T-15: Mapping Items in the Case Report Form to CDISC SDTM Standards Using Pre-Map Function in the Electronic Data Capture System
Poster Presenter
Toshiki Saito
Director, Department of Regenerative Medicine
National Hospital Organization Nagoya Medical Center Japan
Objectives
To increase efficiency in the mapping of collected data to datasets in the CDISC SDTM format, we aim to setup the workflow to map items in the case report form (CRF) to SDTM standards in advance while developing the CRFs, instead of mapping the items to SDTM after the data collection.
Method
We added a pre-mapping function to our EDC system, "Ptosh". Ptosh adopts a database that can handle schemaless structured data, which allows the system to easily accept the following metadata: CDISC SDTM domains, information on variables, and controlled terminology, in the input field of a CRF.
Results
We have developed functions to map items in the CRF to SDTM standards in the EDC system, keeping the consistency of the structure defined in the SDTM specification and its implementation guidelines. Those functions were used in five clinical trials, where all the input fields of the CRFs were mapped to SDTM variables. Since the system sets mapping information at the time of creating an electronic CRF (eCRF), no mapping is required following data collection. The system is also able to export a draft of define.xml which is consistent to the SDTM-premapped data. The SDTM-premapped data can be exported anytime during the trial in the SDTM format and derived-variables can be easily generated with programs that can be shared within multiple trials. Although direct statistical comparison of the cost is difficult to perform, it was obvious that the creation of an eCRF using a CDISC-SDTM-compatible field significantly reduced the cost of data output in the SDTM format, compared to that for mapping in the same format following data collection. As an extra advantage, the setting of a field on an eCRF was rationalized because field setting during the creation of an eCRF can only be implemented when the structure of the system allows data acquisition in the CDISC SDTM format.
Conclusion
Nowadays, when applying for approval of a drug to the Food and Drug Administration (FDA), clinical trial data must be submitted in the CDISC SDTM (Clinical Data Interchange Standards Consortium, Study Data Tabulation Model) format. The standardization of clinical trial data has a number of advantages, in addition to that associated with data submission to the FDA, and it is also being actively promoted in Japan, too. However, currently, clinical trial data in many different formats are collected as they are and mapped in the SDTM format following the completion of a trial. This process requires significant effort, and it is often difficult to map data completely. As far as we know, there is no EDC (electronic data capture) system that can implement this process completely automatically. We have developed functions in our original EDC that force you to collect data in SDTM-compatible manner, and shown that those functions are usable in real clinical trials in hemato-oncology area, in which clinical data are rather complicated due to the long-treatment period with combination chemotherapy.
