Marmaras_edit3.qxp 5/6/07 12:03 pm Page 12
Optimising eClinical
Data Integration – Business Requirements and Implications for Data Standards
a report by
Sharon Marmaras
Director, Clinical Trial Support and Life Sciences Data Hub (LSH) Business Project Leader, Boehringer Ingelheim
The integration, analysis and reporting of data collected in clinical easily access the data they required. In addition, the automation of some
trials is one of the critical paths in drug development. Challenges of the process would be an extra benefit.
arrive from integrating data from multiple sources. Furthermore, the
analysis of this data must be controlled and must deliver reproducible To achieve this, BI wanted to establish a unified environment for
results. Data from clinical trials are captured, cleaned, processed, clinical data building, analysis and reporting at a global, cross-trial and
reported and archived using appropriate computerised tools. These individual trial level – effectively, a central global repository for data,
tools support various aspects of the research process, such as trial which would integrate data from multiple sources and help to move
design, data entry, source document tracking, data cleaning, analysis BI’s processes in the direction in which the industry is headed in terms
and reporting. of eClinical development.
Problems arise when multiple data sources need to be shared. Often, BI put out a set of user requirements that the company felt were
problems linked to sharing data between different systems are essential from the system and put up several requests for proposals to
addressed by several one-to-one links between the various systems. This different groups. BI’s main considerations were for a central global
results in considerable maintenance effort and version dependencies. repository that allowed users to be able to work in a secure location. BI
Typically, data transformation and reporting programs are developed by also wanted to integrate SAS (the skill-set of the organisation) to
copying files from a file system, which are then edited, possibly trial by promote the use of standards and reusability by use of library functions.
trial. Managing version control and the testing cycle is a major challenge
in such an environment. Accordingly, integrating data across the BI also wanted its medical personnel to be able to review data easily with
systems is time-consuming. Because data are dispersed, end-users a user-friendly interface – the current data management system is often
cannot access crucial information without the involvement of a considered too complicated for non-technical personnel. The new system
technically skilled person. Such delays during regulatory submission also needed to support regulatory requirements such as US Food and
preparation are very costly. Drug Administration (FDA) 21 CFR part 11 compliance. Other key
requirements identified by BI included:
Boehringer Ingelheim (BI) currently uses Oracle Clinical (OC) as its
clinical data management system (CDMS) to perform clinical data • enabling a configurable workflow, e.g. to support users in the
management functions. Data in OC are depicted in a relational validation of programs;
structure, which does not implicitly support reporting, but does allow • facilitating the import/export of data, together with associated
flexible study set-up and data entry. However, in OC it is difficult to metadata and programs, to send to regulatory authorities and
handle cross-trial data and to perform the sophisticated data contract research organisations (CROs); and
transformations required for statistical analysis. Reporting and analysis • providing archiving functions.
are currently supported by an SAS-based system developed in-house,
CARE (Clinical data Analysis and Reporting Environment). CARE was BI had originally contracted a company to build a customised system.
developed to reduce global programming effort within BI, by providing However, Oracle approached BI with a view to building a tier 1 product
a standard environment for generating the tables, listings and figures in collaboration with a partner from the pharmaceutical industry. A
for clinical trial reports. CARE promotes a harmonised SAS environment consortium was formed – the current partners being BI, Oracle and
across the corporation, through standardised OC database structures, IBM – and development began on the Life Sciences Data Hub (LSH) in
SAS data libraries, a library of standard SAS programs and the use of an 2000. The LSH is not intended to specify the standard reporting
SAS-based report appendices generator (RAGe) to apply headers, programs, but to provide an interface for creating, maintaining and
footers and a table of contents to the SAS output. Using CARE, data running these programs. BI will develop standard programs in parallel.
managers and statisticians create their own tables, listings and figures
for the reporting and analysis of clinical data. The analysed data are Data Loading
then collated and published using Documentum®. To allow clinical data reporting in the LSH, the data needs to be made
available from the external source systems by loading them into the LSH.
The Life Sciences Data Hub Loading data is handled through predefined adapters. Adapters for OC,
BI needed a solution where data could be loaded by end-users from a Oracle tables, SAS datasets and ASCII files are shipped with LSH. Data
variety of sources (e.g. OC, SAS), where programs and reports could be loading may be accomplished by physically copying data into LSH or by
built from a library of re-usable standards, and all authorised users could just defining views onto the source system. There are two components to
12 © TOUCH BRIEFINGS 2007
Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26