By Deborah Borfitz 

February 25, 2025 | A panel of experts from inside several major pharma companies discussed regulatory initiatives around pragmatic trials in depth at the recent Summit for Clinical Trial Ops Executives (SCOPE). “I got a fever and the only prescription is more pragmatic trials,” quipped Henry Wei, M.D., head of development innovation at Regeneron and a major fan of TransCelerate’s recent “tabletop exercise” with the U.S. Food and Drug Administration (FDA) as well as the agency’s CDER Center for Clinical Trial Innovation (C3TI) demonstration project focused on the topic. 

Wei was joined by panelists Gracy Crane, international regulatory policy lead for real-world data at Roche Products Limited; Melodi J. McNeil, director of regulatory policy and intelligence at AbbVie; Stephanie Derbyshire, clinical operations portfolio leader at Roche; and Kevin Bugin, associate vice president of global regulatory policy and intelligence at Amgen and former director of C3TI, launched last year by the Center for Drug Evaluation and Research (CDER) to support innovative approaches to clinical trials. 

First up was Crane, making the point that pragmatic trials are intended to be a bridge between clinical research and clinical practice. Randomized controlled trials have been the backbone of clinical research programs for a long time and are good at assessing the effectiveness of a therapy in a tightly controlled environment, she says. The question remains whether the therapy then works in the clinic in less homogenous groups of patients. 

Pragmatic trials are an attempt to make the result of a trial applicable to a broad real-world population, Crane continues. “Pragmatism itself has many different domains”—nine, by the FDA’s count, each alluding to one aspect of a trial that could be changed such as the eligibility criteria or trial setting. “You can have all of [the pragmatic elements] or you can have some of them, so you pick, and you choose.” 

The reason regulators are interested in pragmatic trials is due to the possibility that they can answer regulatory questions and thereby enable drug development, she says. “We want to keep it simple... to be able to integrate research into clinical practice by enrolling a diverse population or... utilizing the data that is already existing in the environment and also to better understand how these products work.” 

The “big but” is that pragmatic trials need to provide substantial evidence for a labeling claim, adds Crane. The sponsor’s perspective is that they’re a great idea with patient-centric characteristics. 

C3TI Activities 

Bugin was filling in for FDA speakers who couldn’t join the conversation due to the “communication pause” put in place by the Trump administration. Pragmatic approaches to clinical trials aren’t new, he says, and have been championed by various centers of the FDA. 

He points to several key programs where the FDA partnered with sponsors on a regulatory-grade trial to generate substantial evidence of effectiveness that the agency could rely upon to make changes to labeling or support new indications. On the list were the Complex Innovative Trial Designs, Model-Informed Drug Development, Real-World Evidence, and Rare Disease Endpoint Advancement pilot programs. 

The FDA understands that pragmatic trials are operationally challenging, but they are not something the agency has had much experience with, says Bugin. The intent was to disseminate learnings from the pilot programs across the entire ecosystem. 

Knowing it was time to evolve, the FDA wanted to help pull together different innovative approaches to incorporating pragmatic design elements into clinical trials to not only share the lessons learned but perhaps expand opportunities for innovation in other areas in lieu of waiting for the enabling legislation from Congress, Bugin says. C3TI provided the means to immediately launch new pilots and programs when the time was right and provide sponsors with needed support for implementing pragmatic approaches and designs. 

To that end, C3TI launched four key activities last year, says Bugin, including the C3TI Demonstration Program where sponsors collaborate “in a real-time fashion” with CDER review and policy division staff. C3TI also launched a searchable knowledge repository, known as the C3TI Compass, which is endeavoring to pull together resources—guidances, case studies, toolkits, and information about demonstrations and pilot programs—from the FDA and its partners. 

Other activities of C3TI seek to address “the confusion that might exist with all the various clinical trial innovation activities,” by giving sponsors a single point of contact for direction, says Bugin. C3TI also intends to host road shows and other educational events for both internal and external audiences. 

Demonstration Program

With the C3TI Demonstration Program, “the idea is to support [pragmatic] trials through implementation so they can... help spur additional adoption and implementation,” he continues. It also creates “precedence,” meaning the agency was able to make a regulatory decision based off those examples. 

The Bayesian Supplementary Analysis demonstration project is focused on the statistical approach used in a phase 3 efficacy, safety, or non-inferiority trial that is not using an adaptive design, says Bugin. “The proposal might be to use Bayesian analysis to supplement the primary analysis,” such as for subgroup analysis. “This would sort of de-risk the use of Bayesian statistics and build that sort of skillset and toolkit for folks using that approach.” 

A second demonstration, known as Selective Safety Data Collection, relates to the collection of safety data within clinical trials, Bugin says. This offers the potential to “streamline data collection in late stage [pre- or post-marketing] trials,” particularly those where a lot of information is already in the safety database. 

In clinical trials involving products that have already been labeled, the idea is to limit the burden on sites in having to collect information on every single adverse event, including those that are “known and already considered to be related to the study drug,” he says. This would also serve to limit the burden on sponsors and CROs to report those events and overall “decrease the noise within the safety database.” 

The Streamlined Trials Embedded in Clinical Practice (STEP) demonstration project aims more broadly to “promote the adoption of any pragmatic design element within a clinical trial that might improve our ability to integrate a randomized controlled trial into routine clinical practice to help sponsors scale adoption of these innovations,” says Bugin. In general, trials later in pre-market development would be most ideally suited to the pragmatic approach, since drugs by then have a well-understood safety profile. That would likewise be the case for a large but simple post-approval trial where sponsors are looking to expand a drug label to new patient populations, or to add new trial procedures or endpoints.   

Each of the demonstrations enables sponsors to submit a proposal which, if accepted, leads to a meeting with FDA policy and review division experts within 45 days to talk through pragmatic design elements and determine the best way to implement them, he says.  

Tabletop Exercise

McNeil, in addition to being a regulatory policy expert at AbbVie, serves as the regulatory lead for TransCelerate’s embedded pragmatic trials workstream. She is particularly excited about a “tabletop exercise” with FDA representatives last December.  

Embedded pragmatic trials were already an “increasing priority” at the FDA when she and her TransCelerate peers began having conversations with FDA in February 2024, culminating with the decision to have the tabletop exercise on the topic. Much of last year was spent collaboratively planning for the discussion-based activity. 

The focus was on the sponsor oversight piece, says McNeil, and the objectives were to identify and educate the FDA about some of the barriers sponsors face when doing embedded pragmatic trials at scale. Pragmatic studies have been done—the Pragmatica-Lung (cancer) and Salford Lung (asthma) studies, for example—but not for regulatory purposes, she notes. 

Detailed public sharing of the learnings and outcomes from the tabletop exercise are planned, says McNeil. TransCelrate also identified some pragmatic trial elements that its member companies could pilot. 

Operational Hurdles 

Derbyshire homed in on the many operational considerations when implementing pragmatic elements into trials. “We really just need to think differently,” she says, including about ways to manage some of the unintended consequences that come from doing that. 

Among her examples was finding a broad and diverse patient population that is reflected through the sites selected to run the trial. Explanatory trials generally default to known, experienced sites, but “patients are going to many different healthcare providers for the diseases that we’re looking at.” Patients whose disease is less severe may be going to their family practitioner, while those whose disease is more severe may be seeing specialists at an academic institution. 

To have the broad range of sites representing differences in disease severity and treatment means “realistically we will have to approach research-naïve sites,” says Derbyshire. Those sites are going to need training and likely different levels of it as well as sponsor-supported initiatives to reduce patient and site burden, many of which (e.g., transportation services) are already in place. “But not all countries and not all sites are comfortable using these services.” 

Perhaps local family physicians could be used to support pragmatic studies, by doing things such as blood pressure assessments and general physicals and reporting the results back to the investigator site, Derbyshire continues. Devices and support for completing patient-reported outcomes (PROs) might also help limit needed patient trips to sites just to complete their PROs.  

There may even be a role for mobile nursing, although there has been “spotty uptake depending on the country,” she says. Efforts may be needed to help improve the comfort level of patients and sites if the practice is to be utilized. 

Thinking differently creates complexities, such as how to understand the patient's journey and the time and resources needed to do that, continues Derbyshire. “Do we need to start studies from an operational standpoint even earlier to do these feasibilities and where is this information available and how can we use that data to move our studies forward?” she rhetorically asks. 

Challenges would likewise emerge around data exchange between sites if local family physicians were brought into the mix, Derbyshire says. Physicians would also need to get paid for the work done, lest patients be asked to pay out of pocket.   

“We don’t want to increase the complexity for the investigators that we have contracted to the study, so how do we manage that, and then how is training managed and organized for different levels of expertise?” Some sites will require a lot of support, she adds, so one consideration is whether resources are available for that as well as different levels of training. 

Beyond operations, another key to success of embedding pragmatic elements into studies is “a simpler protocol...  and minimizing the number of procedures to make it more realistic or closer to what standard practice is,” says Derbyshire. By using shorter consent forms and either minimizing the amount of data collected or having information from electronic health records automatically populate electronic case report forms, potentially sites with less research experience could more easily incorporate trials into their daily routine. 

A simplified protocol would also spare patients what can sometimes be a day-long ordeal at a study site to get all their procedures done, Derbyshire later adds. “It’s wonderful that they’re willing to do it, but it’s also kind of not fair.” 

Value Prop

The safety net in American healthcare is breaking, says Wei, and “the last thing we need is something in the name of quality having a backseat driver.” He was referring to sponsors’ propensity to send many different monitors to study sites and “harassing us to get a bunch of paper sorted out... to prove to somebody else that you’re doing your job to prove to somebody else that they’re doing their job.” 

It is said that cost, flexibility, and speed come after quality, he says, but “they’re not mutually exclusive.” Kaiser Permanente, for example, is using ambient artificial intelligence scribes to help doctors reduce documentation burden and “bring back the joy of practicing medicine.” Use of such technologies will generate more data and precision of information from patient encounters than was once thought possible. 

Wei also believes that translational science and research needs a bridge to everyday primary care doctors like himself. General practitioners are under tremendous pressure to deliver on increasingly thin profit margins and, in some cases, are operating at a loss. These “legions of talent” could be used to help find new cures for the benefit of all given modern infrastructure, courage, and leadership, he says. 

“The value proposition is significant when it comes to the prospect of doing trials that are more simple, more streamlined, more accessible to more people, producing more results that are more broadly generalizable,” agrees McNeil. But it would be helpful to elevate the conversation to the global level, since most clinical trials are run by companies at that scale. 

“We don’t have a common lexicon... so, getting all of us speaking the same language and using the same terms will help move things along, McNeil says. “We need to be very deliberate and intentional about who we bring into these conversations, making sure that we are reaching all stakeholders,” including most notably institutional review boards. 

Learnings from all the pilots and demonstration projects also need to have an audience with FDA’s review division staff, she adds, where trials with pragmatic elements tend to get a more tepid reception compared to the reception of agency leadership. “It is just going to take time, like everything... but we need to be very mindful about accelerating that.” 

Wei says that if he had a magic wand, he would be “very disciplined about removing resourcing” to force people to do more with less and focus on the job at hand. He points to the best-in-breed emergency physicians in New York who were trained to function in the absence of seemingly crucial tools, including endotracheal tubes and urinary catheters—as well as the creative rise of DeepSeek after being deprived of GPUs, “despite OpenAI having millions of dollars of market cap at their disposal.” 

Change Agents 

The reason study sites, and potential investigators, find themselves strapped for time has a lot to do with what is happening in the overall healthcare system, points out Bugin. They find it burdensome and duplicative to participate in a clinical trial despite simplified processes and better technologies because of payer requirements. 

For oncology patients to opt for a clinical trial as their treatment option, which might give them the best possible care, means that they first must fail everything else that’s available, Bugin says. Yet patients are asking to participate in specific studies based on research they’re done on ClinicalTrials.gov. “So, I think there are certainly opportunities to simplify our clinical trials, but then there are opportunities to simplify the healthcare system that we’re trying to run these trials in.” 

A sustainable model for supporting pragmatic trials is going to require some changes at the policy level, says Wei. As has been shown in the Medicaid and Medicare segments of the country’s three-tiered insurance system, 30% to 40% per-episode savings is possible by treating patients with solid tumors in rather than outside a clinical trial due to the cost of drugs, imaging tests, and treatment visits. One central problem is that the U.S. healthcare insurance system generally reimburses for one visit reason at a time. 

The National Institutes of Health, for its part, has been looking to identify opportunities to run large-scale pragmatic clinical trials through community-based practices, says Bugin. Last year, for instance, the NIH Pragmatic Trials Collaboratory launched an initiative designed to teach researchers how to select the most appropriate study design for a pragmatic clinical trial.