A Harvard University professor defended research suggesting the FDA was too hasty in approving drug, which the agency has criticized as mistaken, The Wall Street Journal writes. The professor, Daniel Carpenter, acknowledged that the FDA had identified mistakes in his study, which was published by the New England Journal of Medicine in March.

But he said that the FDA’s own records also had errors, and that after he made necessary corrections, his original finding still held: Drugs the FDA has approved soon before a regulatory deadline have been more likely to develop severe safety problems later on, the paper continues.

Between 1993 and 2005, the 88 drugs approved near the FDA’s deadline had a 15 percent chance of being flagged for severe safety problems, either a Black Box warning or a complete withdrawal for safety reasons. For 226 other drugs, that rate was just 5 precent, according to Carpenter’s revised analysis, which the New England Journal published this week.

The discrepancy is statistically significant but less dramatic than his original finding: a 14 percent rate of severe safety problems for drugs approved soon before a regulatory deadline, and 3 percent for other drugs, the Journal writes.

The agency has maintained that its own internal database showed only a modest increase in the rate of safety problems for such drugs, and is too small to be statistically significant, the paper adds. “We don’t think their conclusions will hold up,” an agency economist, Clark Nardinelli, said at the time. The New England Journal published a letter* from Nardinelli this week.

But Carpenter stood by his conclusions, releasing a 151-page memo and his raw data on the Internet, according to the Journal. After a detailed comparison, both sides’ records turned out to have been imperfect. “I’ll frankly admit that we omitted four or five of the black-box warnings,” Carpenter tells the Journal, adding that the FDA “missed a whole bunch of black-box warnings,” as well.

The findings suggest that the agency may have been under stress to meet deadlines imposed by agreements with the pharmaceutical industry since 1992, the Journal writes. In exchange for agreeing to quick approval of drug applications, the FDA gained the legal authority to collect substantial fees from companies with business before the agency.

Industry fees now account for about half of the $680 million the FDA will spend on reviewing new drugs this year. Currently, the agency promises to reach a decision on 90 percent of new drug applications within 10 months after a company submits an application, the Journal notes.

* a subscription to the NEJM may be required