Barohn, Richard J.
SECTION 3: BIOMEDICAL INFORAMTICS
A. INTRODUCTION AND SPECIFIC AIMS
Biomedical Informatics accelerates scientific discovery and improves patient care by converting data into
actionable information. Pharmacologists and biologists receive improved molecular signatures; translational
scientists use tools to determine potential study cohorts; providers view therapeutic risk models individualized
to their patient; and policy makers can understand the populations they serve. Informatics methods also lower
communication barriers by standardizing terminology describing observations, integrating decision support into
clinical systems, and connecting patients with providers through telemedicine.
The Heartland Institute for Clinical and Translational Research (HICTR) has catalyzed the development and
integration of informatics capabilities to specifically support translational research in our region and to address
issues and needs like those identified above. Our vision is to provide rich information resources, services, and
communication technologies across the spectrum of translational research. We will: a) adopt methods which
facilitate collaboration and communication both locally and nationally, b) convert clinical systems into
information collection systems for translational research, c) provide innovative and robust informatics drug and
biomarker discovery techniques, d) work with state and regional agencies to provide infrastructure and data
management for translational outcomes research in underserved populations, and e) measure clinical
information systems’ ability to incorporate translational research findings. Our specific aims complement the
novel methods and technologies of our region and target translational needs requiring further investment.
Specific Aims:
1. Provide a HICTR portal for investigators to access clinical and translational research resources, track usage
and outcomes, and provide informatics consultation services.
2. Create a platform, HERON (Healthcare Enterprise Repository for Ontological Narration), to integrate clinical
and biomedical data for translational research.
3. Advance medical innovation by linking biological tissues to clinical phenotype and the pharmacokinetic and
pharmacodynamic data generated by research in phase I and II clinical trials (addressing T1 translational
research).
4. Leverage an active, engaged statewide telemedicine and Health Information Exchange (HIE) to enable
community based translational research (addressing T2 translational research).
B. BACKGROUND
The University of Kansas has a wide range of informatics capabilities, developed over decades of research
and service. Through the National Center for Research Resources (NCRR) IDeA Networks for Biomedical
Research Excellence (INBRE), the Kansas’ KINBRE has created a ten campus network of collaborative
scientists using common bioinformatics resources. Kansas has pioneered the use of telemedicine since the
early 1990s, providing inter-state connectivity at over 100 sites and conducting thousands of clinical
consultations and hundreds of educational events for health care professionals, researchers, and educators.
The Center for Health Informatics’ Simulated E-hEalth Delivery System, a jointly funded program between the
University of Kansas and Cerner Corporation, creates an equitable partnership to fully integrate applied clinical
informatics into an academic setting and supports over 40 international academic clients. These existing efforts
are complemented by more recent investments in clinical research informatics and medical informatics.
B.1. Medical Informatics
In 2010, KU committed $3.5 million dollars over the next 5 years towards medical informatics academics and
service and recruited Russ Waitman, PhD to lead the Division of Medical Informatics in the Department of
Biostatistics, to integrate clinical research informatics, and to provide overall leadership for the HICTR
Biomedical Informatics program. Dr. Waitman was previously responsible for directing the development,
implementation, commercialization, and operation of the inpatient clinical systems used at Vanderbilt University
Medical Center––both internally developed applications by the Computerized Provider Order Entry team (the
WizOrder project, medication reconciliation) and commercial nursing documentation and barcode medication
administration systems. Dr. Waitman conducted research to extend education and decision support into
1
Barohn, Richard J.
clinical systems as well as NIH funded research to discover medication-laboratory result relationships,
measure the utility of preadmission medication data for post-marketing surveillance of adverse drug events,
and build surveillance systems for acute kidney injury. Over time, his team’s clinical research database
combined ten years of clinical, administrative, social security death index, and user activity log data to support
three National Library of Medicine (at NIH) funded grants (R01 LM007995; R01 LM009965; R01 LM007995).
Concidental with Dr. Waitman’s recruitment, this year the University of Kansas Hospital will complete their fiveyear, $50 million investment to implement the Epic Electronic Medical Record, concluding with Computerized
Provider Order Entry in November. This effort has transformed the inpatient environment. The University of
Kansas Physicians have now begun adoption of Epic across all ambulatory clinics.
Since arriving in February 2010, Dr. Waitman has conducted extensive interviews across our campuses to
define a vision for biomedical informatics with clinical and translational research at the forefront. He has
worked closely with Gregory Ator, MD (Chief Medical Information Officer, Senior Medical Director, and
Professor of Otolaryngology), Chris Hansen (hospital Chief Information Officer), James Bingham (KUMC Chief
Information Officer) and his Information Resources organization, Judith Warren, PhD, RN (Center for Health
Informatics), Matthew Mayo, PhD, Chair of the Department of Biostatistics and his project management staff,
and other informatics staff to establish medical informatics and a clinical data repository described in Aim 2
below. Dr. Ator participates in the Epic Corporation’s Research Council, ensuring our efforts are aligned with
other academic medical center’s using the Epic clinical information system. Arvinder Choudhary is project
director responsible for our clinical data repository. Mr. Choudhary was previously with the Medical College of
Georgia and the University of Pittsburgh and has over 17 years of software engineering, database
administration, and project management experience in healthcare and biomedical informatics. He previously
constructed bio-surveillance data warehouses. Daniel Connolly is the biomedical informatics software
engineer responsible for establishing the development environment and integration. For over 15 years, Mr.
Connolly chaired the working groups defining technical standards for web and prototyped semantic web
technologies at the Massachusetts Institute of Technology and the World Wide Web Consortium (W3C). Four
additional personnel or faculty will be recruited as part of establishing the medical informatics division. Since
April, the team has deployed a data repository with i2b2 integration, contributed to the i2b2 community code for
integration with the Central Authentication Service (an open source project used by many universities for
“single sign-on”), worked with other CTSA institutions to use the UMLS as a source terminology for i2b2, and
established a software development environment for the team using Trac and Mercurial.
B.2. Clinical Research Informatics
Starting in 2005, KU selected Velos e-Research for our Comprehensive Research Information System (CRIS).
CRIS is a web application, supports Health Level 7 messaging for systems integration, and complies with
industry and federal standards (CFR Part 11) for FDA regulated drug trials. It consists of a management
component for patients, case report forms, and protocols, and allows financial management for conducting
studies with administrative dashboards and milestones for measuring research effectiveness. KU has invested
over $2 million, with another $750,000/yr investment for support and staffing (6 research staff), to not only
support KU investigators but also to provide enterprise licensing for HICTR investigators to conduct multicenter, cooperative group, and investigator-initiated research with CRIS across unlimited participating locations
without additional license fees. CRIS has been deployed throughout our institution and the team has
experience supporting multicenter trials as outlined in Table 3.1. In fiscal year 2010, CRIS was used to
register 199 new studies, enroll 2331 new subjects, and enter data on 11,098 electronic case report forms.
The CRIS team reports to Dr. Waitman and consists of three clinical information specialists, three clinical
application administrators, and a information technology project manager. Approximately 2.5 FTE is devoted
towards initiatives that expand clinical research informatics capabilities. Examples include deploying Velos
financial management modules this year and working as a development partner with Velos to transform
Institutional Review Board (IRB) review and approval into a paperless process. Investigators’ applications are
validated in real time and then routed through the appropriate review mechanisms. This successful pilot will be
integrated into the Velos product release in 2011 and our contribution to this eIRB initiative will benefit other
organizations that have adopted Velos software. With the incorporation of financial and IRB tracking, CRIS
provides greater capabilities to account for all research projects conducted in the enterprise. The KU
Department of Biostatistics, in which the Division of Medical Informatics and CRIS reside, is also an alpha
partner with Velos in developing a web-based project management system (eProject) for project registrations,
tracking statuses, and creating budgets. Further details on CRIS and eProject operations are provided in the
2
Barohn, Richard J.
Biostatistics section (Section 4). Additional planned CRIS initiatives (clinical data, administrative systems, and
biological tissue repository integration) are discussed in the specific aims of this section below.
Table 3.1. 2010 Examples of Trials using CRIS for Multi Site, Multi Sponsor, and Community Outreach
Title
PI
Description (T1, clinical, T2)
Phase II Trial of
Methotrexate in
Myasthenia Gravis
Breast Cancer Prevention
by Letrozole
R. Barohn,
Neurology
Secoisolariciresinol (SDG)
for Breast Cancer
Prevention
Enhancing Tobacco Use
Treatment for African
American Light Smokers
Safety and Tolerability
Trial of Arimoclomol for
Sporadic Inclusion Body
Myositis
Azacitidine in newly
diagnosed Acute
myelogenous leukemia
C. Fabian,
Oncology
Determine if oral methotrexate is an effective therapy for
myasthenia gravis patients who are prednisone dependent.
(FDA R01 FD003538-01) Multisite study
Assess the potential efficacy of Letrozole as a
chemopreventive agent in high risk postmenopausal women
on hormone replacement therapy. (NIH CA122577-01A1)
SDG (flax seed) as a prevention strategy for pre-menopausal
women at high risk for developing breast cancer. Conducted
in eight universities. (Komen Promise Grant)
Evaluate the efficacy of bupropion plus health education
versus placebo plus health education for smoking. Have
enrolled over 540 participants. (NIH R01 CA091912-09)
Assess the safety and tolerability of arimoclomol as compared
with placebo over four months of treatment in patient with
Inclusion Body Myositis. Multisite Kansas and London.
(Sponsor: CytRx)
Evaluate response rates of Azacitidine in elderly veteran
population with untreated acute myelogenous leukemia (AML)
where intensive chemotherapy is not appropriate. Department
of Veterans Affairs multi-site study. (Sponsor: Celgene)
Evaluate the pharmacokinetics, maximally tolerated dose,
toxicities, and in a preliminary manner for any anti-tumor
activity using RECIST criteria. (Sponsor: Crititech, Inc.)
Create a registry of smokers in rural hospitals and examine the
impact of a disease management program for smoking
cessation. Conducted in 30 rural critical access hospitals.
(NIH R01 CA101963-07)
Phase I Study of
Intraperitoneal
Nanoparticle Paclitaxel
Centralized Disease
Management for Rural
Smokers (KANQUIT II)
C. Fabian,
Oncology
L. Cox, KUMC
Preventive Med
Y. Wang,
Neurology
S
Kambhampati,
KU Cancer
Center
S. Williamson,
KU Cancer
Center
E. Ellerbeck,
Preventive Med
B.3. Bioinformatics
Bioinformatics is a critical technology for translational (T1) research that systematically extracts relevant
information from sophisticated molecular interrogation technologies. Analytical techniques–such as microarray
experiments, proteomic analysis and high throughput chemical biology screening–can probe disease etiology,
aid in development of accurate diagnostic and prognostic measures, and serve as a basis for discovering and
validating novel therapeutics. HICTR institutions benefit from strong molecular bioinformatics research
expertise distributed across three synergistic entities: the Center for Bioinformatics at KU-Lawrence (CBi), the
Bioinformatics and Computational Life Sciences Research Laboratory (BCLSL), and the K-INBRE (Kansas
IDeA Network for Biomedical Research Excellence) Bioinformatics Core (KBC).
CBi (an internationally recognized structural biology modeling research program, focusing on computational
structural biology) and BCLSL (a research program, led by the School of Engineering at KU-Lawrence and
focusing on development of sophisticated algorithms for mining biological data) will be important HICTR
collaborative partners. Examples of nationally funded research led by these two centers include:
- Integrated Resource for Protein Recognition Studies, PI: I. Vakser; Improve understanding of
fundamental properties of protein interaction and facilitate development of better tools for prediction of
protein complexes. NIH R01GM074255-06, and
- Mining Genome-wide Chemical-Structure Activity in Emergent Chemical Genomics Databases, PI: J.
Huan; Advance the underlying theoretical and computational principles of data mining in the emergent
chemical genomics databases. NSF IIS 0845951
3
Barohn, Richard J.
Bioinformatics activities within HICTR will be coordinated by the K-INBRE KBC (an NCRR-funded multiinstitutional research support core serving basic and translational biomedical research in Kansas via
computational resources, informatics consultation, collaboration and education). The KBC administers a broad
range of bioinformatics software and data management utilities for biomedical researchers and strong
computational hardware holdings are already in place. The BCLSL and KBC collaboratively led a successful
$4.6 million NCRR G20 proposal (1G20 RR031125) aimed at enhancing a regional life sciences research data
center, for which HICTR is a key target client.
Gerald Lushington, PhD, who leads the KBC, will be the assistant director for bioinformatics within HICTR.
Housed at KU-Lawrence, KBC has faculty-level satellite directors at KU medical center and Kansas State
University. It employs three full-time PhD-level bioinformatics specialists for research support, plus a full-time
bioinformatics outreach coordinator. KBC also partially supports five other staff positions and multiple graduate
research assistants for bioinformatics research support. KBC has a strong background of successful
collaboration with other molecular-oriented HICTR services such as those in the Translational Technologies
Resource Center (Section 8) and the Institute for Advancing Medical Innovations (Section 6), an extensive
collaborative network with experimental biologists, and a growing suite of clinical interactions. The KBC, in
partnership with other bioinformatics entities, has a successful history of fostering biomedical program projects
(see Table 3.2) and is well positioned to address T1 translational research needs for the HICTR.
Table 3.2. Federally Funded Bioinformatics Program Projects and Centers Supported by the K-INBRE
Bioinformatics Core and Collaborating Centers.
Title
PI
COBRE Center for Cancer
Experimental Therapeutics
B. Timmerman,
School of Pharmacy
COBRE Center for
Molecular Regulation of Cell
Development and
Differentiation
COBRE Center for Nuclear
Receptors for Liver Health
and Disease
D. Abrahamson,
Anatomy & Cell
Biology
Kansas IDeA Network for
Biomedical Research
Excellence
Kansas Intellectual and
Developmental Disabilities
Research Center
KU Center for Chemical
Methodology & Library
Development
J. Hunt, Anatomy &
Cell Biology
KU Program Project on
Aging
E. Michaelis, School
of Pharmacy
C. Klaassen,
Pharmacology,
Toxicology &
Therapeutics
P. Smith, Molecular &
Integrative Physiology
J. Aubé, School of
Pharmacy
Description of informatics provided to the
research program
collaboration with high throughput screening (HTS)
Lab to develop a screening database (NIH P20
RR015563)
collaboration to support post-microarray pathway
identification, and bioinformatics services in SNP and
copy number variations, epigenetics and microRNA
(NIH P20 RR024214)
bioinformatics support of microarray experiments and
pathway analysis; chemical informatics support of
high throughput screening experiments and lead
optimization (NIH P20 RR021940)
broad bioinformatics support of developmental and
cell biology research across a network of 9 Kansas
and 1 Oklahoma universities (NIH P20 RR016475)
Informatics support of microarray experiments and
pathway analysis (NIH P30 NICHD HD 002528)
data mining techniques for prioritizing chemicals
according to bioavailability, bioactivity against specific
pharmacological targets, and data management (P50
GM069663)
collaboration to facilitate prototype file sharing and
tissue sample tracking systems in a newly funded
program project grant (NIH NIH P01AG012993)
B.4. Health Informatics
The University of Kansas Center for Health Informatics (KU-CHI), established in 2003, is an interdisciplinary
center of excellence designed to advance health informatics through knowledge integration, research, and
education of faculty and students in the expanding field of biomedical science and information technology.
Like many funded CTSA programs, KU is a founding member of the American Medical Informatics
Association’s (AMIA) Academic Forum. Graduate health informatics education, continuing education, AMIA
10x10 program, consultation, and staff development workshops/seminars designed to advance health
4
Barohn, Richard J.
informatics are sponsored or co-sponsored through KU-CHI. Graduate education in health informatics began
in 2003 in the School of Nursing and is now a multidisciplinary master’s degree program offered by the Office
of Graduate Studies and managed by the KU-CHI. Helen Connors, PhD, RN, FAAN (executive director of KUCHI) chairs the State of Kansas E-Health Advisory Council that oversees the development of the state’s health
information exchange strategic and operational plans and is a board member of the Kansas City Bi-state
Health Information Exchange, the metropolitan area’s Regional Health Information Organization. Dr. Waitman
serves as the director of medical informatics for the KU-CHI.
Under the direction of Dr. Ryan Spaulding, The University of Kansas Center for Telemedicine & Telehealth
(KUCTT) is a leader in telehealth services and research. The program began in 1991 with a single connection
to a community in western Kansas. The Kansas telehealth network now is used to connect 60 health facilities
in Kansas each year. It also is the basis for several national and international collaborations, demonstrating
significant potential of telehealth technologies to eliminate distance barriers. Over the last 19 years, nearly
30,000 clinical consultations have been conducted across numerous allied health, nursing and medical
specialties, making the KUCTT one of the earliest and most successful telehealth programs in the world. In
2009, over 5000 Kansas patients benefited from telemedicine services that included clinical consultations,
community education, health screenings and continuing education for health professionals. Through the
Midwest Cancer Alliance (MCA) alone, the KUCTT supported numerous second opinion cancer consultations,
multidisciplinary tumor board conferences, chemotherapy administration courses and routine patient
consultations. Other specialties facilitated by the KUCTT include cardiology, psychology, psychiatry,
neurology, wound care, etc. The KUCTT also has been very active integrating numerous technologies to
support clinical and research activities, including telehealth systems, electronic health records, data registries
and patient management systems. Examples of telemedicine support for clinical research is highlighted in
Table 3.7 and in specific aim C.4.2. These core competencies will enhance HICTR and national CTSA activity.
Judith Warren, PhD, RN, BC, FAAN, FACMI (Director of Nursing Informatics, the Center for Health Informatics)
will serve as assistant director of health informatics for HICTR. She has extensive experience developing
health informatics standards and terminologies, teaching informatics, and developing national policy for health
information technology. Dr. Warren, a member of the National Committee on Vital and Health Statistics
(NCVHS) since 2004, co-chairs the NCVHS Standards Subcommittee. She served on the task forces that
developed the recommendations on the initial technical and functional requirements for the Nationwide Health
Information Network (NHIN), recommendations on E-Prescribing, and observations on Meaningful Use. Dr.
Warren also serves on the International Health Terminology Standards Development Organization’s Quality
Assurance Committee and has been an active member of Health Level 7 (HL7) since 2000. Dr. Warren’s
secondary appointment in the Department of Biostatistics has focused on aligning our clinical trials
management with National Cancer Institute terminology standards defined by the cancer Biomedical
Informatics Grid (caBIG). Dr. Warren’s involvement with the federal policy structure and extensive
collaboration with Drs. Connors and Spaulding positions HICTR to exploit existing telemedicine capabilities
and regional health information exchange developments, and to apply appropriate standards for
communicating research findings nationally. Dr. Warren will lead our strong focus on use of standards to
promote data integration and reuse.
C. PROGRAM DESCRIPTION
C.1. Aim #1: Provide a HICTR portal for investigators to access clinical and translational research
resources and provide informatics consultative services.
C.1.1. Establishing the portal and integrating resources.
As shown as our first aim in Figure 3.1, we are developing a web application to link all HICTR research
resources and foster communication among our investigators. It will provide investigators with direct access to
the Translational Technologies and Resource Center, the Participant & Clinical Interactions Resources, the
Community Partnership for Health, and the Clinical Research Development Office services. It also will allow
investigators to identify potential collaborators, obtain targeted internal and extramural funding opportunities,
address regulatory and institutional approval processes, request biostatistics support and clinical research
database construction, request and access informatics resources, access national CTSA resources, and
explore educational offerings. The user focused portal will be complemented by tools to help HICTR
management track projects, distribute resources, and facilitate evaluation of the HICTR by analyzing user
5
Barohn, Richard J.
activity similar to efforts at Vanderbilt University1. We will expand upon the web-based project management
system (eProject2) that was been developed within the Comprehensive Research Information System
(CRIS) to manage project registrations, track status, create budgets, and perform other management activities
for biostatistics and clinical research database development. For further details on eProject capabilities, see
the Biostatistics section (Section 9) of this application. Other components of the portal will build upon the
Ingeniux content management system used by the medical center’s Information Resources and other
collaborative tools such as wikis currently used by medical informatics for project management and the
Personalized Medicine and Outcomes Center for www.cvoutcomes.org (Section 7). An overview of planned
functionality is in Table 3.3. We will conduct yearly surveys of investigators to prioritize portal development.
While developed by biomedical informatics, system hardware and administration is provided in-kind by KU
Information Resources under the direction of James Bingham, Chief Information Officer.
Figure 3.1 Conceptual Model of Biomedical Informatics and Specific Aims.
In addition to the existing capabilities of the CRIS, there are several informatics specific components currently
under development. We have integrated authentication between the portal, the campus Central Authentication
Service, CRIS, and i2b2 to streamline access to eIRB and our clinical data repository described in Aim C.2
below. We are desiging secure methods for the process for submitting data requests and receiving data and
complementary methods to support Data Request Oversight Committee activities. Additionally, the K-INBRE
Bioinformatics Core (KBC) will help to promote awareness of regional bioinformatics resources and expertise
by developing and maintaining a Bioinformatics Resources section of the portal. The Bioinformatics
Resources section will clarify which core resource is best equipped to meet clinical and translational research
needs and simplify the process of requesting bioinformatics services. Finally, we will provide a request
management system for the informatics consult service described below.
C.1.2. Create an informatics consultation service to identify investigators’ information needs.
6
Barohn, Richard J.
We will match investigators’ needs against current information available in HICTR resources, specifically the
new repository described in Aim 2. With CTSA funds we will add a clinical informaticist who understands
clinical research hypotheses and knows what relevant data and clinical processes are in the repository.
He/she will assist users with tools like i2b2 and will recommend alternative approaches when required data are
not integrated. He/she will be knowledgeable about data resources described in the Personalized Medicine
and Outcomes Center (Section 7). Early experience from front-line consulting with translational researchers
will drive prioritizing additional data sources, terminology and ontology development (Aim C.2.3), and will guide
the adoption of new technologies for data retrieval and analysis. The informaticist also will support
translational (T2) research outlined in Aims C.4.3 and C.4.4. As our capabilities mature, the service will be
staffed, and needs addressed, by the combined biomedical informatics personnel. Requests for bioinformatics
and computational biology capabilities will be facilitated by the Bioinformatics Resources section of the portal
and support by Dr. Lushington and the K-INBRE Bioinformatics Core. Personal interaction with the
investigator will identify the best approach to their educational goals (formal educational offerings versus
existing web based informatics educational materials developed by existing CTSA programs).
Table 3.3. Examples of existing and planned modules for HICTR portal
Module
General HICTR
Information
Education
Collaborator
Connection
Find Funding
Pilot funding
Request TTRC
Resources
Request CTSU
Resources
CRIS
Biostatistics
eIRB
DROC
HERON
Informatics
Consult Service
Bioinformatics
Resources
IT support
Telemedicine
National CTSA
Resources
Description
Maintained in Ingeniux by Clinical Research Development Office
Status
Planned
Connect to offerings provided by the Clinical and Translation
Research Education Center and other Centers (e.g. K-INBRE)
Integration with Dykes Library “meet the experts” Bibapp application
to allow investigators to identify potential collaborators locally and
NIH VIVO project for national networking for scientists.
Collect and organize funding announcements. Allow investigators
to subscribe to keyword tailored updates.
Submit pilot projects to HICTR for review and funding.
Integration with Translational Technologies Resources Center
(TTRC) request management systems as well as NIH EAGLE-I
project for national resources.
Integration with existing web site to request Clinical Translational
Science Unit (CTSU) services and resources.
Access Comprehensive Research Information System for accessing
protocol management system, case report forms, and reporting.
Request collaboration, obtain status, and management by eProject
Submit protocols and applications to institutional review board.
Investigators can also check on the status of their application.
Request data access to de-identified clinical repository.
de-identified clinical data repository
Request assistance with information needs; using HERON, CRIS,
and Bioinformatics Resources
Detailed inventory of capabilities and expertise in the region with
links to request systems used by the individual cores.
Request network attached storage, wiki or Sharepoint from
Information Resources
Request telemedicine services for research purposes
Researchmatch.org, www.vivo.org, www.eagle-i.org.
Planned
Existing
Planned
Planned
Existing
Existing
Existing
Existing
Existing
Planned
Existing
Planned
Planned
Existing/
Planned
Existing
Existing
C.2. Aim #2: Create a platform, HERON (Healthcare Enterprise Repository for Ontological Narration), to
integrate clinical and biomedical data for translational research.
7
Barohn, Richard J.
Transforming observational data into information and knowledge is the cornerstone of informatics and
research. The opportunity for data driven knowledge discovery has exploded as health care organizations
embrace clinical information systems. However, when dealing with confidential patient information, science
and technology are dependent on law, regulation, organizational relationships, and trust. HERON provides
HICTR with secure methods for incorporating clinical data into standardized information aligned with national
research objectives. It facilitates hypothesis generation, allows assessing clinical trial enrollment potential, and
minimizes duplicate data entry in clinical trial data capture, outcomes research, and evaluation of T2
interventions for translating research into practice. While initially focused on integrating data between the
KU Medical Center, the KU Hospital, and KU affiliated clinics, our methods are designed to
subsequently integrate information among all HICTR network institutions.
C2.1. Develop business agreements, policies, data use agreements and oversight.
Beginning in April 2009, the HICTR started piloting a participant registry in the clinics which allows patients to
quickly consent to be contacted by clinical researchers. The registry contains the consented patients’
demographics, contact information, diagnoses, and procedure codes. From April to July 2010, the medical
center, hospital and clinics reviewed comparable practices4,5,6,7 and drafted a master data sharing agreement
between the three organizations that was signed September 6, 2010. The HERON executive committee is
composed of senior leadership (e.g. chief operating, financial, executive officers and chief of staff) from the
hospital, clinic and medical center and provides governance for institutional data sharing. Establishing
business processes and servicing research requests is conducted by the Data Request Oversight
Committee(DROC) which reports to the HERON executive committee. The repository’s construction, oversight
process, system access agreement, and data use agreement for investigators were approved by the
Institutional Review Board. Since HERON is currently funded by the KU and contains medical center and
affiliated clinics and hospital data, access requires a medical center investigator. As additional institutions
and health care organizations provide support and contribute data, they will be incorporated with
multi-institutional oversight provided by the HERON executive committee and DROC.
As agreed upon by the HERON executive committee, HERON has four uses: cohort identification, de-identified
data, identified data, and participant contact information:
1. After signing a system access agreement, cohort identification queries and view-only access is allowed
and activity logs are audited by the DROC.
2. Requests for de-identified patient data, while not human subjects research, are reviewed by the DROC.
3. Identified data requests require approval by the Institutional Review Board prior to DROC review. After
both approvals, medical informatics staff will generate the data set for the investigator.
4. Investigators who request contact information for cohorts from the HICTR Participant Registry have
their study request and contact letters reviewed for overlap with other requests and adherence to
policies of the Participant and Clinical Interactions Resources Program Data Request Committee.
C.2.2. Implement open source NIH funded (i.e. i2b2) initiatives for accessing data.
The utility of data for clinical research is proportional to the amount of data and the degree to which it is
integrated with additional data elements and sources. However, as additional data are integrated and a richer
picture of the patient is provided, privacy concerns increase. HERON receives and preserves data into an
identified repository, links and transforms those data into de-identified, standardized concepts, and allows
users to retrieve data from this separate de-identified repository (Figure 3.2). We recognize that certain
research requires access to identified data but our approach streamlines oversight when needs are met with
de-identified data. We have adopted the NIH funded i2b28 software for user access, project management, and
hypothesis exploration. We have chosen Oracle as our database because of existing site licensing and i2b2
compatibility. Transformation and load activities are written in the python language and deployed onto SUSE
LINUX servers. Servers are maintained in the medical center and hospital’s data center which complies with
HIPAA standards for administrative, technical and physical safeguards.
Access to the identified repository is highly restricted and monitored. Data exchange between source systems,
HERON, and user access is handled via security socket layer communications, https with digital certificates,
and 128-bit or higher public key cryptography (ex: SSH-2). We will continue to monitor federal guidance
regarding appropriate security measures for healthcare information14. User access, activity logs, and project
management for the de-identified repository is managed via the i2b2 framework and is integrated with the
medical centers’ open source Jasig Central Authentication Service and the HICTR portal. Account creation
8
Barohn, Richard J.
and access control is provided by the medical center. Audit logs are maintained for both the identified and deidentified repositories. Intrusion detection mechanisms are used and monitored by medical center Information
Resources security personnel who also conduct HIPAA Security Rule reviews of the systems.
Current literature highlights challenges with providing anonymity after de-identifying data9,10,11,12. While we
remove all 18 identifiers to comply with HIPAA Safe Harbor, the de-identified repository is treated as limited
data set, reinforced by system access and data use agreements. Dr. John Keighley, Assistant Professor in
Biostatistics, will provide statistical guidance regarding de-identification methods and re-identification risk. He
has done similar work with the Kansas Cancer Registry. Initially, HERON will only incorporate structured data.
We will stay abreast of best practices (e.g., De-ID, the MITRE Identification Scrubber Toolkit versus only
extracting concepts) prior to incorporating narrative text results which have greater risk for re-identification9,13.
Figure 3.2. HERON Architecture
Pilot use of HERON and i2b2 for the HICTR Participant Registry began in April 2010. The current architecture
was deployed in August 2010 and populated with production data from IDX and Epic in September 2010.
Initial extract, load, and transform processes obtain data files from the source system via secure file transfer
methods. Over time, we will increase the frequency for incorporating source data, such as event driven HL7
listeners. This will facilitate data integration with the Comprehensive Research Information System.
C.2.3. Transform data into information using the NLM UMLS Metathesaurus as our vocabulary source.
From its inception, HERON has benefitted from i2b2 community expertise and will continue to work with other
academic medical centers to align our terminology with CTSA institutions. Led by Dr. Warren, our approach is
to rely on the National Library of Medicine Unified Medical Language System (UMLS) as the source for
ontologies and to use existing or develop new update processes for the i2b2 Ontology Management Cell and
other systems. The National Center for Biological Ontologies will augment the UMLS to bridge clinical data
towards bioinformatics domains and to incorporate more recently authored ontologies that are not fully
supported by the NLM. As we develop new mappings for going from UMLS into i2b2 and other systems, we
will share our code and processes with the i2b2 community. Using i2b2 and the UMLS will streamline the
HICTR’s ability to participate in national interoperable data exchange initiatives. Current and planned activities
are shown in Table 3.4. We anticipate refining ontologies based on investigator feedback. We are currently
evaluating an approach outlined by several CTSA organizations that shifts mapping inside i2b2 instead of in
database transformation and load processes15. This preserves the original data side by side with the
transformed information allowing easier review by investigators and domain experts.
Standardizing information for research strengthens our relationship with the hospital and clinics. Reusing
information for clinical quality improvement also requires standardization. Translational research using
HERON highlights areas needing terminological clarity and consistency in source systems. Research and
quality improvement goals also align with achieving “meaningful use”16 and Health Information Exchange
activities. Dr. Warren will partner with Dr. Ator, Chief Medical Information Officer, and the University of Kansas
9
Barohn, Richard J.
Hospital’s Organizational Improvement (which has two of their leaders on the DROC) to develop a data driven
process for improving standardization. This process will be aided by Dr. Ator and the hospital’s foresight in
purchasing and integrating IMO® Problem(IT) into the Epic Electronic Medical Record to support accurate
collection of observations based on both ICD-9 and SNOMED CT.
Table 3.4. Terminology activities
Source terminology
Demographics: i2b2
Extract
Status
Complete
Completed
planned
April 2010
Notes
caBIG: CTEP
Synced
October 2010
Diagnoses: ICD9
Procedures: CPT
Complete
Complete
April 2010
June 2010
Lab terms: LOINC
Medication terms:
RxNorm
Planned
Planned
November 2010
November 2010
Medication ontologies:
NDF-RT
Nursing Observations
Planned
December 2010
Ongoing
July 2010-
Pathology: SNOMED CT
Planned
February 2011
Clinical narrative
Planned
2012
National Center for
Biological Ontology
Planned
2013
Using i2b2 hierarchy. Restricted search
criteria to geographic regions (> 20,000
persons) instead of individual zipcodes
Clinical Research forms for demographics,
adverse event reporting, and disease
progression. (Drs. Warren, Mayo, Waitman)
Using i2b2 hierarchy
UMLS extract scripts developed with UTHSC
at Houston
Plan to use i2b2 hierarchy
UMLS derived. Used to map between
Medispan/FDB, NDF-RT, and Multum used by
current outcomes databases (Section 7).
Physiologic effect, mechanism of action,
pharmacokinetics, and related diseases.
NDNQI pressure ulcers mapped to SNOMED
CT to evaluate automated extraction of self
reported activity. (Drs. Dunton and Warren.)
Providing coded pathology results and patient
diagnosis is a critical objective for defining
cancer study cohorts in Aim 3.
As the hospital restructures clinical narrative
documentation to use Epic’s SmartData (CUI)
concepts, will determine appropriate standard.
In support of Aim 3 focus on bridging clinical
and bioinformatics to advance novel methods.
C.2.4. Link clinical data sources to enhance their research utility.
Data sharing is a very sensitive concern among healthcare providers. We will gradually build relationships by
initially focusing on linking data sources from KU Hospital and affiliated clinics. As capabilities of the team
develop we will expand to other HICTR affiliated network institutions, state agencies, and our rural practice
networks (outlined in C.4.4). This may require centralized and/or distributed data integration17,18. For example,
Dr. Theresa Shireman, head of the Large Database Analysis Core of our Novel Methods 2 section, conducts
research with Medicaid administrative claims data. We will work with her, the Kansas Health Policy Authority,
the Kansas Health Information Exchange, and the Kansas City Bi-state Health Information Exchange to link
clinical databases and state administrative data for research. KU institutional support will be used to construct
and maintain HERON with planned data sources (see Table 3.5 which also shows preliminary statistics on the
volume of data currently extracted). Data from additional sources will be integrated depending on research
interest and support.
Table 3.5 Timeline for Incorporating Clinical Data Sources into HERON
Data source
Source
System
HICTR Participant Registry
IDX
System
“Go-Live”
Date
2009
10
Extract
Status
Complete
Data extraction target
and
prelim statistics
April 2010
3173 volunteers
Barohn, Richard J.
Data source
Source
System
Inpatient/Emergency demographics,
“ADT” locations & services
Epic
Inpatient Diagnoses (ICD9)
Epic
2006
In progress
Laboratory Results and other Discrete
Results (inpatient/outpatient)
Epic/Misys
2002
backfilled
In progress
Electronic Medication Administration
Epic
2007
Planned
Outpatient visits services, diagnoses,
procedures (beyond HICTR Registry)
IDX/Epic
2002
In progress
for IDX
Vital signs, inputs, outputs and
discrete nursing observations
Epic
2007
Planned
Clinical Research Information System
Integration
Provider Order Entry
Other Elements of Patient Problem List
Microbiology, Cardiology, Radiology
CRIS
2007
Planned
Epic
Epic
Epic/Misys
Theradoc
Epic
NDNQI
2010
2009
2006
Planned
Planned
Planned
2007
2001
Planned
Planned
2011
2011
2011
637,000 micro orders
2011
2011
Epic
ORSOS
SSDI
Phillips and
Datacaptor
KHPA
2009
2005
Na
2007
Planned
Potential
Potential
Potential
2012
2012
2013
2013
2005
Potential
2013
Medication reconciliation
Nursing Unit
Quality Indicators
Provider Notes (discrete elements)
Perioperative schedule and indicators
Social Security Death Indicator
Dense physiologic signals
Medicaid databases integration
System
“Go-Live”
Date
2006
Extract
Status
In progress
Data extraction target
and
prelim statistics
October 2010
1.8 million patients
October 2010
215,000 inpatient
November 2010
1.4 million serum K
67 million overall
December 2010
1.2 million outpt orders
2.5 million inpt orders
November 2010
214,000 outpt ICD9
from Epic
2011
1 million patient days
265 million total
6 million pulse
2011
C.3. Aim #3: Advance medical innovation by linking biological tissues to clinical phenotype and the
pharmacokinetic and pharmacodynamic data generated by research in phase I and II clinical trials
(addressing T1 translational research).
We see two points where HICTR informatics must bridge between basic science resources and clinical
activities to enhance our region’s ability to conduct translational research. Our region provides a wealth of
biological and analytical capabilities but struggles at times to understand if our clinical environments see
enough patients to support research. This leads to clinical trials which fail to accrue enough subjects19.
Accurately annotating biological tissues and routine clinical pathology specimens with the patients’ phenotypic
characteristics (such as diagnosis and medications from the clinical records) will provide a more accurately
characterized clinical research capacity. Our second targeted area is designed to support the Drug Discovery
and Development activities of the HICTR Novel Methods 1 program (Section 6) in managing information in
phase I and II clinical trials. Molecular biomarker (Section 8) and pharmacokinetics/pharmacodynamics
research activities (Section 9) provide high quality analysis but their results are not easily integrated with other
records maintained in clinical research information systems. Reciprocally, if clinical characteristics could be
provided with samples in a standardized manner it would allow improved modeling and analysis by the core
laboratories. Collaboration between the bioinformatics and biomedical informatics specifically targeted at
11
Barohn, Richard J.
pharmacokinetic/pharmacodynamic results and clinical trials will provide a foundation for subsequent
integration and our final goal of providing molecular bioinformatics methods.
C3.1. Incorporate clinical pathology and biological tissue repositories with HERON and CRIS to
improve cohort identification, clinical trial accrual, and clinical trial characterization.
Patients’ tissue specimens are a fundamental resource that links the majority of basic biological science
analysis to clinical relevance. We need to improve our characterization of these samples to maximize our
investment in maintening them for clinical research. By integrating routine clinical pathology and research
specimens with clinical information in HERON researchers can quantify samples belonging to patients (e.g.
aggregating all breast cancer diagnoses using the i2b2 a clinical concept “malignant neoplasm of the breast”
and treated with the adjuvant therapy trastuzumab). Similarly, integration between the Biological Tissue
Repository (BTR) in the Translational Technologies and Resource Center (Section 8) and the Comprehensive
Research Information System (CRIS) will enhance analysis and characterization of data from clinical trials.
The BTR builds upon the existing KU Cancer Center Biospecimen Shared Resource and plays a vital role in
collecting and distributing high quality human biospecimens essential to research. Improving specimen
management for clinical research has been identified as a high priority initiative for KU. Working with Ossama
Tawfik, MD, PhD, director of the BTR, we will evaluate the new capabilities provided by Velos eSample and the
National Cancer Institute’s caBIG caTissue initiative to modernize information management in the BTR. This
work will be performed in-kind as a biomedical informatics initiative.
We also will work with Dr. Tawfik in his role as the director of surgical pathology to integrate clinical pathology
reports into the HERON repository. Source systems for this information will be Cerner CoPath and the Epic
clinical information systems. We will exploit current SNOMED CT encoding of pathology results as our
standard terminology. Additionally, the K-INBRE bioinformatics core will interact closely with biomedical
informatics and biostatistics to evaluate the viability of proposed biomarker studies according to fundamental
criteria such as whether enough participating patients or tissue samples exist under a suitable range of
conditions or phenotypic characterization to form an adequate basis for meaningful analysis.
C3.2. Support IAMI clinical trials by integrating and standardizing information between the
bioinformatics and pharmacokinetic/pharmacodynamic (PK/PD) program and CRIS.
As described in the Novel Methods 1 and PK/PD sections of this application (Section 6, 9), the Institute for
Advancing Medical Innovation (IAMI) builds upon the region’s significant drug discovery capabilities and
creates a novel platform for medical innovations. Dr. Gerald Lushington, Dr. Mahesh Viswanathan, and the KINBRE Bioinformatics Core (KBC) provide bioinformatics analysis across the spectrum of target discovery,
compound screening, lead identification and optimization, and preclinical development. In developing these
bioinformatics resources, KBC adheres to international data standards (e.g., Minimum Information About a
Microarray Experiment, MIAME) to ensure consistent data format and content. These resources include
references to publicly available annotation sites (e.g., as provided by National Center for Biotechnology
Integration) in order to amplify the informational content of the stored data. As candidates successfully
transition to a new investigational drug application, we will continue bioinformatics support into clinical trials.
The KBC, the IAMI, the PK/PD program, the CRIS team, and the new clinical research building will provide the
region with a state of the art environment to advance drug discovery. Collaboration between bioinformatics
and the CRIS team will improve the integration of data collected during trials by the clinical research team and
the analytical results generated from histology and chemistry studies. This work will be catalyzed by Drs.
Timothy Welty, Greg Reed, and Maxine Stolz from the PK/PD program. Dr. Lushington and the KBC will
ensure that the data and knowledge associated from the above studies is integrated into standardized format
and made available for general use by both clinical and basic science collaborators.
C.3.3. Apply molecular bioinformatics methods to enhance T1 translational research in areas such as
molecular biomarker discovery efforts and pharmaceutical lead optimization.
Establishing stronger bridges between analytical resources and clinical information allows phenotypic variables
to be exploited by bioinformatics expertise in data mining by the K-INBRE Bioinformatics Core (KBC) and the
Bioinformatics and Computational Life Sciences Research Laboratory. We will augment HICTR T1
translational research efforts by providing data mining algorithms that extract key information from dataintensive wet lab studies and information specification standards that enable systematic and comprehensible
reporting of such information for clinical applications. KBC also has developed a number of tools that facilitate
basic and clinical scientists’ ability to interact and experiment with the data. For example, KBC developed
12
Barohn, Richard J.
OmicsMiner20 which encodes a broad range of machine learning and statistical techniques for feature
selection, phenotype classification and molecular biomarker validation into an intuitive (i.e., non-intimidating)
but comprehensive first-approach data processing engine. OmicsMiner is an open-source, platformindependent, context-insensitive analysis suite that has been applied to microarray profiling, miRNA analysis,
proteomic peak assignment, bioavailability predictions, and next generation sequencing read counts analysis.
This bridge between molecular-level analysis and clinical studies is a potent model for biomarker prediction
and validation, PK/PD studies, and the prospective refinement of personalized medicine. KBC has already
collaborated with the K-INBRE Partnership Core to promote the availability of our translationally-oriented
systems biology skills to collaborative teams of basic and clinical scientists. Both the K-INBRE Partnership
Core and the HICTR Translational Discovery Forums provide rich opportunities for bringing basic and clinical
scientists together, a key HICTR goal. In its current K-INBRE-funded mission, the KBC employs a combination
of explicit salary support from grants for large-scale collaborative or developmental efforts and fee-for-service
cost recovery for small-scale activities. This model will be extended for the HICTR, with HICTR salary support
provided for development of the PK/PD integration and fee-for-service for efforts specifically targeted toward a
distinct translational project (e.g., biomarker discovery, chemical screening data analysis, basic consultation,
etc.) While nationally recognized for achievement in basic biological sciences, the KBC has increasingly
focused on translational activities. As such, the KBC has led the way in developing, acquiring, and
implementing many of the bioinformatics resources currently available to HICTR investigators. A selection of
recent (i.e., 2010) biomarker identification efforts and IAMI-supported projects are shown in Table 3.6.
Table 3.6 Recent Biomarker identification and IAMI-supported projects
Title
PI
MiRNA Regulation of
Ovarian Function
L. Christenson KUMC;
co-I: O. Tawfik,
KUMC
Human Lung Cancer
Biomarker Identification
Using Informatics and
Clinical Samples
M. Visvanathan KBC;
co-I: G.S.
Sittampalam, IAMI;
co-I: O. Tawfik, KUMC
Development of BRCA1
expression promoters for
breast cancer therapy and
prevention
L. Harlan Williams KU
Cancer Center
Oxidative defense against
CYP2e1 toxicity via Keap1Nrf2 interaction blockers
C. Klaassen KUMC
Pharmacology,
Toxicology &
Therapeutics
J. Barycki; U.
Nebraska Dept. of
Biochemistry
K. Peterson; KUMC
Department of
Biochemistry
G.S. Sittampalam;
IAMI
Prostate cancer treatment
via UGDH modulation
Sickle cell anemia
therapeutic development via
HbF upregulation promoters
Colon cancer treatment via
Wnt – Beta Catenin
modulators
Early Detection of Shock
using Biomarker predictors
C. Van Way University
of Missouri Kansas
City
Description of informatics services provided to
the research program
Applies miRNA analysis and advanced machine
learning algorithms to identify molecular biomarkers
relevant to female reproductive health disorders
based on clinical ovarian tissue and serum samples
R01HD61580
KUCC Pilot Award for collection of clinical samples
addressing four distinct lung carcinoma phenotypes,
followed by genomic profiling and lung cancer
molecular biomarker identification and validation via
sophisticated machine learning algorithms
high throughput screening (HTS), data mining,
quantitative structure-activity-relationship (QSAR)
modeling, analog identification and lead optimization,
leading to promising lead and pending patent
(20090170842)
HTS data mining, QSAR modeling, analog
identification and early-stage lead optimization
HTS data mining and hit prioritization
HTS data mining, QSAR modeling, analog
identification and early-stage lead optimization
hit analog identification and prioritization
Microarray technology to identify early changes in
gene expression in white blood cells that will predict
favorable and unfavorable outcomes. DoD: ONR
13
Barohn, Richard J.
N00014-01-1-0151; USAMRIC, W81XSH-06-1-530
We anticipate connections cultivated by the HICTR that prompt integration of bioinformatics and clinical
informatics will also promote collaboration between clinical and basic researchers with data mining expertise.
C.4. Aim #4: Leverage an active, engaged statewide telemedicine and Health Information Exchange
(HIE) to enable community based translational research (addressing T2 translational research).
2010 will be seen as a landmark year for KU, the region, and the nation for health information technology
adoption. The University of Kansas Hospital will have completed a five-year plan to implement the Epic
Electronic Medical Record. Through the HICTR Personalized Medicine and Outcomes Center (Section 7), the
Patient Risk Information Services Manager (PRISM; see C.4.6) developed by John Spertus, MD, and
colleagues at St. Lukes Mid-America Heart Institute will be available for widespread use. Kansas and Missouri
have established health information exchanges and regional extension centers. The Kansas Health Policy
Authority has invested in and deployed a Data Analytic Interface (DAI) maintained by Thomson Reuters that
provides a portal to more than five years of claims data from Medicaid, state employee health plans, and other
private insurers. Cerner Corporation has increased the number of participating institutions in their HealthFacts
repository to encompass 24 million lives, 178 million medication orders on 5,945 brand name drugs, and 1.5
billion lab results from general lab and microbiology. HealthFacts has also integrated EMR data elements such
as vital signs, chief complaint, symptoms, and pain assessment with existing medication and laboratory data.
These regional informatics capabilities are complemented by a long history of engaging the community through
outreach via telemedicine, continuing medical education, and our extensive community research projects in
urban rural and frontier settings.
C.4.1. Provide health informatics leadership to ensure state and regional healthcare information
exchange (HIE) and health information technology initiatives foster translational research.
Under the leadership of Dr. Helen Connors, the Center for Health Informatics has been instrumental for
advancing HIE for Kansas and the Kansas City Bi-state Health Information Exchange (KCBHIE). As chair of
the eHealth Advisory Council, Dr. Connors guided the process to establish the Kansas HIE (KHIE)21; a nonprofit corporation incorporated by the governor in response to the federal HITECH Act. On September 2, 2010,
the governor appointed Dr. Connors as a founding member of the KHIE board of directors. Allen Greiner, MD,
director of the Community Partnership for Health (Section 5), participates in selecting electronic health records
for the Kansas Regional Extension Center. Dr. Russ Waitman participates in coordinating exchange activities
for outcomes research in underserved populations managed by the Kansas Health Policy Authority
(responsible for Kansas Medicaid). In-kind participation in these activities by Drs. Connors, Greiner and
Waitman will keep the HICTR engaged in state and regional infrastructure capabilities for translational
research and will facilitate identifying other collaborating organizations.
C.4.2. Promote existing telemedicine and clinical research informatics capabilities to conduct research
in critical access hospitals and rural regional referral centers.
As described in our background, and in the community engagement section (Section 5, Figure 5.2), the Center
for Telemedicine and TeleHealth (KUCTT) provides enormous reach into diverse communities for
implementing clinical trials and other research across the state. Telemedicine has figured prominently in a
number of clinical research programs (e.g. studies of diagnostic accuracy, cost effectiveness, and patient and
provider satisfaction with telemedicine) as well as studies that employ telehealth for randomized control trial
interventions (e.g. for home parenteral nutrition and sleep apnea.) Complementing KUCTT, the Midwest
Cancer Alliance (MCA) is a membership-based organization bringing cancer research, care and support
professionals together to advance the quality and reach of cancer prevention, early detection, treatment, and
survivorship in the heartland. The MCA links the University of Kansas Cancer Center with the hospitals,
physicians, nurses and patients battling cancer throughout Kansas and Western Missouri. The MCA advances
access to cutting-edge clinical trials as well as professional education, networking, and outreach opportunities.
These clinical trials use our Comprehensive Research Information System (CRIS). Notably, the MCA critical
access and rural referral centers are also the most active participants in telemedicine. The medical director of
the MCA, Dr. Gary Doolittle is a pioneer user and has published telemedicine research with Dr. Ryan
Spaulding, the director of the telemedicine network22,23. The network has been successfully used to facilitate
clinical research and to support special needs populations (see Table 3.7). Especially in the area of smoking
cessation, studies are already benefiting from the complementary technologies of telemedicine and CRIS. In
14
Barohn, Richard J.
addition to statewide integration, as a national leader in telehealth operations, technology and research, Dr.
Spaulding is well-positioned to support CTSA research activities and collaborations across the U.S. and world
using the latest telehealth technologies. Dr. Spaulding has previously partnered with Internet2 leadership on
the Federal Communication Commission’s Rural Healthcare Pilot Project and will leverage these relationships
to advance research activities. Internet2 is a nationwide, high bandwidth network linking universities for
research and data-sharing capabilities. It will be available to HICTR leadership via Dr. Spaulding’s efforts.
Biomedical informatics will work with KUCTT to provide a telemedicine clinical trials network connected by
CRIS for remote data collection via study personnel or direct data entry through a patient portal. KUCTT also
will provide in-kind support for HICTR investigators using telemedicine for meetings across the region.
Table 3.7 Grant Examples using Telemedicine and the Comprehensive Research Information System
Title
PI
Grant/Agency
CRIS
used?
Describing and Measuring Tobacco
K. Richter
R21DA020489, National Institute on
Yes
Treatment in Drug Treatment
Drug Abuse
Telemedicine for Smoking Cessation
K. Richter
R01HL087643, NIH National Heart, Lung No
in Rural Primary Care
and Blood Institute
Using CBPR to Implement Smoking
C. Daley
R24MD002773, National Center on
Yes
Cessation in an Urban American
Minority Health and Health Disparities
Indian Community
Centralized Disease Management for
E. Ellerbeck
R01CA101963, National Cancer Institute Yes
Rural Hospitalized Smokers
Pediatric epilepsy prevalence study
D. Lindeman
RTOI # 2008-01-01 AUCD, National
No
Center for Birth Defects and
Developmental Disabilities, CDC
Kansas Comprehensive Telehealth
E-L Nelsen, L.
Health Resources and Services
No
Services for Older Adults
Redford
Administration Office for the
Advancement of Telehealth
C4.3. We will “wire” the University of Kansas Hospital and clinic information systems to provide a
laboratory for translational informatics research.
Figure 3.4 provides a timeline for University of Kansas Hospital and associated clinics Epic clinical information
system implementation. Working with Dr. Gregory Ator, we will build upon this investment by consulting with
HICTR investigators to optimize the use of clinical systems for disseminating translational evidence and
recording measures to verify adoption. This work is also foundational to medical informatics and will allow the
evaluation of native clinical information system “signals” against manual processes used to report measures to
government, regulatory agencies, and national registries. Working with the hospital to develop capabilities to
acquire user activity data to measure the systems’ impact on workflow and clinical decision making (e.g. drugdrug interaction overrides, time to complete medication administration task) will also help us overcome
translational research’s last mile: implementation into clinical workflow. Preliminary analysis indicates that
many measures already exist in Epic Clarity audit tables, though we realize this goal may require developing
custom messages using Epic “Bridges” functionality.
15
Barohn, Richard J.
Figure 3.4. Epic rollout timeline expands T2 capabilities
C.4.4. We will develop methods to directly engage with community providers regarding EHR adoption
and explore their systems capabilities as a platform for translational research.
While the University of Kansas leadership is partnering with state agencies to shape regional policy regarding
health information exchange and electronic health record adoption, those activities primarily focus on adoption,
incentives, and external requirements as opposed to investigator-initiated research. However, there is a
unique alignment between existing health services research in underserved populations with the core mission
of the Kansas Health Policy Authority (KHPA is responsible of Medicaid in Kansas). Implementing a medical
home for all Medicaid recipients is a primary goal for the State Medicaid HIT Plan (SMHP). We will work with
Dr. Allen Greiner (director of the HICTR Community Partnership for Health Program) and Dr. Theresa
Shireman (leading the Medicaid databases section of the HICTR Personalized Medicine and Outcomes
Center), the state Regional Extension Center, and members of the Kansas Physicians Engaged in Practice
Research (KPEPR) Network24 to pilot connections between rural clinical systems and HERON to evaluate
clinical research in rural settings. This will benefit rural practices as our combined expertise will help tune
electronic health records (EHR) for meaningful use. We will determine the adopted clinical systems’ capacity
to deliver evidence-based medicine and promote the medical home approach to primary care. We also will
provide i2b2 to researchers and clinicians working with underserved populations. Our targeted evaluation will
help define research requirements in rural practices and inform the Kansas and bi-state health exchanges.
C.4.5. We will enhance data maintained by the state, national registries, and regional collaborators.
Informatics will stay abreast of the database resources in our Novel Methods 2 section. We will work with
those resources and investigators to create capabilities to enrich each data source by linking to additional
clinical information from HERON. Kansas has made significant progress with its multi-payer database (Data
Analytic Interface or DAI) - an ambitious technology infrastructure initiative to consolidate and manage health
care data for several public and private insurance programs. DAI includes the Medicaid Management
Information System (MMIS), the State Employee Health Program (SEHP), and the Kansas Health Insurance
Information System (KHIIS.) It allows analysis of health care based on episodes of treatment, disease
management, predictive modeling, and cost and outcome effectiveness measures. The Medicaid and SEHP
modules have been successfully integrated and launched in January 2010 with KHIIS integration expected to
be completed by first quarter 2011. Preliminary comparative pricing and utilization reports on pharmacy,
durable medical equipment, physician services, hospital inpatient services, and dental services have been
publicly shared with the Kansas Health Data Consortium – a multi-stakeholder advisory committee of key
government agencies, hospitals, physicians, insurers, purchasers, and consumers convened by the Kansas
Health Policy Authority to leverage the state’s data for health reform via data-driven policy recommendations.
We will adopt methods for distributed data integration between such datasets and clinical information systems
as outlined in C.2.4 above. This will further three objectives: (a) national data may provide outcome measures
lacking in acute care clinical information systems, (b) national registries, such as the National Database of
Nursing Quality Indicators, can evaluate the degree to which manually abstracted measures might be
16
Barohn, Richard J.
automatically derived from a mature clinical information system, and (c) data integration will allow hypotheses
to be explored that suggest methods for improving care.
C4.6. We will collaborate with the HICTR Personalized Medicine and Outcomes Center to provide
complex risk models for decision support to a variety of clinical specialties.
As described in the Novel Methods 2: Personalized Medicine and Outcomes Center (Section 7), regional
investigators have developed the Patient Risk Information Services Manager (PRISM): software that translates
complex risk models into fully functional decision-support tools for physicians as well as personalized
educational and informed consent documents for patients. Though initially deployed and evaluated in
cardiology (R01 HL096624 Transforming PCI Informed Consent into an Evidence-Based Decision Making
Tool), this approach has direct applicability to other clinical specialties. As investigators develop validated
statistical models, biomedical informatics will work in consultation with Dr. Spertus and his colleagues to deploy
PRISM into practice and evaluate its adoption and effectiveness. In year three, our inpatient clinical
informatics capabilities will have matured so that we may evaluate integration of PRISM risk modeling into
other clinical workflows such as inpatient Computerized Provider Order Entry provided by Epic.
D. STRUCTURE, GOVERNANCE, AND CTSA INTEGRATION
D.1.Structure and governance.
Russ Waitman, PhD will direct the overall biomedical informatics key function. Drs. Lushington and Warren will
lead and coordinate bioinformatics and health informatics activities respectively, as assistant directors. Their
expertise and qualifications were described in the Background section. Dr. Waitman will meet with the
assistant directors on a bi-weekly basis to discuss issues and go over project statuses. He will determine
support for each component through an annual review with the assistant directors. Dr. Waitman will report to
the HICTR director and be an active member of the HICTR Leadership Team. If he cannot attend, Dr. Warren
or Dr. Lushington will attend in his place. Dr. Waitman will represent HICTR at the CTSA Informatics Steering
Committee. Biomedical informatics requires collaboration across multiple groups at the University of Kansas
medical center, KU hospital, and HICTR affiliated network institutions. The HERON Executive Committee
(shown in Table 3.8) will provide guidance for institutional data sharing. Our rapid success implementing
change has been due to Dr. Waitman’s leadership and work with this group. The HERON Executive
Committee will engage the HICTR Leadership Team, Deans Council, Community Council, and Health Systems
Leadership Council regarding expansion. To achieve our informatics objectives for the medical center, Dr.
Waitman will use a working group composed of: Shelley Gebar, MPH, Senior Associate Dean for Operations;
Karen Blackwell, MS, Director of our Human Research Protection Program; Matthew Mayo, PhD, Chair of the
Department of Biostatistics; Richard Barohn, MD, HICTR Director; Gregory Ator, MD, hospital Chief Medical
Information Officer and Senior Medical Director; Chris Hansen, hospital Chief Information Officer; and James
Bingham Associate Vice Chancellor for Information Resources and Chief Information Officer.
Table 3.8 HERON Executive Committee
Name
Title
Organization
Paul Terranova
Shelley Gebar
Vice Chancellor for Research
Chief of Staff and Senior Associate
Dean for Operations
Director of Medical Informatics
Executive Vice President and Chief
Operating Officer
Executive Vice President and Chief
Financial Officer
Senior Vice President for Ambulatory
Services and Chief Information Officer
President
Chief Executive Officer
Chief Operating Officer
University of Kansas Medical Center
University of Kansas Medical Center
Russ Waitman
Tammy Peterman
Scott Glasrud
Chris Hansen
Kirk Benson
Jim Albertson
Teresa Neely
D.2. Data and Software Sharing.
17
University of Kansas Medical Center
University of Kansas Hospital
University of Kansas Hospital
University of Kansas Hospital
University of Kansas Physicians
University of Kansas Physicians
University of Kansas Physicians
Barohn, Richard J.
The HICTR will adhere to NIH guidelines on data sharing. We will make scientific data available as widely as
possible while safeguarding the privacy of our participants and protecting confidential or proprietary data. As
described in Aim 2, we have developed data sharing agreements at an organizational level to streamline
requests for de-identified data. Our repository’s use of i2b2 and UMLS terminologies should facilitate national
collaboration. Because of the risk for re-identification using even de-identified data, our system access and
data use agreements reinforce the investigator’s responsibilities as custodians of patient data to prevent
improper disclosure. The informatics consultation service will assist investigators with data sharing and with
providing meta-data to identify the data and the methods used in data processing. Data generated from
microarray experiments will be posted by investigators on publically accessible databases (e.g. NCBI with
MIAME compliant meta-data). All new software developed using support from this grant will be available to
others using open source tools and licenses. For work internal to i2b2, modifications will follow the terms of
the Brigham and Women’s Hospital Inc. i2b2 Open Source Software License. While not open source, we also
anticipate that our partnership with Velos regarding the eIRB module will prove beneficial to other CTSA and
research organizations using Velos.
E. INTERACTION AND INTEGRATION OF INFORMATICS WITH THE OTHER HICTR COMPONENTS.
Clinical & Translational Research
Education Center (CTREC)
Participant & Clinical Interaction
Resources Program (PCIRP)
Community Partnership for Health
Program (CPH)
Novel Methods 1: Institute for
Advancing Medical Innovations
(NM:IAMI)
Novel Methods 2: Personalized
Medicine and Outcomes Center
Pilot and Collaborative Studies
Funding Program
Translational Technologies Resource
Center (TTRC)
Biostatistics
Regulatory Knowledge and Support
In aim C.1, the portal will promote access to educational offerings. Dr.
Lushington coordinates educational offering from CBI and the School
of Engineering which may be relevant for certain translational
investigators. Dr. Warren will coordinate educational offerings (ex:
AMIA 10x10) from the Center for Health Informatics and provide
guidance regarding educational simulations.
CRIS and HERON support for PCIRP research (e.g, study form
development using standardized terminology, potential study cohort
size and hypotheses generation). PCIRP provides oversight for
HICTR Participant Registry via the Data Request Committee.
We will work with CPH leadership and KUMC Outreach to develop and
evaluate methods for T2 translation of research into practice in
partnership with the rural KPEPR network in aim C.4.
Bioinformatics provides extensive informatics consultation for IAMIsupported high throughput screening and lead optimization projects (8
distinct projects within the past year alone). We specifically extend
bioinformatics expertise to integrate pharmacokinetics and clinical
trials in aim C.3.
In aim C.1, we will advance the adoption of web-based tools for
collaborative research and on-line communities such as CRIS/Velos
and www.cvoutcomes.org. In aim C.4, we will collaborate with Dr.
Spertus on applying PRISM to other workflows. In aim C.4 we will
collaborate with Drs. Shireman and N. Dunton regarding Medicaid and
nursing quality databases respectively.
Research funded by pilot programs will use informatics resources. In
aim C.1, the portal will be used to track Pilot program requests.
Many of the TTRC resources are informatics technologies. This is
facilitated by the fact that the informatics co-director, Dr. Lushington,
also directs the TTRC Molecular Biomarker Core. In aim C.3,
informatics will work with the Biological Tissue Repository to link
specimens to the clinical information and improve specimen
management and with the Pharmacokinetic/Pharmacodynamics
(PK/PD) program to integrate findings with the Comprehensive
Research Information System and information from the clinical record
for phase I clinical trials.
Biostatistics and Informatics work closely regarding maintenance and
expansion of data-related resources and data warehouses for the
HICTR; compliance with national standards and innovation in
developing interfaces between biostatistics and informatics will be
ongoing. The CRIS team is led by the Biomedical Informatics director
who is a faculty member in the Department of Biostatistics.
The portal enhances research users’ access to regulatory materials
and processes. Informatics is directing the electronic IRB to improve
18
Barohn, Richard J.
Program
Ethics Program
regulatory oversight and ease access for investigators. Regulatory
support uses informatics methods to track initial project requests,
research monitoring, and clinical research coordinator services.
The Ethics program will be engaged in oversight regarding ethical and
privacy concerns in the development of databases especially HERON
(Aim 2). Our recent capability to provide electronic IRB processes and
use HERON to determine if we have the patient population to carry out
studies will inform the ethical oversight of research protocols.
F. EVALUATION, IMPLEMENTATION AND MILESTONES
The evaluation plan and identification of specific milestones for the Biomedical Informatics key function are in
the Evaluation Section (Section 14) and the Implementation and Milestones Section (Section 15) of this
application. Figure 3.5 shown below also provides an overview of our specific aims in relation to the in-kind
current and projected staffing and additional personnel requested. Where specific aims are aligned with
other institutional objectives for informatics and the biomedical informatics team, they will be provided in-kind.
19
Barohn, Richard J.
Key Personnel: RW Russ Waitman; GL Gerry Lushington; JW Judy Warren; AC Arvinder Choudhary; DC Dan Connolly; MV Mahesh Visvanathan
Teams: kbc K-INBRE Bioinformatics Core; bmi Biomedical Informatics, telemed Kansas University Center for Telemedicine and Telehealth
Note: Bar height indicates relative effort
Combined Teams
Full Time
Equivalents (FTE) for
Clinical and
Translational
Research
5.5
3.7
Existing
CRIS
Initiatives
and
Services
Research Budgeting
0.8 FTE (bmi)
eIRB rollout
1.2 FTE (bmi)
9.5
9
8.3
7.5 post CTSA funding
Research Budgeting maintain and refine 0.2 FTE (bmi)
eIRB adv. process
mgt. 1.0 FTE (bmi)
eIRB maintain and refine 0.3 FTE (bmi)
CRIS & demographics and lab
integration 1 FTE (DC, bmi)
CRIS improved integration maintain and refine 0.3 FTE (bmi)
C4.6 ePRISM -> EPIC integration 1 FTE
(RW, JW, bmi)
C.4.5 HERON<-> Medicaid/Medicare dataset
integration 0.6 FTE (AC, RW, bmi)
Aim 4
C.4.3 wire EPIC & systems for T2 research with
HERON 1 FTE (JW, RW, DC, bmi)
C.4.1 HIE planning (HC, RW)
Community EMR adoption & HIE established
Maintain 0.2 FTE (bmi)
Maintain intervention monitoring 0.3 FTE (bmi)
C.4.4 Community EHRs and HIE as platform for research
1.7 FTE (JW, RW, bmi)
C4.2 CRIS and Telemedicine in kind/fee for service (RS, bmi, telemed)
C3.3 BI, CRIS,HERON integration 0.7
FTE (GL, MV, RW, kbc, bmi)
C3.2 PK/PD and CRIS 0.7 FTE (GL, MV, RW,
kbc, bmi)
C3.1 Clinical Pathology and HERON 1
FTE (bmi, RW)
Aim 3
C3.1 Biological Tissue Repository and CRIS or
caTissue 1 FTE (bmi, RW)
Maintain (bmi) 0.1 FTE
Maintain, refine Clinical Pathology integration 0.2 FTE (bmi)
Biological Tissue Repository integration maintain, refine 0.3 FTE (bmi)
Bioinformatics in kind/fee for service (GL coord 0.05 FTE, kbc)
CRIS/caBIG, start ontology
HERON: ontology mgt enhanced 0.7 FTE
(JW, bmi)
HERON: ontology mgt maintain 0.5 FTE (JW, GL, bmi)
Aim 2
HERON Construct 2.5 FTE
(AC, DC, RW, bmi)
Consults (RW)
HERON: maintain + ETL/messaging + feeds
3 FTE (AC, DC, RW, bmi)
HERON: maintain, refine, add feeds
2.5 FTE (AC, DC, RW, bmi)
Staffed Informatics Consult Service Established
0.7 FTE (JW, bmi)
Informatics Consult Service Team Based
0.7 FTE (RW, JW, bmi)
Aim 1
Portal: foundation 0.5 FTE
(DC, RW, bmi)
2011
Portal: HICTR focused 0.6 FTE (DC, RW, bmi)
Bioinformatics Resources 0.05 FTE (GL)
2012
2013
Portal maintain, refine 0.5 FTE (bmi, DC, kbc)
2014
2015
Jul 2010
2016
Jul 2016
Figure 3.5. Biomedical Informatics Aims and Team Resources Timeline
G. REFERENCES
1. Pulley JM, Harris PA, Yarbrough T, Swafford J, Edwards T, Bernard GR. Acad Med. 2010 Jan;85(1):164-8.
An informatics-based tool to assist researchers in initiating research at an academic medical center: Vanderbilt
Customized Action Plan.
20
Barohn, Richard J.
2. Parks M, Anderson W. E-Project: A Web-based System to Enhance the Pure Matrix of Project
Management for Biostatistical Informatics. The Society for Clinical Trials 31st Annual Meeting, Baltimore
Maryland, May 17, 2010.
3. Kriete, A and Eils, R. Systems Biology. Academic Press, Burlington MA, 2006.
4. Sears CS, Prescott VM, McDonald CJ. The Indiana Network for Patient Care: A Case Study of a
Successful Healthcare Data Sharing Agreement. American Bar Association Health eSource, September 2005.
5. Weber GM, Murphy SN, McMurry AJ, Macfadden D, Nigrin DJ, Churchill S, Kohane IS. The Shared Health
Research Information Network (SHRINE): a prototype federated query tool for clinical data repositories. J Am
Med Inform Assoc. 2009 Sep-Oct;16(5):624-30.
6. Roden DM, Pulley JM, Basford MA, Bernard GR, Clayton EW, Balser JR, Masys DR. Development of a
large-scale de-identified DNA biobank to enable personalized medicine. Clin Pharmacol Ther. 2008
Sep;84(3):362-9. Epub 2008 May 21.
7. Pulley J, Clayton E, Bernard GR, Roden DM, Masys DR. Principles of human subjects protections applied
in an opt-out, de-identified biobank. Clin Transl Sci. 2010 Feb;3(1):42-8.
8. Murphy SN, Weber G, Mendis M, Gainer V, Chueh HC, Churchill S, Kohane I. Serving the enterprise and
beyond with informatics for integrating biology and the bedside (i2b2). J Am Med Inform Assoc. 2010 MarApr;17(2):124-30.
9. Friedlin FJ, McDonald CJ. A software tool for removing patient identifying information from clinical
documents. J Am Med Inform Assoc. 2008 Sep-Oct;15(5):601-10. Epub 2008 Jun 25.
10. Loukides G, Denny JC, Malin B. The disclosure of diagnosis codes can breach research participants'
privacy. J Am Med Inform Assoc. 2010 May 1;17(3):322-7.
11. Benitez K, Malin B. Evaluating re-identification risks with respect to the HIPAA privacy rule. J Am Med
Inform Assoc. 2010 Mar 1;17(2):169-77.
12. Karp DR, Carlin S, Cook-Deegan R, Ford DE, Geller G, Glass DN, Greely H, Guthridge J, Kahn J, Kaslow
R, Kraft C, Macqueen K, Malin B, Scheuerman RH, Sugarman J. Ethical and practical issues associated with
aggregating databases. PLoS Med. 2008 Sep 23;5(9):e190.
13. Yeniterzi R, Aberdeen J, Bayer S, Wellner B, Hirschman L, Malin B. Effects of personal identifier
resynthesis on clinical text de-identification. J Am Med Inform Assoc. 2010 Mar 1;17(2):159-68.
14. SP 800-66 Rev 1 Oct 2008 An Introductory Resource Guide for Implementing the Health Insurance
Portability and Accountability Act (HIPAA) Security Rule http://csrc.nist.gov/publications/nistpubs/800-66Rev1/SP-800-66-Revision1.pdf
15. Wynden R, Weiner MG, Sim I, Gabriel D, Casale M, Carini S, Hastings S, Ervin D, Tu S, Gennari J,
Anderson N, Mobed K, Lakshminarayanan P, Massary M, Cucina R. Ontology mapping and data discovery for
the translational investigator. Presented at the 2010 AMIA Summit on Clinical Research Informatics meeting,
March 12-13 2010, San Francisco, CA.
16. Blumenthal D, Tavenner M. The "Meaningful Use" Regulation for Electronic Health Records. N Engl J
Med. 2010 Jul 13.
17. Diamond CC, Mostashari F, Shirky C. Collecting and sharing data for population health: a new paradigm.
Health Aff (Millwood). 2009 Mar-Apr;28(2):454-66.
18. Berman JJ. Zero-check: a zero-knowledge protocol for reconciling patient identities across institutions.
Arch Pathol Lab Med. 2004 Mar;128(3):344-6.
19. Nass SJ, Moses HL, Mendelsohn. A National Cancer Clinical Trials System for the 21st Century:
Reinvigorating the NCI Cooperative Group Program. Committee on Cancer Clinical Trials and the NCI
Cooperative Group Program; Institute of Medicine. Washington, DC: National Academies Press, 2010.
20. Zhou, Z., Lee, I.-H., Lushington, G.H., Visvanathan, M. ISMB 2010, submitted.
21. Parkinson M. Executive Order 10-06 Kansas Health Information Exchange, Inc. Governor of the State of
Kansas. http://governor.ks.gov/issues-a-initiatives/executive-orders/717-executive-order-10-06
22. Spaulding RJ, Russo T, Cook DJ, Doolittle GC. Diffusion theory and telemedicine adoption by Kansas
health-care providers: critical factors in telemedicine adoption for improved patient access. J Telemed
Telecare. 2005;11 Suppl 1:107-9.
23. Doolittle GC, Spaulding RJ. Defining the needs of a telemedicine service. J Telemed Telecare.
2006;12(6):276-84.
23. Cox LS, Cupertino AP, Mussulman LM, Nazir N, Greiner KA, Mahnken JD, Ahluwalia JS, Ellerbeck EF.
Design and baseline characteristics from the KAN-QUIT disease management intervention for rural smokers in
primary care. Prev Med. 2008 Aug;47(2):200-5.
21