Notified body perspective: Evaluating uses for real-world data

BRUSSELS – Real-world data (RWD) can provide useful information that may be included in clinical data submissions, but the greatest value may be during the postmarket follow-up period, Richard Holborow, head of clinical compliance at BSI, said at the 2023 RAPS European Clinical and Risk Management Conference.

“Premarket data are obviously key to getting the device to market because it’s done in a controlled, methodological manner, following certain rules,” Holborow told Focus. “Once you’ve got your CE certificate, because that device is going out to all physicians to be able to use in Europe, then we expect that there are activities to capture that data.”

Holborow provided an overview of RWD from the perspective of a notified body and noted that, under the EU Medical Device Regulation (MDR), manufacturers are expected to commit to postmarket clinical follow-up (PMCF). This is where RWD and real-world evidence (RWE) can provide critical information on how the device performs during actual clinical use, he said.

Definitions of RWD and RWE vary slightly depending on the document or regulatory body, but RWD is generally described as data to that relates to patient health status or the delivery of health care, which is routinely collected from a variety of sources, often outside of a clinical study. RWE, which is derived from analysis of the RWD, is the clinical evidence of usage, and the potential benefits and risks.

There has been a surge in the use of RWD in recent years, sources of which include disease- or device-based registries and administrative data, Holborow said, particularly among manufacturers that have legacy devices and need to prove the clinical value under MDR, even if the device is currently considered to be “standard of care.”

Another reason for the increase in RWD use may be the difficulty in obtaining sufficient data for less common situations or indications, or where there are multiple variants of a device, such as spinal or hip replacement parts, or cardiac stents. Manufacturers are increasingly relying on RWD to plug the gaps, he said.

“The reality is, somebody needs those devices, otherwise, a manufacturer wouldn't be making them,” Holborow said. “But how do we get sufficient data of quality and quantity for those particular extreme variants?”

On the premarket side, RWD can be useful to help inform the types of clinical investigations that manufacturers should be doing in terms of safety and performance objectives, Holborow said. On the postmarket side, RWD can be used to help confirm data, support new indications, assess benefit-risk in subgroups, and look at the long-term performance of a device, among other uses.

RWD can be more representative of a target population, as well as provide an efficient and cost-saving way to get results, Holborow said. Additionally, RWD provides an ethical way to gain information on established devices and it can be less burdensome on already overstretched healthcare professionals since it is collected retrospectively.

But there are downsides too, he said. RWD are likely to be observational, making the data open to greater bias than clinical trials, and there may be limited intervention of the data source. Data entry may lack detail outside of a controlled environment, and data may be focused more on safety than device performance. Plus, capturing longer-term data may be difficult to capture without some form of intervention, he said.

When considering the use of RWD to support a conformity assessment, Holborow recommended several considerations:

Can the data reflect performance, as well as safety, and does it capture this in the long term?
Can the data demonstrate the lifetime of the device(s)?
Is the data able to be stratified to support all indications, variants and sizes?
Is a multiple approach required, such as using registries and surveys?
How have patient reported outcomes been validated?

“We've got to be very careful in terms of how we manage this data as a community, to make sure that we don't end up with data overload,” Holborow said. “We have to think about this simply, about what we're trying to prove, and where is the best source of data?”

Notified body perspective: Evaluating uses for real-world data

Related topics

Recommended content

Welcome to the new RAPS Digital Experience