The evolving challenges and new opportunities from the latest wave of change in clinical data management practice.
Clinical data management underwent a seismic change when the industry evolved from trials conducted with paper case re-port forms (CRFs) to using electronic data capture (EDC) systems to collect clinical trial data. The impact of this change was rapidly apparent across the industry-it changed clinical data management entirely. A new culture of real-time data, risk as-sessment and mitigation customized clinical monitoring strategies emerged and with it the expectation of faster timelines for clinical trial data delivery. The clinical data management landscape is changing again, albeit less obviously than the move to EDC, but it is still changing.
The level of clinical trial protocol complexity is increasing year on year. In efforts to maximize the value of trials and reduce their spiralling cost, sponsors are turning to adaptive trials. An adaptive design clinical trial can help to create a stronger value proposition for the sponsor’s product by reducing development costs and accelerating time-to-market. However, implementation of these complex protocols within current EDC systems can be challenging and introduce several data management difficulties into the process.
Some EDC systems have complex functionality that enable comprehensive data-collection forms, complex logic-check programming, and dynamic functionality for smart form/visit structures to meet today’s complex protocol design demands However, several EDC systems do not have all these features supported. Features such as tracking and reporting overall development progress, approval of CRFs, tracking edit-check programming and validation processes, management of review-reported issues, management of user acceptance testing, and approval to go-live are not always available. Data management across several tools and systems can contribute to an exponentially increased effort to reach EDC go-live dates. The ideal scenario would be to have one central EDC system where the entire process is managed, with one single source of truth and status.
The intent of data standards, CRF standards, and edit-check standards is to reduce the effort involved in getting an EDC system live to collect the trial protocol data and ultimately deliver the data and data analysis. Implementation of standards also has the advantage of re-use of downstream programming code for data
listings; study data tabulation model (SDTM) and analysis data model (ADaM) datasets; tables, figures, and listings (TFLs), etc. Data standards are widely available, strongly governed, and well-intentioned. However, the benefit of standardization is rarely fully realized. Deviation from data standards is still accommodated and in many cases is legitimate, but in 2020 should we still be debating the content of a vital signs form? No.
Apart from the EDC set-up challenges due to protocol complexity, there is a growing challenge from the increasing number of “external” data sources. These sources, generated in systems other than the EDC database, can include IRT/IxRS data, laboratory data, speciality lab data, ECG data, actigraphy data, and electronic clinical outcome assessment (eCOA) data. The task of managing these external sources throughout a trial is growing significantly. Managing the volume of data and associated transfers requires investment in system architecture. Process changes may be required and investment in training is vital. Blind faith in technology cannot be accepted due to the critical nature of the data collected. Systems and processes development will ensure that data quality and integrity is maintained.
Automation has long remained in software development testing or on the factory floor but it is now emerging as an option in the clinical trial industry. As there exists so much scope for automation, perhaps there is an opportunity to challenge the overall clinical trial process and evaluate if those steps that are rule-based-and thus can be easily automated-are actually adding value. Is time best spent on other tasks?
Rhona O’Donnell, senior director, Data Management, ICON plc
Related Content:A Closing Thought