In this post, we handed the virtual microphone to SDC's Adam Hamm, Ph.D. and Vice President of Biostatistics and Strategic Consulting to address a commonly asked question of our biostatistics team.
Whenever I am asked about the best time to bring a biostatistician into a study, my answer is always "It's never too soon." It's important to know that every successful clinical research endeavor is a
well-articulated collaboration between a sponsor organization, their CRO, and their biometrics partner, and has many moving parts.
The very best partnerships are established at the beginning starting with the clinical development plan, through to study implementation, study conduct, and finally, study closure.
In my experience, almost every rescue study has the same two themes and they are completely preventable:
#1 - the pitfall of not involving a seasoned and experienced statistician early, or#2 - the failure to understand how the data are being captured, queried, and ultimately analyzed in your trial
When statisticians are given the opportunity to engage right away with protocol, for instance, it can ensure the correct study questions are being asked, that they are consistent with the sponsor's objectives and that the study hypotheses are being tested appropriately. I know this seems intuitive, but this is often overlooked if the statisticians aren’t involved from the outset of your trial design.
The most successful efforts I have seen that led to regulatory approval had an early integrated and highly involved statistician.
This level of involvement can even mean asking our statisticians to program the analysis data themselves to give them the highest level of understanding of it.
Once we agree on the right study questions, ensuring data integrity and analysis starts with producing proper case report forms (CRF). Ask yourself if the CRFs capture all the information needed to answer your study questions. From there, the data manager and statistician should regularly review data, perhaps even involving statistical programming, and proactively share any concerns with the client. I have seen study database locks delayed due to a data review that started too late in the study’s life cycle. An early review involving these parties can also help identify if there could be potential issues in data analysis.
Furthermore, high-quality submissions should include team members with expertise in CDISC data formats and leverage statisticians on DMS planning, regulatory meeting representation, and support.
Ensuring data quality in any organization is never easy, however, data services partners are uniquely equipped with the tools to focus on where it matters most, that is clean high-quality data and appropriate analysis techniques. Data services should be integrated early and often, and the leads from data management and biostatistics should be key and influential members of the clinical team. If study teams recognize this, our trials should run smoothly, without delay, and with trusted results that provide beneficial treatments to the populations that need them.
If you would like to connect with Adam directly or anyone from our SDC team, please email us here.
Todd Bonta, SDC's Vice President of Clinical Data Management discusses the biggest bottlenecks or “barriers” to speed and delivery of an EDC database...
SDC's Adam Hamm, Senior VP of Biostatistics and Strategic Consulting discusses How to Guide Clinical Research Sponsors Through an Effective ISS/ISE...
SDC leaders in biostatistics and clinical data management address the biggest mistakes clinical research sponsors and CRO's make with their...