Biomedical Engineering Reference
In-Depth Information
of Diseases, 10th revision (ICD-10), clinical trial observations from CDISC
SDTM [15], and a dictionary of curated inhibitors. Our curators created an
internal J&J drugs dictionary, clinical trials metadata dictionary, and a cell line
dictionary.
16.6
DEVELOPMENT METHODOLOGY
The development took about 12 months after the fi rst demonstration proto-
type using the agile software development methodology. A team of scientists,
bioinformatics professionals, and software engineers designed and built the
system. The implementation required several distinct efforts. First, we needed
to develop appropriate rules and regulations for adding data to the system and
granting access to users. Second, we needed to design a system for effi ciently
storing and querying data. Finally, we needed to develop a system for securely
accessing stored data.
A set of appropriate policies for adding data to the system and granting
access to data is necessary to ensure cooperation from the data owners and to
avoid compliance issues. We modeled the system on processes that were used
in academic medical centers in the United Sates. Academic medical centers
usually restrict data access to qualifi ed researchers. Researchers are granted
access only to data for specifi c studies and are only allowed to access the data
after approval by an institutional review board. Additionally, clinical research
data are protected by law through HIPAA (Health Insurance Portability and
Accountability Act) and HITECH (Health Information Technology for
Economic and Clinical Health Act) regulations. In addition to these restric-
tions, pharmaceutical companies need to be also bound by U.S. Food and Drug
Administration (FDA) regulations and health care compliance.
To develop the appropriate policies for tranSMART, we formed a data
governance working group and created a new data steward role in the orga-
nization. The data steward is responsible for the data acquisition process,
including negotiating with the source data owners, ensuring safe transfer of
the data from source systems to the staging area of tranSMART, and the
security and integrity of the data in transit.
Publication rules govern the publication of fi ndings resulting from mining
the data warehouse. Finally, the data governance working group led an effort
to develop guidelines for user training and ethics training which are prereq-
uisite to access. All users are required to attend the offi cial training before
being given access to the system.
Two parallel efforts were undertaken for software implementation. First, a
team of application developers designed and built a Web-based application to
provide graphical user access to the data. The application developers worked
closely with research scientists to design the system.
Second, a team consisting of data curators and ETL (extract, transform, and
load) engineers worked with research scientists, biostatisticians, and informat-
Search WWH ::
Custom Search