How do you define the importance of data quality in clinical trials? An accurate reflection of a subject’s experience during a trial? Uniformity and completeness of the data collected? Frequency of deviations? All of the above. The need to define quality standards and implement structured data collection methods is imperative in order to reduce variability in the analysis and improve outcomes of clinical trials.
The initiatives underway in accordance with the ICH E6 R2 regulation and setting Quality Tolerance limits now play an important role in harmonizing and establishing quality standards across the industry. The quality of the data collected is directly reliant upon the data collection tool. If the correct datapoints are not collected, or the design of the collection instrument is poor and causes inconsistencies in the data collected, a meaningful analysis may not be possible. It follows, therefore, that the design, development and quality assurance of such an instrument are of utmost importance. Guidelines for imposing standard data formats and uniformity of data collection from the outset are defined in the Clinical Data Acquisition Standards Harmonization (CDASH) Model. The CDASH Model provides naming conventions for the CDASH Implementation Guide (CDASHIG) variables along with additional metadata to help facilitate mapping of collected data to their respective SDTM. It defines how questions should be formulated for data collection within the eCRF, making use of standard CDISC controlled terminology. Implementing the CDASH standards during the eCRF and study database setup facilitates the development of structured data collection routines. Whereas site staff have previously had to complete eCRFs which use a variety of different formats, replacing non-standard CRFs with the CDASH Case Report Forms vastly improves the situation by promoting the benefits of familiarity and training in the use of standard eCRFs. The speed at which the data is entered and the quality of data is greatly improved, because the eCRFs are easier to complete and therefore generate fewer data queries.
Using CDASH standards also allows the ability to re-use programmed data validation checks across studies, promoting efficiencies and savings both at the study set up phase and during the data review and cleaning processes.
The latest release of the CDASH Model aligns with and is structured similarly to the SDTM Model. The CDASH Model organizes data into classes, which represent meaningful groupings of data in clinical research. It defines CDASH metadata for identifier variables, timing variables, general observation class variables (Events, Interventions, and Findings), domain-specific variables, and special-purpose domain variables. This alignment of CDASH with SDTM promotes faster, more efficient production and traceability of submission data.
Clearly defining and categorizing data quality together with implementing structured data collection formats are essential for improving and managing clinical trial outcomes.
