Can We Truly Harness Real-World Evidence for Medical Device Safety and Effectiveness? - September Newsletter Feature

Using real-world data (RWD) to generate real-world evidence (RWE) appears to be on everyone’s mind these days.  There are numerous publications, webinars, conferences (and newsletters :>) focused on this topic.  There are plenty of sources for information about RWD/RWE and the potential it has to change the way we establish the safety and effectiveness of medical devices.  But, one question I do not see addressed very often (so, not at all) is this…why is it so incredibly difficult for us to generate RWE about medical devices in the United States?

Surely, I am not the only one who has noticed that we are working really hard to gather relevant, comprehensive, quality data that can be used to establish safety and effectiveness.  And, the reality is, we’ve been working really hard for a long time to do so.  Between the Medical Device Epidemiology Network (MDEpiNet),  FDA’s Sentinel Initiative, and the National Evaluation System for health Technology Coordinating Center (NESTcc) we’ve been working to find a way to harness available data on the real-world use of medical devices for over a decade.  As an epidemiologist, I am deeply familiar with the challenges associated with integrating observational data form various sources and trying to interpret some generalizable meaning from it.  This is not an easy task, but in the U.S. we have made it particularly hard to do on a large scale.

The quick answer to the question above, is our health care system in the United States is decentralized and lacks consistency, which forces us to find ways to cobble together RWD from several sources, none of which were designed to work together.  Outside of the U.S., countries with centralized, single-payer health systems can enforce consistency in quality, content, and type of data reported.  Registries abound in places like the Netherlands, Scandinavia, and EU.  There are still limitations, but for researchers and regulators accessing and analyzing the various clinical data sources in single-payer system, the barriers to aggregating real-world data are minimal.  I am not using this as justification for single-payer health system in the U.S. I’ll leave my political views at the door.  But it is worth noting that the way we collect clinical data in the U.S. has essentially hamstrung our ability to harness that data for use in regulation and protecting the public health.

A quick look at the three main real-world programs noted above reveals the extent of the efforts necessary to gather meaningful data.  Sentinel is designed to get a jump on adverse event trends before they become a public health crisis.  MDEpiNet, once the main observational research program within the Center for Devices and Radiological Health (CDRH), has moved on to be primarily a public-private partnership focused on integrating infrastructure, developing methodologies, and conducting studies. I used to hear frequent updates about MDEpiNet, but it seems to have never fully achieved its aim of creating a framework for integrating RWD to generate RWE. When was the last time you heard about it?  Instead, we now have NESTcc, the brain child of the Duke-Margolis Center for Health Policy and funded by FDA.  Everyone is talking about NESTcc the same way they talked about MDEpiNet a few years ago.  In fact, the two programs are very similar in their vision, goals, and partnerships.  Somehow, though, NESTcc appears to have more traction than MDEpiNet ever did.  

The limitations of working with observational data are not new.  In 2007 their textbook, Medical Device Epidemiology and Surveillance, Roselie Bright and S. Lori Brown describe the many challenges to understanding real-world use of medical devices.[1]  These include the iterative nature of devices, lack of identifying codes in medical records, and reusable devices that may be used on multiple patients.  The result is that real-world medical device data is inherently difficult to work with; trying to aggregate these data from multiple sources across the care spectrum increases the difficulty by orders of magnitude.

The programs currently underway to address the need for real-world evidence of safety and effectiveness for medical devices have merit and momentum.The hope is they can overcome the various challenges to create programs that really work to enhance our understanding of real-world use of medical devices.It’s just a shame that we must work so hard to get to that end goal.

[1] Bright RA and Brown SL. (2007)  Medical Device Epidemiology (Chapter 3) in Medical Device Epidemiology and Surveillance. Ed. Brown SL, Bright, RA, and Tavris, DR. Rockville, MD. Wiley and Sons.