Diagnostics & Imaging Week Washington Editor

WASHINGTON – As the FDA struggles to keep pace with the growth of the medical device and pharmaceutical industries, and growing sophistication and complexities of new products, it says it is continuing to speed its current review procedures with some self-examination and a variety of collaborative efforts with industry.

However, climbing out of the present-day quagmire into a leaner, meaner future will not be easy. The Center for Devices and Radiological Health (CDRH) currently is charged with keeping an eye on 15,000 companies in the device and diagnostics industries, up from 6,000 just nine years ago, and more than 100,000 different types of hardware and software as compared to 84,000 in 1997.

Attempting to make things happen on a more up-to-the-minute schedule, the FDA recently unveiled its Medical Device Innovation Initiative (MDII), hoping, it said, "to ensure that clinical trials, product reviews and approvals, and manufacturing processes are conducted in the most efficient and effective ways."

But MDII would appear to be mostly a substantial repackaging – trumpeting the effort at ongoing "improvements" of programs already under way, such as the Critical Path Initiative (CPI).

Miriam Provost, PhD, the deputy director at the Office of Device Evaluation, in e-mail responses to questions from Diagnostics & Imaging Week, acknowledged that MDII "pulls together some of the things that we are working on – and that we plan to do more of – and some programs that were just begun this year." Provost acknowledged that the CPI is one of the components of the MDII.

The MDII announcement points out that devices "that would have been considered the stuff of science fiction just a few years ago are quickly becoming the standard of care," including "smart technologies" and other developments that are feeding "a progressively decentralized healthcare delivery system."

Nothing new here. The agency has been bannering the difficulties of its new scientific challenges for several years.

Among the "critical steps" described in the MDII announcement is the use of Bayestian Statistics in Medical Device Clinical Trials. However, the guidances for this effort are in development "but not yet final," according to Daniel Schultz, MD, CDRH director.

In unveiling MDII, Schultz said that the Bayesian statistical methodology "is a very exciting move in the direction of expediting new technology to the public." The Bayesian statistics idea, which was itself the subject of a separate announcement, is said to offer industry "the option of using prior, legally available information" to design leaner clinical trials without sacrificing scientific rigor.

A key question will be whether the industry itself has the expertise to use these new methods and if the promised guidances will offer the necessary, understandable "how-to's."

Overall, the other programs bannered by the MDII announcement tend to be expansions of previously reported efforts rather than new initiatives. Following are some of these:

• CDRH states that it is putting into place "a new process for prioritizing guidance document development" that will render the review process more transparent. Some of the guidances to be released will address drug-eluting stents, combination products for cartilage repair and tests for genetics and pharmacogenomics.

These guidances, of course, have been long promised, and D&IW will continue to report on the timeliness of the promised roll-outs.

• MDII states that CDRH will work to improve the science behind some high-need areas of medical science, including virtual systems that mimic the human circulatory system for adults and children, and statistical models to forecast the long-term behavior of cardiac stents. Interesting ideas – not new, but interesting when actually used.

• CDRH says it is seeking to leverage its lab expertise to help product development move along, citing a recent instance in which its experience in fluid dynamics helped a device sponsor avoid problems with proposed post-approval change to a pediatric left ventricular assist device. It also says it is keen on computer-assisted diagnosis, which has been applied to mammography as well as to lung cancer and CT colonoscopy.

• The agency says it is turning its quality review scope on itself, putting into place "an ongoing quality review program for premarket submissions that evaluates" the agency's review "in several key scientific areas."

However, the document goes into no specifics on this subject. CDRH's first round of analyses of its reviews of several cross-cutting areas, namely biocompatibility, sterilization and statistical analysis, are complete, but the results were not included in this document.

• The MDII announcement notes that the FDA is working on an IT system that will improve "monitoring and tracking of premarket submissions" as well as on a "Center-level tracking system that helps monitor and manage workload and . . . ensure that appropriate scientific expertise is used to review each document." This, we can safely assume, is an effort that has been ongoing – or should have been ongoing – for years.

Other areas of interest in IT include the use of electronic filings for PMAs and 510(k)s. CDRH will assemble "integrated, searchable electronic databases that will allow reviewers to identify and use information from applications previously sent to the agency" faster than plodding the hard way through hard copy.

However, the task of analyzing submissions and notifications will still rest wholly on eyeballs and gray matter. "We are not currently pursuing the development of expert systems that would review or analyze data in the place of a human reviewer," Provost commented in her e-mail message.

However, Provost said that "our staff are passionate about their jobs and believe strongly in our public health mission" despite the pay differential. Retention is abetted, she said, by "providing them with opportunities for training and other developmental experiences."

The agency will, like private industry, rely on information technology (IT) to boost productivity.

• CDRH says it is using MDUFMA user fee monies to expand its medical device fellowships, which has been a boon in evaluations of products such as "a first-of-a-kind neuro-interventional device for retrieval of clots" for a patient with ischemic stroke. Another product class that it says has benefited from the additional fellowship savvy is the long-term left ventricular assist device for patients who are not candidates for transplant.

According to Provost, CDRH "typically offer[s] a 13-month term, which may be extended for another 13 months" to these fellows. However, some fellowships are structured for shorter terms, such as for half a year.

The increasing workload at FDA could not be without implications for the continuity of permanent staff. At the May 22 Radiological Devices panel meeting held in Gaithersburg, Maryland, Thomas Gross, MD, the director of the Division of Postmarket Surveillance at CDRH, said that the difficulty of keeping up with post-market study reviews was due to "tremendous turnover" in reviewers. CDRH's analysis of its post-market study tracking program disclosed that it had lost or reassigned the lead reviewers in 18 of 45 post-marketing studies for products approved between January 1998 and December 2000. Is it likely that this trend won't continue?

Overall, the rollout of MDII would seem to offer a key message to the industry, which is: "Stay tuned" – to assess actual results.