FDA MEDICAL DEVICE 510(k) "NONAPPROVAL" DECISION

FDA MEDICAL DEVICE 510(k) "NONAPPROVAL" DECISION will stop the "endless loop" of FDA supplemental data requests for deficient 510(k) submissions, FDA Center for Devices and Radiological Health Director Bruce Burlington told executives at a March 30 session of the Health Industry Manufacturers Association annual meeting in Palm Beach, Florida. Burlington said that FDA will deem a 510(k) "nonapprovable" if, after comprehensive review, it "needs a major fix such as product redesign or a major study." In his first public speech, Burlington explained that "in the past, we didn't really have a mechanism to say no on a 510(k) application." About 2% of 510(k)s are "not substantially equivalent" because a premarket approval application should be filed instead, he noted; however, for products that are "still in the 510(k) arena," the agency has not had "a clear mechanism to say...this application is just so poorly worked up...Go back, redo it, and come in with a new 510(k)." In some cases, manufacturers have withdrawn 510(k)s, Burlington noted, but "sometimes [the firm] and we in FDA were just in a closed loop. We ask for...more information, and you really don't understand enough about the process to provide real fixes for the application." He maintained that the nonapproval decision "will be more efficient for us and...more efficient for" industry. "Wouldn't you rather have straight talk from the agency instead of an endless loop of requests for information that really is a 'no' in disguise?" The nonapproval decision is one of a number of steps CDRH is taking "to make the best use of its resources" as it works to "improve predictability and consistency" in device reviews and reduce delays in the review process, Burlington asserted. The CDRH director and outgoing Office of Device Evaluation Acting Director Alan Andersen also told the HIMA execs about resource reallocation plans, as well as other resource-saving initiatives (see stories, p. 3 & 5). Burlington emphasized that the nonapproval decision would apply to 510(k)s that have made it through an initial "filing screen" and have undergone a full-scale review. The agency is planning to set up a filing review system for 510(k)s that will be similar to its existing PMA filing program. "We are going to quit babysitting low-quality applications through the review process," Burlington said. Therefore, at an early stage in application review, FDA will determine if the submission is "administratively complete and if there's a reasonable chance of getting through the process for full evaluation. If not, we intend to return it right away and tell you why." He maintained that "refusing to review applications that have little chance of being approved is critical to using our resources more efficiently." Burlington, a physician who became CDRH director on Feb. 22 after working for several years in FDA's biologics and drug programs, also told the device industry topsiders about his philosophy regarding the need for clinical data in device applications. "Some of you have worried that an MD from the Center for Drugs would expect a 2,000-patient, double-blind, randomized controlled trial for every device...or would edge up to that sort of standard. That's what I've heard called the drugs model." He asserted that CDRH does not intend "to require controlled clinical trials for devices" nor does it "expect 2000-patient double-blind randomized controlled clinical trials for devices." He added that "in fact, clinical trials aren't necessary for most devices. However, when clinical trials are required, FDA is going to demand that they are "good ones -- not necessarily big ones, just good ones. They must be well-designed, ask the right questions," use "good scientific methods" and provide "high- quality data on safety and effectiveness." The recently released report of FDA's Committee for Clinical Review ("the Temple group") concluded that the sponsors of the approval applications it assessed had failed to adequately consider the use of randomized trials ("The Gray Sheet" March 8, p. 1). Burlington told HIMA that he believes that the Temple report "is an important assessment of some...applications, with excellent recommendations." However, he stressed that the findings, which were based on a review of several problematic applications, cannot be "generalized." The study also was done from "a drugs perspective," Burlington noted, adding that "we need to determine what...a device model of product development" is. The device center director said that he has asked Robert Eccleston, special assistant to the director, and Philip Phillips, director of ODE's program operations staff, to conduct a followup "targeted, retrospective review" of a limited number of device applications to see if the problems observed in the Temple report are more widespread.