Medical device industry representatives said Thursday the U.S. Food and Drug Administration's device approval process is burdensome and difficult to navigate, hindering the ability of startups to thrive and driving away medical innovation from the United States.

Six witnesses testified before the U.S. House Energy & Commerce Committee's Subcommittee on Health during the "Impact of Medical Device Regulation on Jobs and Patients" hearing held in Washington.

One of the panelists was Dr. Jeffrey Shuren, director of FDA's Center for Device and Radiological Health, who was grilled by committee members on the differences between the medical device approval process in the U.S. and Europe and the length of time it takes the FDA to run a device through the approval process.

Shuren defended the strength of the FDA's approval pathways. "Some would say that, despite the record of growth and prosperity in the U.S. device industry, the European regulatory system is better for industry and patients," he said. "It is difficult to make direct comparisons between the U.S. and European systems given their fundamental differences, including, at the most basic level, differing approval standards -- the EU lacks the requirement in U.S. law that a device be shown to be effective."

In their testimonies, the panelists addressed different concerns about the FDA's device approval process, including its impact on medical innovation, America's competitive edge and patient safety.

Trouble for medical device startups

In his submitted testimony, Dr. Josh Makower, a consulting professor of medicine with Stanford University and CEO of ExploraMed Development, said that over the years, the FDA's approval process has become "more time consuming, more costly and less predictable to navigate."

In response to industry concerns, Makower and a colleague conducted a survey of 200 medical technology companies to generate data on their experiences with the FDA.

He noted that although the U.S. has a number of medical device giants, more than 80 percent of med-tech companies have fewer than 50 employees. "These small startups are the engine that fuels the developments of innovative new devices, which are often acquired by the larger companies," Makower said.

Makower's study found that it took companies up to two years longer to navigate the FDA approval process than the European regulations for low- and moderate-risk devices. "For higher-risk devices, the discrepancy was greater -- in the U.S. it took three and a half years, or five times as long as Europe, to grant approval," he said.

In addition, half of the companies said FDA personnel responsible for reviews of products changed throughout the process and one-third reported that appropriate staff were absent from the meetings between the agency and the company when the review process was discussed.

"Given that it takes longer and costs more money to launch a product in the U.S., a reasonable question is what is gained from the additional time and costs that result from the FDA process," Makower said.

He acknowledged that the FDA was within its right to ask for clinical data when it deemed it necessary but changing the requirements without transparency or justification is a major obstacle for startups.

In his opening remarks, Henry A. Waxman (D-Calif.) criticized Makower's findings, saying that the study sample was too small to produce meaningful results. In addition, the companies included in the study had little experience dealing with the FDA's device approval process, he said.

Another panelist, Mark Deem, a partner at the medical device company incubator The Foundry, also expressed concern about new ventures in his testimony, saying that, "the most fragile end of the medical device ecosystem, the startups, are struggling."

Deem cited data from PricewaterhouseCoopers LLP and the National Venture Capital Association, which found that between 2007 and 2010, venture capital investment in the medical device sector declined by more than 37 percent. Funding for new startups dropped from 118 new companies in 2008 to 60 in 2010, according to data from Dow Jones Venturesource.

"We exist for our patients, but we live on venture capital," said Deem.

He said the slowness and inconsistency of the FDA approval process are the predominant threats to investment in the sector. "Given the lack of predictability in the U.S., and the relative stability and predictability of the CE mark system, we are no longer structuring, staffing or operating our companies for first commercial release in the U.S.," said Deem. "We develop our products here, and then run the same large, multicenter randomized trials we would otherwise have conducted in the U.S. overseas."

A clash of opinions: is the 510(k)approval process safe?

In a study that appeared online Monday in the Archives of Internal Medicine, researchers looked at 113 devices that the FDA recalled between 2005 and 2009 that posed serious health risks.

Examining FDA data, they determined if the recalled devices were approved using the premarket approval process, the 510(k) process or were exempt from review.

The researchers found that 21 out of the 113 devices were approved through the PMA process, while 80 were cleared using the 510(k) process (8 devices were exempt from any sort of FDA regulation).

In the study, the authors reasoned that the FDA has relied on the 510(k) process more than PMA because it's considerably cheaper for the CDRH to review thousands of devices on an annual basis using the less stringent approval procedure. In 2005, it cost the FDA an average of $18,200 to review a 510(k) submission, while a review of a PMA submission cost the agency $870,000, according to the study.

"The Congress has not appropriated sufficient funds to the CDRH to use the more expensive PMA process for most devices, and this large cost differential creates an incentive for CDRH to rely heavily on the 510(k) process," the authors wrote. "The FDA is partially supported by industry user fees, but the FDA charges much smaller user fees to review medical devices than it charges to review prescription drugs, even for the largest companies."

One of the study's authors, Dr. Steven E. Nissen, served as a panelist at the hearing. Nissen is the chairman of the department of cardiovascular medicine with the Cleveland Clinic. In his testimony, Nissen said the study's finding that 71 percent of the high-risk recalls involved devices cleared using the 510(k) process, "represents a paradox."

"Federal regulations require devices used to support or sustain life to undergo a full PMA approval process," he said. "By law, these life-sustaining devices should not be cleared using the 510(k) provision. Yet, 71 percent of recalls for defects that could 'cause serious health problems or death' were originally approved using the 510(k) pathway."

Ralph F. Hall, a professor and practitioner of law with the University of Minnesota Law School and CEO of MR3 Medical LLC, a medical device startup working on cardiac rhythm devices, also examined the FDA approval process in the same study period. However, he ended up with different results.

He examined the overall safety profile of medical devices approved or cleared by the FDA from 2005 to 2009 by using Class I safety recall data.

Hall found that more than 99.5 percent of the 510(k) medical submissions assessed during the study didn't result in a Class I safety recall. Hall also found than less than 9 percent of the Class I recalls involved potential undiscovered clinical risks. "As such, increased pre-approval clinical testing would not have any meaningful impact on reducing the number of Class I recalls," he said in his testimony.

"This study demonstrates that very few 510(k) medical device submissions -- less than 0.5 percent -- become the subject of a Class I safety recall," Hall said.

Subcommittee Chair Joe Pitts (R-Pa.) asked Hall to explain why his findings were different from Nissen's results and the study's. After the hearing, Hall told DOTmed News he has four issues with the study.

Hall said it's inaccurate to just look at the numerator to see how the whole system is performing, explaining that Nissen's study didn't take the total number of devices going through the FDA's approval process into consideration. "Are there more Class I recalls for 510(k) products? Of course there are," said Hall. "Why? Because every year there are more than 3,000 510(k) submissions and 20 to 40 PMA submissions. Does it surprise you that there are more for the 510(k)?"

Another problem with the study is that researchers didn't look at the root causes of device recalls. "If you want to try to solve a problem, you need to know what caused the problem. And in my study, we looked at the reason for the recall and more than half of all the recalls were post-market problems, having nothing to do with the 510(k) system," Hall said.

"Even within the ones that may have existed at the time of clearance, the vast majority of those are quality system problems, not a lack of clinical data," he added.

Hall also said the researchers confused the classification of a product for its approval pathway with the classification of its recall. He used an example of a tongue depressor -- it's a low-risk product. If it's contaminated with deadly bacteria it's a high-risk recall. "They say that if it's a high-risk recall, it should be a high-risk PMA product. Those are two different things," Hall said.

Hall also said the study has "a number of misstatements about the law."

And the Medical Imaging & Technology Alliance, a major industry trade group, agrees with Hall. In a statement following the meeting, MITA expressed its support for improving the 510(k) approval process and criticized the study published in the Archives of Internal Medicine.

"The study uses a flawed analysis based on a simple count of the number of FDA recalls to draw inaccurate conclusions about the 510(k) process that would actually exacerbate the problem of timely patient access to innovative medical technologies," Dave Fisher, MITA's executive director, said in prepared remarks.

Innovate, but not at the expense of safety

Another panelist, Dr. Rita Redberg, said although the focus on medical innovation is important, the safety and effectiveness of the devices should be the leading factors of the FDA's device approval process.

In her testimony, Redberg, director of women's cardiovascular services, division of cardiology, with the UCSF Medical Center and editor of Archives of Internal Medicine, argued that the 510(k) process, conceived in 1976, is too outdated to handle today's complex technologies, especially in the fields of cardiology and orthopedics. "In contrast to most devices in the 70s, the newer products pose substantially greater risks -- even life-threatening risks -- to patients," she said.

Redberg said that Congress has mandated that high-risk devices go through the more stringent PMA process (or be reclassified as lower-risk products), but according to a 2009 report by the General Accountability Office, the FDA has failed to do so.

Among her recommended improvements to the device approval process, Redberg suggested that high-risk devices cleared through the 510(k) pathway should be supported by significant clinical data. A 2009 study by Redberg and colleagues found that only 27 percent of all premarket approvals were based on a randomized trial and only 14 percent of these trials were blinded.

And after the device is approved, there should be a system of "post-marketing surveillance that would detect serious adverse events in a timely fashion," she said.

"I appreciate greatly the value of innovation in helping me to take excellent care of my patients," Redberg said. "But it is not in the interest of my patients, or the American public, to allow medical devices to get to market without the benefit of high quality clinical studies to demonstrate that they will actually help patients to feel better and even live longer, without causing harm."

Throughout the hearing, Shuren acknowledged that the 510(k) process needs to be improved and emphasized the need for additional funding and resources for the agency to function properly. In response to criticisms that the agency takes too long to approve a device, Shuren said the industry hasn't been pulling its weight -- companies often submit poor clinical data and ignore FDA's recommendations throughout the review process.

When it comes to comparing U.S. and EU's device approval processes, Hall told DOTmed News he agreed with Shuren. The two systems are difficult to compare, but that doesn't mean it shouldn't be done. He said he's working with the FDA on finding potential ways to design a study that will resolve the methodological difficulties.

Click here to read an exclusive report on FDA's search for balance between safety and innovation.