Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
PCORI Methodology Standards:
Academic Curriculum
© 2016 Patient-Centered Outcomes Research Institute. All Rights Reserved.
Prepared by Hadi Kharrazi, MD, PhD
Dan Ford, MD, MPH
Presented by Hadi Kharrazi, MD, PhD
Module 9: Case Study (Others)
Category 7: Data Networks as Research-Facilitating Structures
DN-1 (requirements for the design and features of data networks) is discussed in
Modules 3, 4, 5, 6, and 7
A. Data Integration Strategy—Module 4
B. Risk Assessment Strategy—Module 6
C. Identity Management and Authentication of Individual Researchers—Module 6
D. Intellectual Property Policies—Module 7
E. Standardized Terminology Encoding of Data Content—Module 4
F. Metadata Annotation of Data Content—Module 4
G. Common Data Model—Module 5
DN-2 (selection and use of data networks) is discussed in Modules 8 and 9
Data Network Methodology Standards Mapping With Content
4
Mini-Sentinel
DARTNet: Distributed Ambulatory Research in Therapeutics Network
HMORN: HMO Research Networks
SCANNER: SCAlable National Network for Effectiveness Research
SHRINE/i2b2: Shared Health Research Information Network
eMERGE: Electronic Medical Records and Genomics
caBIG: Cancer Biomedical Informatics Grid
Regenstrief
Other Data Networks: Examples
4
Mini-Sentinel is a pilot project sponsored by the U.S. Food and Drug Administration
(FDA) to create an active surveillance system
The Mini-Sentinel project aims to inform and facilitate development of a fully
operational active surveillance system for monitoring the safety of medical products
Mini-Sentinel uses preexisting electronic healthcare data from multiple sources
Data partners for this network include (#18) Group Health, Harvard Pilgrim Health Care
Institute, Health Partners, Henry Ford, Lovelace, Marshfield, Myers / Fallon, Kaiser
Permanente (Colorado, Hawaii, North Carolina, Northwest, South Carolina),
Vanderbilt, Healthcore / Wellpoint, Humana, and Aetna
Information obtained through Mini-Sentinel is intended to complement other types of
data and information compiled by FDA scientists
Mini-Sentinel
5 Source: http://minisentinel.org.
“Mini-Sentinel uses a distributed data approach in which Data Partners maintain
physical and operational control over electronic data in their existing environments.”
“The Mini-Sentinel Common Data Model standardizes administrative and clinical
information across Data Partners.”
“Data Partners execute, within their own institutions’ firewalls, standardized
computer programs (e.g., modular programs) provided by the Operations Center or
project workgroups.”
“Data Partners then share the output of these programs with the Operations Center
and project workgroups, typically in aggregated form.”
“A key benefit of the distributed approach is that it minimizes the need to share
identifiable patient information.”
Mini-Sentinel: Architecture
6 Source: Mini-Sentinel Data Activities. Available at: http://www.mini-sentinel.org/data_activities/. Accessed September 2, 2015.
178 million members (assuming no double counting due to health plan change)
358 million person-years of observation time
48 million individuals currently enrolled, accumulating new data
35 million individuals have over 3 years of data
Age distribution:
35% 22-44 years
26% 45-64 years
<10% for other age ranges
3.9 billion dispensings
Accumulating over 45 million dispensings per month
4.1 billion unique encounters, including 42 million acute inpatient stays
Accumulating over 51 million encounters per month
30 million members with at least 1 laboratory result
Mini-Sentinel: Data Specification
7 Source: Mini-Sentinel Distributed Database “At A Glance.” Available at: http://www.mini-sentinel.org/about_us/MSDD_At-a-Glance.aspx.
Accessed September 2, 2015.
Mini-Sentinel: Additional Information
8
Distributed Database
Architecture and
Specification
http://www.mini-
sentinel.org/work_products/Data_Activities/Mini-Sentinel_Year-4-
Distributed-Database-Summary-Report.pdf
Common Data Model
http://www.mini-
sentinel.org/work_products/Data_Activities/Mini-
Sentinel_Common-Data-Model.pdf
Data Quality Review http://minisentinel.org/work_products/About_Us/Mini-
Sentinel_SOP_Data-Quality-Review-and-Characterization.pdf
Governance, Principles,
and Policies
http://www.mini-sentinel.org/work_products/About_Us/Mini-
Sentinel-Principles-and-Policies.pdf
“DARTNet is a rapidly growing collaboration of PBRNs [practice-based research
networks] that are working to build a national collection of EHR [electronic health
record] data, claims data, and patient-reported outcomes data”
“The networks seek to blend quality improvement, effectiveness, and translational
research with a data-driven learning system”
“The learning system includes advanced performance measures and assistance with the
development and deployment of clinical decision support systems”
“All DARTNet members benefit from the DARTNet Learning Community, which provides
benchmarked data to practices to identify high performers in clinical care”
DARTNet pilot studies include the comparative-effectiveness research (CER) and safety
of oral medications for type 2 diabetes, MRSA, therapies for major depression and
chronic kidney disease
Distributed Ambulatory Research in Therapeutics Network (DARTNet)
9 Source: Distributed Ambulatory Research in Therapeutics Network (DARTNet). Available at: http://www.aafp.org/patient-care/nrn/studies/all/dartnet.html.
Accessed September 2, 2015.
DARTNet is a federated network of EHR data and other clinical information from
multiple organizations
Databases reside in multiple member practices and are linked through a secure
web-based system so they can be searched and queried
Three models of data extraction and sharing:
1. OMOP XML Data Extract
• Import via ROSITA, which handles patient index matching, removing HIPAA PHI,
harmonization and standardization of the data, and minimal outputs for local
querying
2. Third-party EHR data extraction tools
• Local federated databases + ROSITA
3. Manual data extraction programs
• Local federated databases + ROSITA
DARTNet Project: Architecture
10 Source: DARTNet Institute. Data extraction. Available at: http://www.dartnet.info/technology.htm. Accessed September 2, 2015.
The 12 distinct research networks that make up DARTNet Institute offer access to
approximately 12.5 million patient visits per year, 5 million patient lives, and
approximately five billion data points
Partners types include:
Academic Partners: Northwestern University, Ohio State University, University of
Alabama at Birmingham, University at Buffalo, University of California San Diego,
University of Colorado Denver, University of Kansas, University of Minnesota,
University of Texas San Antonio, University of Vermont, and University of
Washington
Private Partners: ABC Crimson Care Registry, AAFP, and QED CINA Inc
CTSA Partners: 9 university-based Clinical and Translational Science Award centers
Network Partners: AppNet, eNQUIRNet, CoNNECT, CCPC, FREENET, LA_NET,
MAFPRN, NCNC, SAFINet, STARNet, UNYNet, and WPRN
DARTNet Project: Data Specification
11 Source: DARTNet Institute. Networks. Available at: http://www.dartnet.info/networks.htm. Accessed September 2, 2015.
Architecture and Data
Extraction Specification
http://www.dartnet.info/Technology.htm
Common Data Model (OMOP) http://www.dartnet.info/OMOPwork.htm
Reusable OMOP and SAFTINet
Interface Adaptor (ROSITA)
http://www.dartnet.info/Rosita.htm
The Scalable Architecture for
Federated Translational
Inquiries Network (SAFINet)
http://www.ucdenver.edu/academics/colleges/medicalschool/
programs/outcomes/coho/saftinet/Pages/default.aspx
DARTNet Datasets http://www.dartnet.info/datasets.htm
DARTNet Project: Additional Information
12
HMORN brings together the research departments of a number of healthcare systems
Collectively, HMORN represents more than 1,400 scientists and research staff with
methodological and content expertise from an array of disciplines
HMORN will be renamed as Health Care Systems Research Network (HCSRN)
HMORN’s VDW Data Area Workgroups:
Define, maintain, and interpret data file specifications
Propose new variables
Identify site-specific issues with data standards
Provide scientific input for each data area
Health Maintenance Organization Research Network (HMORN)
Source: Health Care Systems Research Network. Available at: http://www.hmoresearchnetwork.org. Accessed September 2, 2015. 13
HMORN Members
Baylor Scott & White Center for Applied
Health Research
Catholic Health Initiatives
Essentia Institute of Rural Health
Geisinger Center for Health Research
Group Health Research Institute
Harvard Pilgrim Health Care Institute
HealthPartners Institute for Education and
Research
Henry Ford Health System Research
Centers
Kaiser Permanente Division of Research
Kaiser Permanente Institute for Health
Research
Kaiser Permanente Center for Clinical and
Outcomes Research
Kaiser Permanente Department of Research
& Evaluation
Maccabi Institute for Health Services
Research
Marshfield Clinic Research Foundation
Meyers Primary Care Institute
Mid-Atlantic Permanente Research Institute
Palo Alto Medical Foundation Research
Institute
The Center for Health Research, Kaiser
Permanente Hawaii
The Center for Health Research, Kaiser
Permanente Northwest
14
HMORN Members
15
The VDW is a federated database in which each site stores its data locally in identical
data structures
Each site creates a series of datasets across selected content areas based on common
definitions and concepts
The purpose is to enhance efficiencies in performing multisite research
Each site retains ownership and control of its respective data
Periodic data quality checks “look at ranges, cross-field agreement, implausible data
patterns, and cross-site comparisons”
“Each institution’s VDW data remain at their site until a study-specific need arises.
The minimum necessary required data are extracted after ethical, contractual and
HIPAA requirements are met”
HMORN Virtual Data Warehouse (VDW) Architecture
16 Source: HCSRN. Data Resources. Available at: http://www.hcsrn.org/en/About/Data/. Accessed September 2, 2015.
More info: http://www.hcsrn.org/en/Tools%20&%20Materials/VDW/
VDW data include …
Enrollment/demographics
Census
Pharmacy
Utilization (procedure and diagnosis codes; inpatient and outpatient events)
Vital signs
Laboratory
Tumor registry
HMORN Virtual Data Warehouse (VDW) Data Specification
17
HMORN Virtual Data Warehouse (VDW) Data Specification
18
HMORN VDW Data Types
“Patient-centered SCAlable National Network for Effectiveness Research (pSCANNER) is
one of 11 clinical data research networks that comprise PCORnet”
“pSCANNER is designed to be a stakeholder-governed federated network that will
utilize a distributed architecture to integrate data from three existing networks
covering over 21 million patients”
Data sites in California include Veterans Health Administration system; University of
California, Davis; University of California, Irvine; University of California, Los Angeles;
University of California, San Diego; University of California, San Francisco; AltaMed;
Queenscare Clinics; and the Children’s clinics
Task forces include Data Harmonization, Technical and Analytics, IRB Coordination and
Streamlining, and Patient Engagement and PPRN
pSCANNER
Source: http://pscanner.ucsd.edu. Accessed September 2, 2015. 19
“The vision of pSCANNER is to provide a scalable and flexible distributed infrastructure
for collaborative comparative effectiveness research, while engaging patients and
clinicians”
“pSCANNER will be used for conducting multiple studies within the same computer
network, where each study might focus on different clinical domains, have varying
data needs and data models, and might have different data-sharing policies”
“SCANNER is comprised of a set of SSL-encrypted Web Services that allow a user to
perform distributed statistical analysis on data hosted on remote sites. Multiple
statistical analysis methods are supported via a plug-in architecture …”
pSCANNER uses OMOP CDM v4
pSCANNER Architecture
Source: http://pscanner.ucsd.edu/about/architecture. Accessed September 2, 2015. 20
pSCANNER
Architecture
Source:
http://pscanner.ucsd.edu/about/architecture.
Accessed September 2, 2015. 21
pSCANNER: OMOP
Mapping (Sample
UCSD Data)
Source:
http://pscanner.ucsd.edu/about/semantic-
interoperability. Accessed September 2, 2015. 22
Shared Health Research Informatics Network (SHRINE) is a federated query tool for
i2b2 databases
“… Harvard Medical School (HMS) proposed creating a web-based sotware network
(based on the prototype model SPIN, Shared Pathology Informatics Network) that
would alllow the participating Harvard hospitals to link their respective i2b2 instances
for the sharing of obfuscated, aggregated counts of patients meeting selected inclusion
and exclusion criteria for demographics, diagnoses, medications, and labs”
“It was envisioned that this network, called SHRINE (Shared Health Research
Informatics NEtwork), would greatly enable population-based research, assessment of
potential clinical trials cohorts, and hypothesis formation for followup study by
combining the EHR assets across the hospital system”
SHRINE/i2b2
Source: https://www.i2b2.org/work/shrine.html. Accessed September 2, 2015. 23
SHRINE/i2b2
“i2b2 (Informatics for Integrating Biology
and the Bedside) is an NIH-funded
National Center for Biomedical
Computing based at Partners HealthCare
System”
“The i2b2 Center is developing a scalable
informatics framework that will enable
clinical researchers to use existing
clinical data for discovery research and,
when combined with IRB-approved
genomic data, facilitate the design of
targeted therapies for individual patients
with diseases having genetic origins”
Source: https://www.i2b2.org/. Accessed September 2, 2015.
Central
“aggregator”
broadcasts
query to local
hospital
“adaptors”,
which return
aggregate
counts only
24
SHRINE/i2b2
25
i2b2 interface running on SHRINE
“eMERGE is a national network organized and funded by the National Human Genome
Research Institute (NHGRI) that combines DNA biorepositories with electronic medical
record (EMR) systems for large scale, high-throughput genetic research in support of
implementing genomic medicine”
“Each center participating in the consortium is uniquely situated to provide critical
resources to this highly collaborative and productive network”
“Each site combines a biobank or study cohort with extensive genomic data and access
to clinical data derived from electronic medical records”
“Sites are geographically dispersed and have diverse patient populations, including two
sites focusing specifically on pediatrics”
Electronic Medical Records and Genomics (eMERGE)
Source: https://emerge.mc.vanderbilt.edu. Accessed September 2, 2015. 26
Electronic
Medical Records
and Genomics
(eMERGE)
27
Electronic
Medical Records
and Genomics
(eMERGE)
Source:
https://emerge.mc.vanderbilt.edu/emerge-sites/.
Accessed September 2, 2015. 28
Other Data
Networks:
Available
Information
29
Summary of Research Networks: Architecture
30
Practice caBIG DARTNet HMORN Mini-Sentinel SCANNER Regenstrief SHRINE
i2b2 eMERGE
Architecture
Paradigm Distributed Distributed Distributed Distributed Distributed Central Distributed Unknown
Query
distribution/
data
request
process
Federated
queries
with local
storage
Federated with
local storage in
common data
model format
Ad hoc
(until
2011)
Publish and
subscribe with
local storage in
common data
model format
Multiple
Direct from
central
Repository
Federated Unknown
Data
integration
Strategy
Multiple
shared
data models
Common Data
Model
Common
Data Model
Common Data
Model Multiple
Common
Data Model
I2b2 Star
Schema Ad hoc
Security
standards
Tight Element-
based Access
Defense in
Depth Ad hoc
Defense in
Depth
NIST Level
3 Unknown Unknown Unknown
Summary of Research Networks: Governance
31
Guideline caBIG DARTNet HMORN Mini-
Sentinel SCANNER eMERGE
Timely research data sharing Yes Yes Yes Yes Yes Yes
Identity management and authentication of
individual researchers Yes Yes Unknown Unknown Yes Unknown
Healthcare and research network audits Unknown Unknown Unknown No Yes Unknown
Specific consent for data sharing with
informed consent for research Sometimes Unknown Yes NR
Per IRB
request Yes
Network trust, business associate
agreements
Contract
Agreements Unknown Yes Yes Yes Yes
Data use agreement Yes Unknown Yes No Yes Yes
Intellectual property policies No Unknown Yes Yes No Unknown
Regenstrief and SHRINE were not reviewed.
Summary of Research Networks: Governance
32
Guideline caBIG DARTNet HMORN Mini-Sentinel SCANNER eMERGE
Governing body NCI Board of Dir. Gov. Board FDA Board No Steering C.
Data access committees No Unknown Unknown Proposal
Review No Yes
Stakeholder engagement Workgroups Unknown Stakeholder
Council Privacy Panel
Patient
Focus Panel
Stakeholder
Engagement
Workgroup
Centralized coordinating center NCI University No FDA University University
Other committees and
workgroups for operational
performance
Architecture,
Clinical Trial,
Research,
Training…
Admin,
Technical,
Research…
Asset
Stewardship,
IRB
Coordinate,
KM group
Safety,
Operations,
Project
Management
Architec-
ture,
Policy, CER
Genomics,
Informatics,
Consent,
Oversight
Regenstrief and SHRINE were not reviewed.
Summary of Research Networks: Semantic Interoperability
33
Guideline caBIG DARTNet HMORN Mini-Sentinel SCANNER
Standardized
terminology encoding of
data content
VCDE Workspace
Terminologies
ICD9-CM,
RxNorm,
SNOMED-CT…
CPT, HCPCS, ICD9-
CM, Insurance
Claims…
ICD9-CM,
HSPCS/CPT,
NDC
ICD9-CM,
LOINC,
RxNorm, CPT…
Metadata annotation of
data content
ISO/IEC 11179
Standards
Continuity of
Care Records
(CCR)
No Limited meta-
data No
Common data model
Biomedical
Research
Integrated
Domain Group
(BRIDG)
No
Virtual Data
Warehouse (VDW)
specifies data
structure
Mini-Sentinel
Common Data
Model (MSCDM)
Observational
Medical Outcomes
Partnership
(OMOP)
*Regenstrief, SHRINE, and eMERGE were not reviewed.