In Respond Africa, ensuring accurate data collection has been a priority. Here Erik van Widenfelt gives overview of our Electronic Data Capture system (EDC), how it is currently working and what it provides our META III Trial.
Background of the EDC
“Clinicedc” is a publicly available open-source clinical trial data management project currently maintained by the RESPOND Africa group. As we have done for our other clinical trials, we developed a protocol-specific software layer for the META Phase III trial that operates on top of the “Clinicedc”. The result is the META3-EDC. This cloud-based online data system is deployed at all the study sites and securely serves all aspects of the trial’s data collection and management needs in real-time at the point-of-care.
The power of real-time data collection and validation
The META3-EDC, or just the EDC as the trial team refers to it, captures data in real-time at the point-of-care. Almost all case report forms are considered as “source documents”; that is, we avoid transcribing wherever possible. For example, screening information for a potential participant is captured and submitted at the point-of-care in the presence of the patient and the eligibility outcome calculated immediately by the EDC. Any issues with the data bring up prompts to be resolved on the fly. Similar data validation checks are embedded in case report forms and other data collection forms. These checks can be configured to validate dates, test results, AE grading and any value that would be best rectified by clinic staff as they enter the data.
EDC and Meta III – why is it important?
Meta III is a large trial, with over 2000 participants expected to be randomised. Participants will also be followed up for 3 years, and this will generate a lot of data that needs to be clean for analysis in a timely manner. Working together across all study sites we will be able to closely monitor in real-time each participant’s trial status, adherence to the data collection schedule, laboratory results, data completeness, ill visits, medication, adverse events, and so on. Using the EDC ensures that we clean data as an ongoing process so that by the time the study ends, we will be able to conduct analyses without weeks or months of delay due to data quality issues.
How can the EDC help us do this?
The EDC has been designed to empower each study site – the EDC provides trial sites with immediate and secure access to their own data. Sites can implement quality assurance and other data monitoring strategies by scanning through individual participant documents, searching the comprehensive audit trail, or using the EDC’s many QA features. Similar resources enable country and global level quality assurance, monitoring, and reporting. For those granted special access, raw data, less personally identifiable data, can be accessed from anywhere in the world for import into custom reporting tools and statistical packages. For example, a statistician at LSTM in Liverpool can review data that was collected just minutes ago in Tanzania.
What if we find issues in the EDC?
Any well-functioning EDC is the result of an iterative process; the team tested the EDC comprehensively before the trial began and we keep an active watch for any unforeseen circumstances or inconsistencies that may crop up as the trial proceeds. The EDC will continue to improve over the course of meta III. Staff have been trained to report issues to the EDC development team on our issue tracking system. Issues are discussed in team meetings, solutions agreed upon and if a software update is required, an updated EDC is versioned, released and deployed. Every data point recorded in the EDC is tagged with the release number to link it to the system source code used at the time of data capture. Since the system is cloud-based, all sites have access to the most recent versions of the EDC at the same time.
Weekly reports will soon be provided to all study sites!
For the immediate future, the team is working to design weekly reports for each study site. These reports will be provided an overview of recruitment rates across all study sites, both in absolute numbers but also by key factors such as recruitment milestones and participant characteristics such as gender. On top of this, we will capture and report on missed visits, participant withdrawals and reasons for withdrawals as the study progresses. This will be a valuable source of information that we can use inform the study where necessary and possible. Moreover, whilst serious adverse events have their own reporting structure, we want to closely monitor the occurrence of grade 1 and 2 events throughout the study and will be able to create up-to-date reports on grade 1 and 2 adverse events across the course of the study.