CLINICAL AND TRANSLATIONAL SCIENCE INFORMATICS
INFRASTRUCTURE
A Framework and Case Study
Arkalgud Ramaprasad, Annette L. Valenta
University of Illinois at Chicago, Chicago, IL 60607, USA
Ian Brooks
National Center for Supercomputing Applications, Urbana, IL 61801, USA
Keywords: Clinical and Translational Science, Informatics Infrastructure.
Abstract: This paper presents a comprehensive socio-technical framework for the design and development of a
Clinical and Translational Science Informatics Infrastructure (CTSII). Based on our experience with
developing and applying the framework we present a case study to illustrate the issues that arise in the
creating a CTSII, and how possibly these issues can be resolved. The framework is presented as a menu
with six columns, each column representing a dimension of the framework. The categories within each
dimension can be concatenated, with the conjunctive phrases/words between the columns, to form sentences
that describe all the functions of the CTSII. Elucidation of all the combinations will provide an exhaustive
list of all the possible functions of CTSII.
1 INTRODUCTION
In 2002, the National Institutes of Health (NIH) in
the US charted a roadmap for this century to identify
opportunities and gaps in biomedical research in
order to make the biggest impact on the progress of
medical research (NIH Office of Communications,
2003). The roadmap seeks to foster new pathways to
discovery, to develop innovative research teams of
the future, and to reengineer the clinical research
enterprise (NIH Office of Portfolio Analysis and
Strategic Initiatives, 2006). It seeks to create a new
discipline called Clinical and Translational Science
(CTS) to reduce the time-to-practice of biomedical
scientific discoveries, and the time-to-research of
clinical and community health care issues (NIH
Office of Portfolio Analysis and Strategic Initiatives,
2007).
Clinical and Translational Science (CTS) by
definition is interdisciplinary; however, it is difficult
to foster interdisciplinary research cutting across
basic, animal, clinical, and public health disciplines.
One barrier to such research is disciplinary silos that
often manifest themselves in the form of
departments, colleges, journals, and conferences. A
well designed CTS Informatics Infrastructure
(CTSII) can help break these barriers.
It is natural for people to know more about the
research and researchers in their discipline than in
others. Disciplinary research is the foundation of
academic advancement, at least in the short run. The
incentives systems in universities are woven around
disciplinary productivity and the performance is
evaluated by peers in the discipline. Consequently,
the silos foster relationships within their boundaries
rather than across them. While disciplinary research
is necessary, it is also necessary to cut across these
silos to develop CTS. How can CTSII help?
There is a disconnection between the availability
and the use of informatics tools and techniques.
Many popular consumer informatics tools
demonstrate immense potential. Our objective is to
import these tools and techniques and apply them to
create an effective CTSII. Metaphorically, the ideal
CTSII is a combination of Google™, Facebook™,
Amazon™, and Orbitz™. It should have the global
indexing, ranking, and search capabilities of
Google™; the social networking capabilities of
213
Ramaprasad A., L. Valenta A. and Brooks I. (2008).
CLINICAL AND TRANSLATIONAL SCIENCE INFORMATICS INFRASTRUCTURE - A Framework and Case Study.
In Proceedings of the First International Conference on Health Informatics, pages 213-218
Copyright
c
 SciTePress

Facebook™; the data mining, cataloging and
customer [researcher] relationship management of
Amazon™; and the complex scheduling [chaining]
capabilities of Orbitz™. Analogues of these four
systems, which have revolutionized consumer
informatics, will serve as excellent bases for the
design of CTSII (Valenta et al., 2007).
2 CTSII FRAMEWORK
To break the silos of research, while simultaneously
advancing science, the CTSII should facilitate back
and forth translation of information between basic
researchers, animal researchers, clinical
investigators, and public health researchers. (We
use information to generically connote data,
information, and knowledge.) It must support the
translation of information between the sub-
disciplines of each group as well (Valenta et al.,
2007).
The quality and quantity of information
translation will determine the effectiveness of CTS.
In the following sections we present a systematic
framework to analyze and design CTSII. We are
currently using the framework to develop the CTSII
at our university. The framework incorporates,
integrates, and extends the ideas from the CTS
proposals that have been funded by NIH. It has been
presented to and discussed with a large group of
researchers across the campus from a wide range of
disciplines – including medicine, nursing, applied
health sciences, engineering, business
administration, public health, and pharmacy. We
will discuss the issues that have been raised during
our discussions, and how we plan to address them.
CTSII is not just a technological infrastructure,
but also a social, psychological, organizational, and
educational one – a fact that can be easily
overlooked. The proposed system will, by its very
design, restructure workgroups, causing stress to the
organization, its social groups, and individuals.
Appropriate education, consultation, training,
change management, evaluation, and assessment
mechanisms will be critical for the success of CTSII.
2.1 CTSII Menu
We present our CTSII framework as a menu with six
columns (Figure 1), each representing a dimension
of the framework. The six dimensions represent: (a)
the different types of integration central to
translation, (b) the different areas of research that
have to be translated, (c) the resources available for
translation, (d) the diseases that form the focal point
of translational research, (e) the methodological
steps in any research (including translation
research), and (f) the tools for translation. In fact, the
menu is a method of representing a matrix with six
dimensions; each dimension being represented by a
column. The categories within each dimension can
be concatenated, with the conjunctive phrases/words
between the columns, to form sentences that
describe all the functions of the CTSII. Some
example combinations follow:
• Lateral integration of basic research databases
related to HIV/AIDS for theory construction
using scientist relationship management.
• Temporal integration of public health
researchers related to asthma for empirical
testing using scientific workflow management.
It can be seen that even with the abbreviated list
of entries in the columns, the total number of
combinations is very large, indicating the
complexity of CTSII. Elucidation of all the
combinations will provide an exhaustive list of all
the possible functions of CTSII. It would be
difficult, if not impossible, to incorporate all of them
in one system – they have to be prioritized. The
following provides a description of the six
dimensions and a sample of the categories within
each.
2.1.1 Integration Dimension
Integration is one of the major driving forces behind
CTS. It has been a somewhat elusive but important
goal sought through earlier initiatives in
interdisciplinary and multi-disciplinary research.
The objective of CTS is to substitute serendipitous
integration with systematic integration.
Research
Basic
research
Animal
research
Human
research
Public health
research
Resources
Databases
Knowledgebases
Researchers
Tissue banks
Ani mal model
banks
Subject banks
Registries
Methods
Theory
construction
Hypotheses
development
Empirical testing
Clinical
application
Community
application
for
using
Tools
Logical data
warehousing (LDW)
Data extraction,
mining, and
visualization
(DEMV)
Statistical analysis
and modeling
(StAM)
Simulation and
modeling (SiAM)
Scientists
relationship
management (SRM)
Scientific workflow
management
(SWM)
Scientists social
networking (SSN)
Scientific knowledge
management (SKM)
Interdisciplinary
learning
management (ILM)
Diseases
HIV/AIDS
Asth ma
Obesity
Diabetes
Cancer
related to
Integration
Lateral
(cross-silo)
Vertical
(within-silo)
Temporal
Geographical
integration of
Figure 1: CTSII Menu.
HEALTHINF 2008 - International Conference on Health Informatics
214

The integration dimension has four categories.
They are: (a) Lateral integration (cross-silo); (b)
Vertical integration (within-silo); (c) Temporal
integration (over time); and (d) Geographical
integration (across many locations).
CTS requires integration across and within silos
of basic research, animal research, human research,
and public health research. To be effective, the
research also has to be integrated longitudinally –
over time, and across many geographical locations
where the research resources may be located. Hence
the four categories of the integration dimension.
Each of the four types of integration imposes a
different set of requirements on the CTSII. In
addition to the purchase and installation of the
hardware, software, and networks, the participants
will have to be informed, and trained to use the new
infrastructure, and the processes of scientific
collaboration will have to be reengineered to utilize
the new infrastructure.
2.1.2 Research Dimension
CTS encompasses four types of research. They are:
(a) Basic research; (b) Animal research; (c) Human
research; and (d) Public health research.
Each of these phases includes many components;
for example, the human research phase includes
human trials, treatment modalities, and clinical
practice. Similarly, public health research includes
dissemination of the results to the public and
community.
These four phases encapsulate the concept of
moving basic research to the patient’s bedside and
the public – the central tenet of CTS. While these
four phases are commonly presented as a
progressive sequence, research ideas may originate
in any phase and move across these phases in any
order. Thus, research ideas may originate from
basic research and may be fed-forward directly into
human research; or, they may originate in public
health research and may be fed-back directly to
animal research.
One of the major concerns of CTS is that each
phase tends to be a silo. These silos are reinforced
by norms of the disciplines and associated
incentives. The silos inhibit feed-forward and feed-
back. Consequently, the movement across the phases
tends to be slow and not smooth. A significant body
of research may accumulate in a phase without any
impact on the subsequent phases through feed-
forward or on prior phases through feed-back. When
this happens, both the creative and corrective value
of feed-forward and feed-back is lost. Streamlining
the feed-back and feed-forward mechanisms using
CTSII, on the other hand, can improve both the
efficiency and effectiveness of translation
(Ramaprasad, 1979, 1982, 1983). Similarly, silos
within silos can inhibit feed-in. Streamlining feed-in
using CTSII can lead to improvement in the quality
of feed-forward and feed-back.
2.2 Resources Dimension
CTS requires integration of a large number of
resources. They are: (a) Databases – central,
homegrown, relational, flat files, etc.; (b)
Knowledgebases – structured, unstructured, text,
formal, informal, etc.; (c) Researcher databases –
directories, résumés, profiles, etc.; (d) Tissue banks;
(d) Animal model banks; (e) Subject banks –
deidentified subjects, identified subjects, volunteers,
etc.; and (f) Registries
Under each of the above categories, there is
likely to be a large number of subcategories, and
ultimately a larger number of actual resources.
Developing an inventory of these resources will be a
key step in developing the informatics infrastructure.
Researchers often focus exclusively on the
integration of databases as a requirement of the
CTSII. While databases are important, integration of
information about other resources is equally
important. A clinical researcher, for example,
probably does not need access to the genomic
database used by a basic researcher to discover a
gene marker for breast cancer, but needs information
about the marker and how to test for it. The clinical
and basic researchers need to know which other
researchers can help them develop a reliable test for
the marker, how they can obtain a panel of subjects
for a trial, and the tools to evaluate the results of the
trial.
The CTSII should, ideally, replace the usually
ad-hoc processes for accessing these resources with
more efficient and effective processes. By making
the resources visible and accessible to all, the CTSII
should improve both the quality and utilization of
these resources.
2.3 Diseases Dimension
Different diseases may require different
combinations of capabilities in the CTSII. While
gene-based diseases may require the ability to
handle large volumes of genomic data,
environmentally induced diseases may require the
ability to manage disparate public health data.
Similarly, some diseases may require the ability to
CLINICAL AND TRANSLATIONAL SCIENCE INFORMATICS INFRASTRUCTURE - A Framework and Case Study
215

educate and interact with the public health workers
and the community physicians and nurses.
The CTS in an institution may be focused on a
few or a large number of diseases. Following is an
illustrative list of diseases suited to the current
research at our university: HIV/AIDS, asthma,
obesity, diabetes, and cancer.
Focusing on the above will fit our university’s
expertise as well as its mission as a premier urban
research university. Other institutions may have
other priorities based on their vision, mission,
strategy, and environment. The design of the CTSII
will naturally depend upon the different informatics
required for the management of the diseases in
question.
2.4 Methods Dimension
The informatics requirements of different stages of
research are different. The CTSII should support all
the stages.
The first three categories in the methods
dimension are standard stages in research
methodology. Carrying them forward to clinical and
community application, the last two categories, are
an essential part of translation. Thus, the five
categories are: (a) Theory construction; (b)
Hypotheses development; (c) Empirical testing; (d)
Clinical application; and (e) Community application.
The type of translation required for translational
theory construction may be quite different from that
required for translational clinical application. The
former may require metaphorical translation from
one discipline to another; the latter may require a
methodological translation. (Please see section 3
below for a more detailed discussion of the different
types of translation.) Thus, for the metaphorical
translation, the CTSII may have to facilitate the
social networking of theoreticians from the different
disciplines, while for methodological translation, the
methodologists from the disciplines may have to be
brought together. These groups, in turn, may need
access to different types of knowledge.
2.5 Tools Dimension
CTS requires integration using many tools. This
dimension articulates the metaphor we used earlier
for CTSII as a combination of Google™,
Facebook™, Amazon™, and Orbitz™. Under each
of the following categories, there is likely to be a
large number of tools. In fact, many tools may span
multiple categories. Developing an inventory of
these tools will be a key step in developing the
informatics infrastructure. The categories of tools
include:
• Logical data warehousing tools;
• Data extraction, mining, and visualization tools;
• Statistical analysis and modeling tools;
• Simulation and modeling tools;
• Scientist relationship management tools;
• Scientific workflow management tools;
• Scientist social networking tools;
• Scientific knowledge management tools; and
• Interdisciplinary learning management tools.
The above is not an exhaustive list of the types of
tools. There are many other types of tools, and many
more are likely to emerge in the future. There will
also be many more tools within each type. The
difficulty is not one of the availability of tools, but
of their application to CTS and in developing
workflow management capabilities to integrate the
tools. The cross-fertilization of the application of
these tools across traditional CTS disciplines, and
between non-CTS disciplines (for example:
marketing, production and operations management,
semiotics, computer science, and library science)
will be facilitated by the CTSII.
3 CTSII CASE STUDY
In this section, we will present the key issues that
have arisen as we have tried to adopt and apply the
CTSII framework described above over the past
year. These issues are unlikely to be unique to our
institution. It is intrinsic to the nature of
transformation that CTS is trying to engender. In
addition to illuminating the process of application,
the case also highlights the importance of
considering the psychological, social, organizational,
and educational aspects of the CTSII.
3.1 Not Just Databases
To many, the term informatics appears to connote
only databases. A number of early meetings focused
exclusively on developing an inventory of the
databases and making them easily accessible to other
researchers. Perhaps it reflected the participants’
primary concern with their research. It took many
meetings to convince the participants (a) that
databases were only one type of informatics
resources, and (b) that informatics should focus on a
broader range of functions than simply creating,
integrating, and providing access to the databases.
HEALTHINF 2008 - International Conference on Health Informatics
216

3.2 Oracle™ and Google™ Mindsets
Related to equating informatics with databases, there
was a strong tendency to think exclusively in terms
of structured databases: simple flat Excel files and
more complex relational databases. This could be
called an ‘Oracle™’ mindset to distinguish it from
the ‘Google™’ mindset – storage and search based
on unstructured repositories and documents. Part of
the bias appears to be due to lack of knowledge of
today’s information systems’ ability to index,
search, and access large volumes of unstructured
data in documents and other sources. Even though
researchers used Google™ and similar search
engines, few have knowledge of how search engines
work. The bias was compounded by an unrealistic
equating of the cost of shrink-wrapped databases
with the cost of building a structured database for
unstructured data.
3.3 Thinking Outside the SiloS
Often, informatics requirements presented by the
other core groups involved in the CTS effort appear
to be focused on research within their silos, and not
across silos. They do not seem to be focused on the
truly translational processes of feed-forward and
feed-back, but focused on feed-in. While many of
the informatics requirements will no doubt facilitate
research, it is not clear how they will facilitate
translational research. It will perhaps take some
more time for the researchers to change their
framework through education, consultation, and
experience with the proposed CTSII.
3.4 Foundation and Frontier
Requirements
A consequence of the issues discussed in the above
sections is that most of the requirements presented
tend to be what we have called ‘foundation’
requirements; for example: web-accessible databases
and an interactive directory of researchers as a
baseline requirement. They are necessary for CTS,
and for that matter, any research. They are unlikely,
however, to transform the CTS research at the
institution and to distinguish one institution from
another. That requires the ‘frontier’ capabilities.
These capabilities reflect some of the best practices
across different types of organizations and may have
to be adapted to CTS. While many of the
researchers are aware of these capabilities and have
used them, they do not often see their application to
CTS. The barriers to the transfer of these best
practices are many, one being the inaccurately
perceived high cost of performing certain functions.
The cost of full-text search is an example of the
incorrect perception. The complexity and difficulty
of social network analysis is another example of
such perception.
It will likely be a while before the researchers
will start using CTSII as a tool for CTS research
instead of just a service to improve their current
research. Until then, the stated requirements are
likely to be at the foundation end of the spectrum
rather than at the frontier end. Movement of the
researchers’ thinking along the spectrum will be an
important part of the transformation. When, in fact,
they move into the frontier, there will likely be a
sudden cascade of new and innovative requirements
of CTSII.
3.5 Metaphoric Success
The metaphor ‘CTSII = Google™ + Facebook™ +
Amazon™ + Orbitz™’ was very successful in
communicating our ideas and to reframe their
thinking. Many people, given their age, were not
familiar with Facebook™, although some reported
their children were using it or a similar system. They
were less familiar with Orbitz™ than with
Expedia™ and Travelocity™. We chose Orbitz™ as
a result of the recent experience of one of the
authors with booking a complex international trip.
Of course, most researchers were familiar with and
had used Google™ and Amazon™.
Almost everybody liked the metaphor once it
was explained to them. Many, especially those in
informatics related disciplines, intuitively grasped its
meaning, and immediately liked it. We still had to
explain the application of the metaphor to the design
of CTSII, however. The barriers to application of the
metaphor were similar to those explained in 3.3
above. The metaphors and the menu have been the
anchors of all the recent discussions on CTSII.
3.6 Feed-forward, Feed-backward, and
Feed-in
In our eagerness to adhere to the spirit and letter of
CTS, we initially proposed feed-forward and feed-
back mechanisms as the bases for translation
between the four types of research. Feed-in was
added in response the expressed need of improving
the informatics within each discipline, too; however,
the feed-concepts did not appear to resonate with the
CTS researchers. As a consequence, we renamed
“feed-forward and feed-back” to “two-way
CLINICAL AND TRANSLATIONAL SCIENCE INFORMATICS INFRASTRUCTURE - A Framework and Case Study
217

translation of information”. We continue to believe,
however, that the feed-concepts are central to CTS,
and part of the transformation will be to understand
and apply these concepts systematically.
In a sense, the CTS paradigm has close parallels
to the Continuous Improvement Paradigm that
emerged in the context of the Total Quality
Management (TQM) movement (Söderholm, 2004).
The feed-concepts will likely be as critical to CTS as
they have been to Continuous Improvement.
3.7 Technological Bias
Given the focus of most researchers on the technical
aspects of informatics – the databases and data
warehouses, it took time to incorporate the socio-
technical view in the discussions. Clearly, help
desk, consultation, education, training, and other
user services must be part of the CTSII. The idea of
little or no human intervention in this transformation
is not realistic.
3.8 Local and Global Transformation
The major impetus for CTS is to transform research.
Some of this transformation can be in the CTSII, and
some using CTSII. The CTSII transformation can be
local or global. Local transformation is one which is
innovative within an institution, but not globally.
Other institutions may have adopted the innovations
earlier. Global transformation is one which is
innovative within the institution and without. It is
the first of a kind, anywhere.
Both local and global CTSII transformations are
necessary for the success of CTS efforts in an
institution. Many local transformations may be
necessary to bring an institution on par with other
institutions; at least a few global transformations
will be necessary to provide a competitive advantage
to the institution.
4 CONCLUSIONS
An informatics infrastructure for clinical and
translational science (CTS) can be complex. We
conceptualize the information flows for CTS as a
bidirectional flow of feed-forward and feed-back
between basic researchers, animal researchers,
clinical researchers, and public health researchers,
and feed-in among researchers within a discipline.
We present a six-dimensional framework as a menu
to deconstruct the complexity and help specify the
requirements of the clinical and translational science
informatics infrastructure (CTSII) for an institution.
The framework provides a simple concise way of
enumerating all the functions required of a CTSII.
Last, we present a case study of our experience in
using the framework at our institution. The case
study illustrates some of the barriers to the
application of the framework and how these barriers
can be overcome.
REFERENCES
NIH Office of Communications. NIH News: NIH
announces strategy to accelerate medical research
progress. September 30, 2003. Available at
http://www.nih.gov/news/pr/sep2003/od-30.htm.
Accessed May 31, 2007.
NIH Office of Portfolio Analysis and Strategic Initiatives.
Re-engineering the clinical research enterprise. April
7, 2006. Available at http://nihroadmap.nih.gov/
clinicalresearch/. Accessed May 31, 2007.
NIH Office of Portfolio Analysis and Strategic Initiatives.
Re-engineering the clinical research enterprise. March
1, 2007. Available at http://nihroadmap.nih.gov/
clinicalresearch/overview-translational.asp Accessed
May 31, 2007.
Ramaprasad, A., 1979. The Role of Feedback in
Organizational Change: A Review and Redefinition.
Cybernetica, 22 (2), 105-113.
Ramaprasad, A., 1982. Revolutionary Change and
Strategic Management. Behavioral Science, 27 (4),
387-392.
Ramaprasad, A., 1983. On the Definition of Feedback,
Behavioral Science, 28 (1), 4-13.
Ramaprasad, A. and Rai, A., 1996. Envisioning
Management of Information. Omega: The
International Journal of Management Science, 24(2),
179-193.
Söderholm, P., 2004. Continuous Improvements of
Complex Technical Systems: a Theoretical Quality
Management Framework Supported by Requirements
Management and Health Management. Total Quality
Management, Vol. 15, No. 4, 511–525.
Valenta, A.L., Brooks, I., Laureto, R.A., and Ramaprasad,
A., 2007. Informatics to Support Clinical and
Translational Science. Journal of Health Information
Management (forthcoming).
HEALTHINF 2008 - International Conference on Health Informatics
218