More than 60 percent of current drug submissions to the FDA include real-world data (RWD) in some form, and agency officials are saying the FDA needs to expedite honing its skill set in regulating the use of it in trials.

Amy Abernethy, principal deputy commissioner for food and drugs at the FDA, said the agency needs to strengthen those skills as soon as possible. “RWD is becoming progressively more common and important,” she said.

Abernethy said the FDA is learning how to analyze real-world datasets better and how to incorporate them as part of an overall evidence package. But she said the agency still needs to develop a comfort and familiarity with working with RWD to create an inclusive approach that weighs the totality of the evidence.

“You may think that we’ve known the scientific methods for observational research for a very long time, but in fact these datasets are different,” Abernethy told attendees of the Pharma Clinical 2021 Conference last week. “We need to develop definitions for variables within the datasets that are common and well-understood across different datasets. We need to develop ways of working with datasets when the information is rapidly accumulating and also when the day-to-day delivery of care is adjusting across time. And we really need to be able to rapidly convey the methods as they are being developed so that they can be cross-checked by other researchers and also shared for more general-purpose use.”

In developing a framework, Abernethy said the FDA is weighing whether RWD are fit for use, whether the clinical trial or study design used to generate real-world evidence can provide adequate scientific evidence to answer questions by regulators, and whether the data and analytic combination can satisfy specific regulatory requirements. She said the agency was exploring the use of RWD for a variety of situations, including as:

• A supplement to an overarching package to describe an underlying population;
• An information set for secondary indications or label expansions; and
• A potential substitute for a control arm in a clinical trial.

The pandemic helped the agency crystalize the importance of being able to learn from many different types of datasets, Abernethy said, adding that the FDA’s partnership with the Reagan Udall Foundation in launching the COVID-19 Evidence Accelerator helped to explore how RWD could be put to use. Regulators analyzed electronic health record systems as well as in silico and synthetic datasets.

In other presentations by FDA officials last week, Robyn Bent, director of the Patient-Focused Drug Development Program at the agency’s Center for Drug Evaluation and Research, said the program is moving forward on publishing its four-guidance series on collecting patient experience data and other relevant information from patients and caregivers to inform drug trial design. Following publication of the first guidance on methods to collect patient experience data that are accurate and representative of the intended patient population in June 2020, the agency is close to releasing a final version of its guidance on approaches to identify what is most important to patients with respect to their experience as it relates to burden of disease and burden of treatment, which it published in draft form in October 2019.

Next up will be two new draft guidances: one on selecting, modifying, developing and validating clinical outcome assessments to measure outcomes of importance to patients in clinical trials, and another on methods, standards and technologies for collecting and analyzing clinical outcome assessment data for regulatory decision-making. Bent says the agency is almost finished writing and will be ready to release in draft form. “We are working hard to get those out soon,” Bent said.