The COVID-19 pandemic has brought forth a new era of collaboration between health care data providers, biopharma, and regulators, especially as vaccines enter the market. But with this collaboration comes an increased awareness of how unwieldy these data have become, and of the work needed to use and aggregate high quality real-world data (RWD).
At the second annual Aetion Summit, “Evidence at the Apex,” a panel of health care data leaders met to discuss the changing RWD landscape that has emerged from the COVID-19 pandemic, and the strategies we must take to approach novel challenges.
The panel convened Andrew Kress, Co-Founder and CEO of HealthVerity, Gilles Paubert, Global Head of Health Data at Cegedim, and Ryan Fukushima, Chief Operating Officer at Tempus. Here, read takeaways from our discussion, including learnings thus far from working with COVID-19 RWD, and how the pandemic has impacted industry’s needs around data privacy and standards.
COVID-19 demands speed, agility, and rigor in the approach to RWD studies.
The panelists expressed that COVID-19 has both accelerated the trend toward increased reliance on RWD to answer key questions, and highlighted the urgent need for real-time data.
“We’ve found that agility is key: What do we need to do to help our clients and our team better understand the data so we can act faster?” Gilles asked.
The pandemic also posed unique challenges to working with RWD, in part because it is a novel disease. But for Cegedim, because most COVID-19 symptoms were already captured in their data set, researchers were able to validate diagnoses and identify potential COVID-19 cases early on in the pandemic. To ensure these data were recent and published in a timely manner, Cegedim changed their data delivery frequency from monthly to daily.
This demand for speed and agility also requires increased focus on conducting high quality studies.
As Ryan noted, “we must be cognizant of all the elements that go into high quality RWD research if we want to create a model that works beyond this pandemic.” These considerations include data provenance and quality, sample bias, methodology, and transparency.
Data sampling questions are top of mind as biopharma and regulators look to RWD to study post-approval vaccine safety and effectiveness.
Some of the key challenges to assessing vaccine safety and effectiveness will pertain, according to Andrew, to data sampling and analysis methods.
When looking for vaccine safety and efficacy measures in a data set, Andrew shared that one of the essential first steps is ensuring that the data source fully captures the exposure of interest—that the patient was vaccinated—and is able to capture the outcome event. It’s also important to ensure that the studied cohorts are representative of real-world populations, including those who may be at higher risk—for example, pregnant women.
“We still aren’t sure how COVID-19 vaccines will manifest in the data, so we must take a flexible approach and act quickly to figure out what that signal is, when it happens,” Andrew said.
Data privacy regulations may be modified to support COVID-19 RWD research.
Concerns around data privacy regulations and their potential impact on RWD analyses have become particularly apparent in COVID-19 research. For example, Andrew explained that due to the sparse case numbers and extensive media reporting at the beginning of the pandemic, early COVID-19 data is particularly identifiable. This raises privacy concerns, but he argued that there are ways to modify the current regulations to support COVID-19 RWD research.
Working with fully anonymized data sets is an important starting point, Ryan and Gilles shared. Andrew also called upon regulators to issue guidance around potential exceptions to how we approach de-identification for specific research protocols.
Another approach, which empowers patients to consent to sharing their data for the purpose of RWD research, can also support this work. “Patients are more willing to consent to sharing their data if they know it can be used to improve treatment,” Ryan explained. This can grant researchers access to key data attributes that may otherwise be obfuscated due to privacy restrictions.
While industry awaits FDA guidance around RWD quality and transparency, biopharma and data partners can collaborate to inform regulatory standards.
The panelists agreed that the FDA is likely to soon release guidance regarding data quality, including standards around transparency. However, Andrew noted that this guidance may evolve over time, or present as a spectrum to allow stakeholders to work with potentially less rigorous data for certain applications if it gives a quicker read on key data points.
Ryan expects to see standards around data provenance, especially to trace the sources of data used in external control arms. He encouraged biopharma to require this information from data partners to help establish credibility with regulators. “When we work on external comparator projects, we spend much time and effort fostering relationships with health care providers to understand exactly where the data came from,” Ryan said. “That way, we can provide this information to FDA reviewers instead of asking them to ‘trust us.’”
Gilles echoed these points, and shared the importance of making a “continuous investment in working with physicians and helping them understand the value of the data for their own practice” to ensure data quality.
Ultimately, through existing RWD collaborations between data providers, biopharma, and regulators like the COVID-19 Evidence Accelerator, Andrew expressed that “we should already have a good set of criteria to evaluate data quality.” Therefore, in anticipation of the FDA’s formal guidance, the panelists called upon industry and its partners to lead the way in developing standards for high quality RWD.
Carolyn Magill, CEO of Aetion, echoed this point: “We need to lead the way toward developing standards. There are places where we are confident that we can drive change and consensus, and we should focus there to help the rest of the industry move forward.”