| CRF(Case Report Form) is the primary tool for data capture in clinical trials. It is considered the very crucial document in clinical trials. The quality of CRF designed directly affects the quality of data captured in clinical trials, which determines success or failure of clinical projects.CDASH(Clinical Data Acquisition Standards Harmonization), as one of standards developed and established by CDISC (Clinical Data Interchange Standards Consortium), defines the basic standard for data capture in clinical trials and is used to simplify and standardize data collection in clinical trials. CDASH is based upon SDTM (Study Data Tabulation Model) recommended as the standard for clinical data reporting and submission by FDA(Food and Drug Administration) and contains the most common used CRF domains. Under these domains, CDASH specifies variables, variable types, question prompts, code lists, CRF completion instruction, all of which offer detailed standards and guidance for CRF design.At present, big pharmaceutical companies still apply their own data standards to CRF design in clinical trials. However, especially the recommended SDTM by FDA tells us it is obvious that a uniform data-capture standard across companies will be implemented in the near future.In this context, based on CRF design practice for a clinical study, the study established three indexes and their application methods for CRF evaluation. Then the three indexes were applied to evaluate CRFs of Pharma A to explore the gap between current-used CRFs in big pharmaceutical companies and CDASH. The findings provide a meaningful reference for development of high-quality CRF in pharmaceutical industry and also for self-optimization of CDASH in the future.OBJECTIVEFrom the perspective of CRF developing toward standardization and high quality, the study explored and evaluated the Pharma A current-used CRFs based on conformance rules of CDASH, the well-recognized data-capture standard. It was expected that suggestions for CRF’s development in pharmaceutical companies and CDASH’s development would be raised from the evaluation results in the practice and exploration in the thesis. Key points and considerations in CRF design would be summarized and CDASH comformant CRF templates would be established to provide valuable guidance and reference for standardized and high-quality CRF design in clinical trials.METHODAccording to protocol, CRFs were developed for a clinical trial. With the practical experience,3 indexes and application methods for CRF evaluation regarding complete data collection, verbatim-used question prompts and mapped code lists were respectively established. Then the indexes and methods were applied to evaluate CRFs from 7 studies of Pharma A. The evaluation was done by fields and by domains. The results were further analyzed. Based on the analysis, a set of CRF templates covering gap between Pharma A CRF and CDASH as well as conformant with CDASH were developed.RESULTDesigned CRF contained 28 modules, covering most of the data capture modules required for the clinical trial. According to CDASH conformance rules,3 indexes and detailed application method regarding complete data collection, verbatim-used question prompts and mapped code lists were established. Then the 3 indexes were applied to Pharma A CRFs for evaluation. The results showed that the percentages of complete data collection (for Highly Recommended fields) and mapped code lists are 100% and 83.85%, respectively. The percentages of complete data collection across different fields conform to the corresponding CDASH requirements. However, a certain gap in verbatim-used question prompts (31.62%) does exist between Pharma A CRF and CDASH. The evaluation results by domain demonstrate that the percentage of verbatim-used prompts or mapped code lists for some domains is relatively low (<20%), which should be paid more attention in future CRF design. Analysis of the evaluation results from the view of CDASH indicates that some reliable code lists could be added in domain of adverse event, inclusion/exclusion criteria and physical examination. Based on findings, CDASH conformant CRF templates (15 safety domains) into which strengths of Pharma A CRF are considered are established.CONCLUSIONFor big pharmaceutical company represented by Pharma A in this study, indexes regarding complete data collection and mapped code lists well meet the corresponding CDASH conformance rules, while there is a gap in the index regarding verbatim-used prompts between Pharma A CRF and CDASH. The results indicate that pharmaceutical companies should take actions in response to the trend that CDASH is developing towards industry data-capture standard. Modifying and adjusting their own data standards to achieve the compatibility with CDASH could be one of effective actions. For CDASH, some potential to-be-added code lists also are shown in the study. These code lists suggest that CDASH, as a standard developed based on big pharmaceutical companies, should keep its communication with these companies and absorb essence from them to further improve itself. Pharmaceutical companies and CDASH both should make efforts for the agreed goal of standardized and high-quality CRFs in pharmaceutical industry. |
