Originally published March 1996
James G. Dickinson
When it comes to regulatory gray zones, a company's natural tendency to avoid what it considers needless expense can produce a penny-wise, pound-foolish result--a hazard that may be diminished if FDA's new culture of openness is successful. A recent case in point is the trouble and embarrassment that 31-employee Spirometrics Medical Equipment (Auburn, ME) encountered when it decided not to submit an expensive 510(k)--and thus suffer the attendant sales-losing delay--when it developed a new, compact version of its Flowmate pulmonary function spirometer. The labyrinthine FDA bureaucracy's apparent inability to advise the firm when agency enforcers contradicted the no-file decision ultimately led to public condemnation of the company's product as "misbranded."
In making its flawed decision, Spirometric's management first consulted the Center for Devices and Radiological Health's (CDRH) April 8, 1994, draft guidance, "Deciding When to Submit a 510(k) for a Change to an Existing Device," and, as many firms might do in such circumstances, decided to err on the side of its own economic best interests, wishfully interpreting the guidance as meaning that the company only needed to document the device as altered for its own files.
The original Flowmate operated on either 120- or 240-V current, while the compact FlowmateLTE was designed for nickel- cadmium batteries. The company apparently neglected to take full note of the fact that the governing statute (21 USC 807.81 (a)(3)(i)) requires a new premarket notification (510(k)) for "a significant change or modification in . . . energy source," a requirement embellished in FDA's April 1994 draft guidance, which shows a flowchart that reaches the same conclusion for a change in "energy type" and gives as an example "a change from ac to battery power."
The same guidance also says that changes in "ergonomics" might also require a 510(k) in the event of a change in operating instructions. This was the case with the LTE, which came with its own 80-page users manual. In addition, the size of the device was reduced and a few reporting functions were deleted. All these modifications required more consideration in applying the guidance than the company apparently gave it. (On the company's side, the guidance conceded that the adjective "significant" can "lead to subjective interpretations," and did not make clear that the burden is on the manufacturer to justify a decision not to file. However, it also advised that when doubt exists, "manufacturers are invited to discuss these decisions with the Agency's Offices of Compliance and Device Evaluation.")
Spirometric's first inkling that its decision not to file was in error came in a July 1994 letter from Karen Stutsman of the Office of Compliance at FDA's Rockville, MD, headquarters, asking why no FlowmateLTE 510(k) had been filed. The company answered in two weeks with a detailed explanation of its analysis under the April 1994 guidance.
FDA did not respond to Spirometric's explanation for 10 months. Then, in May 1995, an inspector arrived in Auburn to take statements about the FlowmateLTE. Management saw trouble and hired a lawyer, Deborah Brand Baum of Shaw, Pittman, Potts & Trowbridge (Washington, DC), but her efforts--a telephone tour she described to FDA in later correspondence--were entirely unavailing. She first reached Edward Kuntze at the Office of Compliance and was referred to Michael Gluck, a spirometer expert. His name led inexorably into the Office of Device Evaluation's voice information processing system, but in fact he had left the agency. Another potential FDA contact, Frank Gessing of the Boston district office, was spending the summer helping prosecute a lengthy federal criminal case. Attorney Baum then wrote to David Zier, at the time another respiratory-device evaluator in Rockville, to restate the problem and request a meeting. They never met.
Then, on August 28, 1995, acting as if Spirometrics had made no efforts at all to address FDA's concerns, the agency's Boston office issued a warning letter signed by district director James Rahto, telling the firm--and all others who might see the letter under the Freedom of Information Act--that the LTE was misbranded for lack of a 510(k) "or other information." For a small company with General Services Administration contract approval, not a happy day.
Spirometrics has since solved its problems by filing the 510(k). Further, FDA's April 1994 guidance has been clarified by an August 1, 1995, revision acknowledging earlier confusion and flatly stating: "Manufacturers have the responsibility for determining when a change to their product is significant enough to merit submission of a premarket notification." The revision also makes it much clearer that a 510(k) will be the "likely result" of a change in power source, this time detailing the reasons--namely, that a change from ac to battery power often means portability, new environments, and new instructions.
FDA's regulatory policy can sometimes be hard to pin down. Take, for instance, the aggressive instruction to investigators in subchapter 150 of its Investigations Operations Manual (IOM): "All regulatory procedures are designed to discover and develop evidence of violations with the ultimate aim of assuring compliance with the law and protection of the public." With an attitude like that, an industry representative demanded at a recent grassroots regulatory partnership follow-up meeting at FDA headquarters, how can FDA expect to nurture a cooperative response from companies? "You're absolutely right," commiserated agency chief counsel Margaret Jane Porter, as others (including Commissioner David Kessler) nodded their agreement that this attitude must be fixed.
It turned out, however, that at least this particular expression of the attitude was actually "fixed" three years ago, when the offending words were expunged in a massive streamlining of the IOM. But nobody ever told anybody about it.
Subchapter 150 (Regulatory Procedure), which contained the offending passage, has not existed in the IOM since the 1992 edition was superseded in August 1993 with a staple-bound edition that reflected wholesale deletions and reorganization as well as rewriting of instructions to the field. In each new edition since then, the subchapter has failed to return. "We probably took it out because, as I read it, it didn't make sense to me," Office of Regulatory Affairs compliance policy director David Haggard explained in December 1995, struggling to remember the IOM editing process as he supervised it three years earlier as director of field investigations. The whole subchapter, minus the offending words and possible redundancies, was absorbed into chapter 7 (Regulatory).
"I don't think we were writing new policy," Haggard reflected, confessing that in one chapter alone he probably deleted 40% of the words. "If you went out to investigators all around the country and asked, 'Are all regulatory procedures designed to discover and develop evidence?' they would say, 'What do you mean, "all regulatory procedures"?' There are a lot of regulatory procedures--for instance, recalls--that are not designed to develop evidence against a firm. It sounds like a grand statement, but when you really analyze it from a practical viewpoint, it doesn't say anything to the investigator."
However, those words, which Haggard speculated might have rested largely unnoticed in the IOM for 20 or 30 years, said a lot to industry. They explained every belligerent attitude and posture of FDA investigators and cemented a whole industry mindset regarding the hostility and untrustworthiness of FDA as an institution. Perhaps the time has come for FDA to tell everyone (especially its investigators) that the words were expunged, and why.