(Originally published May 2nd on Applied Clinical Trials)
________________________________________________________________________________________________________
ACT: How can feasibility studies be used to safeguard against trial complexities?
Nambisan: So I touched on this just briefly in my prior response. But I think the purpose of feasibility is to understand if the design that's defined in a protocol can be feasibly accomplished in a set amount of time with a set amount of budget in the real world, right? And so there's this interplay between the science defining that I want to recruit patients of this specific type in this disease with these characteristics across these countries. And I want them to be able to come into the site this many times for data collection, and maybe have some remote assessments as well. And I want the site to be able to handle all of that data and handle all those participants so that I can prove that this therapy is safe and efficacious to a particular cohort of patients.
The challenge often is that you want the trial to be very narrowly defined. So you have the best in the participant population, especially to be narrowly defined, so you have the best chance at proving an effect of that therapeutic. But the challenge with that is operationally as you filter the patient populations down, so we're not looking at diabetes type two anymore, we're looking at diabetes, type two comorbid, with chronic kidney disease patients, and metformin, so you're filtering on filtering and filtering. It's much harder to find access. And, frankly, enroll those participants in a trial. Not to mention, if you're actually doing a much more complex study, with many more assessments tha]en risk of discontinuation, the risk of noncompliance increases at the site as well. And so I think that there are opportunities for us to look at, first of all, the inclusion/exclusion criteria, parse that and say, Is this first of all realistic to assume that we can recruit 500 or 600 patients in this particular condition with these inclusions and exclusions on the participant population? When you do the analysis on the real-world data, for example, in the US, you may find out that actually, you only have about 150 patients of that ilk here, right? So you shouldn't expect that this is operationally feasible. At the same time if there are specialized data collection that needs to be done in very specialty labs, and now you're asking participants to travel hundreds of miles, and that's specified in the protocol itself as one of the assessments that needs to be done. And you would recognize that this is a somewhat rare indication in the sense that participants are dispersed and not all centrally located that specialty lab. That may not be feasible to expect them to do this; we all have our lives, we have our work, etc. It may not be feasible to expect them to do this, and to be able to get good quality, high quality, statistically powered data. So there's a number of different aspects related to this. And I think there's always this tension in some senses, a healthy tension between the science and the real world operations that needs to figure this out. And the more data, the more analysis you can bring upstream in the process. And when I'm saying the more data, the more analysis, I also mean, what is realistic, what is actually happening in the real world, not just how many patients do I need to meet this statistical power, which is a theoretical exercise, at least at the outset.
The more you can bring upstream, the better it is for your downstream steady progress operations cost budget, and frankly, when we talk to patient centricity, it's much better for the participants as well. You're asking these number of participants, volunteers to take a therapy that is neither proven to be fully safe nor efficacious yet. And if you're going to ask them to do this, shouldn't we do the lion's work of the planning and the analysis and make sure that we streamline the operations to move this therapy into the patient's hands for approval as fast as possible? So I think it's not just about making sure that we can run things more efficiently on behalf of the pharmaceutical developers, it's also making sure we can get these therapies efficiently into the hands of the patients in need.