Enabling the Healthcar Enterprise for Discovery Research with i2b2
Transcript of Enabling the Healthcar Enterprise for Discovery Research with i2b2
![Page 1: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/1.jpg)
Enabling the Healthcare Enterprise for Discovery
Research with i2b2
Shawn Murphy MD, Ph.D.
![Page 2: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/2.jpg)
High Throughput Method for supporting Research
� Set of patients is selected from medical record data in a high throughput fashion
� Investigators work with the data of these patients using new i2b2 tools and a specialized team, both developed to work specifically with medical record data
�Using the Crimson system, tissues of these patients can be made available for genomic and biochemical analysis
� Automated discovery can be created from these projects to support further hypothesis-driven research
![Page 3: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/3.jpg)
High Throughput Method for supporting Research
� Set of patients is selected from medical record data in a high throughput fashion
� Eugene Braunwald � Henry Chueh � John Glaser � Diane Keogh
![Page 4: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/4.jpg)
Research Patient Data Registry exists at Partners Healthcare
Query construction in web tool1) Queries for aggregate patient numbers
-Warehouse of in & outpatient clinical data- 4.6 million Partners Healthcare patients- 1.2 billion diagnoses, medications, procedures, genomics, laboratories, & physical findingscoupled to demographic & visit data- Authorized use by faculty status- Clinicians can construct complex queries- Queries cannot identify individuals, internally
can produce identifiers for (2)
Z731984X Z74902XX... ...
De-identified
Data Warehouse
Encrypted identifiers
2) Returns identified patient data OR0000004 2185793- Start with list of specific patients, usually from (1) ...
- Authorized use by IRB Protocol ...
- Returns contact and PCP information, demographics, providers, visits, diagnoses, medications, procedures, laboratories, microbiology, reports (discharge, LMR, operative, radiology, pathology, cardiology, pulmonary, endoscopy), and images into a Microsoft Access database and text files.
0000004 2185793 ... ...
Real identifiers
![Page 5: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/5.jpg)
Security and Patient Confidentiality of Step 1
� All patients at Partners are added � HIPAA notification that their data may be used for research upon registration.
� RPDR data is anonymized at the Query Tool. � Aggregated numbers are obfuscated to prevent identification of individuals;
automatic lock out occurs if pattern suggests identification of an individual is being attempted.
� Queries done in Query Tool available for review by RPDR team, a user lock out will specifically direct a review.
� De-identified data warehouse is a “Limited Data Set” by HIPAA � Medical record numbers are encrypted and obvious identifiers are removed from
data.
� Concept of “established medical investigator” is promoted by classification as a faculty sponsor.
![Page 6: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/6.jpg)
Security and Patient Confidentiality of Step 2
� Only studies approved by the Institutional Review Board (IRB) are allowed to receive identified data.
� Queries may be set up by workgroup member, but faculty sponsor on IRB protocol must directly approve all queries that return identified data.
� Special controls exist when distributing data regarding HIV antibody and antigen testresults, substance abuse rehab programs, and genetic data, due to specific state and federal laws.
� Queries that return identified data are reviewed (retrospectively) by the IRB.
![Page 7: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/7.jpg)
Getting data into the RPDR from clinical systemsSource Systems User Interfaces
LMR WEB
PACS Services Results Review
ED CPOEMGH Services
Faulkner Hospital BWH Services Data
VB/COM WebServices
Services Cache’ Layer
Enterprise Integration
Engine
CDR LMR
HPCGG RPDR Results
EMPI Loinc Manager Reference Lab1Reference Lab2 (Cache’)
![Page 8: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/8.jpg)
Organizing data in the Clinical Data Warehouse
Star schema Concept DIMENSION Patient DIMENSION concept_key patient_keyconcept_text patient_id (encrypted) search_hierarchy Patient-Concept FACTS sex
agepatient_key birth_date concept_key race start_date deceased end_date ZIP
Encounter DIMENSION practitioner_keyencounter_key
encounter_key value_typeencounter_date numeric_value hospital_of_service textual_value Pract . DIMENSION abnormal_flag
practitioner_key name service
.14 4.6 150 .06
1200 million
Binary Tree
start search
![Page 9: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/9.jpg)
Query items Person who is using tool
Query construction
Results - broken down by number distinct of patients
![Page 10: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/10.jpg)
![Page 11: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/11.jpg)
![Page 12: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/12.jpg)
![Page 13: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/13.jpg)
![Page 14: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/14.jpg)
![Page 15: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/15.jpg)
![Page 16: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/16.jpg)
![Page 17: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/17.jpg)
![Page 18: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/18.jpg)
![Page 19: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/19.jpg)
![Page 20: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/20.jpg)
![Page 21: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/21.jpg)
![Page 22: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/22.jpg)
![Page 23: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/23.jpg)
![Page 24: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/24.jpg)
![Page 25: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/25.jpg)
![Page 26: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/26.jpg)
![Page 27: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/27.jpg)
![Page 28: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/28.jpg)
![Page 29: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/29.jpg)
![Page 30: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/30.jpg)
![Page 31: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/31.jpg)
![Page 32: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/32.jpg)
![Page 33: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/33.jpg)
![Page 34: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/34.jpg)
![Page 35: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/35.jpg)
![Page 36: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/36.jpg)
![Page 37: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/37.jpg)
Identified data is gathered Output files placed in special directory
Data is gathered from RPDRand other Partners sources
Files include a Microsoft Access Database
![Page 38: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/38.jpg)
2007’s usage of RPDR
� 1,580 registered users, 332 new in 2007
� 294 teams gathering data for research studies
� 815 identified patient data sets returned to these teams, containing data for of 8.8 million patient records.
� From a survey of 153 teams � Importance of the data received from the RPDR was
evaluated in relation to the study it was supporting. � The adequacy of the match of a patient profile that could
be obtained through the RPDR query tool was estimated.
� $94-136 million total research support critically dependent on RPDR from patient data received throughout life of funding.
� ~300 data marts were created to support hospital operations, representing about 80 million patient records
Usefulness of Detailed Data 106 Total Responses
Critical 43%
Useful 42%
Not Useful 15%
% of Patients Who Fit Required Profile 105 Total Responses
50% - 75% 22%
25% - 50% 26%
> 75% 33%
< 10% 19%
![Page 39: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/39.jpg)
Enterprise-wide repurposing and distribution of medical record data for research
Data Repository
(CRC)
File Repository
Identity Management
Ontology Management
Correlation Analysis
De -Identification
Of data
Natural Language Processing
Annotating Genomic Data #1
Project Management
Workflow Framework
PFT Processing
Annotating Genomic Data #2
Annotating Imaging
Data
� Enable high performance collection of medical record data for querying and distribution � Enterprise web client � Create patient cohorts for further investigation
� Enable discovery within data on enterprise wide scale � Relevance networks � Health Surveillance
![Page 40: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/40.jpg)
Enterprise web client http://services.i2b2.org/webclient/
![Page 41: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/41.jpg)
Technical Overview
� Formed as a collection of interoperable services provided by i2b2 Cells
� Loosely coupled �Makes no assumptions about proximity �Connected by Web services � Activity can be directed manually or automatically
![Page 42: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/42.jpg)
i2b2 Cell: Canonical Hive Unit
Business Logic i2b2 Data Access
Data Objects HTTP XML(minimum: RESTful, others
like SOAP optional)
Programmatic Access
![Page 43: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/43.jpg)
i2b2 project portal
i2b2 Environment
A B
C
shared / central
Clinical Research Chart
“Hive” of software services and core CRC cells
remote
local
![Page 44: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/44.jpg)
![Page 45: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/45.jpg)
CommunityUnited States International � Beth Israel Deaconness Hospital, Boston, MA � Boston University School of Medicine, Boston, MA � Brigham and Women's Hospital, Boston, MA � Children's Hospital, Boston, MA � Denver Children's Hospital, Denver, CO � Cincinnati Children's Hospital, Cincinnati, OH � Cleveland Clinic, Cleveland, OH � Weil Medical College of Cornell, NYC, NY � Group Health Cooperative � Harvard Medical School, Boston, MA � Massachusetts General Hospital, Boston, MA � Maine Medical Center, Portland, ME � Marshfield Clinic, Wisconsin � Morehouse School of Medicine, Atlanta, GA � Oregon Health & Science University, Portland, OR � Ohio State University Medical Center, Columbus, OH � Philadelphia Children's Hospital, Philadelphia, PA � Renaissance Computing Institute, Chapel Hill, NC � Tufts New England Medical Center, Boston, MA � University of California Davis, Davis, CA � University of California San Francisco, SF, CA � University of Massachusetts Medical School, Worcester, MA � University of Michigan Medical Center, Ann Arbor, MI � University of Pennsylvania School of Medicine, Philadelphia, PA � University of Rochester Medical Center, Rochester, NY � University of Texas Health Sciences Center Houston, Houston, TX � University of Texas Health Sciences Center San Antonio, SA, TX � University of Texas Health Sciences Center Southwestern, � Utah Health Science Center, Salt Lake City, UT � University of Washington, Seattle, WA
� Georges Pompidous Hospital, Paris, France � University of Goettingen, Goettingen, Germany � University of Pavia, Pavia, Italy � University of Seoul, Seoul, Korea
![Page 46: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/46.jpg)
SHRINE (Shared Research Informatics Network) = Distributed Queries
SHRINE adapter
SHRINE adapter
SHRINE adapter
Query Aggregator
i2b2
i2b2 caBIG
Central “aggregator” broadcasts query to local hospital “adaptors”, which return aggregate counts only
![Page 47: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/47.jpg)
SHRINE
![Page 48: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/48.jpg)
High Throughput Method for supporting Research
� Set of patients is selected from medical record data in a high throughput fashion
� Investigators work with the data of these patients using new i2b2 tools and a specialized team, both developed to work specifically with medical record data
� Isaac Kohane � John Glaser � Susanne Churchill � Henry Chueh � Griffin Weber � Michael Mendis � Andrew McMurry � Vivian Gainer � Lori Phillips � Rajesh Kuttan � Wensong Pan
![Page 49: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/49.jpg)
Set of patients is selected through Enterprise Repository and data is gathered into a data mart
EDR
Selected patients Project
Specific Phenotypic
Data
Data directly Data from other Data imported from EDR sources specifically for
project
Automated Queries search for Patients and add Data
![Page 50: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/50.jpg)
Use of medical record data in clinical studies focused upon teamwork and workflow
Data Repository
(CRC)
File Repository
Identity Management
Ontology Management
Correlation Analysis
De -Identification
Of data
Natural Language Processing
Annotating Genomic Data #1
Project Management
Workflow Framework
PFT Processing
Annotating Genomic Data #2
Annotating Imaging
Data
� Repurpose medical record information for research studies � I2b2 Workbench � Natural language processing
� Enable Genomic Studies and Public Heath � Tissue/blood selection � Health Reporting
![Page 51: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/51.jpg)
Data is available through the i2b2 Workbench
![Page 52: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/52.jpg)
Team support for Projects
RPDR
Final Project
DB
RPDR Mart
Local Clinical EDC
Local sources Ex: BICS
Project Manager Biostatistician Analyst
Local data extract analyst Programmer
RPDR Support Programmers
![Page 53: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/53.jpg)
NLP Workflow
I2b2 Project Investigators
NLP Specialists
![Page 54: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/54.jpg)
NLP (and comedy) is not pretty
HOSPITAL COURSE: ... It was recommended that she receive …We also added Lactinax, oral form of Lactobacillus acidophilus to attempt a repopulation of her gut.
SH: widow,lives alone,2 children,no tob/alcohol.
BRIEF RESUME OF HOSPITAL COURSE: 63 yo woman with COPD, 50 pack-yr tobacco (quit 3 wks ago), spinal stenosis, ...
SOCIAL HISTORY: Negative for tobacco, alcohol, and IV drug abuse.
SOCIAL HISTORY: The patient is a nonsmoker. No alcohol.
SOCIAL HISTORY: The patient is married with four grown daughters, uses tobacco, has wine with dinner.Smoker
Non-Smoker
SOCIAL HISTORY: The patient lives in rehab, married. Unclear smoking history from the admission note…
Past Smoker
Hard to pick
Hard to pick
???
![Page 55: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/55.jpg)
NLP Specialists Workstation
NLP Specialists
Export Notes
Import Derived Codes
![Page 56: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/56.jpg)
Investigator Review
![Page 57: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/57.jpg)
Select patients for clinical trials
![Page 58: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/58.jpg)
Workbench display of Mutual Inform. Correlations
![Page 59: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/59.jpg)
Project data can be added back to Enterprise Repository
i2b2 DB [ Enterprise Project 1
Shared Data ]
i2b2 DB Shared data of Project 1
Ontology
Consent/Tracking
Project 2
of Project 2 Security i2b2 DB
of Project 3 Project 3
![Page 60: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/60.jpg)
High Throughput Method for supporting Research
� Set of patients is selected from medical record data in a high throughput fashion
� Investigators work with the data of these patients using new i2b2 tools and a specialized team, both developed to work specifically with medical record data
�Using the BETR/Crimson system, tissues of these patients can be made available for genomic and biochemical analysis
� Lynn Bry � Natalie Boutin
![Page 61: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/61.jpg)
Genotype samples and compare to controls
![Page 62: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/62.jpg)
Cost and time benefit of Instrumenting with Sample Collection for Modest-size Study with 10,000 subjects (cases + controls)
Old vs. New Cost ($)
Time
1 chart review per patient (CP1) $20 15 minutes/subject
High-throughput phenotyping (iP) through RPDR and i2b2
$50K Total
1 month total (conservative high estimate)
Sample acquisition through primary care provider (CP)
$650 3-5 subjects/week1
High-throughput sample acquisition through RPDR and BETR/Crimson.
$20 50-200 subjects /week2
= $6.7 million/study vs. $250 thousand/study
![Page 63: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/63.jpg)
Escalating cost and time benefit of Instrumenting with Sample Collection
Previous model for collecting specimens
New model for collecting specimens
![Page 64: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/64.jpg)
Meeting Expectations
![Page 65: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/65.jpg)
Accrual Rates
![Page 66: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/66.jpg)
High Throughput Method for supporting Research
� Set of patients is selected from medical record data in a high throughput fashion
� Investigators work with the data of these patients using new i2b2 tools and a specialized team, both developed to work specifically with medical record data
�Using the Crimson system, tissues of these patients can be made available for genomic and biochemical analysis
� Automated discovery can be created from these projects to support further hypothesis-driven research
� John Brownstein � Judy Colecchi
![Page 67: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/67.jpg)
First signal: • 1 year after
Celecoxib • 8 months
after Rofecoxib
Brownstein et al., 2007
![Page 68: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/68.jpg)
Effect on patient age
� Negative association between mean age at MI and prescription volume
� Spearman correlation -0.67, P<0.05
![Page 69: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/69.jpg)
Centrally supported Automated Discovery Projects Clinical trials performed in-silico
� Performing an observational, phase IV study is an expensive and complex process that can be potentially modeled in a retrospective database using groups of patients available in the large amounts of highly organized medical data.
� Fundamental problems complicate this approach: � Patients drift in and out of the system. Sophisticated statistical models
using adequate control populations are necessary to compensate. � Confounding variables are not found in the database. Sophisticated
natural language processing is needed to extract the confounders from textual reports to allow these confounders to be controlled.
� Missing data disrupts typical statistical approaches
![Page 70: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/70.jpg)
Odds Ratios for Diseases expressed in comparing Rosiglitazone vs. Pioglitazone
![Page 71: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/71.jpg)
Take-away points from i2b2
� 1) Enable Enterprise use of Patients for Research � 2) Power comes from Numbers of Patients Recruited � 3) Extensible Architecture for Developers � 4) Enable Scientist Workflow � 5) Enable Teamwork between Informaticians and Researchers � 6) Communicate through Visualizations � 7) Enable Natural Language Processing � 8) High throughput Tissue acquisition for Genomic Research � 9) Enable Health Surveillance � 10) Enable Data Sharing
![Page 72: Enabling the Healthcar Enterprise for Discovery Research with i2b2](https://reader030.fdocuments.in/reader030/viewer/2022020704/61fb68942e268c58cd5dd1d8/html5/thumbnails/72.jpg)
References
� Murphy, S. N., Morgan, M. M., Barnett, G. O., Chueh, H. C. Optimizing Healthcare Research Data Warehouse Design through Past COSTAR Query Analysis. Proc AMIA Symp. 1999;892-
� Kimball, R. The Data Warehousing Toolkit. New York: John Wiley, 1997. � Johnson SB, Paut T., Khenima A.; Generic database design for patient management
information. Proc AMIA Annu Fall Symp 1997:22-26. � Nadkarni, P.M., Brandt, C. Data Extraction and Ad Hoc Query of an Entity-Attribute-
Value Database. J Am Med Inform Assoc. 1998; 5:511-7. � Murphy, S.N., Chueh, H A Security Architecture for Query Tools Used to Access
Large Biomedical Databases. AMIA, Fall Symp. 2002, pages 552-556.� Murphy, S.N., Gainer, V.S., Chueh, H. A Visual Interface Designed for Novice Users to
find Research Patient Cohorts in a Large Biomedical Database. AMIA, Fall Symp. 2003: 489-493.
� Murphy, SN; Mendis, M, K. Hackett; et al.; Architecture of the Open-source Clinical Research Chart from Informatics for Integrating Biology and the Bedside; AMIA, Fall Symp. 2007.
� Qing T Zeng, Sergey Goryachev, Scott Weiss, et. al., Extracting principal diagnosis, co-morbidity and smoking status for asthma research: evaluation of a natural language processing system, BMC Medical Informatics and Decision Making, 2006.
� Tannen RL, Weiner MG, Xie D. Use of primary care electronic medical record database in drug efficacy research on cardiovascular outcomes: comparison of database and randomised controlled trial findings. BMJ. 2009 Jan 27;338:b81