Implementation Of CDISC Standards In Drug Clinical Trials

CDISC (Clinical Data Interchange Standards Consortium) is an open, multidisciplinary, non-profit organization committed to the development of worldwide, platform-independent industry standards to enable information system interoperability for medical and biopharmaceutical product development. Over the past years, the Consortium has worked synchronously across several focus areas of clinical trial to establish a set of standard data models to faciliate data collection, analysis, interchange and submission. Among these models, SDTM (Study Data Tabulation Module) is the recommended standard for the submission of clinical and preclinical trial data to the FDA (Food and Drug Administration) in marketing applications, and also the standard for FDA’s clinical data warehouse (JANUS) to perform more advanced, robust analysis from multiple clinical trials to help improve efficacy and safety. CDISC have been successfully used to help FDA better understand industry data by providing a platform of standard data content, which minimize programming and rework of the data during FDA review and greatly facilitate the integration and reuse of data from multiple submissions for broader scientific and medical evaluation. Many companies begin to consider reengineering their internal processes to adopt CDISC standards.Under this background, this paper introduced several ways of obtaining SDTM implementation and sum up the advantages and disadvantages. The most advantageous linear method has been optimized by defining CDASH (Clinical Data Acquisition Standards Harmonization) standard into database on the front end. The new process was modeled using Microsoft Access, illustrated by the internal mapping to SDTM and achieved by extracting SDTM datasets. Taking the above CDISC approach as a reference, this paper analysed the process of getting final SDTM datasets on the basis of current PDS (Pfizer Data Standard) definition in the database.OBJECTIVEThrough the introduction of FDA data submission requirements and CDISC standards, discuss the development direction of clinical data standardization. Through the analysis and illustration of data process in both CDISC approach and PDS approach for getting SDTM datasets, explore the application prospect of CDASH optimized linear method, analyse PDS mapping process to SDTM domains to provide reference for CDISC implementation for companies which have their own well-defined standardMETHODCombined with practical experience in clinical data management, get a deep understanding with PDS and CDISC models and establish data standard application process according to linear method. Simulate clinical database using Access to define variables of four data collection modules, to establish relationships in database, and to entry and export data. Illustrated by3groups of trial data from database definition to getting final SDTM datasets, together with varible attributes, data definition files and SAS mapping process, compare PDS approach with CDISC approach with time as the main indicator.RESULTSDTM compliance datasets were obtained through illustration of CDISC apporach with CDASH optimized linear method. This process enforces standardization of database structures across studies and projects and allows the data to be housed in the format that it will be submitted. It minimizes or in some cases eliminates the need for manipulation of extracted data, since the underlying database tables are already CDISC compliant, and are extracted in the format in which they will be analyzed and reported.PDS raw datasets were obtained by illustration of PDS apporach with linear method process. PDS mapping relationship to SDTM domains were analyzed and compatible results were gotten. SDTM compliance datasets were finaly obtained based on the good mapping relationship and standarized format conversion process. Most of the SAS code can be reused in the following studies, which makes the mapping time greatly reduced and the total time comparable with CDISC approach.CONCLUSIONAn organization should carefully plan its objectives for implementation and perform an assessment of available resources. If companies have their well-established standard with good mapping relationship to SDTM (like PDS), it is feasible for them to keep their existing database definition and develop modular data mapping programs or tools based on that. While for companies struggling for standards and cooperation with other organizations, it is strongly recommended adopting linear CDISC approach with CDASH definiton in database as a considerable choice.