FDA’s Deadline Problem

The Wall Street Journal January 3rd, 2020

More drugs are approved in December than any other time of year—but the burst of activity has an unintended side effect: The end-of-year approvals are later associated with more hospitalizations, life-threatening events and deaths, according to a new study.

“We see about twice as many adverse effects,” said Lauren Cohen, a professor of finance and entrepreneurial management at Harvard Business School who, with colleagues at MIT and the University of Texas at Dallas, documented the pattern.


They found similar bumps, along with the same safety implications, at the end of each month. The surges, they say, are not driven by explicit deadlines, leading the researchers to suspect the increases are motivated by self-imposed “desk-clearing” benchmarks.

The Food and Drug Administration acknowledged the December jump but said the pattern has diminished in recent years. “We were seeing almost 40% of approvals happening in December in the 1980s and ’90s,” FDA spokesman Nathan Arnold said. “If you just look since PDUFA was approved in 1992, it’s a much more even distribution.”


PDUFA—the Prescription Drug User Fee Act—mandates that each drug-approval request is assigned a “goal date,” or deadline, for the FDA to complete its review, Mr. Arnold said. Applications for the approval of standard new drugs must be reviewed within 10 months; priority drugs must be evaluated within six months. Mr. Arnold stopped short of dismissing the study’s findings regarding adverse events, but said the FDA has not detected a related increase.

“While we cannot speak directly to the results on informal deadlines highlighted in this study, FDA has—on multiple occasions—investigated a closely related issue, which is the relationship between formal PDUFA deadlines and postmarket safety of drugs,” Mr. Arnold said. “We have not found evidence of such a relationship.”


Other researchers have made that connection. A decade ago, Daniel Carpenter, a professor of government at Harvard University and author of “Reputation and Power: Organizational Image and Pharmaceutical Regulation at the FDA,” demonstrated that new drugs approved just before FDA deadlines were three to five times as likely to encounter measurable safety problems once on the market. “Compared to non-deadline approvals, deadline approvals were much more likely to be withdrawn later, have a black-box warning or have dosages quietly discontinued,” Dr. Carpenter said.


Black-box warnings are the strictest cautions that appear on the package insert of prescription drugs. In later studies, Dr. Carpenter and his research partners found that deadline approvals were also associated with more safety alerts and higher levels of adverse events, which the FDA defines as any untoward medical occurrence linked to the use of a drug. Dr. Carpenter also observed what he called the “December effect,” a swell of approvals in the final month of the year.


Now, according to the FDA, only 12% of approvals for “new molecular entities”—drugs that have never previously been approved in any form—occur in December. These drugs are a subset of what Dr. Cohen and his research partners examined. They found that 15% of all new drugs are approved in December, an increase of around 80% relative to the average month.


To arrive at their conclusions, they analyzed all new drug applications approved by the FDA between 1980 and 2016—a total of 3,312 drugs. For international comparisons, they used drug approvals obtained from the data analytics firm Clarivate Analytics Cortellis. The analysis is based on absolute counts of approvals and adverse effects. The researchers don’t know how many people used each drug, but they control for the disease it is meant to treat, market size and priority-review status.


To help test their desk-clearing theory, they checked to see whether drug approvals also surged just before the holidays. In the U.S., the number of approvals increased ahead of Thanksgiving. In China and Japan, approvals rose ahead of the Lunar New Year. “That’s a nice check,” Dr. Carpenter, who has read the paper, said of the holiday test. “Thanksgiving matters in the U.S. where you think it would, but does it matter in other countries? It’s negative in Japan. You wouldn’t expect it to matter there.”


The researchers were careful to say their findings don’t mean drugs approved right before a deadline should not have been given the green light since they don’t take into account how many people benefited from having access to the medicines. But they concluded that drugs approved during end-of-month or end-of-year surges should receive additional scrutiny. In particular, they suggested December decisions could be reviewed in the coming year before providing final approval. That prescription might prove to be a bitter pill to swallow.


It would slow things down, when one goal of the FDA drug-approval process is to speed things up.