Clinical Trial Feasibility Shouldn’t Wait Until the Protocol is Done

Here’s a scenario I have watched play out more times than I can count. A protocol team writes a protocol. They tell leadership they are a week from final. Then feasibility happens, and suddenly there are questions that need to be answered, assumptions that need to be revisited, and cross-functional tension that nobody wanted. The team that thought it was done is not done and someone has to tell the CEO that the timelines have shifted. 

This is what happens when feasibility is treated as a checkpoint rather than a process. And it’s one of the most avoidable sources of delay and frustration in clinical trial design today. 

AI-enabled tools are changing this, and not in a vague, theoretical way. What we walked through in a recent webinar showed specifically how feasibility can move from a retrospective validation step to an ongoing, iterative part of protocol development. The difference in outcomes is significant. 

Why Late Feasibility Creates Problems

When feasibility is conducted after the protocol is written or near final, the team is essentially being asked to validate decisions that have already been made. Assumptions about cohort sizes, outcome rates, and eligibility criteria are locked in. If those assumptions turn out to be wrong, you have a problem. 

What does that look like in practice? Timeline delays. Protocol amendments, which are costly and time-consuming. Sites telling you after outreach that the study criteria are too restrictive. A company leadership team that thinks the trial is ready when it’s not. 

The tension this creates across a study team is real. And it’s largely preventable. 

The Shift: Feasibility as an Ongoing Process

The core idea is straightforward even if the execution requires the right tools. Instead of asking, “Can we recruit this study?” after the protocol is written, you ask that question continuously throughout the design process. Feasibility becomes part of the protocol development itself, not something that happens to a finished document.  

When feasibility is woven into design from the start, teams can evaluate assumptions against real-world data (RWD) as criteria are being set, not after they are locked in. That continuous feedback loop is what allows study teams to make informed decisions while they still have the flexibility to act on them. Waiting until a protocol is final means leaving time and certainty on the table. 

What AI Enables That Wasn’t Possible Before 

Moving feasibility upstream has always been conceptually appealing. The barrier was practical. Running cohort queries required deep knowledge of medical coding systems. ICD-10 codes, LOINC lab terms, structured electronic health record (EHR) data, and the nuances between them aren’t accessible to everyone on a study team. Feasibility work got bottlenecked because only a few people had that specialized knowledge to use the data effectively. 

Natural language interfaces change this. With TriNetX, users can enter eligibility criteria in plain language and get back editable and reviewable clinical code sets that can quickly be used to generate patient counts. You don’t need to know the specific ICD-10 codes for RSV to run a query on pediatric RSV patients. You type what you’re looking for, and the system translates it into well-defined and reviewable code sets. 

During the webinar, we walked through a live demonstration of exactly this. A pediatric RSV cohort was built in plain language in minutes. Then, in a type 2 diabetes query, adjusting a single BMI exclusion criterion changed the cohort size by 27,000 patients instantly, with no analyst required. That kind of real-time, visual feedback on the impact of protocol decisions is exactly what teams need during design, not after. 

This is what democratization of study design looks like. It’s not just about giving more people access to a platform. It’s about making data-driven design decisions accessible to everyone involved in building a protocol, regardless of their technical background. 

A Better Way to Think About the Process

Feasibility isn’t a black-and-white science. There are always variables you cannot fully anticipate, whether that’s competitive landscape shifts, regulatory changes, or late-breaking safety data that reshapes a trial’s direction. 

But moving feasibility upstream gives you more pieces of the puzzle before you have to make consequential decisions. It doesn’t eliminate uncertainty. It gives you better information to navigate it. 

That’s the case for AI-enabled feasibility tools, and it’s why this conversation feels so timely. The technology exists to do this now. Teams that continue treating feasibility as the last step in protocol development are doing so by habit, not because it’s the better approach. 

Watch the Full Webinar 

There’s a lot more to unpack here, including the live platform demo, a deeper discussion of how to handle eligibility criteria that aren’t well-captured in structured EHR data, and audience Q&A. I encourage you to tune in to the full webinar recording to see the tools and workflows in action. 

Or if you would like to see the TriNetX LIVE™ platform in action for your own research needs, request a demo. 

About Jeffrey Brown, PhD 

With more than 25 years of experience in research and consulting, Jeff is an internationally recognized expert in the use of RWD to support the evidentiary needs of regulatory agencies and medical product sponsors and an expert in the assessment of data quality of RWD resources.