Volume 10, Number 6, June 2012

Key Points:

• Congress has just re-authorized the Medical Device User Fee Act for five years. This continues a decade-long program whereby the medical-device industry supplements the Food and Drug Administration’s appropriated budget with user fees.
• Although the FDA’s budget for licensing new medical devices has grown by almost half, in real terms, from 2003 through 2011, it has not increased its approvals of new medical devices. One simple estimate of productivity shows a drop of 29 percent.
• The FDA’s failure to make productive use of its significantly increased financial resources threatens the future of research and development in medical innovation in the United States.
• Americans should use the reauthorization of the Medical Device User Fee Act to demand that the FDA finally make good on its promises of improved regulatory effectiveness.

Congress recently passed, almost unanimously,
S.3187, the Food and Drug Administration Safety and Innovation Act. Title II of the Act comprised the third reauthorization of user fees for the medical-device industry, for fiscal years 2013 through 2017. A product of lengthy negotiations between the medical-device industry and the Food and Drug Administration, the amendments commit the FDA to a number of new performance goals.

It has long been recognized that the FDA’s process for permitting new medical devices is unsatisfactory. This explains why the medical-device industry initially accepted the policy of paying user fees to the FDA, starting in 2003. These user fees increased the budget available for reviewing applications, above and beyond that which was appropriated by Congress from taxpayers. In return, the FDA committed to meeting certain targets for reviewing applications for new medical devices in a timely and transparent way.

In 2004, the Government Accountability Office (GAO) issued a report in the wake of the first year of results from the Medical Device User Fee and Modernization Act of 2002. The sub-title tells the story: Data to Measure the Timeliness of Reviews of Medical Device Applications Are Limited.1 Of course, the program was only one year old, so it is understandable that data was still lacking.

21 U.S.C §360 defines three classes of risk for medical devices. Class I devices (for example, adhesive tape or bedpans) pose minimal risk; Class II (for example, syringes or hearing aids) pose median risk; and Class III (for example, pacemakers and heart valves) pose the highest risk. Most Class I devices are exempt from prior approval; whereas 80 percent of Class II products are approved via the so-called 510(k) system; and Class III devices must go through a much more rigorous process called Premarket Approval (PMA). These standards were updated by the Food and Drug Administration Modernization Act of 2007, which included the first reauthorization of the user-fee program.2

There are far more 510(k) applications than PMAs, because manufacturers frequently roll out new models of Class II devices. For example, if a manufacturer makes a new model of hearing aid that includes components from new suppliers, it may have to undertake a new 510(k). Manufacturers must also pay annual fees to license the establishments, either domestic or foreign, where they make the devices. There are also a number of post-marketing regulations that the FDA can impose on manufacturers.

Medical devices in use in the U.S. are remarkably safe. Recalls are extremely infrequent. Furthermore, “recall” in medical-device regulation does not always mean what it does to laymen. It can be limited to a warning to users, rather than actually pulling a device off the market.3 Confusingly, recalls are also defined by class, but Class I refers to the highest risk, that is, a significant risk of serious health consequences. Class II recalls involve generally remote risk; and Class III minimal or no risk to patients.

An expert review of Class I safety recalls over the six years from 2005 through 2009 concluded that over 99.5 percent of both 510(k) submissions and PMA submissions during this period did not result in a Class I safety recall. Furthermore, less than 9 percent of the Class I recall actions involved possible undiscovered clinical risks. That is, in over nine of ten cases, increased clinical testing would not have revealed the problems. Over half of Class I recalls had to do with issues that came up after approval, e.g., a manufacturing problem three years after approval.4

So, the FDA is not being slack in approving medical devices without adequate oversight. And yet, there are still widely recognized problems with the FDA’s ability to allow new medical devices onto the market in a timely and effective manner. Last summer, the prestigious Institute of Medicine (IOM) published the report of an expert committee, which concluded that: “It is unclear - and the committee concludes that it is indeterminable, given current information - whether the 510(k) process over the last 35 years has had a positive or negative effect on innovation.”5

A 2011 survey by the National Venture Capital Association (NVCA) reported that the FDA’s ineffectiveness was causing U.S. venture capitalists to reduce their investments in U.S. medical innovation in favor of Europe and Asia, as well as shifting capital towards non-FDA regulated investments. 85 percent of them expected to seek regulatory approval outside the U.S. first. These respondents accounted for 92 percent of the NVCA members’ invested capital, and they had invested $10 billion in health deals over the previous three years.6
In response to concerns about the 510(k) approval process, the FDA issued a “plan of action” last year to address 25 shortcomings that had been brought up by its 510(k) working group.7 Nearly all of these address inadequate communications between the FDA and manufacturers about what is expected. Last April, the FDA reported that it had met its deadlines in making progress towards redressing these shortcomings.8

The FDA appears to desire, in good faith, to improve its processes. Last April, the Center for Device and Radiological Health (CDRH) announced an update to its so-called Innovation Pathway 2.0. The goal is to shorten the time to approval by improving collaboration between innovators and regulators very early in the research and development cycle.9

On the other hand, these challenges were already supposed to have been addressed by the 2007 re-authorization of user fees. At that time, the Administration committed to timely reviews of medical-device applications in exchange for the industry’s signing on to the renewal. For example, the FDA committed to review 90 percent of 510(k)s within 90 days, and 98 percent within 150 days.10

According to the FDA’s self-reporting, these targets have largely been met.11 However, the GAO’s latest report on medical-device regulation challenges these claims. “FDA days” only include days when the FDA is actually reviewing an application. The clock stops if the FDA asks the manufacturer for more information. The GAO discovered that the FDA has been stopping the clock in this way more frequently in recent years. Thus, while the average “FDA review time” for 510(k)s has always been shorter than 80 days, the actual time from application to final decision was between about 100 to 120 days from 2000 to 2006, but had shot up to about 160 days by 2010.12

The medical-device industry’s leading trade association is hopeful that this month’s legislation will lead to improved communications between applicants and the FDA, and a firm commitment from the FDA to reduce the total time that it takes to approve new medical devices, including periods where the FDA previously stopped the clock by asking manufacturers for more information.13

However, to achieve a significant improvement in results, the FDA will have to change the way it uses its ever increasing funding for medical-device regulation. In the three years before user fees were introduced (2000 through 2002), the average annual number of 510(k) submissions was 4,300. The average annual number of PMA submissions was 77. The figures for the first five years of MDUFMA (2003 through 2007) were down to 3,869 and 55; and for the next four years after the 2007 reauthorization (2008 through 2011) they were only 3,979 and 48.14

Nevertheless, the amount of money spent on medical-device regulation increased dramatically, from $217 million in 2003 to over $378 million in 2011 – an increase of three quarters.15 Even accounting for inflation, the increase in funding was 45 percent – just under half – over the period.16 Furthermore, the number of full-time equivalent employees at the FDA who are dedicated to medical-device regulation increased from 1,485 to 1,902, an increase of 28 percent – over one quarter.17 Both these data series are shown in Chart 1.

Endnotes