A Medical Device Daily

It is no surprise that private industry is not always happy with how the FDA’s accountants handle their calculations of cost. And a recent report by the Government Accountability Office (GAO) makes clear that industry does indeed scrutinize this aspect of FDA’s bookkeeping, but that GAO thinks the agency’s practices are up to snuff.

The cover letter to a June 25 GAO report notes that FDA is required to file annual reports on the device user fee program and that the agency “supplemented these MDUFMA reports by reporting more detailed information on the estimated average cost of reviewing medical device applications.”

GAO states, however, that “industry has challenged the appropriateness of the methodologies FDA used to identify the total cost of the process for reviewing medical devices,” thus questioning the average cost as well as the “degree to which staffing of the device review process has increased.”

GAO said its personnel met with staff from FDA’s Office of Management and Systems “who are responsible for preparing the annual MDUFMA reports.”

The GAO says its staffers did not attempt to establish whether the raw numbers were accurate, but it determined that “FDA’s methodologies for identifying its annual costs of reviewing device applications and for allocating the costs used in calculating the average ... are consistent with federal cost accounting standards.”

The report does not specify the nature of the conflict between FDA and industry on the cost calculation methods, but a footnote indicates that “costs represent obligations recorded at the end of the fiscal years — regardless of whether related expenditures have been made — because cost information” is the basis for MDUFMA reports. It says that the FDA believes that “obligations represent a reasonable estimate of cost because FDA’s financial records have historically shown that more than 81% of obligated amounts are expended within one year, and 96% within two years.”

The GAO says that FDA hired “a private contractor to develop a methodology to calculate the average cost of different application types for fiscal years 2003 through 2005.” It said that the FDA’s work met federal accounting standards which went into force in 1998.

The GAO report also says that it was not within its purview to “determine whether the amounts of the user fees for medical device applications are appropriate.”

As to device review staffing, FDA started the review period, FY02, with 917 full-time equivalents (FTEs), increasing to 943 the following year. In both those years, almost 90 of those personnel were contract employees.

In FY04, 102 of the 1,017 FTEs, about 10%, dedicated to device applications were contract employees. The following fiscal year, 158 of the 1,192 FTEs for device review were contract employees, about 13% of the total.

The total FTEs fell slightly the following fiscal year to 1,181, with 161 contract employees. The report notes that “FDA imposed a hiring freeze on CDRH,” triggered by the possibility that FDA “would not have had the authority to continue assessing and collecting user fees” if appropriations excluding user fees did not meet the amounts required to allow the agency to continue to collect user fees.

FDA hitches warning to Staar

A June 26 warning letter to Staar Surgical (Monrovia, California) was probably one of the more ironic warnings in recent memory, given FDA’s widely known preference for randomized clinical trials. The agency cited the company for allowing one of the study sites involved in an IDE trial for the implantable contact lens known as the Toric Phakic to conduct a randomized trial with controls getting photorefractive keratectomy. However, the approved IDE trial was not a randomized trial.

According to the warning letter, the clinical investigator at the site told FDA that his work was done “by and with the consent” of the sponsor, which gained approval for the non-randomized study from the institutional review board in 2003.

The company’s April 5 response to the findings apparently indicated that “this issue was identified by the Staar clinical team” in a February 2006 visit to the site. Staar was said to be “in active discussion with [the] site monitor contractor” regarding “a very specific schedule of activities to be undertaken” to fix the problem, but FDA said that response lacked specifics regarding how the firm would prevent future occurrences, and asked for documentation to that end.

Staar also tripped on a checklist appearing on case report forms (CRFs). The CRFs used at one study site called for “direct and retro illumination” photographs prior to surgery, but the company’s response to the 483 argued that this was not a requirement of the approved study protocol. FDA’s response was that “if the requirement ... was not a part of the study, it should not have been included on the CRF.”

The warning letter also cited Staar for failure to report unexpected adverse events within 10 days in the case of a study subject who experienced a subcapsular cataract in a treated eye in 2004. Another patient experienced an enlarged pupil and a darkened iris in January 2003. The warning letter stated that the medical reviewer did not review the event until the following January.