By Alan Huang, Senior Database Programmer, Firma Clinical Research
When the Tufts Center for the Study of Drug Development released their 2017 eClinical Landscape Study and “revealed” that database build delays can create downstream delays of up to a month, few were surprised. The complexity and scope of clinical trials has been steadily rising for years and innovative technological advances in data collection have multiplied the number of potential data streams that need to be taken into consideration.
The time teams have to craft databases which can capture and analyze the now voluminous amounts of data, however, remain the same. Sponsors and CROs need data management teams that can handle more data of a higher complexity from more sources and still make timeline.
So what are the keys to achieving a quality clinical trial data build in such an environment? And what opportunities arise from optimizing the database design process with standards and systems that can anticipate and mitigate potential delays? Here are the four steps exceptional data management and biostatistics experts rely on.
#1 Plan for Protocol Amendments
Protocol amendments pose one of the greatest obstacles to achieving rapid database build. If data has already been collected, protocol amendments can mean sites have to re-enter that information. At the very least a reassessment of how data is being collected and how it should be collected moving forward are required.
Common protocol amendments include:
- Modifications made to the eligibility criteria and demographics of the patient population
- Availability of new safety information
- Adjustments in the number/types of safety assessment procedures
- Edits and revisions made to general protocol information
- Revisions to the safety endpoints based on FDA’s comments
A database that has been designed with protocol amendments in mind, allows sites to enter the most critical data quickly and efficiently without spending too much time. While amendments will likely always be a challenge study teams will have to face, a database build that is focused on both efficient and accurate input and output will make the process smoother.
Additionally, look for a data team that will, should a protocol amendment be required, provide a complete migration plan that includes what needs to be changed, how the migration will be produced, and a reasonable timeline. Even if the EDC can’t be changed, your data team should be willing to provide alternative solutions for gathering the data you need.
#2 Involve Biostatistics Team Early
Involving the biostatistics team early in the clinical trial strategy process ensures data collection is being designed in lock step with the study design in the most efficient way possible. As mentioned above, a quality clinical trial database build isn’t just about input, it has to consider output as well and this is where the biostats team is critical.
Early involvement ensures the right data for both primary and secondary endpoints are being captured and allows the biostats team to weigh in on things like:
- Study design considerations, including whether an adaptive trial design might be suitable
- Potential interim testing and analysis strategies
- Data Safety Monitoring Board (DSMB) discussions
The biostatistics team can also help defend your planned analysis of study data at regulatory meetings to further reduce the likelihood of protocol amendments.
#3 Create Timeline to Release Database Prior to FPFV
Building and releasing a database quickly is important to maintain the larger project timeline. On average, clinical trial teams take 68 days to build and release a typical study database, risking downstream delays reaching regulatory submission and approval.
Create a timeline with your data services team to ensure database release occurs prior to first patient, first visit (FPFV). This includes:
- Production of all EDC screens
- Finalization of all validated screens
- Production of edit check document
- Finalization of all validated edit checks
- Finalization of user acceptance testing (UAT)
- Completion of all data-processing requirements
Ideally, your data team should plan for system release 1 to 2 weeks prior to FPFV and provide system training for both the site and your internal team to ensure everyone is familiar with the EDC system.
#4 Ensure EDC Can Capture All Data Streams
In a recent survey, more than three-quarters of industry professionals noted they have issues loading data into their EDC application, and most (66%) claimed that EDC system or integration issues are the primary reasons they are unable to load study data.
Today, researchers collect data from many different sources, including wearable technology, handheld devices, and social media platforms. Loading data from multiple applications can be challenging to complete in a timely manner, leading to greater delays and costs.
A quality clinical trial database build performed by an experienced data services team plans for all data streams and includes tactics such as intelligent guided entry, skip logic and well-designed pop-ups to make data entry as simple as possible for each entry point.
How do you ensure your data team has the appropriate experience? Ask the following critical questions:
- What EDC platform should your study implement? What advantages does that platform provide compared to others?
- Will all desired data be captured and easily integrated with the EDC platform?
- Can the team ensure a CDISC-certified SAS database for regulatory submission? What is the strategy for implementing the requirements?
- Will the database be US 21 CFR Part 11 and EU Annex 11 compliant upon completion?