patient care.
Key ‘means’ to that
goal:
› Collaboration
› Research
› Innovation
Introducing Optum Labs
Health services
United Health Group (UNH)
Health benefits
Context within United Health Group
Optum Labs: Bringing together diverse perspectives of health care research and innovation
• Contribute to the discussion and the direction of research at Optum Labs
• Build collaborative relationships with other key stakeholders
• Participate in high-visibility research designed to improve patient care
Health Care Research and
Innovationev
Health care
research
and
innovation Academic
institutions
Technology
leaders
Life sciences
companies
Commercial
payers/
Employers
Government
researchers
Health care
providers
Patient/
Consumer
organizations
Partners:
Partnership Building Blocks
Research
Participation Individual Research Collaborative Projects Research Funding
Data Contribution No Data Proprietary Data in
Sandbox Contribute Data
Access to Data Project Based Bring your own
Technology and Data Exploratory Access
External Visibility Partnership
Announcement Ongoing joint PR
Ongoing Branding
Opportunities
Research Impact Journal publication Participate in
Translation Drive Translation
Event Participation Invited to relevant
Symposia and Events
Leadership in
Symposia and Events
Sponsorship /
Branded Symposia
and Events
Increasing Partnership Intensity
>100 million Administrative
Building expanding data resources
2,000+ data fields: • Medical claims
• Pharmacy claims
• Lab claims and results
• Health risk assessments
• Standardized costs of care
• Race
• Income
• Education level
• Language preference
• Household
• Geography
• Mortality
Tests,
Treatments 315 million U.S. population
>30 million
Clinical
Mayo
Expanded insights with deeper clinical context
500+ additional data fields: • Encounters
• Vitals
• Labs
• Medication orders
• Procedures
• Admissions, discharges and transfers
• Patient appointments
• PHQ-9
• Patient-provided information
>100 million
Administrative
>30 million
Clinical
Data growth through partnership
315 million U.S. population
Mayo Health System
2
Health Plan 1
Health Plan 1
Health System
3
Consumer
Genomic
Privacy protection and linkage of data sources
Optum Labs employs
certified de-identified
data sets, together
with a hashing
methodology to
enable matching
individuals from
multiple sources, yet
preserving statistical
de-identification and
confidentiality.
OPTUM LABS
Shared Salt Code
(same for all contributors)
PHI is then hashed by
contributors at their site.
Uses Optum Labs’
Confidential Salt Code
Statistically de-identified views accessed via data enclaves
Name
Address
Birthdate
SSN
Phone,
etc
Source 1 (e.g. EMR /
Clinical)
Name
Address
Birthdate
SSN
Phone,
etc , .
Source 2 (e.g. Insurer
claims)
Primary
Hash
XXXXX
Primary
Hash
XXXXX
Secondary
Hash
YYYYY
De-identification
Research process
A research approach: ‘Constellations’
“Constellations”: Broadly important, high-
impact health care problems that can attract
multiple research groups from various parts
of the health care ecosystem working on
many different aspects of the problem Example
Constellation:
Complex
comorbidity
Economic
implications
Innovative
practice
patterns
Novel
approaches to
performance
measurement
Innovative
processes
of care
Patterns of
comorbidity
Successful
existing
practice
patterns
• Technical Requirements
– Secure storage
– Speed
• Computation
• Data exploration
– Linkage methodology
– Multiple de-identified versions of
the database
– Facilitated but controlled access to
the appropriate de-identified
database version (Data
enclaves/”sandboxes”)
– Analytic tools
– Outputs
• Business Requirements
– Privacy protection for patients,
providers, and partners
– Governance
– Sustainability
– Research proposal review and
improvement process
– Linkage between researchers and
research users
– Partner recruitment and relationship
facilitation
Architecting a multi-partner solution…
• Emerging, but incomplete:
– Governance models
– Sustainability models
• Key issues that may be framed as competitive and/or complementary:
– Personalization and privacy
– Public and commercial
– Distributed and federated data models
– Experimental and observational methods
– Involvement of IRBs
– Research and translation
– (Research and quality improvement)
Challenges, Gaps, and Opportunities