Automation consistently comes up to help tackle today’s complexity in clinical trials, but there are different ideas of what it is and where it is useful. We asked several leaders in trial operations, digital and data innovation to give us their views on automation.
How do you describe automation in clinical trials?
Taras Carpiac, Executive Director, Innovation - Global Development Operations, Amgen / Sheryl Jacobs, VP, Global Development Operations, Amgen: Automation is the use of technology to perform (often repetitive and/or standardizable) tasks that we would previously use humans to do.
William Illis, Global Head, Collaboration and Technology Strategy, Novartis: There are currently many manual steps in clinical operations where information is transmitted via largely unstructured documents which require manual intervention to read, interpret and process, and each of these is a potential automation target. When implemented, automation can successfully speed up cycle times, improve quality and reduce re-work leading to overall increased productivity and throughput.
Stephen Lutsch, Senior Director Clinical Trial Digital Innovation, Genmab: Automation means leveraging Robotics Process Automation, standard languages like Python, cloud services and AI like chatGPT and CoPilot to automate processes or eliminate processes across the clinical development lifecycle.
We could even look at automating the rules, enforcement and decision-making so our people can work the difficult exceptions and have the capacity to innovate further, and focus on sites, patients and new research.
Deirdre BeVard, SVP, R&D Strategic Operations, CSL: Automation is applying technology to repeatable, labor-intensive, error-prone tasks to reduce the demand on our talent and accelerate our processes. I put it in the context of doing more with the same. It can free up resources to spend more time on higher value activities and the work that brings them more joy and learning.
Kimberly Barnholt, PhD, Executive Director, Evidence Generation Leader, Genentech: The growing volume and complexity of data and data sources have become too much to manually ingest, track, curate and visualize in any reasonable time frame or resource investment.
Automation minimizes time wasted on repetitive tasks, standard processes and rework from manual errors, allowing us to “do more with less”.
"When implemented, automation can successfully speed up cycle times, improve quality and reduce re-work leading to overall increased productivity and throughput."
Where is automation poised to make a difference in clinical trial operations?
Taras Carpiac / Sheryl Jacobs: One is in clinical data management and statistical programming, in terms of discrepancy management, data management, production of statistical outputs, etc.
Another is helping in structured content management-related areas, such as using a digitized protocol to automate the production of associated study-related collateral.
Finally, AI can help with routine day-to-day automation for clinical operations staff, e.g., in producing meeting minutes and summarizing information.
Amy Cramer, MMCi, BSN, CPHQ, Founder and Leader, Vulcan FHIR Accelerator: Data transfer; feasibility and pre-screening would benefit from technology assistance; and appointment/visit reminders and other prompts to ensure protocol compliance.
William Illis: Site selection and trial feasibility. There is a fast-growing increase in good quality and consumable data that can be leveraged to effectively automate parts of these processes. The adoption of protocol digitalization is one of the most promising developments which can make a big difference to enable automation. Creating a machine-readable, executable, digital protocol in a standard format can drive automation of many downstream clinical operations processes.
However, attention needs to be paid, as automating poorly designed processes can simply lead to execution of ill-conceived processes faster, and in these cases there may be more to be gained by considering how a process could be done differently versus blindly automating it.
Stephen Lutsch: Process automation: digitalization of manual processes in clinical development, automation in near-real time of oversight and document review, and streamlining data integrations. This way, sponsors can meet the sites where they are instead of bending them to our will.
Deirdre BeVard: Data entry, management and reconciliation areas. Ideally, it could improve the start-up activities across multiple areas, such as document completion, downstream system development and more.
Kimberly Barnholt, PhD: Automation will help accelerate routine activities, improve data flows, optimize our ability to meet the compliant use of our data, and provide more real-time access to data for decision making. Ideally, this will translate into faster trials with better outcomes for our patients.
Victoria Chiou, MD, Head, Clinical Excellence and Innovation, AstraZeneca: With increasing advances in emerging technology, automation provides an opportunity for leveraging data simplification and efficiency to support workstream delivery.
"Attention needs to be paid, as automating poorly designed processes can simply lead to execution of ill-conceived processes faster, and in these cases there may be more to be gained by considering how a process could be done differently versus blindly automating it."
Do you think we're in a place to take real advantage of automation? And if not, what do we need to do to get to that place?
Taras Carpiac / Sheryl Jacobs: Automation can generate efficiencies that let us focus staff attention on higher value-add activities. The technology is there and increasingly so are the frameworks that allow us to take advantage of automation in a consistent, standards-based way. One of the areas where Amgen is already seeing real success is in the Regulatory Operations space, and we increasingly are seeing the same in Clinical Operations
To take even better advantage of what automation has to offer, sponsors must make the upfront investment in data standards and underlying technical infrastructure (like rules engines, workflow tools, etc.) that enable more and more automation use cases
Amy Cramer, MMCi, BSN, CPHQ: The technology is there, so change management is the focus.
William Illis: Many partial point solutions which offer some incremental benefits are possible. However, larger gains and lasting, sustainable improvements require collaboration across multiple stakeholders in clinical research, namely, pharma study sponsors, CROs, sites and site management organizations, technology providers, regulators and standards development organizations.
In advancing clinical research, there is a constant tension between the science of medicine, which often involves doing something for the very first time, and the science of operations, which involves optimizing repeatable activities. In clinical operations we have to continue seeking the efficiency gains through automation without constraining the unique research leading to therapeutic advances.
Deirdre BeVard: I do think so, yet there is a slow adoption rate of technology in many areas of clinical development. To be better positioned, I think we need those on the forefront to share the challenges and failures in addition to the wins. We also need some mindset and cultural shifts from one of activity-based/volume value to recognizing impact and insights more.
Kimberly Barnholt, PhD: We are at a place to take advantage of automation, but not yet to where we can maximize its full potential.
We are currently automating discrete steps rather than the full end-to-end clinical trial process. This next level will require a strong data culture and supporting ecosystem to drive use of standards, digitized documents and metadata capture. I believe the gains from end-to-end automation will be exponential, driving efficiencies not only in the data flows, but also in the surrounding governance, access, safety reporting and insights generation.
"To take even better advantage of what automation has to offer, sponsors must make the upfront investment in data standards and underlying technical infrastructure."
