At the end of August, pharmacoepidemiology enthusiasts from across the globe gathered in Philadelphia for the 35th International Conference on Pharmacoepidemiology and Therapeutic Risk Management (ICPE). The conference, centered on “using real-world data (RWD) and designs to optimize decisions,” brought together FDA officials, biopharma leaders, and accomplished researchers.
Five days of short courses, podium sessions, and poster presentations—including 23 presentations by the Aetion team—yielded five key points about the future and scope of real-world evidence (RWE), and action items for biopharma:
Read on for our takeaways, with insight from the Aetion science team at ICPE:
1. Agencies continue to call for examples of regulatory-grade RWE
To kick off the conference, Jacqueline Corrigan-Curay, J.D., M.D., director of the Office of Medical Policy within the Food and Drug Administration’s (FDA) Center for Drug Evaluation and Research (CDER), presented the keynote, “The role of real-world evidence In regulatory decision-making.”
She re-articulated a need for more examples of RWE used in regulatory decision-making, an ask FDA leadership has consistently made of industry, and cited a variety of demonstration projects as important steps forward.
One such project, RCT DUPLICATE, a collaboration between the FDA, Brigham and Women’s Hospital, and Aetion, seeks to reproduce the results of 30 published randomized controlled trials (RCTs) and predict the results of seven ongoing trials, was cited by multiple panelists as a key exploration of where RWE can help regulators make decisions—and where it cannot.
The European perspective was expressed by Miriam CJM Sturkenboom, Ph.D., professor in the Department of Global Health at the University Medical Center Utrecht’s Julius Center and expert to the European Medicines Agency (EMA). She acknowledged the willingness of the EMA to join this shared learning process and the wealth of opportunity to explore RWE’s utility in Europe. With health technology assessment (HTA) agencies in Europe’s interest in real-world drug efficacy, RWE could be a natural fit for decision-making there.
2. All stakeholders continue to define and explore fit-for-purpose data
Several presentations addressed the importance of validated, fit-for-purpose data. Asking “How do we assess when electronic health records or claims databases are fit for a specific research or regulatory purpose,” a panel including members of the Patient-Centered Outcomes Research Institute (PCORI), RTI Solutions, and the FDA emphasized the importance of assessing data quality, completeness, and provenance before choosing a data set for an analysis. As real-world data sources expand to include wearable devices and other data collected out of the health care system, this will become increasingly important, said Dr. Corrigan-Curay.
Panelists also addressed the importance of understanding data sources, especially global ones, as different populations may result in various health outcomes depending on culture and policy differences.
3. Decision-makers increase focus on principled database epidemiology and transparent analyses
Echoed throughout the sessions was the importance of careful study design. Thoughtful planning—with traceability—for fit-for-purpose data sources, appropriate methods, control for confounding, and other considerations can make the difference between regulatory-grade RWE and an untrustworthy study.
One way to ensure validity of a study is to use a traceable, transparent platform to run analyses, that way regulators can test methods to ensure trustworthiness of the data. In a presentation on RCT DUPLICATE, for example, RCT DUPLICATE’s lead investigator, Jessica Franklin, Ph.D., provided examples of registering all her team’s RWD analyses on clinicaltrials.gov. To enable FDA and the public to monitor their work in real time.
In the session “Can and should we establish thresholds for validity?” David Martin, M.D., associate director for real-world evidence in FDA’s OMP, spoke to three requirements of regulators to ensure validity: transparency before a study is run: detailed reporting, and regulators’ access to the data used in the study.
4. Keep an eye on future applications of machine learning in pharmacoepidemiology
Machine learning emerged this year as a hot topic, and was highlighted in several sessions and the special interest group “Digital Epidemiology.” In these sessions, presenters explored methods in machine learning, including automated selection for covariates, and how they can fit into the pharmacoepidemiologic world. Panelists explored how those implementing machine learning can gain acceptance of their methods from regulators. Aetion President and Chief Science Officer Jeremy Rassen explored the topic in his presentation, “Where should I submit my manuscript? Applying machine learning to enhance communication of medical research.”
5. Expect a gradual process for RWE acceptance and adoption
In the session “Replicating randomized trial data using real-world data,” Dr. Sturkenboom, of the University Medical Center Utrecht and EMA consultant, spoke of validating, accepting, and adopting RWE more widely.
According to Dr. Sturkenboom, RWE acceptance exists on a spectrum—some are fully on board, others are skeptical, others are not likely to ever accept RWE as a valid source of information. Addressing the room, she made the assumption that those gathered were in the “fully on board” category, and encouraged RWE advocates to focus energy on proving its utility to skeptics- rather than trying to convince the inconvincible.
Pointing to the wealth of research demonstrating the trustworthiness high-quality RWE derived from well-designed analyses on fit-for-purpose data, she said that the question isn’t about the validity of RWE, but rather in the perception of RWE. Continuing to conduct transparent analyses will reassure skeptics of the value of this new tool in finding and evaluating new treatments and improving outcomes for patients.