This report originally appeared in the November 2010 issue of DOTmed Business News

On the eve of the U.S. Food and Drug Administration’s 510(k) reform, industry fears the agency will keep life-saving products off the shelves, while consumer advocates worry the latest changes don’t do enough to protect patients.

After two years plagued by scandal, the FDA’s device division is now looking to toughen its 510(k) process, a clearance process used for tens of thousands of supposedly lesser-risk medical devices.

But as the division, known as the Center for Devices and Radiological Health, prepares to overhaul the nearly 35-year-old process, the agency faces a choice. In essence, it has to decide what sort of mistake it wants to make.

“We have an old, old tension here between which type of error or injury the FDA is willing to risk,” Walter Olson, a scholar with the libertarian think-tank the Cato Institute, told DOTmed News.

While the reform is driven partly by fears that slack regulation has allowed shoddy or untested devices to enter the market, overly burdensome regulations carry a risk of their own, even if it can be harder to quantify. A proliferation of red tape, costly to cut through, could scare away potential innovators, leaving life-saving products off U.S. shelves.

The need to balance innovation and safety was recognized early by the agency. In the 1970s, David Link, the chief of the Bureau of Medical Devices, the precursor to the FDA’s current device regulating section, noted this when he said he thought too many agency people were antagonistic to industry.

“[We] recognize that the bureau had the responsibility to get safe and effective products to the market as much as it was to keep unsafe and ineffective products off the market,” he said in a statement posted on the FDA website.

‘Fast-track’ process under review
At issue is a commonly used process for approving medical devices sometimes dubbed the “fast-track” process, as clearance usually takes less than three months.

This 510(k) process applies to most moderate-risk, Class 2 devices that are “substantially equivalent” to devices currently on the market, called “predicate” devices. The category, borne out of 1976 legislation, covers everything from relatively low-tech motorized wheelchairs to state-of-the-art MRI scanners.

Very low-risk products, like tongue depressors, fall under the Class 1 category. They, and even some Class 2 devices, are exempt from the 510(k) process. The riskiest devices, such as pacemakers, come under the much tougher Class 3 rules, which generally require a premarket application (or PMA), often involving extensive and expensive clinical trials.

Scandal hits CDRH
The move to overhaul the 510(k) process was set in motion almost two years ago, when whistleblowers within the FDA sent letters to members of Congress, telling of pressure from above to approve devices they thought were unsafe or unproven. Although the eight or nine people involved in the campaign were a small fraction of those at the agency, it painted a disturbing picture of a department in which device approval owed as much to political concerns, as to scientific ones. The whistle-blowing would culminate in FDA’s decision in September 2009 to fund a $1.3 million review of the 510(k) process, handled by the respected Institute of Medicine, an independent group whose 2000 report To Err Is Human led to sweeping changes in adverse event reporting.

But problems continue to crop up. This summer, the FDA was rocked by further scandal, as a former reviewer accused the agency of sacking him because he refused to approve a CT scanner for virtual colonoscopies.

An investigation into some of these complaints is being handled by the Office of the Investigator General. (It’s looking for administrative, not criminal violations. A criminal investigation by the OIG of the U.S. Department of Health and Human Services was dropped in February). And the results of IOM’s FDA investigation are due in March. What it will decide is anyone’s guess. When contacted by DOTmed News, IOM spokeswoman Christine Perrel said it would be “premature” to talk about the report before it’s released next year.

Industry fears changes could cause more delays
But the FDA is not sitting idle as it waits for the IOM’s report. It has already issued draft guidelines that some think might utterly revamp the 510(k) process.

Among the 65-odd proposed changes released this summer are some related to increased transparency, such as requiring device-makers to publish submission materials on publicly accessible databases, as well as to offer more surveillance of devices after they reach the market.

But perhaps the most notable is the creation of a Class 2b category for devices riskier than those normally cleared under Class 2 rules, and which could require clinical trials. By press time, the FDA had not clarified what devices would fall under this category, although hints dropped at meetings have worried trade groups, such as the Advanced Medical Technology Association (AdvaMed).

Although in principle AdvaMed said it supports subjecting a subset of Class 2 devices to enhanced requirements, it fears the FDA could use too broad a brush for the category. For instance, suggested criteria for making a device Class 2b could include products permanently implanted in the body. But AdvaMed points out this would include a device most agree is reasonably low risk: sutures.

“We made a point that just because a device met those criteria, that it might not fall into [the higher-risk category], because it might have been on the market for a long time and demonstrated a good safety profile historically, so there would be no need to put it into a 2b,” Janet Trunzo, executive vice president for technology and regulatory affairs for AdvaMed, told DOTmed News.

In its comments on the proposed changes submitted to the FDA in early October, AdvaMed said while it supported close to one-third of the proposed changes, its main fear was that cumulatively, many of the new changes could increase the burden on companies, making it that much harder and more expensive to get devices approved.

Already, the industry is grumbling that despite increases in user fees – money collected from device companies to help the FDA pay for its operations – the clearance process has gotten slower and slower. Largely, this has come about because of the agency ordering increasing numbers of review cycles, according to AdvaMed, which can add months to the allotted time the agency has to review a product submission.

“What happens is, a review cycle is when FDA stops the review clock and asks for more information, and then upon receipt of that they restart the clock,” Trunzo explained.

She said over the years it went from 1.4 review cycles per submission to now close to two per submission.

“We don’t have any data or analysis from FDA as to why they’re stopping the review clock,” she said. “But stopping the review clock does lengthen the review process.”

Safety under scrutiny
As part of its opposition to some of the changes, AdvaMed points out that the 510(k) process works, and that its safety has been upheld by numerous studies.

An analysis by Battelle Memorial Institute commissioned by the group this summer found of the 47,000 devices cleared by the process since 1998, only 0.16 percent were involved in a Class 1 recall – that is, a recall in which the products involved had the potential to cause serious injury or death.

The group says this echoes two other recent studies, one from Ralph Hall of the University of Minnesota and one from Dr. William Maisel of Beth Israel Deaconess Medical Center. Hall's study found that the rate of Class I recalls for 510(k)-cleared products was less than one out of 500. Maisel's study looked at all classes of recalls and found a recall rate in the range of one out of 100, to one out of 67.

"The 510(k) process has an extraordinary safety record," AdvaMed’s senior executive vice president David Nexon told reporters in a call last month. "It's hard to imagine that any other regulatory process has as good a record at weeding out products before reaching patients."

But not everyone is convinced by these findings. The National Research Council for Women & Families, a Washington, DC-based advocacy group long critical of the 510(k) process, dismissed the industry’s interpretations.

“The issue is not what percentage of 510(k) devices are recalled,” Dr. Diana Zuckerman, the group’s founder, told DOTmed News. “The issue is how many people are harmed. So, you can have one recall that affects 20 million people.”

As an example, she mentions the 2000 recall of more than one million bottles of contact lens solution. Some of the bottles, contaminated by a fungus, led to eye injuries and even in some cases, blindness.

Further, she said the reports aren’t accurate because no one knows what the “real” denominator is for the 510(k) process. Many cleared devices are never sold, she argues, either because the company decided to discard the product or because it only got approval for the device to use it as a “bridge” product. That is, to use it as a predicate device for which to clear another product.

“There should be zero high-risk recalls,” she said. “If a product by 510(k) is, by law, not a high-risk product, it should not end up with a high-risk recall. If it’s a high-risk recall, the product was high-risk to begin with.”

Public Citizen, a corporate watchdog group founded by Ralph Nader, also finds the process lacking. In the July 2010 article “Left to Their Own Devices: Breakdowns in United States Medical Device Premarket Review,” published online in the peer-reviewed journal PloS Medicine and co-authored by Dr. Sidney Wolfe, director of Public Citizen’s Health Research Group, the organization noted that device approvals were generally laxer than drug ones.

“In practice, [New Drug Applications] typically contain two or more well-controlled clinical studies, whereas for PMA applications, a single study is the norm and most 510(k)s contain no clinical data,” the authors wrote. “Thus, data that would never be sufficient to support the approval of a drug can result in the approval of a device used to treat the same condition, potentially diverting patients from effective drugs to less-effective devices.”

As an example, the authors cite the approval of a vagus nerve stimulator, intended to treat depression, using evidence that an FDA psychopharmacology expert said would not suffice for an antidepressant drug.

Even so, PMAs get almost five times more attention than 510(k)s, according to the paper. The average review time – circa 2006 – for a 510(k) submission is only 54 days, whereas it’s nearly 283 days for PMAs.

“The 510(k) premarket approval process has failed to keep dangerous and ineffective medical devices from the market,” Public Citizen concludes on its website.

Revolving door?
But 510(k) safety data aren’t at the heart of consumer advocates’ concerns. Many feel that FDA’s proposed changes don’t answer one of their biggest worries: what’s seen as a cozy relationship between device companies, lawmakers and regulators. This has real-world consequences for patients, as uncovered by a Wall Street Journal investigation last year. In a January 2009 article, the paper argued that pressure from local politicians and the then-head of the FDA’s CDRH led the agency to clear through the 510(k) process a collagen meniscus device made by the company ReGen, despite objections from some reviewers that it was not “substantially” equivalent to an existing product and that it had safety issues. Ultimately, the Centers for Medicare and Medicaid Services denied coverage to the product, saying the evidence wasn’t in its favor. The FDA is now reconsidering its clearance of the device.

Regardless of whether the device ultimately proves to be safe, the brouhaha brought increased media scrutiny to the way commissioners, usually political appointees, influence the decision-making process at the agency and ties between industry and division directors.

In fact, virtually all previous directors of the CDRH have taken jobs in the industry after leaving the agency, becoming consultants for big drug firms like Amgen or Wyeth-Ayerst, launching their own consulting firms or even working for lobbies like AdvaMed.

“It’s very rare for a high-level FDA person to retire and never work again,” Zuckerman said. “Definitely there’s a huge revolving door to industry, and that’s where they almost always go.”

Still, this is not to say she’s accusing anyone of corruption, as many of the CDRH directors spent decades working at the FDA or other public agencies, earning much less than they could have in the private sector, Zuckerman said. While the salary of the FDA commissioner is around $153,000 a year, and some employees can earn more than $191,000 under Title 42 policies designed to retain specialists, it’s still far less than they could be making if they were in the industry, in which salaries for many of these officials would be half a million dollars or more, according to The Wall Street Journal.

“Almost anybody who heads CDRH…probably could get a job that pays a whole lot more if they went to industry, and certainly some of them don’t go, or at least stay at FDA for decades, getting paid much less than they could,” Zuckerman said.

And some think the problem with the revolving door is less the corruptibility of FDA staff and more the dysfunction it reveals in our system, that the industry thinks it needs someone intimately familiar with the Byzantine halls of a federal agency in order to get devices approved.

“There’s something disturbing that in order to get a lot done in that area they feel they have to hire a big Washington establishment figure and find people inside the agency who know the otherwise impenetrable corridors,” Cato’s Olson said. “It also works to the detriment of someone who has a great idea but not a Washington budget.”

Olson also thinks the standards argued by groups like Public Citizen create an impossible choice, one he says is reminiscent of the dilemma faced by Stendhal’s hero in the “Red and the Black”– in which at a young age, ambitious, middle-class strivers must pick either the church or the army.

“You can either choose to deal with influencing the federal government, or you can choose to influence the world, but choose at age 21,” he joked. “Do you force people to be mandarins and never sully their hands and never launch a thing, or totally be on the profitable side and never contribute usefully to how the regulatory process should be structured?”

“You’ve got lots of biases in both directions,” he said, “and somehow you have to sort through the biases in a way that does not deny the fact that scientific talent is scarce and you can’t write off 99 percent of it.”